iSeries
Backup and Recovery
Version 5
SC41-5304-05



iSeries
Backup and Recovery
Version 5
SC41-5304-05
 Note
Before using this information and the product it supports, be sure to read the information in
“Appendix I. Notices” on page 893.

Sixth Edition (May 2001)

This edition applies to version 5, release 1, modification 0 of IBM® Operating System/400® (product number
5722-SS1) and to all subsequent releases and modifications until otherwise indicated in new editions. This edition
applies only to reduced instruction set computer (RISC) systems.
This edition replaces SC41-5304-04.
© Copyright International Business Machines Corporation 1997, 2000, 2001. All rights reserved.
US Government Users Restricted Rights – Use, duplication or disclosure restricted by GSA ADP Schedule Contract
with IBM Corp.
 Contents
Figures . . . . . . . . . . . . . . xiii Chapter 2. Saving your Server . . . . 15
Saving your server with the GO SAVE command . . 15
Tables . . . . . . . . . . . . . . . xv Overview of the GO SAVE command menu
options . . . . . . . . . . . . . . . 17
Change Save menu defaults with GO SAVE:
About Backup and Recovery,
Option 20 . . . . . . . . . . . . . . 18
SC41-5304-05. . . . . . . . . . . . xix Save your whole server with GO SAVE: Option
Who should read this book . . . . . .. xix . . 21 . . . . . . . . . . . . . . . . 18
Prerequisite and related information . . .. xix . . Save system data with GO SAVE: Option 22 . . 19
Operations Navigator . . . . . . . . . . . xx Save user data with GO SAVE: Option 23 . . . 19
Installing Operations Navigator . . . . . . xx Saving parts of your server with other GO SAVE
Operations Console . . . . . . . . . . xxi command menu options . . . . . . . . . 20
How to send your comments . . . . . . . . xxi Using GO SAVE: Options 21, 22, and 23 . . . . 20
Using the ObjectConnect/400 Function . . . . . 29
Summary of Changes to Backup and Components of ObjectConnect/400 . . . . . 30
Recovery . . . . . . . . . . . . . xxiii Setting Up Your System to Use
Moving to online media... . . . . . . . . . xxiii ObjectConnect/400 . . . . . . . . . . . 30
How the System Runs an ObjectConnect
Command . . . . . . . . . . . . . . 30
Part 1. Designing Backup, Recovery, Using the ObjectConnect Commands . . . . . 32
and Availability . . . . . . . . . . . 1 Investigating ObjectConnect Problems . . . . 33
CPFAD84 Error Codes . . . . . . . . . . 33
Chapter 1. Options for Backup,
Recovery, and Availability–Overview . . 3 Part 3. Recovering Information on
Save and Restore Operations–Overview . . . . . 3 Your System . . . . . . . . . . . 35
Tape Units–Overview . . . . . . . . . . . 4
Automated Tape Library Systems–Overview . . . . 4
Alternate Installation Device-Overview . . . . . 4
Chapter 3. Restore Procedures–General
Optical Media Support–Overview . . . . . . . 4 Information . . . . . . . . . . . . . 39
Journal Management–Overview . . . . . . . . 5 The Relationship Between Save and Restore
Access-Path Protection–Overview . . . . . . . 5 Commands . . . . . . . . . . . . . . 41
System-Managed Access-Path Protection . . . . 5 What Happens When You Restore Objects . . . . 42
Explicit Journaling of Access Paths . . . . . . 6 Sequence for Restoring Related Objects . . . . . 45
Commitment Control–Overview . . . . . . . . 6 Putting Your System in a Restricted State . . . . 45
Auxiliary Storage Pools–Overview . . . . . . . 6 Reclaiming Storage . . . . . . . . . . . . 46
Mirrored Protection–Overview . . . . . . . . 7 How to Reclaim Storage . . . . . . . . . 46
Device Parity Protection–Overview . . . . . . . 8 Controlling Restoration of Security-Sensitive Objects 50
Comparison of Disk Protection Options . . . . . 8 How to Set the QMCHPOOL, QBASPOOL, and
Uninterruptible Power Supply–Overview . . . . . 9 QPFRADJ System Values for a Different Processor
Dual Systems–Overview . . . . . . . . . . 9 or Memory . . . . . . . . . . . . . . 53
ObjectConnect–Overview . . . . . . . . . . 10 Locked Objects While Restoring . . . . . . . 53
OptiConnect for OS/400–Overview . . . . . . 10 How to Verify That Objects Are Restored
Backup Recovery and Media Services for Successfully . . . . . . . . . . . . . . 54
iSeries–Overview . . . . . . . . . . . . 10 Recovery from an Unsuccessful Restore Operation 56
Tivoli Storage Manager–Overview . . . . . . . 11 Recovering from an Error While Restoring
Content Manager OnDemand for iSeries Libraries . . . . . . . . . . . . . . 56
Server–Overview . . . . . . . . . . . . 11 Recovering from an Error While Restoring DLOs 57
| Business Continuity and Recovery How to Perform a Normal IPL . . . . . . . . 58
| Services–Overview . . . . . . . . . . . . 11 Parallel Restore Operations . . . . . . . . . 59
Data Migration Service Offerings–Overview . . . 12 Recovery Considerations for Cryptographic Access
Provider . . . . . . . . . . . . . . . 60

Part 2. Saving Information on Your

Chapter 4. Selecting the Right
System . . . . . . . . . . . . . . 13 Recovery Strategy . . . . . . . . . . 61
Some Common Recovery Terminology . . . . . 61

© Copyright IBM Corp. 1997, 2000, 2001 iii

 Recovery Procedure for a Power Failure . . . . . 62 Using Options 22 and 23 from the Restore
Recovery Procedure for a System Failure . . . . 63 Menu–Checklist 26 . . . . . . . . . . 114
Recovery Procedure for a Program Failure or Recovering User Information Using Tapes from
Human Error . . . . . . . . . . . . . . 63 Operational Assistant Backup–Checklist 27 . . 117
Choosing the Recovery Procedure for a Disk Failure
or Disk Errors . . . . . . . . . . . . . 63 Chapter 5. Recovering the Licensed
Actions for load source disk unit Internal Code . . . . . . . . . . . 121
failure–Checklist 1 . . . . . . . . . . . 65
How to Prepare for Loading the Licensed Internal
Actions for load source disk unit
Code . . . . . . . . . . . . . . . . 122
failure–Checklist 2 . . . . . . . . . . . 66
Task 1–Getting Ready to Load the Licensed
Actions for load source disk unit
Internal Code . . . . . . . . . . . . 122
failure–Checklist 3 . . . . . . . . . . . 67
Task 2–Powering Down the System . . . . . 123
Actions for load source disk unit
Task 3a–Preparing the System to Perform an IPL
failure–Checklist 4 . . . . . . . . . . . 68
from an Alternate Device . . . . . . . . 124
Actions for load source disk unit
Task 3b-Preparing a Logical Partition to Perform
failure–Checklist 5 . . . . . . . . . . . 72
an IPL from an Alternate Device . . . . . . 125
Actions for non-load source disk unit failure or
Task 4–Loading the Licensed Internal Code from
disk units in user ASP disk failure–Checklist 6. . 76
Media . . . . . . . . . . . . . . . 125
Actions for non-load source disk unit
How to Load the Licensed Internal Code . . . . 130
failure–Checklist 7 . . . . . . . . . . . 77
How to Recover Your Logical Partition
Actions for non-load source disk unit
Configuration . . . . . . . . . . . . . 134
failure–Checklist 8 . . . . . . . . . . . 78
How to Recover Your Disk Configuration After
Actions for non-load source disk unit
Installing the Licensed Internal Code and
failure–Checklist 9 . . . . . . . . . . . 79
Initializing the System . . . . . . . . . . 135
Actions for non-load source disk unit
| How to Recover Your Disk Configuration Using
failure–Checklist 10 . . . . . . . . . . . 82
| Operations Navigator at DST . . . . . . . . 138
Actions for a failure in a user ASP disk
How to Recover Your Disk Configuration . . . . 142
unit–Checklist 11 . . . . . . . . . . . 86
How to Start Your System After Restoring Licensed
Actions for a failure in a user ASP disk
Internal Code . . . . . . . . . . . . . 145
unit–Checklist 12 . . . . . . . . . . . 86
Disabling and Enabling the High-Speed Feature on
Actions for a failure in a user ASP disk
the 2440 Tape Unit . . . . . . . . . . . 145
unit–Checklist 13 . . . . . . . . . . . 88
Actions for non-load source disk unit
failure–Checklist 14 . . . . . . . . . . . 90 Chapter 6. Restoring the Operating
Actions for non-load source disk unit System . . . . . . . . . . . . . . 147
failure–Checklist 15 . . . . . . . . . . . 91 Choosing the Right Procedure for Restoring the
Actions for non-load source disk unit Operating System . . . . . . . . . . . . 148
failure–Checklist 16 . . . . . . . . . . . 92 How to Load the Operating System Using a
Actions for independent ASP disk Manual IPL . . . . . . . . . . . . . . 149
failure–Checklist 17 . . . . . . . . . . . 93 How to Restore the OS/400 Licensed Program . . 149
| Actions for a failure in an independent ASP disk Task 1–Starting to Restore the Operating System 150
| unit–Checklist 18 . . . . . . . . . . . 93 Task 2–Selecting the Installation Options . . . 154
| Actions for a failure in an independent ASP disk Task 3–Selecting IPL Options . . . . . . . 158
| unit–Checklist 19 . . . . . . . . . . . 94 Task 4–Setting Major System Options . . . . 160
Recovering your entire system after a complete Task 5–Defining or Changing the System at IPL 160
system loss–Checklist 20 . . . . . . . . . 96 Task 6–Completing the IPL . . . . . . . . 162
Recovering your entire system after a complete Recovering from SRC A900 2000 . . . . . . . 164
system loss including independent Creating a Configuration for 34xx Tape Units 165
ASPs–Checklist 21 . . . . . . . . . . . 98 Creating a Configuration for Other Tape Units 166
| Recovering your entire system after a complete
| system loss including independent ASPs, when Chapter 7. Starting the System After It
| Service Tools Network Interface is Ends Abnormally . . . . . . . . . . 167
| configured–Checklist 22 . . . . . . . . . 102 What Happens When Your System Stops . . . . 167
Restoring a Logical Partition to Another Logical Using the Disk Configuration Error Report
Partition—Checklist 23 . . . . . . . . . 105 Display . . . . . . . . . . . . . . 168
Choosing the Procedure to Recover User Using the Main Storage Dump Manager
Information . . . . . . . . . . . . . . 107 Occurred Display . . . . . . . . . . . 168
Recovering User Information Using How to Restart Your System . . . . . . . . 169
Commands–Checklist 24. . . . . . . . . 108 Task 1–Performing an Attended IPL . . . . . 169
Using Option 21 from the Restore Task 2–Using the Edit Rebuild of Access Paths
Menu–Checklist 25 . . . . . . . . . . 112 Display . . . . . . . . . . . . . . 171

iv OS/400 Backup and Recovery V5R1

 Task 3–Using the Edit Check Pending Chapter 9. The Restore Menu . . . . 207
Constraints Display . . . . . . . . . . 173 What the Restore Menu Options Do . . . . . . 208
Task 4–Recovering from Damaged Objects and How to Use Restore Menu Options 21, 22, and 23 208
Unreadable Sectors . . . . . . . . . . 174
Chapter 10. How to Restore Specific
Chapter 8. Recovering Information in Types of Information . . . . . . . . 213
a User Auxiliary Storage Pool . . . . 181 Recovering System Information . . . . . . . 213
Describing the Contents of Your User Auxiliary Sequence for Restoring Security Information . . . 213
Storage Pools . . . . . . . . . . . . . 181 Restoring User Profiles . . . . . . . . . . 214
Choosing the Procedure to Recover User ASPs . . 182 What Happens When You Restore User Profiles 215
How to Recover a User ASP After Recovering the What You Should Know About Restoring User
System ASP . . . . . . . . . . . . . . 182 Profiles . . . . . . . . . . . . . . 216
Task 1–Reclaiming Storage . . . . . . . . 183 How the System Establishes Ownership for
Task 2–Restoring User Profiles. . . . . . . 184 Restored Objects . . . . . . . . . . . 218
Task 3–Restoring the Configuration . . . . . 185 How the System Establishes the Authorization
Task 4–Recovering Journals and Journal List for a Restored Object . . . . . . . . 218
Receivers in the QRCL Library . . . . . . 185 How the System Establishes the Primary Group
Task 5–Restoring Libraries to the System for Restored Objects . . . . . . . . . . 218
Auxiliary Storage Pool . . . . . . . . . 186 Restoring Object Authorities . . . . . . . . 219
Task 6–Restoring Document Library Objects to Overview of Restoring Authorities . . . . . 219
the System Auxiliary Storage Pool . . . . . 187 Restoring Authority On a System in a
Task 7–Restoring User-Defined File Systems to Non-Restricted State . . . . . . . . . . 220
the User Auxiliary Storage Pool . . . . . . 187 Restoring Authority On a System in a Restricted
Task 8–Reclaiming Document Library Objects 188 State . . . . . . . . . . . . . . . 225
Task 9–Recovering Save Files from the QRCL What the System Does When You Restore
Library . . . . . . . . . . . . . . 188 Authority. . . . . . . . . . . . . . 225
Task 10–Associating Journal Receivers with How to Restore Configuration Objects . . . . . 227
Journals . . . . . . . . . . . . . . 189 Correcting Problems with the System Resource
Task 11–Restoring Object Ownership . . . . 190 Management Information . . . . . . . . 228
How to Recover An Overflowed User Auxiliary Recovering When You Change the Console Type 230
Storage Pool . . . . . . . . . . . . . . 191 Recovering the System/36 Environment
Resetting An Overflowed User Auxiliary Configuration . . . . . . . . . . . . 230
Storage Pool without an IPL . . . . . . . 192 Restoring Logical Partitions . . . . . . . . 231
Resetting An Overflowed User Auxiliary Restoring Libraries . . . . . . . . . . . 231
Storage Pool during an IPL . . . . . . . . 193 Restoring a Library From a Previous Release 232
How to Delete Overflowed Objects during Restoring Multiple Libraries . . . . . . . 232
Recovery . . . . . . . . . . . . . . . 196 Considerations and Restrictions . . . . . . 233
How to Recover a Damaged User Auxiliary How to Restore All Libraries from a Single Save
Storage Pool . . . . . . . . . . . . . . 196 Operation . . . . . . . . . . . . . 233
Task 1–Restoring User Profiles. . . . . . . 196 How to Restore All Libraries from Multiple
Task 2–Determining the Contents of the Lost Save Operations . . . . . . . . . . . 234
Auxiliary Storage Pool . . . . . . . . . 197 How to Restore Objects . . . . . . . . . . 234
Task 3–Determining Tasks to Restore Objects 197 | Restoring Objects That Are Journaled . . . . . 235
Task 4–Restoring Libraries to the User Auxiliary | What Happens When You Restore Journaled
Storage Pool . . . . . . . . . . . . . 198 | Objects to a Different Library . . . . . . . 235
Task 5–Restoring Journals to the User Auxiliary Restoring Database Files. . . . . . . . . . 236
Storage Pool . . . . . . . . . . . . . 198 Comparing File Attributes during a Restore
Task 6–Restoring Documents to the User Operation . . . . . . . . . . . . . 238
Auxiliary Storage Pool . . . . . . . . . 200 How the System Matches File Members during
Task 7–Restoring User-Defined File Systems to a Restore Operation . . . . . . . . . . 241
the User Auxiliary Storage Pool . . . . . . 200 Restoring Members to a File . . . . . . . 241
Task 8–Restoring Journal Receivers to the User Restoring Logical Files . . . . . . . . . 242
Auxiliary Storage Pool . . . . . . . . . 201 How the System Restores Access Paths . . . . 242
Task 9–Restore Save Files to the User Auxiliary How the System Restores Files with Shared
Storage Pool . . . . . . . . . . . . . 201 Formats . . . . . . . . . . . . . . 245
How to Remove a Failed Disk Unit from the How the System Restores Files with Referential
System ASP . . . . . . . . . . . . . . 202 Constraints . . . . . . . . . . . . . 245
Task 1–Access Dedicated Service Tools . . . . 202 How the System Restores Files with Triggers 247
Task 2–Delete the Auxiliary Storage Pool Data 203 Steps Before Deleting a Physical File . . . . 248
Task 3–Remove the Disk Unit from the Restoring Journals and Journal Receivers . . . . 248
Auxiliary Storage Pool Configuration . . . . 204

Contents v
 Restoring Journals . . . . . . . . . . . 248 Task 3–Determining Whether You Need to Apply
Steps before Deleting a Journal . . . . . . 249 Journaled Changes . . . . . . . . . . . 282
Restoring Journal Receivers. . . . . . . . 250 Task 4–Determining What Journal Receivers to Use 282
Steps before Deleting a Journal Receiver . . . 251 Task 5–Applying Journaled Changes for User
How the System Restores Programs . . . . . . 252 Journals . . . . . . . . . . . . . . . 284
Restoring Programs to a Different Release . . . 253 Task 6–Applying Journaled Changes for the
Restoring OPM Program Objects . . . . . . . 254 QAOSDIAJRN Journal . . . . . . . . . . 286
Restoring Save File Data. . . . . . . . . . 255 Task 7–Restoring Changed Documents and Folders 287
Restoring Spooled Output Files . . . . . . . 255
Restoring Licensed Programs . . . . . . . . 255 Chapter 12. Mirrored Protection
Restoring Documents and Folders . . . . . . 255 Recovery Actions . . . . . . . . . 289
RSTDLO Command Options . . . . . . . 255
System Actions for Permanent Errors . . . . . 289
Using Multiple Concurrent DLO commands . . 256
Suspending Mirrored Units. . . . . . . . . 290
Output from the RSTDLO Command . . . . 256
Resuming Mirrored Units . . . . . . . . . 291
Considerations and Restrictions . . . . . . 256
Replacing a Mirrored Unit . . . . . . . . . 292
Restoring Folders . . . . . . . . . . . 258
Using Spare Nonconfigured Units for
Renaming Documents When Restoring . . . . 258
Replacement. . . . . . . . . . . . . 294
Restoring OfficeVision/400 Mail and
Mirrored Protection Recovery Actions
Distribution Objects . . . . . . . . . . 258
Performed by the Service Representative . . . 296
How the System Restores Descriptive
Other Recovery Considerations for Mirrored
Information for DLOs . . . . . . . . . 259
Protection . . . . . . . . . . . . . 297
How the System Restores Authority and
Mirrored Protection Disk-Error Handling . . . 297
Ownership for DLOs . . . . . . . . . . 259
Missing Disk Units . . . . . . . . . . 298
When to Run the Rename Directory
Saving a Unit . . . . . . . . . . . . 299
(RNMDIRE) Command . . . . . . . . . 260
Restoring a Unit . . . . . . . . . . . 299
When to Run the Rename Document Library
Active Mirrored Load Source Failure . . . . 299
Object (RNMDLO) Command . . . . . . . 260
Unknown Unit 1 Status . . . . . . . . . 301
Recovery of Text Index Files for Text Search
Display Incorrect Licensed Internal Code Install 302
Services . . . . . . . . . . . . . . . 260
Restoring Objects in Directories . . . . . . . 261
Completing Recovery for the iSeries Integration for Chapter 13. How to Restore Your
Windows Server Product . . . . . . . . . 263 System Using Operational Assistant
Recovery for save performed with Integrated Tapes . . . . . . . . . . . . . . . 305
xSeries Server varied off . . . . . . . . . 263 How to Restore Your Libraries. . . . . . . . 306
Recovery for save performed with Integrated How to Restore Libraries That You Saved by Using
xSeries Server varied on . . . . . . . . . 264 a Backup List . . . . . . . . . . . . . 307
| Recovering Linux in a Partition . . . . . . 264 How to Restore Changed Objects That You Saved
Recovery Steps for OS/400 Enhanced by Using Operational Assistant . . . . . . . 308
Integration for Novell NetWare . . . . . . 264
Recovering a Domino™ Server. . . . . . . . 265 Chapter 14. How to Restore the
Recovering an entire Domino server . . . . . 265
System from the Save Storage Media . 311
Recovering Domino mail . . . . . . . . 266
Task 1–Powering Down the System and Loading
Recovering specific Domino databases . . . . 267
the Licensed Internal Code . . . . . . . . . 311
Restoring changed objects to a Domino server 267
Task 2–Restoring the Save Storage Tapes . . . . 312
Restoring a Windows server . . . . . . . . 270
Task 3–Responding to Messages . . . . . . . 314
Restoring the NWSD and disk drives for
Task 4–Completing the Restore Storage Operation 315
Windows server on iSeries . . . . . . . . 270
Task 5–Restoring Additional Information . . . . 318
Recovering Windows server files . . . . . . 273
Task 6–Restoring Program Temporary Fixes (PTFs) 318
Examples: How to address parts of Windows
How to Resume the Restore Storage Operation . . 318
server . . . . . . . . . . . . . . . 274
Restrictions When Using the Restore Command 275
How to Restore Program Temporary Fixes. . . . 278 Part 4. Release-to-Release Support 321

Chapter 11. How to Restore Changed Chapter 15. Release-to-Release

Objects and Apply Journaled Support . . . . . . . . . . . . . . 323
Changes . . . . . . . . . . . . . 279 Current Release-to-Previous Release Support . . . 323
Task 1–Restoring Changed Objects . . . . . . 280 Creating the Object for the Previous Release . . 324
Restoring Changed Objects by Library . . . . 280 Saving the Object for the Previous Release . . 325
Restoring Changed Objects Individually . . . 280 Testing the Object on the Current Release . . . 330
Task 2–Restoring Changed Objects in Directories 281

vi OS/400 Backup and Recovery V5R1

 Restoring and Using the Object on the Previous Commands and APIs for Journal Receivers . . 386
Release . . . . . . . . . . . . . . 331 Commands and APIs for Journal Entries . . . 386
Restrictions for Current Release-to-Previous Commands and APIs for Starting and Ending
Release Support . . . . . . . . . . . 331 Journaling of an Object . . . . . . . . . 387
Previous Release-to-Current Release Support . . . 331 | Commands and APIs for Getting Information
Considerations when Moving System | about Journaled Objects . . . . . . . . . 387
Customization Information . . . . . . . . 332 Commands for Recovering an Object by Using
Restoring previous release user data to a new the Journaled Changes . . . . . . . . . 388
system . . . . . . . . . . . . . . 332 Planning Journal Management. . . . . . . . 388
Restrictions when going from previous release | What Objects Should You Journal? . . . . . 389
to current release . . . . . . . . . . . 348 Should You Journal Objects That the System
Does Not Journal? . . . . . . . . . . . 390
Chapter 16. System Should You Use Access Path Journaling? . . . 391
Synchronization-Planning and How Should Objects Be Assigned to Journals? 391
Should You Journal Before-Images? . . . . . 392
Procedures . . . . . . . . . . . . 351 Should Your Journal Receivers Be in a User
Synchronization Methods: Overview . . . . . 352 ASP? . . . . . . . . . . . . . . . 393
Moving Changed Objects . . . . . . . . . 353 Journaling with the Save-While-Active Function 394
Steps for Saving Changed Objects . . . . . 354 Journaling and System Performance . . . . . . 394
Steps for Restoring Changed Objects . . . . 355 Journaling and Auxiliary Storage . . . . . . . 395
Problems When Restoring Changed Objects . . 357 How to Estimate the Size of a Journal Receiver 396
Moving Entire Libraries . . . . . . . . . . 358 Journaling Features That Increase the Journal
Considerations for Moving Entire Libraries . . 360 Receiver Size . . . . . . . . . . . . 397
Moving Individual Objects . . . . . . . . . 360 Setting Up Journaling . . . . . . . . . . 398
Applying Journaled Changes . . . . . . . . 361 How Journal Receivers Are Created . . . . . 399
Refreshing Your new system . . . . . . . . 363 How Journals Are Created . . . . . . . . 402
Additional Synchronization Tips . . . . . . . 363 Creating Journals and Journal
Receivers–Examples . . . . . . . . . . 409
Part 5. Considerations for Merging Starting Journaling . . . . . . . . . . . 411
Two or More Systems. . . . . . . 367 How to Start Journaling for Physical Files . . . 411
How to Start Journaling for DB2® Multisystem
Files . . . . . . . . . . . . . . . 412
Chapter 17. Tips for Merging Two How to Start Journaling for Access Paths . . . 412
RISC Systems Into a Single RISC | How to Start Journaling for IFS objects . . . . 413
System . . . . . . . . . . . . . . 369 | How to Start Journaling for Data Areas and
Guidelines for Restoring Information from the | Data Queues . . . . . . . . . . . . 414
Development System . . . . . . . . . . . 369 Why You Must Save Objects after Starting
Journaling . . . . . . . . . . . . . 414
Tasks for Managing Your Journaling Environment 415
Part 6. Journaling and Keep Records of Journaled Objects . . . . . 415
Commitment Control . . . . . . . 371 Evaluate How System Changes Affect
Journaling . . . . . . . . . . . . . 416
Chapter 18. Protecting Access Paths How to Manage Journals and Journal Receivers 416
Using System-Managed Access-Path Ending Journaling . . . . . . . . . . . . 423
How to End Journaling for DB2 Multisystem
Protection . . . . . . . . . . . . . 375
Files . . . . . . . . . . . . . . . 423
| Using SMAPP with Independent ASPs . . . . . 375
Managing IBM-Supplied Journals. . . . . . . 424
Benefits of SMAPP . . . . . . . . . . . 376
How SMAPP Works with Access Path Journaling 376
How the System Chooses Which Access Paths to Chapter 20. Working with Journal
Protect . . . . . . . . . . . . . . . 377 Entries, Journals, and Journal
How SMAPP Affects System Performance and Receivers . . . . . . . . . . . . . 427
Auxiliary Storage Use . . . . . . . . . . 377 About Journal Entries . . . . . . . . . . 427
Working with Recovery Times for Access Paths . . 378 Contents of a Journal Entry. . . . . . . . 428
How SMAPP Handles Changes in the Sending Your Own Journal Entries . . . . . . 429
Configuration of Auxiliary Storage Pools . . . . 380 Displaying and Printing Journal Entries . . . . 429
Output for Journal Entries Directed to a Work
Chapter 19. Planning and Setting Up Station . . . . . . . . . . . . . . 430
Journaling. . . . . . . . . . . . . 383 Output for Journal Entries Directed to a
Journal Management Commands and APIs . . . 385 Database Output File . . . . . . . . . . 431
Commands and APIs for Journals . . . . . 385 Using the Receive Journal Entry Command . . . 432

Contents vii
 Exit Program for Receiving Journal Entries . . 433 Auxiliary Storage Considerations . . . . . . . 494
Using the Retrieve Journal Entry Command . . . 436 Main Storage Considerations . . . . . . . . 494
Using the Retrieve Journal Entries Example Remote Journal Function Environments 494
(QjoRetrieveJournalEntries) API . . . . . . . 437 Data Replication Environment . . . . . . . 494
Working With Pointers in Journal Entries . . . . 437 Hot-Backup Environment . . . . . . . . 497
| Working with entries which contain minimized Displaying Remote Journal Function Information 499
| entry-specific data . . . . . . . . . . . . 438 Managing the Remote Journal Function . . . . 499
| Data Area Considerations . . . . . . . . 439 Keeping Records of Your Remote Journal
| Database Physical File Considerations . . . . 439 Network . . . . . . . . . . . . . . 499
Using Journaling to Provide an Audit Trail . . . 439 Evaluating How System Changes Affect Your
Comparing Journal Images . . . . . . . . . 440 Remote Journal Network . . . . . . . . 499
Journal Receiver Chains . . . . . . . . . . 440 Maintaining Journal Receivers . . . . . . . 500
Using the Work with Journal Attributes Command 442 Save and Restore Considerations . . . . . . 500
Working with the Receiver Directory . . . . 442 Working with Journal Entries and the Remote
Inoperable Journal Receivers . . . . . . . 444 Journal Function . . . . . . . . . . . . 505
Applying and Removing Journaled Changes . . . 444 Retrieving Journal Entries from a Remote
How Applying and Removing Journaled Journal with Library Redirection . . . . . . 506
Changes Works with Referential Constraints . . 445 System Dependent Information . . . . . . 506
How Applying and Removing Journaled Retrieving Journal Entries from a Remote
Changes Works with Trigger Programs . . . . 446 Journal During the Catch-up Phase . . . . . 506
Applying Journaled Changes . . . . . . . 446 Retrieving Journal Entries with Commitment
Apply Journaled Changes (APYJRNCHG) Control Considerations . . . . . . . . . 507
Command Examples . . . . . . . . . . 448 Confirmed and Unconfirmed Journal Entries 508
Removing Journaled Changes . . . . . . . 450 IPL Considerations . . . . . . . . . . . 509
Remove Journaled Changes (RMVJRNCHG) Main Storage Considerations during IPL . . . 510
Command Examples . . . . . . . . . . 451 Journal Receiver ASP Considerations . . . . . 511
Actions of the APYJRNCHG or RMVJRNCHG Example: Remote Journal Function Recovery . . . 511
Command by Journal Code. . . . . . . . 452 Example Scenario Description . . . . . . . 513
Work with Journal (WRKJRN) Command Options 455
Recovery Options . . . . . . . . . . . 456 Chapter 22. Commitment Control . . . 525
Recovery of a Journaled Object Using Journaled Commitment Control Introduction . . . . . . 525
Changes . . . . . . . . . . . . . . 462 Terms Used with Commitment Control . . . . . 526
Recovery after Abnormal System End . . . . 462 Commitment Control Startup . . . . . . . . 527
Recovering When a Journal Is Damaged . . . 464 Lock-Level Parameter . . . . . . . . . 527
Recovering When a Journal Receiver Is Notify Object Parameter . . . . . . . . . 530
Damaged . . . . . . . . . . . . . . 465 Commit Scope Parameter . . . . . . . . 533
Commit Text Parameter . . . . . . . . . 539
Chapter 21. Remote Journal Function 467 Default Journal Parameter . . . . . . . . 539
Introduction to the Remote Journal Function . . . 467 Omit Journal Entries Parameter . . . . . . 540
Benefits of the Remote Journal Function . . . 467 Two-Phase Commit–Overview. . . . . . . . 540
Remote Journal Function Configuration Placing Resources Under Commitment Control . . 541
Examples . . . . . . . . . . . . . . 468 Types of Committable Resources . . . . . . 541
Supported Communications Protocols . . . . 470 Location of Committable Resources . . . . . 543
Release to Release Considerations . . . . . 471 The Commit Protocol of a Committable
Planning for the Remote Journal Function . . . . 472 Resource . . . . . . . . . . . . . . 544
Determining Which Journals are Good The Access Intent of a Committable Resource 544
Candidates for Remote Journals . . . . . . 472 Files Being Journaled under Commitment Control 545
Determining Which Communications Protocol Sequence of Journal Entries Under Commitment
and Delivery Mode to Use . . . . . . . . 472 Control . . . . . . . . . . . . . . . 545
How Do the Journal Attributes Affect the Commit Cycle Identifier . . . . . . . . . . 547
Remote Journal Function? . . . . . . . . 473 Commit and Rollback Operations . . . . . . 547
Setting up the Remote Journal Function . . . . 473 Roles in Commit Processing . . . . . . . 548
Preparing to use Remote Journal Function . . . 473 How the System Performs a Commit Operation 549
Adding a Remote Journal . . . . . . . . 474 How the System Performs a Rollback Operation 550
Activating and Inactivating a Remote Journal 479 Flow of Commit Processing . . . . . . . 551
Activating and Inactivating a Local Journal . . 487 Errors During Commit Processing . . . . . 554
Summary of the Journal Type, Delivery Mode, Changing the Commitment Definition to Not
and Journal State Interactions . . . . . . . 488 Wait for Outcome . . . . . . . . . . . 554
Removing a Remote Journal . . . . . . . 490 Changing the Commitment Definition to Not
Troubleshooting Errors . . . . . . . . . 491 Select a Last Agent . . . . . . . . . . 557
Performance Considerations . . . . . . . . 492

viii OS/400 Backup and Recovery V5R1

 Changing the Commitment Definition to Install and Recovery Implications for Models
Indicate OK to Leave Out . . . . . . . . 558 600, 620, 720, and S20 . . . . . . . . . 635
Changing the Commitment Definition to Allow How to Set Up an Alternate Installation Device 635
Vote Read-Only . . . . . . . . . . . 560 How to Disable an Alternate Installation Device 638
Vote Reliable Affect on Flow of Commit Recovering Your System Using an Alternate
Processing . . . . . . . . . . . . . 562 Installation Device. . . . . . . . . . . 639
Retrieving Commitment Control Information . . . 564
States of the Logical Unit of Work (LUW) . . . . 564
Part 8. Disk Configuration and
XA Transaction Support . . . . . . . . . . 566
Commitment Control Recovery during Initial Protection — Procedures . . . . . 641
Program Load . . . . . . . . . . . . . 569
Making Heuristic Decisions and Cancelling Chapter 24. Procedures for
Resynchronization . . . . . . . . . . . . 570 Configuring Disks and Disk Protection 643
Implicit Commit and Rollback Operations . . . . 573 Choosing the Right Procedure for Configuring
Commitment Control Status . . . . . . . . 576 Disks . . . . . . . . . . . . . . . . 643
End Commitment Control (ENDCMTCTL) Configuring Disks on a New System–Checklist 1 644
Command . . . . . . . . . . . . . . 579 Adding Disk Units without Device Parity
Commitment Control during Activation Group End 580 Protection–Checklist 2 . . . . . . . . . 645
Commitment Control during Normal Routing Step Adding Disk Units to a 9337 Disk Array
End . . . . . . . . . . . . . . . . 581 Subsystem–Checklist 3 . . . . . . . . . 646
Commitment Control during Abnormal System or Adding a New 9337 Disk Array Subsystem and
Job End . . . . . . . . . . . . . . . 581 Starting Device Parity Protection on It–Checklist
Commitment Control during a Save-While-Active 4 . . . . . . . . . . . . . . . . 647
Operation . . . . . . . . . . . . . . 583 Adding a New 9337 Disk Array Subsystem That
Commitment Control Considerations and Has Device Parity Protection Already
Restrictions . . . . . . . . . . . . . . 584 Started–Checklist 5 . . . . . . . . . . 648
The Size of a Transaction . . . . . . . . 584 Adding Disk Units to an Existing Input/Output
Record Locking. . . . . . . . . . . . 585 Processor–Checklist 6. . . . . . . . . . 649
Minimizing Locks . . . . . . . . . . . 586 Adding a New Input/Output
Commitment Control for Batch Applications . . 588 Processor–Checklist 7. . . . . . . . . . 650
Performance Considerations for Commitment Moving Disk Units Between Non-Mirrored
Control . . . . . . . . . . . . . . 588 Auxiliary Storage Pools–Checklist 8 . . . . . 652
Miscellaneous Considerations and Restrictions Moving Disk Units Between Mirrored Auxiliary
for Commitment Control . . . . . . . . 591 Storage Pools–Checklist 9 . . . . . . . . 653
Commitment Control Errors . . . . . . . . 592 Deleting an Auxiliary Storage Pool–Checklist 10 654
Error Conditions . . . . . . . . . . . 592 Removing Disk Units Without Device Parity
Non-Error Conditions . . . . . . . . . 594 Protection–Checklist 11 . . . . . . . . . 655
Monitoring for Errors after a CALL Command 595 Removing Disk Units with Device Parity
Error Messages to Monitor for Commitment Protection (9337) from an ASP Without Mirrored
Control . . . . . . . . . . . . . . 595 Protection–Checklist 12 . . . . . . . . . 656
Normal Commit or Rollback Processing . . . 598 Removing Disk Units That Have Device Parity
Commit or Rollback Processing During Job End 599 Protection(9337) from an ASP with Mirrored
Commit or Rollback Processing During IPL . . 599 Protection–Checklist 13 . . . . . . . . . 657
Example of Using Commitment Control . . . . 599 Removing Disk Units That Have Device Parity
Example of Transaction Logging File . . . . . 602 Protection from an ASP Without Mirrored
Starting Application Programs Using a Notify Protection–Checklist 14 . . . . . . . . . 658
Object . . . . . . . . . . . . . . . . 608 Removing Disk Units That Have Device Parity
Using a Unique Notify Object for Each Program 609 Protection from an ASP With Mirrored
Using a Single Notify Object for All Programs 614 Protection–Checklist 15 . . . . . . . . . 659
Using a Standard Processing Program . . . . 614 Using System Service Tools and Dedicated
Commitment Control Practice Problem . . . . . 622 Service Tools . . . . . . . . . . . . 660
Steps Associated with Logic Flow . . . . . 631 How to Display Your Disk Configuration . . . 663

Part 7. Alternate Installation Chapter 25. Working with Auxiliary

Device . . . . . . . . . . . . . . 633 Storage Pools . . . . . . . . . . . 669
How to Add Disk Units to an Auxiliary Storage
Chapter 23. Using an Alternate Pool . . . . . . . . . . . . . . . . 669
How to Change the Storage Threshold for an
Installation Device . . . . . . . . . 635
Auxiliary Storage Pool . . . . . . . . . . 672
Alternate Installation Device-Overview . . . . . 635

Contents ix
How to Change the Storage Threshold for the What the System Does When You Start Mirrored
System Auxiliary Storage Pool. . . . . . . . 673 Protection . . . . . . . . . . . . . 722
How to Move a Disk Unit to a Different Auxiliary Mirrored Protection Configuration Errors . . . . 722
Storage Pool . . . . . . . . . . . . . . 676 How to Stop Mirrored Protection . . . . . . . 723
How to Remove a Disk Unit from an Auxiliary
Storage Pool . . . . . . . . . . . . . . 678 Chapter 28. Working with Disk
How to Delete an Auxiliary Storage Pool . . . . 680 Compression . . . . . . . . . . . 725
How to Calculate Space Requirements for an
Introduction to Disk Compression . . . . . . 725
Auxiliary Storage Pool . . . . . . . . . . 681
Restrictions and Considerations . . . . . . 725
Task 1–Calculating the Current Storage Used 682
Disk Compression and Capacity . . . . . . 726
Task 2–Calculating Storage Needed . . . . . 683
Disk Unit Full Considerations . . . . . . . 728
How to Display the Objects in a User ASP . . . 684
How The System Responds to Disk Unit Full 728
Balancing an Auxiliary Storage Pool . . . . . . 685
SRC Code A6xx 0277 . . . . . . . . . . . 729
Capacity Balancing . . . . . . . . . . 685
User Action 1 . . . . . . . . . . . . 730
Usage Balancing . . . . . . . . . . . 685
User Action 2 . . . . . . . . . . . . 731
Hierarchical Storage Management (HSM)
User Action 3 . . . . . . . . . . . . 731
Balancing. . . . . . . . . . . . . . 685
User Action 4 . . . . . . . . . . . . 731
Transferring Objects between Auxiliary Storage
Examples of A6xx 0277 . . . . . . . . . 732
Pools . . . . . . . . . . . . . . . . 686
How to Start Disk Compression . . . . . . . 732
How to Move Authorities to a Different ASP 686
How to Stop Disk Compression . . . . . . . 735
How to Transfer a Library to a Different ASP 687
Procedural Sequences for Configuring Disks and
How to Transfer a Folder to a Different ASP . . 687
Protection . . . . . . . . . . . . . . 737
How to Transfer Journals and Objects to a
Adding a New I/O Compression-Capable
Different ASP . . . . . . . . . . . . 687
Storage Controller . . . . . . . . . . . 737
How to Create Objects in a Library User ASP 688
Adding Disk Units to an Existing
How to Place Journal Receivers in a User ASP 689
Compression-Capable Storage Controller . . . 738
How to Move Journal Receivers From an
Moving Disk Units from the System ASP to a
Overflowed User ASP . . . . . . . . . 690
User ASP . . . . . . . . . . . . . . 739
How to Reset a Journal with a Status of
Recovering from Error Codes . . . . . . . . 740
Overflowed . . . . . . . . . . . . . 691
Recovering from SRC 6xxx 7051 . . . . . . 740
How to Work with Nonlibrary User ASPs . . . . 692
Recovering from SRC 6xxx 7052 . . . . . . 741
Creating Objects in a Nonlibrary User ASP . . 692
Transferring an Object to a Nonlibrary User ASP 693
Transferring a Journal to a Nonlibrary User ASP 693 Chapter 29. Managing Auxiliary
Storage Pools . . . . . . . . . . . 743
Chapter 26. Working with Device Working with ASP Trace and ASP Balance. . . . 743
Capacity Balance . . . . . . . . . . . 744
Parity Protection . . . . . . . . . . 695
Hierarchical Storage Management (HSM)
Starting Device Parity Protection . . . . . . . 695
Balance . . . . . . . . . . . . . . 745
How to Start Device Parity Protection for a 9337
Usage Balance . . . . . . . . . . . . 745
Disk Array Subsystem . . . . . . . . . 695
ASP Trace . . . . . . . . . . . . . 746
How to Start Device Parity Protection for an
Determining adequate disk storage . . . . . . 746
Input/Output Processor . . . . . . . . . 699
Stopping Device Parity Protection . . . . . . 701
How to Stop Device Parity Protection for a 9337 Chapter 30. Creating Logical
Disk Array Subsystem . . . . . . . . . 701 Partitions . . . . . . . . . . . . . 747
How to Stop Device Parity Protection on an Preparing for Logical Partitions . . . . . . . 747
Input/Output Processor . . . . . . . . . 704 Hardware and Software Requirements for
How to Include a Disk Unit in Device Parity Logical Partitions . . . . . . . . . . . 747
Protection . . . . . . . . . . . . . . 705 Before You Partition Your System. . . . . . . 748
How to Exclude a Disk Unit from Device Parity Creating a Logical Partition. . . . . . . . . 749
Protection . . . . . . . . . . . . . . 707
How to Display Device Parity Status . . . . . 708 Part 9. Backup and Recovery
How to Enable Disk Units Attached to the MFIOP
to Use Device Parity Protection . . . . . . . 710 Tools and Techniques . . . . . . 757

Chapter 27. Working with Mirrored Chapter 31. Techniques and

Protection . . . . . . . . . . . . . 719 Programming Examples for Backup
Mirrored Protection–Configuration Rules . . . . 719 and Recovery . . . . . . . . . . . 759
How to Start Mirrored Protection. . . . . . . 719 Job Recovery Considerations . . . . . . . . 759
Interactive Job Recovery . . . . . . . . . 759

x OS/400 Backup and Recovery V5R1

 Batch Job Recovery . . . . . . . . . . 760 Calculating the Time to Add Disk Units to the
Considerations for Using Save Files . . . . . . 761 System ASP . . . . . . . . . . . . . 796
Using Control Language (CL) Commands for Calculating the Time to Move Extents . . . . 797
Save Files . . . . . . . . . . . . . 762 Determining the Time to Synchronize the Disk 797
Save File Security . . . . . . . . . . . 762 9406 Model 500 Using the 6602 Disk Unit
Opening a Save File . . . . . . . . . . 763 Configuration . . . . . . . . . . . . . 798
Input and Output Operations on a Save File . . 763 Calculating the Time to Add Disk Units to the
File-Dependent Attributes for a Save File . . . 764 System . . . . . . . . . . . . . . 798
Damage to a Save File . . . . . . . . . 764 Calculating the Time to Move Extents . . . . 799
Clearing a Save File . . . . . . . . . . 764 Determining the Time to Synchronize the Disk 799
Sending Network Files . . . . . . . . . 764 Assessment of Single-Points-of-Failure . . . . . 800
Programming Examples for Backup and Recovery 765 System 1 . . . . . . . . . . . . . . 800
Retrieving the Device Name from Save System 2 . . . . . . . . . . . . . . 801
Completion Messages . . . . . . . . . 765 Full Mirrored Protection Versus Partial Mirrored
Displaying Status Messages When Saving . . . 766 Protection . . . . . . . . . . . . . . 803
Using the Retrieve Journal Entry (RTVJRNE)
Command in a Program . . . . . . . . . 766 Appendix D. Journal Entry Information 805
CL Program to Handle Escape Conditions . . . 767 Journal Codes . . . . . . . . . . . . . 806
Writing output to save media using the receive Journal Entries by Journal Code and Type . . . . 807
journal entry command . . . . . . . . . 769 The Fixed-Length Portion of a Journal Entry . . . 815
Variable-Length Portion of a Journal Entry . . . 821
Part 10. Appendixes . . . . . . . 773
Appendix E. How to Create and Use
Appendix A. Licensed Internal Code Output from the Save and Restore
Installation Error Screens . . . . . . 775 Commands . . . . . . . . . . . . 857
Format of the Output. . . . . . . . . . . 857
Appendix B. Example Disaster Header Information Format . . . . . . . 858
Command Information . . . . . . . . . 858
Recovery Plan . . . . . . . . . . . 783
Directory Information . . . . . . . . . 859
Section 1. Major Goals of a Disaster Recovery
Object Link Information . . . . . . . . . 860
Plan–Example . . . . . . . . . . . . . 783
Trailer Information . . . . . . . . . . 861
Section 2. Personnel–Example . . . . . . . . 783
Field Descriptions . . . . . . . . . . . 861
Organization Chart . . . . . . . . . . 784
Section 3. Application Profile–Example . . . . . 784
Section 4. Inventory Profile–Example . . . . . 784 Appendix F. Procedures for
Section 5. Information Services Backup Procedures 785 Recovering the Text Index . . . . . . 867
Section 6. Disaster Recovery Procedures . . . . 786
Disaster Action Checklist . . . . . . . . 786 Appendix G. Recovering your system 871
Recovery Start-Up Procedures for Use after
Actual Disaster . . . . . . . . . . . . 787 Appendix H. Using Operational
Section 7. Recovery Plan–Mobile Site . . . . . 787
Mobile Site Setup Plan . . . . . . . . . 788
Assistant to Save Information . . . . 885
Communication Disaster Plan . . . . . . . 788 Summary of Operational Assistant Commands and
Electrical Service . . . . . . . . . . . 788 Menu Options for Backup . . . . . . . . . 886
Section 8. Recovery Plan–Hot Site . . . . . . 789 Getting Started with Operational Assistant Backup 887
Hot-Site System Configuration . . . . . . 789 Defining What Should Be Saved . . . . . . . 888
Section 9. Restoring the Entire System . . . . . 789 How the System Saves Changed Objects Using
Section 10. Rebuilding Process. . . . . . . . 790 Operational Assistant Backup . . . . . . . . 889
Section 11. Testing the Disaster Recovery Plan . . 790 How to Define Backup Operations . . . . . . 890
Section 12. Disaster Site Rebuilding . . . . . . 792 How to Define When Backup Operations Are Run 891
Vendors . . . . . . . . . . . . . . 793
Floor Plan . . . . . . . . . . . . . 793 Appendix I. Notices . . . . . . . . . 893
Section 13. Record of Plan Changes . . . . . . 793 Trademarks . . . . . . . . . . . . . . 894

Appendix C. Example of Configuration Bibliography . . . . . . . . . . . . 897

Planning for Mirrored Protection . . . 795
9406 Model 500 Using the 9336 Disk Unit Index . . . . . . . . . . . . . . . 901
Configuration . . . . . . . . . . . . . 795

Contents xi
xii OS/400 Backup and Recovery V5R1
 Figures
1. Operations Navigator Display . . . . . . xx | 39. Hot-Backup Environment with Remote
2. Save commands and menu options . . . . 16 | Journal Function, and Application-Code
3. Save Menu–First Display . . . . . . . . 17 | Based Apply . . . . . . . . . . . . 468
4. ObjectConnect Job Flow . . . . . . . . 31 40. Broadcast and Cascade Remote Journal
5. Restore Procedures . . . . . . . . . . 40 Configurations . . . . . . . . . . . 469
6. Save procedures and restore procedures for file | 41. Typical Data Replication Environment with
systems . . . . . . . . . . . . . . 41 | Remote Journal Function. . . . . . . . 495
7. User ASP Configuration Before Failure 181 42. Typical Hot-Backup Environment with
8. User ASP Configuration After Restoring Remote Journal Function. . . . . . . . 498
Operating System . . . . . . . . . . 183 43. Journal Receiver Restore Characteristics and
9. User ASP Configuration After Reclaiming Remote Journal Types. . . . . . . . . 502
Storage . . . . . . . . . . . . . 184 44. Example Remote Journal Environment for
10. User ASP Configuration After Recovering Hot-Backup Recovery . . . . . . . . . 512
Isolated Journal Receiver. . . . . . . . 186 45. More unconfirmed journal entries present in
11. Restore Menu–First Display . . . . . . . 207 BJ1 than known in PJ1. . . . . . . . . 514
12. Sample Job Log for RSTAUT on a System in a 46. Switch-over processing. System S2 is now
Restricted State . . . . . . . . . . . 222 ready to allow applications to run. . . . . 516
13. Expanded Text for Message CPF3736 222 47. System S2 has assumed the role of the
14. Expanded Text for Message CPF3845 223 primary, and DB’ is now being updated. IPL
15. Sample Job Log for RSTAUT on a System in a processing has completed on system S1. . . 518
Non-restricted State . . . . . . . . . 223 48. State of the journals and databases just before
16. Expanded Text for Message CPF3845 224 starting to replay the changes back to the
| 17. Example: Restoring a Journaled Object to a original data, DB. . . . . . . . . . . 520
| Different Library . . . . . . . . . . 236 49. Preparing to let S1 again assume the role of
18. Example of a Database File with Two the primary system. . . . . . . . . . 522
Members . . . . . . . . . . . . . 237 50. Preparing to let S1 again assume the role of
19. Restoring a Copy of a File . . . . . . . 238 the primary system (continued). . . . . . 523
20. Restoring Database Files with Different 51. Overview of Commitment Control Process 527
Creation Dates . . . . . . . . . . . 239 52. Using multiple commitment definitions in a
21. Restoring Database Files with Different job . . . . . . . . . . . . . . . 538
Creation Dates . . . . . . . . . . . 240 53. Roles in Two-Phase Commit
22. Restoring Access Paths . . . . . . . . 244 Processing–Single-Level Tree . . . . . . 548
23. Restoring a Referential Constraint Network 246 54. Roles in Two-Phase Commit
24. An Object with Hard Links–Example 262 Processing–Multi-Level Tree. . . . . . . 549
25. An Object with a Symbolic Link–Example 263 55. Flow of Commit Processing without Any
26. Sample Recovery Time Line. . . . . . . 279 Optimization. . . . . . . . . . . . 552
27. Receiver Directory–Saving Attached Receivers 283 56. Flow of Commit Processing with Last Agent
28. Receiver Directory–Saving Detached Optimization. . . . . . . . . . . . 553
Receivers . . . . . . . . . . . . . 283 57. Wait for Outcome–Yes . . . . . . . . 556
29. How the System Is Saved with Operational 58. Wait for Outcome–No. . . . . . . . . 557
Assistant Backup . . . . . . . . . . 306 59. Flow of Commit Processing without Last
30. Recovery steps for restoring previous release Agent Optimization when agent votes OK to
user data to a new system . . . . . . . 333 leave out . . . . . . . . . . . . . 559
31. Overview of Synchronization Process 352 60. Flow of Commit Processing without Last
32. Edit Recovery for Access Paths Display 378 Agent Optimization when agent votes read
| 33. Journaling Overview . . . . . . . . . 385 only . . . . . . . . . . . . . . 561
34. First Parameter of RCVJRNE 61. Flow of Commit Processing with Vote
Command–Single-Entry Mode . . . . . . 435 Reliable optimization . . . . . . . . . 563
35. First Parameter of RCVJRNE 62. Routing Steps with Files under Commitment
Command–Block Mode . . . . . . . . 435 Control . . . . . . . . . . . . . 600
36. Creating a Journal Receiver . . . . . . . 441 63. Journal Entries for Transfer of Funds from
37. Work with Receiver Directory . . . . . . 443 Savings to Checking . . . . . . . . . 601
| 38. Hot-Backup Environment without Remote 64. Journal Entries for a System That Ended
| Journal Function, and Application-Code Abnormally . . . . . . . . . . . . 601
| Based Apply . . . . . . . . . . . . 468 65. Journal Entries for Rollback Changes 602
66. DDS for Physical File PRDMSTP . . . . . 603

© Copyright IBM Corp. 1997, 2000, 2001 xiii

67. DDS for Physical File ISSLOGP Used by 84. Display Hardware Resource Listing . . . . 664
ISSLOGP . . . . . . . . . . . . . 603 85. Sample Disk Configuration . . . . . . . 682
68. DDS for Logical File ISSLOGL . . . . . . 603 86. Program for Save Completion Messages 765
69. DDS for Display File PRDISSD Used in the 87. Program for Saving Source Files . . . . . 766
Program . . . . . . . . . . . . . 604 88. Program for Retrieving Journal Entries 767
70. Program Flow . . . . . . . . . . . 605 89. Example Program Prompts for Restoring the
71. RPG Program . . . . . . . . . . . 607 Required Receiver For an APYJRNCHG. . . 768
72. CL Program Used to Call RPG Program 90. Program for writing RCVJRNE output to save
PRDISS . . . . . . . . . . . . . 607 media . . . . . . . . . . . . . . 770
73. DDS for Physical File PRDLOCP . . . . . 609 91. 9406 Model 500 Original Configuration 796
74. DDS for Display File PRDRCTD . . . . . 610 92. 9406 Model 500 Mirrored Configuration 796
75. DDS for Notify Object and Externally 93. 9406 Model 500 Original Configuration 798
Described Data Structure (PRDRCTP) . . . 610 94. 9406 Model 500 Mirrored Configuration 798
76. Program Flow . . . . . . . . . . . 612 95. 9406 Model 500 with 9336 Disk Unit
77. RPG Source . . . . . . . . . . . . 613 Configuration . . . . . . . . . . . 800
78. Application Program Example . . . . . . 617 96. System 1–9406 Model 500 Configuration 801
79. Standard Commit Processing Program 620 97. System 2–9406 Model 500 Configuration 802
80. Initial Program Example . . . . . . . . 622 98. Output Sequence 1–SAV and RST Commands 857
81. RPG Source Program for ITMPTCS . . . . 624 99. Output Sequence 2–SAV and RST Commands 858
82. DDS for the Display File . . . . . . . . 629 100. Operational Assistant Save Options . . . . 886
83. Logic Flow of the Practice Problem . . . . 630 101. Change Monthly Backup Options Display 888

xiv OS/400 Backup and Recovery V5R1

 Tables
1. Comparison of Disk Protection Options . . . 9 31. Choosing the Correct Recovery Procedure for
2. Spooled Files Created by the server . . . . 27 User Information . . . . . . . . . . 107
3. ObjectConnect/400 and Associated iSeries Save 32. Checklist for Recovering User Information
and Restore Commands . . . . . . . . 29 Using Commands . . . . . . . . . . 109
4. Relationship Between Save and Restore 33. Checklist for Recovering User Information
Commands. . . . . . . . . . . . . 42 Using Option 21 . . . . . . . . . . 112
5. Restoring Objects with ALWOBJDIF . . . . 43 34. Checklist for Recovering User Information
6. Resolving Problems Detected by the RCLSTG Using Options 22 and 23 . . . . . . . . 114
Procedure . . . . . . . . . . . . . 48 35. Checklist for Recovering User Information
7. Lock Type Needed for Restore Operation 54 Using Operational Assistant Backup Tapes . . 117
8. Choosing the Correct Recovery Procedure for 36. Options from the Install the Licensed Internal
Disk Media Failure . . . . . . . . . . 64 Code (LIC) Menu . . . . . . . . . . 121
9. Recovery Checklist for Disk Failure–Checklist 37. SRC Codes When Loading the Licensed
1 . . . . . . . . . . . . . . . . 66 Internal Code . . . . . . . . . . . 127
10. Recovery Checklist for Disk Failure–Checklist 38. Configuring Disk While Installing the
2 . . . . . . . . . . . . . . . . 66 Operating System . . . . . . . . . . 152
11. Recovery Checklist for Disk Failure–Checklist 39. Recovery for Damaged Objects by Object
4 . . . . . . . . . . . . . . . . 69 Type . . . . . . . . . . . . . . 175
12. Recovery Checklist for Disk Failure–Checklist 40. Object Types That Require Special Procedures
5 . . . . . . . . . . . . . . . . 73 for Deleting . . . . . . . . . . . . 192
13. Recovery Checklist for Disk Failure–Checklist 41. Tasks for Restoring User ASP Objects 197
6 . . . . . . . . . . . . . . . . 76 42. Commands for Changing System Information 213
14. Recovery Checklist for Disk Failure–Checklist 43. How User Profiles Are Restored . . . . . 214
7 . . . . . . . . . . . . . . . . 77 44. Results of Restoring User Profiles . . . . . 216
15. Recovery Checklist for Disk Failure–Checklist 45. Restoring an Object Linked to an
8 . . . . . . . . . . . . . . . . 78 Authorization List . . . . . . . . . . 218
16. Recovery Checklist for Disk Failure–Checklist 46. How Configuration Objects Are Restored 227
9 . . . . . . . . . . . . . . . . 79 47. Methods for Restoring All Libraries–Single
17. Recovery Checklist for Disk Failure–Checklist Save Operation . . . . . . . . . . . 233
10 . . . . . . . . . . . . . . . . 82 48. Methods for Restoring All Libraries–Multiple
18. Recovery Checklist for Disk Failure–Checklist Save Operations . . . . . . . . . . 234
11 . . . . . . . . . . . . . . . . 86 49. Restoring a File Network . . . . . . . 244
19. Recovery Checklist for Disk Failure–Checklist 50. Restoring Files That Have Trigger Programs 247
12 . . . . . . . . . . . . . . . . 87 51. System Actions When Restoring Programs 252
20. Recovery Checklist for Disk Failure–Checklist 52. Restoring Objects That Have Hard Links 262
13 . . . . . . . . . . . . . . . . 89 53. Using the RST Command for QSYS.LIB
21. Recovery Checklist for Disk Failure–Checklist Objects. . . . . . . . . . . . . . 276
14 . . . . . . . . . . . . . . . . 91 54. New Name Options on the RST
22. Recovery Checklist for Disk Failure–Checklist Command–Examples . . . . . . . . . 277
15 . . . . . . . . . . . . . . . . 91 55. Restore Procedures for Changed Objects 280
23. Recovery Checklist for Disk Failure–Checklist 56. Handling Messages When Restoring Storage 315
16 . . . . . . . . . . . . . . . . 92 | 57. Values for TGTRLS Parameter . . . . . . 323
24. Recovery Checklist for Disk Failure–Checklist 58. Language Support for the Target Release
17 . . . . . . . . . . . . . . . . 93 Parameter. . . . . . . . . . . . . 324
| 25. Recovery Checklist for Disk Failure–Checklist | 59. Previous-Release Support by Object Type 326
| 18 . . . . . . . . . . . . . . . . 94 60. Comparison of Synchronization Methods 353
| 26. Recovery Checklist for Disk Failure–Checklist 61. System-Generated Receiver Names . . . . 399
| 19 . . . . . . . . . . . . . . . . 94 62. System-Generated Receiver Names . . . . 400
27. Recovery Checklist for Complete System 63. System-Created Journal Receiver Names 400
Loss–Checklist 20 . . . . . . . . . . 97 64. IBM-Supplied Journals . . . . . . . . 424
28. Recovery Checklist for Complete System 65. Methods for Working with Journal Entries 427
Loss–Checklist 21 . . . . . . . . . . 99 66. Actions by Journal Code and Entry Type 452
| 29. Recovery Checklist for Complete System 67. Journal Types and Associated Characteristics 476
| Loss–Checklist 22 . . . . . . . . . . 102 68. Summary of the Journal Type, Delivery Mode
30. Recovery Checklist for Complete System and Journal State Interactions . . . . . . 488
Loss–Checklist 23 . . . . . . . . . . 105 69. Lock Duration by Lock-Level Parameter 528

© Copyright IBM Corp. 1997, 2000, 2001 xv

 70. Specifying a Commit Identification . . . . 531 107. Hardware requirements for logical partitions 748
71. Commitment Definition Names for a Job 536 108. System Resource Worksheet. . . . . . . 750
72. Additional Examples of Multiple 109. Checklist for Testing the Disaster Recovery
Commitment Definitions in a Job . . . . . 539 Plan . . . . . . . . . . . . . . 790
73. Types of Committable Resources . . . . . 541 110. Journal Entries by Journal Code and Type 807
74. Local and Remote Committable Resources 543 111. Field Descriptions of the Fixed-Length
75. Coexistence Rules for One-Phase Resources Portion of a Journal Entry: *TYPE1 . . . . 815
and Two-Phase Resources . . . . . . . 544 112. Field Descriptions of the Fixed-Length
76. Access Intents for Resource Types. . . . . 544 Portion of a Journal Entry: *TYPE2 . . . . 818
77. Sequence of Commitment Control Journal 113. Field Descriptions of the Fixed-Length
Entries . . . . . . . . . . . . . . 546 Portion of a Journal Entry: *TYPE3 . . . . 819
78. Language Support for Commit and Rollback 114. Field Descriptions of the Fixed-Length
Operations . . . . . . . . . . . . 547 Portion of a Journal Entry: *TYPE4 . . . . 820
79. Logical Unit of Work (LUW) States . . . . 565 115. APYJRNCHG (B AJ, E EQ, F AY) and
80. Commit and Rollback Operations During Job RMVJRNCHG (E EX, F RC) Journal Entries . 822
End . . . . . . . . . . . . . . . 574 116. Change End of Data (F CE) Journal Entry 823
81. Commit and Rollback Operations during 117. CHGJRN (J NR, J PR) Journal Entries 823
Activation Group Ending . . . . . . . 576 118. COMMIT (C CM) Journal Entry . . . . . 823
82. Commit and Rollback Processing for Failures. 598 119. Delete Access Path (R PD) Journal Entry 823
83. Choosing the Right Disk Procedure . . . . 643 120. Delete Receiver (J RD, J RF) Journal Entries 824
84. Configuring Disks on a New System–Tasks 644 121. Force Data to Auxiliary Storage (F FD)
85. Adding Disk Units without Device Parity Journal Entry . . . . . . . . . . . 824
Protection–Tasks . . . . . . . . . . 646 122. INZPFM (F IZ) Journal Entry . . . . . . 824
86. Adding Disk Units to an Existing 9337 Disk | 123. IPL (J IA, J IN) and In-Use (B OI, D ID, E EI,
Array Subsystem–Tasks . . . . . . . . 647 | F IU, Q QI) Journal Entries . . . . . . . 824
87. Adding a New 9337 Disk Array Subsystem 124. Logical Unit of Work (C LW) Journal Entry 824
That Does Not Have Device Parity Protection 125. Logical Unit of Work (C LW) Journal
Started–Tasks . . . . . . . . . . . 648 Entry–Header Record . . . . . . . . . 825
88. Adding a New 9337 Disk Array Subsystem 126. Logical Unit of Work (C LW) Journal
That Has Device Parity Protection Already Entry–Local Record . . . . . . . . . 834
Started–Tasks . . . . . . . . . . . 649 127. Logical Unit of Work (C LW) Journal
89. Adding Disk Units to an Existing Entry–API Record . . . . . . . . . . 835
Input/Output Processor–Tasks . . . . . . 650 128. Logical Unit of Work (C LW) Journal
90. Adding a New Input/Output Processor–Tasks 651 Entry–DDL Record. . . . . . . . . . 838
91. Moving Disk Units Between ASPs–Tasks 652 129. Logical Unit of Work (C LW) Journal
92. Moving Disk Units Between ASPs with Entry–RMT Record . . . . . . . . . 841
mirrored protection–Tasks . . . . . . . 653 130. Logical Unit of Work (C LW) Journal
93. Deleting an Auxiliary Storage Pool–Tasks 654 Entry–DDM Record . . . . . . . . . 845
94. Removing Disk Units That Do Not Have 131. Moving and Renaming Objects (D FM, D FN,
Device Parity Protection–Tasks . . . . . . 655 E EM, E EN, F MM, F MN, F PM, F PN, Q
95. Removing Disk Units with Device Parity QM, Q QN) Journal Entries . . . . . . . 847
Protection from an ASP without Mirrored 132. File OPEN (F OP) and File CLOSE (F CL)
Protection–Tasks . . . . . . . . . . 656 Journal Entries . . . . . . . . . . . 847
96. Removing Disk Units from a 9337 Disk Array 133. Journal Code R, All Journal Entry Types
Subsystem and a Mirrored ASP–Tasks . . . 657 Except IL . . . . . . . . . . . . . 848
97. Removing Disk Units from an IOP and a 134. RGZPFM (F RG) Journal Entry. . . . . . 848
Non-Mirrored ASP–Tasks . . . . . . . 658 135. ROLLBACK (C RB) Journal Entry . . . . . 849
98. Removing Disk Units from an IOP and a 136. Object Restored (B FR, E EL, F MR, J RR, Q
Mirrored ASP–Tasks . . . . . . . . . 659 QZ) and Receiver Saved (J RS) Journal Entries 849
99. Worksheet for Calculating ASP Capacity 681 137. Object Saved (B FS, E ES, F MS, Q QY)
100. Calculating ASP Capacity–Example 2 683 Journal Entries . . . . . . . . . . . 850
101. Calculating ASP Capacity–Example 3 684 138. Start of Save-While-Active (B FW, E EW, F SS,
102. Word formats for SRC codes in V4R5. 729 Q QX) Journal Entries. . . . . . . . . 851
103. SRC code word formats in V4R4 and 139. Start Journal (B JT, E EG, F JM, Q QB) Journal
previous releases. . . . . . . . . . . 730 Entries . . . . . . . . . . . . . . 852
104. Adding a New I/O Storage Controller and 140. License Key Not Valid (L LK) Journal Entry 853
Disk Units . . . . . . . . . . . . 737 141. Usage Limit Changed (L LL) Journal Entry 853
105. Adding Compressed Disk Units to an 142. Usage Limit Exceeded (L LU) Journal Entries 853
Existing Storage Controller . . . . . . . 738 | 143. Update Data Area (E EA, E EB) Journal
106. Moving Disk Units from the System ASP to a | Entries . . . . . . . . . . . . . . 853
User ASP . . . . . . . . . . . . . 739

xvi OS/400 Backup and Recovery V5R1

| 144. Data Queue Cleared, has Key (Q QJ) Journal 151. Directory Information Output–SAV and RST
| Entry . . . . . . . . . . . . . . 854 Commands . . . . . . . . . . . . 859
| 145. Send Data Queue, has Key (Q QK) Journal 152. Object Link Information–Output from SAV
| Entry . . . . . . . . . . . . . . 854 and RST Commands . . . . . . . . . 860
| 146. Received Data Queue, has Key (Q QL) 153. Trailer Information–Output from SAV and
| Journal Entry . . . . . . . . . . . 854 RST Commands. . . . . . . . . . . 861
| 147. Send Data Queue, No Key (Q QS) Journal 154. Recovery for Search Index Services Files 867
| Entry . . . . . . . . . . . . . . 855 155. Type of Index Request Created when Using
| 148. Object Level (D AC, D CG, D CT, D DC, D the Restore Document Library Object
| DT, D GC, D GO, D GT, D RV, D TC, D TD, (RSTDLO) Command . . . . . . . . . 869
| D TG, F DM, F MC) Journal Entries . . . . 855 156. Summary of Operational Assistant Backup
149. Heading Output Information–SAV and RST Options . . . . . . . . . . . . . 887
Commands . . . . . . . . . . . . 858
150. Command Information Output–SAV and RST
Commands . . . . . . . . . . . . 858

Tables xvii
xviii OS/400 Backup and Recovery V5R1
 About Backup and Recovery, SC41-5304-05
This book provides general information about recovery and availability options for
the IBM iSeries server. It describes the options available on the system, compares
and contrasts them, and tells where to find more information about them.

This book release contains minimal information about how to back up your server.
Look for comprehensive information about backing up your server on the iSeries
Information Center at the following Web site:
http://www.ibm.com/eserver/iseries/infocenter.

This book provides information on the following topics:

v Procedures on how to save your system with the Save menu optoins of the GO
SAVE command.
v Restoring to different releases of OS/400
v Selecting the right recovery strategy
v Procedures for restoring information to your system
| v Journal management and object recovery
v Commitment control
v Device Parity Protection procedures
v Mirrored protection procedures
v Uninterruptible power supply
v Backup and recovery programming techniques

Who should read this book

This book is intended for someone who is assigned the responsibilities of backup
and recovery planning and recovering the system after a failure. You should be
familiar with the information contained in the Information Center Website before
using this book. If you know how to operate the system, you should be ready to
use this book.

Prerequisite and related information

Use the iSeries Information Center as your starting point for looking up iSeries
technical information. You can access the Information Center from the iSeries™
Information Center CD-ROM (English version: SK3T-2027) or from one of this Web
site:
http://www.ibm.com/eserver/iseries/infocenter

The iSeries Information Center contains important topics such as logical

partitioning, clustering, Java™, TCP/IP, Web serving, backing up your system, and
secured networks. It also contains Internet links to Web sites such as
iSeriesTechnical Studio. Included in the Information Center is a link that describes
at a high level the differences in information between the Information Center and
the Online Library.

For a list of related publications, see the “Bibliography” on page 897.

© Copyright IBM Corp. 1997, 2000, 2001 xix

Operations Navigator
Operations Navigator is a powerful graphical interface for Windows® clients. With
Operations Navigator, you can manage and administer your servers from your
Windows desktop.

You can use Operations Navigator to manage communications, printing, database,

security, and other system operations. Operations Navigator includes Management
Central for managing multiple servers centrally.

Figure 1 shows an example of the Operations Navigator display:

Figure 1. Operations Navigator Display

This new interface has been designed to make you more productive and is the
only user interface to new, advanced features of OS/400. Therefore, IBM
recommends that you useOperations Navigator, which has online help to guide
you. While this interface is being developed, you may still need to use a traditional
emulator such as PC5250 to do some of your tasks.

Installing Operations Navigator

To use Operations Navigator, you must have Client Access installed on your
Windows PC. For help in connecting your Windows PC to your server consult
Client Access Express for Windows - Setup, SC41-5507-02.

Operations Navigator is a separately installable component of Client Access that

contains many subcomponents. If you are installing for the first time and you use
the Typical installation option, the following options are installed by default:
v Operations Navigator base support
v Basic operations (messages, printer output, and printers)

To select the subcomponents that you want to install, select the Custom installation
option. (After Operations Navigator has been installed, you can add
subcomponents by using Client Access Selective Setup.)
1. Display the list of currently installed subcomponents in the Component
Selection window of Custom installation or Selective Setup.
2. SelectOperations Navigator.
3. Select any additional subcomponents that you want to install and continue with
Custom installation or Selective Setup.

xx OS/400 Backup and Recovery V5R1

 After you install Client Access, double-click the Operations Navigator icon on
your desktop to access Operations Navigator and create a server connection.

Operations Console
Many of the newer iSeries and AS/400e models do not have a Twinaxial
Workstation Controller, but are configured to use Operations Console instead.
Operations Console allows you to use a PC to access and control the system
console and the system control panel. If a system is configured to use Operations
Console, and that system undergoes a backup and recovery cycle, you will have to
perform the following steps:
1. Perform an initial program load (IPL) in Manual mode.
2. Use dedicated service tools (DST) to reconfigure the system so that it will
detect the PC console when you perform an IPL in Normal mode.
Detailed instructions on setting up Operations Console are in the document
Operations Console Setup.

How to send your comments

Your feedback is important in helping to provide the most accurate and
high-quality information. If you have any comments about this book or any other
iSeries documentation, fill out the readers’ comment form at the back of this book.
v If you prefer to send comments by mail, use the readers’ comment form with the
address that is printed on the back. If you are mailing a readers’ comment form
from a country other than the United States, you can give the form to the local
IBM branch office or IBM representative for postage-paid mailing.
v If you prefer to send comments by FAX, use either of the following numbers:
– United States and Canada: 1-800-937-3430
– Other countries: 1-507-253-5192
v If you prefer to send comments electronically, use one of these e-mail addresses:
– Comments on books:
[email protected]
– Comments on the iSeries Information Center:
[email protected]
Be sure to include the following:
v The name of the book or iSeries Information Center topic.
v The publication number of the book.
v The page number or topic to which your comment applies.

About Backup and Recovery, SC41-5304-05 xxi

xxii OS/400 Backup and Recovery V5R1
 Summary of Changes to Backup and Recovery
New and enhanced functionality that was added to the Operating System/400
licensed program for Version 5 Release 1 Modification 0. This additional function
affects backup, recovery, and availability.

Changes are identified by change bars to the left of the information.

Changes to this publication include, but are not limited to, support for the
following:
| v Expanded journaling that includes
| – data areas
| – data queues
| – IFS objects
| – additional journal attributes
| v Save and restore for digitally signed objects
| v Support for independent auxiliary storage pools
| v Support for save and restore of secondary partitions with Linux

Look for performance information that was included in previous versions of this
book in the Performance Capabilities Reference, SC41-0607 book. Access this book on
the Internet at this uniform resource locator (URL) address:
http://www.ibm.com/eserver/iseries/infocenter

Moving to online media...

| The iSeries Information Center contains comprehensive information for backing up
| your iSeries or AS/400e server. The hardcopy book contains basic information on
| how to use the Save menu options with the GO SAVE command. This allows you
| to save all or parts of your iSeries or AS/400e system.

You can download a printable Portable Document Format (PDF) version of each
topic in this section.

The iSeries Information Center contains the following topics for backing up your
system:
v Planning a backup and recovery strategy
v Setting up disk protection for your data
v Controlling system shutdown using a power-handling program.
v Getting your media ready to save your system
v Before you save anything...
v Saving your system with the GO SAVE command (also in “Saving your server
with the GO SAVE command” on page 15).
v Manually saving parts of your system
| v Saving your system while it is active
| v Saving to multiple devices to reduce your save window

Note:

© Copyright IBM Corp. 1997, 2000, 2001 xxiii

 Access the iSeries Information Center at the following Web site:
http://www.ibm.com/eserver/iseries/infocenter

xxiv OS/400 Backup and Recovery V5R1

 Part 1. Designing Backup, Recovery, and Availability
Chapter 1. Options for Backup, Recovery, and
Availability–Overview . . . . . . . . . . . 3
Save and Restore Operations–Overview . . . . . 3
Tape Units–Overview . . . . . . . . . . . 4
Automated Tape Library Systems–Overview . . . . 4
Alternate Installation Device-Overview . . . . . 4
Optical Media Support–Overview . . . . . . . 4
Journal Management–Overview . . . . . . . . 5
Access-Path Protection–Overview . . . . . . . 5
System-Managed Access-Path Protection . . . . 5
Explicit Journaling of Access Paths . . . . . . 6
Commitment Control–Overview . . . . . . . . 6
Auxiliary Storage Pools–Overview . . . . . . . 6
Mirrored Protection–Overview . . . . . . . . 7
Device Parity Protection–Overview . . . . . . . 8
Comparison of Disk Protection Options . . . . . 8
Uninterruptible Power Supply–Overview . . . . . 9
Dual Systems–Overview . . . . . . . . . . 9
ObjectConnect–Overview . . . . . . . . . . 10
OptiConnect for OS/400–Overview . . . . . . 10
Backup Recovery and Media Services for
iSeries–Overview . . . . . . . . . . . . 10
Tivoli Storage Manager–Overview . . . . . . . 11
Content Manager OnDemand for iSeries
Server–Overview . . . . . . . . . . . . 11
| Business Continuity and Recovery
| Services–Overview . . . . . . . . . . . . 11
Data Migration Service Offerings–Overview . . . 12

© Copyright IBM Corp. 1997, 2000, 2001 1

2 OS/400 Backup and Recovery V5R1
Chapter 1. Options for Backup, Recovery, and
Availability–Overview
Today, computer systems and the applications and information that they provide
are at the heart of most businesses. Without the critical functions they
support–customer service, sales, order processing, distribution,
manufacturing–business quickly comes to a halt. You need to plan backup,
recovery, and system availability to protect your ability to maintain crucial
business functions no matter what happens. Planning can mean the difference
between business success and business failure.

IBM iSeries and AS/400e servers offer a wide range of recovery and availability
options. Your hardware or software includes some of the options. Others are
ordered separately. They are intended to help you do the following:
v Make your save operations faster and more efficient.
v Keep your system available for your users.
v Plan and manage your backup and recovery.

This chapter provides an overview of the options. The iSeries Information Center
provides comparisons of the options. Use this information to decide what other
options beyond the basics might be useful, affordable, and practical for your
situation. The overview of each option tells you where you can find more
information if that option appeals to you.

The remaining parts of this book include the following topics:

v Recovering your system.
v Using journaling and commitment control.
v Configuring disks and disk protection.

Save and Restore Operations–Overview

Your Operating System/400 (OS/400) licensed program includes menus and
commands for save and restore. You can use the save operations and restore
operations on the system to do the following:
v Recover from a program or system failure.
v Exchange information between servers.
v Store infrequently used objects offline.

You can use commands and menu options to save individual objects and groups of
objects. You can use some save and restore operations while your system is active.
Other save and restore operations require that no other activity is occurring on the
system.

You can save and restore objects by using diskette, magnetic tape, optical media, or
a save file. You can also use communications capabilities or an optical connection
to save and restore objects with another system.

If your system is busy most of the time, you can use the save-while-active function
to reduce the time period that the system is unavailable while you are performing
save operations.

© Copyright IBM Corp. 1997, 2000, 2001 3

 Comprehensive information about save procedures and the save-while-active
functions are in the Information Center. You can read more about restore
procedures in “Chapter 3. Restore Procedures–General Information” on page 39.

Tape Units–Overview
The iSeries and AS/400e servers offer many different types of tape units to meet a
variety of requirements for performance, capacity, and cost. In many cases, you can
attach enough tape drives with sufficient media capacity to save your entire system
without operator intervention. This allows a complete unattended save operation.

For more information about the tape units that are available to attach to your
system, call 1-800-IBM-CALL (1-800-426-2255). You can read about tape
performance in the Performance Capabilities Reference, ZC41–0607.

Automated Tape Library Systems–Overview

Automated tape library systems are a combination of hardware and software that
store, catalog, and load large numbers of tapes without operator intervention. On
iSeries and AS/400e servers, Backup Recovery and Media Services (BRMS/400)
software supports the use of an automated tape library system.

You can read more about using automated tape library systems with the iSeries
and AS/400e servers in the Automated Tape Library Planning and Management book
or point your browser to the following URL:
| http://www.ibm.com/eserver/iseries/service/brms.htm
|
|
Alternate Installation Device-Overview
Alternate Installation Device support allows you to perform installation and
recovery procedures using a combination of devices. Previously, these types of
activities could only be performed using devices attached to the first system bus.
Now, you can use a combination of devices that are attached on the first system
bus and on additional buses. The alternate installation device is not attached to the
first system bus.

If you use this function, the system uses existing support (a device on the first
system bus) to install or recover enough of the Licensed Internal Code required to
perform an IPL with IPL-type D. Then, using the new alternate installation device
support, the system continues the operation using media in the alternate
installation device. This new function supports installation and recovery from tape
media, such as SAVSYS tapes or distribution tapes which you created, that contains
Licensed Internal Code and may contain the operating system, licensed programs,
and data.

Some models, typically with 3590 tape devices attached, may see a performance
improvement when using an alternate installation device for save operations.

Optical Media Support–Overview

| You can save your entire system, all of your system data, or all of your user data,
| to optical media. The DVD-RAM optical media standalone device supports all save
| and restore commands. It is an economical alternative to tape for save/restore
| operations on entry level systems. The Optical Support, (SC41-5310), book provides
| more information about using optical media.

4 OS/400 Backup and Recovery V5R1

 Journal Management–Overview
| You can use journal management to recover the changes to various objects that
| have occurred since your last complete save.

| You use a journal to define what objects you want to protect with journal
| management. This is often referred to as journaling an object. A journal receiver
| contains the entries (called journal entries) that the system adds when events
| occur that are journaled, such as changes to database files, changes to other
| journaled objects, or security-relevant events. See “Chapter 19. Planning and
| Setting Up Journaling” on page 383 for a list of object types that can be journaled.

You can use the remote journal function to set up journals and journal receivers on
a remote iSeries or AS/400eserver. These journals and journal receivers are
associated with journals and journal receivers on the source system. The remote
journal function allows you to replicate journal entries from the source system to
the remote system.

The main purpose of journal management is to assist in recovery. You can also use
the information that is stored in journal receivers for other purposes, such as:
| v An audit trail of activity that occurs for various objects on the system.
v Assistance in testing application programs. You can use journal entries to see the
changes made by a particular program.

You can read more about planning and setting up journal management in
“Chapter 19. Planning and Setting Up Journaling” on page 383. Information about
the remote journal function is provided in “Chapter 21. Remote Journal Function”
on page 467.

Access-Path Protection–Overview
An access path describes the order in which records in a database file are
processed. A file can have multiple access paths, if different programs need to see
the records in different sequences. If your system ends abnormally when access
paths are in use, the system may have to rebuild the access paths before you can
use the files again. This is a time-consuming process. To perform an IPL on a large,
busy iSeries or AS/400e server that has ended abnormally can take many hours.

You can use journal management to keep a record of changes to access paths. This
greatly reduces the amount of time it takes the system to perform an IPL after it
ends abnormally.

Two methods of access-path protection are available:

v System-managed access-path protection
v Explicit journaling of access paths

System-Managed Access-Path Protection

You can allow the system to determine which access paths to protect. You specify
target recovery times for access paths for the entire system or for auxiliary storage
pools (ASPs). Your system has a default recovery time for access paths for the
entire system of 90 minutes when it is shipped. You can use the Edit Recovery for
Access Paths (EDTRCYAP) command to see and change the target recovery times
for access paths and to see how much space the system is using for
system-managed access-path protection (SMAPP).

Chapter 1. Options for Backup, Recovery, and Availability–Overview 5

 SMAPP provides a simple method to reduce your recovery time after an abnormal
system end. SMAPP manages the required environment for you. You do not need
to use any type of journal management to use SMAPP.

You can read more about SMAPP in “Chapter 18. Protecting Access Paths Using
System-Managed Access-Path Protection” on page 375.

Explicit Journaling of Access Paths

When you use SMAPP, the system decides which access paths to protect, based on
overall target recovery times for access paths. You may want to make sure that
certain access paths are protected because they are critical to your business. You
can use journal management to explicitly protect access paths on your system.

You can use a combination of SMAPP and explicit journaling of access paths. The
system evaluates the protected and unprotected access paths to develop its strategy
for meeting your recovery targets for access paths.

“Chapter 19. Planning and Setting Up Journaling” on page 383 describes explicit
journaling of access paths.

Commitment Control–Overview
Commitment control is an extension of journal management that assists you in
keeping your database files synchronized. A single transaction on your system may
update more than one database file. If your system is in a network, a single
transaction may update files on more than one system.

Commitment control helps ensure that:

v All changes within a transaction are completed for all files that are affected.
v If processing is interrupted before the transaction is completed, all changes
within a transaction are removed.
v Changes that are made during a transaction can be removed when the user
program determines that it is necessary to do so. This is called a rollback
operation.

Without commitment control, recovering data for a complicated program requires

detailed application and program knowledge. Programs that are interrupted cannot
easily be started again. To restore the data to the last completed transaction, you
may need to write a user program or use data file utility (DFU) to reverse the
database updates that are not complete.

The commit (COMMIT) and rollback (ROLLBACK) operations are available in

several iSeries or AS/400e server programming languages including ILE RPG*, ILE
COBOL*, ILE C*, control language (CL), and Structured Query Language (SQL).

You can read more about commitment control in “Chapter 22. Commitment
Control” on page 525.

Auxiliary Storage Pools–Overview

Your system may have many disk units attached to it for auxiliary storage of your
data. To your system, they look like a single unit of storage. The system spreads
data across all disk units. You can use auxiliary storage pools to separate your disk
units into logical subsets.

6 OS/400 Backup and Recovery V5R1

 An auxiliary storage pool (ASP)is a group of units that are defined from all the
disk units that make up auxiliary storage. It is a software definition of how the
disk units are arranged. An ASP does not necessarily correspond to the physical
arrangement of disks.

ASPs allow you to isolate objects on one or more specific disk units. This may
reduce the loss of data due to a disk media failure. In most cases, only data that is
stored on disk units in the affected ASP is lost. However, when a disk unit fails,
the entire system is unusable until the disk unit is repaired, unless the failed unit
is protected by device parity protection or mirrored protection.

ASPs are used to manage system performance and backup requirements:

v You can create an ASP to provide dedicated resources for frequently used
objects, such as journal receivers.
v You can create an ASP to hold save files. Objects can be backed up to save files
in a different ASP. It is unlikely that both the ASP that contains the object and
the ASP that contains the save file will be lost.
v You can create different ASPs for objects with different recovery and availability
requirements. For example, you can put critical database files or documents in
an ASP that has mirrored protection or device parity protection.
v You can create an ASP to place infrequently used objects, such as large history
files, on disk units with slower performance.
v You can use ASPs to manage recovery times for access paths for critical and
noncritical database files using system-managed access-path protection.

| Three types of ASPs are available on the system:

v The system creates the system ASP (ASP 1) and is always configured. It contains
the Licensed Internal Code, licensed programs, system libraries, and temporary
system work space. The system ASP also contains all other configured disk units
that are not assigned to a user ASP.
| v Grouping together a physical set of disk units and assigning them a number (2
| through 32) creates a basic user ASP. ASP 1 is always reserved as the system
| ASP.
| v ASPs 33 through 99 are independent user ASPs, also known as independent
| disk pools in Operations Navigator. You can use clustering software to make
| them switchable ASPs. When you make an independent ASP switchable, you
| can switch it among iSeries 400 servers or logical partitions. If you choose to
| switch these ASPs among iSeries 400 servers, the ASPs must consist of external
| disk units. If you choose to switch these ASPs among logical partitions, the ASPs
| can consist of internal disk units. To use independent ASPs, you must use
| Operations Navigator. You can find more information about independent ASPs
| in the Information Center web site at this URL:
| http://www.ibm.com/eserver/iseries/infocenter/

| You can read more about ASPs in “Chapter 25. Working with Auxiliary Storage
Pools” on page 669.

Mirrored Protection–Overview
If a disk failure occurs, mirrored protection is intended to prevent data from being
lost. Mirrored protection is a software function that uses duplicates of disk-related
hardware components to keep your system available if one of the components
fails. It can be used on any model of the iSeries or AS/400e server and is a part of
the Licensed Internal Code.

Chapter 1. Options for Backup, Recovery, and Availability–Overview 7

 Different levels of mirrored protection are possible, depending on what hardware
is duplicated. You can duplicate:
v Disk units
v Disk controllers
v Disk I/O processors
v A bus

The system remains available during the failure if a failing component and the
hardware components that are attached to it are duplicated.

You can read more about mirrored protection in “Chapter 27. Working with
Mirrored Protection” on page 719.

Device Parity Protection–Overview

Device parity protection is intended to prevent data from being lost if a disk
failure occurs. In many cases, device parity protection can also prevent your
system from stopping when a disk unit fails. Device parity protection provides the
following:
v Technology similar to the RAID-5 (redundant array of independent disks)
technique
v Redundant power
v Concurrent maintenance for single disk failures
v Concurrent maintenance for power supply failures for the 9337 Disk Array
Subsystem

Device parity protection is a hardware function that is available for some disk unit
subsystems and input/output processor features. This includes the 6502, 6512,
6751, and 6532 Input/Output Processors and the 9337 Disk Array Subsystem.

The models with device parity protection use a data redundancy technique to
protect the data. To improve performance the parity information in the disk unit
subsystems with device parity protection is spread across multiple units.

When a failure occurs on a disk unit subsystem that has device parity protection,
the data is reconstructed. The disk subsystem controller automatically reconstructs
the data from the active units in the disk unit subsystem. The system continues to
run while the data is being reconstructed.

When a failure occurs on a disk unit subsystem that does not have device parity
protection or mirrored protection, you cannot use the system until the disk unit is
repaired or replaced.

If possible, you should protect all the disk units on your system with either device
parity protection or mirrored protection. This prevents the loss of information
when a disk failure occurs. In many cases, you can also keep your system running
while a disk unit is being repaired or replaced.

You can read more about device parity protection in “Chapter 26. Working with
Device Parity Protection” on page 695.

Comparison of Disk Protection Options

Table 1 on page 9 compares the disk protection options:

8 OS/400 Backup and Recovery V5R1

 Table 1. Comparison of Disk Protection Options. Yes indicates the component is protected. The system does not stop
when that component fails.
Disk Component Device Parity Protection Mirrored Protection Checksum Protection3

Bus No Yes2 No
IOP No Yes2 No
Controller No Yes2 No
Power supply Yes1 Yes2 No
Disk drive Yes Yes No
1
Only on the 9337 Disk Array Subsystem.

2
Mirrored protection protects this component if your system has sufficient redundant hardware.

3
Checksum protection is no longer available, beginning with V3R6 of the OS/400 licensed program. It is
included for comparison purposes.

Uninterruptible Power Supply–Overview

An uninterruptible power supply providesauxiliary power to the processing unit,
disk units, system console, and other devices that you choose to protect. When you
use an uninterruptible power supply with the iSeries or AS/400e server, you can:
v Continue operations during brief power interruptions.
v Provide normal ending of operations to reduce the recovery time when the
system is restarted. If the system ends abnormally before completing a normal
ending of operations, the recovery time can be significant.
v Protect the system from voltage peaks.

Normally, an uninterruptible power supply does not provide power to all work
stations.You should design your interactive applications to handle the loss of
communications with a workstation. Otherwise, system resources are used to
attempt to do error recovery for workstations that have lost power.

The programming language reference manuals provide examples of how to use the
error feedback areas to handle workstations that are no longer communicating with
the application. The Information Center describes how to develop programs to
handle an orderly shutdown of the system when the uninterruptible power supply
unit takes over.

Look for this information in the Backup, Recovery, and Availability topics at the
following Web site: http://www.ibm.com/eserver/iseries/infocenter.

Dual Systems–Overview
Installations with very high availability requirements use a dual-systems approach.
Some or all data is maintained on two systems. The secondary system can take
over critical application programs if the primary system fails.

| The most common method for maintaining data on the secondary system is
| through the use of journaling. Journal entries from the primary system are
| transmitted to the secondary system. A user-written program on the secondary
| system receives the journal entries and uses them to update journaled objects.

Chapter 1. Options for Backup, Recovery, and Availability–Overview 9

 In this method, the journal entries are transmitted at the application layer by using
the Receive Journal Entry (RCVJRNE) commmand or the Retrieve Journal Entries
(QjoRetrieveJournalEntries) API. Using the Remote Journal function improves this
method. This function allows the primary system to transmit the journal entries to
a duplicate journal receiver on the secondary system at the microcode layer. For
more information on the Remote Journal function, refer to Chapter 21. Remote
Journal Function.

| A third method is to copy journal receivers to tape regularly. The journal receivers
| are then restored to the secondary system. A user-written program uses the journal
| entries to update the objects on the secondary system.

Several software packages are available from independent software vendors to

support dual systems on the iSeries or AS/400e server.

ObjectConnect–Overview
The ObjectConnect function provides the ability to move entire objects between
systems using an optical connection, a communications line, or a local area
network (LAN). The ObjectConnect function provides a set of save and restore
(SAVRSTxxx) commands. For example, you can use the SAVRSTLIB command to
save one or more libraries, send the libraries over an optical connection to a
specified system, and restore the libraries.

The ObjectConnect support is provided at a low level in the operating system. It

provides better performance than other methods of transferring objects, such as
using save files. You can read more about ObjectConnect in “Using the
ObjectConnect/400 Function” on page 29.

OptiConnect for OS/400–Overview

OptiConnect for OS/400 support provides the capability of linking iSeries or
AS/400e servers with a fiber-optic bus.You can use OptiConnect for OS/400 with
dual system products and with ObjectConnect to move objects between systems
quickly and efficiently. You can read more about OptiConnect for OS/400 in
OptiConnect for OS/400, SC41-5414-02.

Backup Recovery and Media Services for iSeries–Overview

The Backup Recovery and Media Services for iSeries (BRMS/400) licensed program
offers a set of functions for defining and performing these tasks:
v Backup
v Recovery
v Archiving
v Retrieval
v Media management

| Starting with V5R1, Backup Recovery and Media Services provides a graphical
| user interface for backup and recovery that is integrated into Operations
| Navigator. You can use Backup Recovery and Media Services to simplify and
| automate your backups and to manage your media.

Backup Recovery and Media Services keeps track of what you have saved, when
you saved it, and where it is saved. When you need to do a recovery, Backup

10 OS/400 Backup and Recovery V5R1

 Recovery and Media Services helps ensure that the correct information is restored
from the correct tapes in the correct sequence.

You can read more about Backup Recovery and Media Services in the Backup
Recovery and Media Services for iSeries, SC41-5345-02 book or on the Backup
Recovery and Media Services website at the following URL:
http://www.ibm.com/eserver/iseries/service/brms.htm

Tivoli Storage Manager–Overview

You can Tivoli Storage Manager to protect data on your workstations and LAN file
servers. The Tivoli Storage Manager can automatically back up critical LAN and
workstation data and archive files that are used infrequently. It provides a disaster
recovery solution for LANs and workstations.

Administer the Tivoli Storage Manager from a client workstation that is attached to
an iSeries server. It can back up data from a variety of workstation platforms. You
can use the Backup Recovery and Media Services (BRMS/400) program to back up
user data to any Tivoli Storage Manager when the server in a client/server
environment. You can use Backup Recovery and Media Services for iSeries to
manage the data that you save on the Tivoli Storage Manager and to manage the
backup of the system data to local media.

You can read more about the Tivoli Storage Manager at the following Web site:
http://www.tivoli.com/support/storage_mgr/ad4serv.htm.

Content Manager OnDemand for iSeries Server–Overview

You can buy the Content Manager OnDemand for iSeries product. You can use
Content Manager OnDemand for iSeries to create search criteria and a search index
for large volumes of data, such as history reports or files, that you want to archive.
You can archive the data either to a folder or to optical media. You can retrieve
specific archived data that meets your search criteria.

| Business Continuity and Recovery Services–Overview

| IBM Business Continuity and Recovery Services (BCRS) offers an extensive
| portfolio of services to help you design, set up, and manage a comprehensive,
| enterprise-wide iSeries 400 or AS/400e business continuity and recovery strategy.
| BCRS can help you minimize the effects of network and system outages and plan
| for their efficient recovery. BCRS consultants can help you do the following:
| v Evaluate critical business functions and applications.
| v Assess your system environment.
| v Design a recovery plan to keep your business running in the event of an
| extended outage.

| BCRS recovery centers are equipped with the latest iSeries and AS/400e server
| technologies. They are staffed with technical experts to assist you in testing your
| recovery plan and performing a recovery in the event of a disaster.

| For more information about Business Continuity and Recovery Services, please call
| 1-800-599-9950, extension 206.

Chapter 1. Options for Backup, Recovery, and Availability–Overview 11

Data Migration Service Offerings–Overview
The AS/400® DASD Migration Service Offering (GIGMIG) is a service offering to
migrate the disk (DASD) on your iSeries server to different disk technology
without reloading your entire system. GIGMIG offerings include the following:
v Replacing all disks on the system.
v Adding new disks and removing some existing disks.
v Moving disks between ASPs.

SYSMIG is a service offering to migrate your data from one iSeries server to
another iSeries server.

For more information about GIGMIG and SYSMIG, please call 1-800-IBM-4YOU
(1-800-426-4968).

12 OS/400 Backup and Recovery V5R1

Part 2. Saving Information on Your System
Chapter 2. Saving your Server . . . . . . . 15
Saving your server with the GO SAVE command . . 15
Overview of the GO SAVE command menu
options . . . . . . . . . . . . . . . 17
Change Save menu defaults with GO SAVE:
Option 20 . . . . . . . . . . . . . . 18
Save your whole server with GO SAVE: Option
21 . . . . . . . . . . . . . . . . 18
Save system data with GO SAVE: Option 22 . . 19
Save user data with GO SAVE: Option 23 . . . 19
Saving parts of your server with other GO SAVE
command menu options . . . . . . . . . 20
Using GO SAVE: Options 21, 22, and 23 . . . . 20
Printing system information . . . . . . . 26
Using the ObjectConnect/400 Function . . . . . 29
Components of ObjectConnect/400 . . . . . 30
Setting Up Your System to Use
ObjectConnect/400 . . . . . . . . . . . 30
How the System Runs an ObjectConnect
Command . . . . . . . . . . . . . . 30
Using the ObjectConnect Commands . . . . . 32
Save/Restore (SAVRST) Command . . . . 32
Save/Restore Object (SAVRSTOBJ) Command 32
Save/Restore Change Objects (SAVRSTCHG)
Command . . . . . . . . . . . . . 32
Save/Restore Library (SAVRSTLIB) Command 32
Save/Restore Document Library Object
(SAVRSTDLO) Command . . . . . . . 32
Save/Restore Configuration (SAVRSTCFG)
Command . . . . . . . . . . . . . 32
Investigating ObjectConnect Problems . . . . 33
CPFAD84 Error Codes . . . . . . . . . . 33
Source System-specific Error Codes from
CPFAD84 Message . . . . . . . . . . 33
Target System-specific Error Codes from
CPFAD84 Message . . . . . . . . . . 33
Source or Target system Error Codes from
CPFAD84 Message . . . . . . . . . . 34

© Copyright IBM Corp. 1997, 2000, 2001 13

14 OS/400 Backup and Recovery V5R1
 Chapter 2. Saving your Server
Look for comprehensive information about how to backup your iSeries server on
the iSeries Information Center on the Internet. See “Prerequisite and related
information” on page xix for information on how to access the Information Center.

If this is your first experience with your iSeries server, use the following
instructions to save all of the information on your iSeries server. Do this with the
GO SAVE menu options. The instructions in the book are the same as the
instructions in the Information Center.

You can browse the Information Center, or print out a copy of the information on
how to backup your entire iSeries server.

Note: To avoid data loss, if you have either of these products installed on your
system, please read this notice before upgrading to OS/400® Version 4,
Release 5 or later releases:
v 5769SA3 - OS/400 Integration for Novell NetWare
v 5769XZ1 - OS/2® Warpserver for AS/400
Once you install OS/400 Version 4, Release 5 or later releases, these
products will no longer work. You cannot use Restore License Programs to
load a previous release of these products onto a server with V4R5 or later
version of the operating system. To avoid losing data that depends on these
products, migrate that data from your server to an accessible location before
upgrading to Version 4, Release 5.
v For assistance in migrating your data to AS/400 Netserver, see the
redbook Advantage AS/400 NetServer, SG24-5196.
v For additional information, read the redpiece: Migration Options for:
OS/2 Warp Server for AS/400 and OS/400 Integration for Novell
NetWare.
v For alternative ideas to help migrate your data, see information APAR
II11689.
r

Saving your server with the GO SAVE command

Using the GO SAVE command is a simple way to make sure that you have a good
backup of your entire server. The GO SAVE command presents you with Save
menus that make it easy to back up your server, no matter what backup strategy
| you decide to use. It is a good idea to use menu option 21 of the GO SAVE
| command right after you install your server.

Menu option 21 of the GO SAVE command is the basis for all save strategies. This
option allows you to perform a complete save of all the data on your server. Once
you have used menu option 21, you can use other menu options to save parts of
the server, or to use a manual save process.

© Copyright IBM Corp. 1997, 2000, 2001 15

 Another save method uses Backup Recovery and Media Services (BRMS/400),
which automates your save processes. BRMS provides a comprehensive and easy
solution for your backup and recovery needs. You can read more about BRMS in
the iSeries Information Center
http://www.ibm.com/eserver/iseries/infocenter

The following figure illustrates the commands and menu options you can use to
save the parts of the server and the entire server.

Figure 2. Save commands and menu options

The following information provides an overview and procedures on how to use

menu options of the GO SAVE command:
v “Overview of the GO SAVE command menu options” on page 17 explains how
to start the GO SAVE command.

16 OS/400 Backup and Recovery V5R1

 v “Change Save menu defaults with GO SAVE: Option 20” on page 18 explains
how to customize the default GO SAVE command menu options.
v “Save your whole server with GO SAVE: Option 21” on page 18 explains how to
use menu option 21 when performing a full save of the server.
v “Save system data with GO SAVE: Option 22” on page 19 explains how to save
your data only after you perform a full save.
v “Save user data with GO SAVE: Option 23” on page 19 explains how to save
your user data only after you perform a full save.
| v “Saving parts of your server with other GO SAVE command menu options” on
| page 20 explains other GO SAVE command menu options.
v “Using GO SAVE: Options 21, 22, and 23” on page 20 provides you with
step-by-step instructions on how to use the GO SAVE command menu options.

Overview of the GO SAVE command menu options

Access the GO SAVE command menu by typing GO SAVE from any command
line. From the Save menu, you see option 21, option 22, and option 23 along with
many more save options. A single plus sign (+) indicates that the option places
your server into a restricted state, which means that the menu option runs without
input from you. A double plus sign (++) indicates that your server must be in a
restricted state before you can run this option.

SAVE Save

Select one of the following:

Save Data
1. Files
2. Libraries
3. Documents and folders
4. Programs
5. Other objects
6. Changed objects only
7. Licensed programs
8. Security data
++ 9. Storage
10. Configuration
11. Objects in directories

Figure 3. Save Menu–First Display

Page down on the Save menu to see additional options:

Save System and User Data

20 Define save system and user data defaults
++ 21 Entire system
++ 22 System data only
+ 23 All user data
Save Document Library Objects
30 All documents, folders, and mail
31 New and changed documents, new folders, all mail
32 Documents and folders
33 Mail only
34 Calendars
Save Libraries
++ 40 All libraries other than the system library
41 All IBM libraries other than the system library

Chapter 2. Saving your Server 17

 42 All user libraries
43 All changed objects in user libraries
Save for Different Systems
50 Save in System/36 format

The following topics describe how to use the menu options of the GO SAVE
command:
v “Change Save menu defaults with GO SAVE: Option 20” explains how to
customize the default GO SAVE command menu options.
v “Save your whole server with GO SAVE: Option 21” explains how to use menu
option 21 when performing a full system save.
v “Save system data with GO SAVE: Option 22” on page 19 explains how to save
your system data only after you perform a full save.
v “Save user data with GO SAVE: Option 23” on page 19 explains how to save
your user data only after you perform a full save.
v “Saving parts of your server with other GO SAVE command menu options” on
page 20 explains other automated GO SAVE command menu options.
v “Using GO SAVE: Options 21, 22, and 23” on page 20 provides you with
step-by-step instructions on how to use the GO SAVE command menu options.

Change Save menu defaults with GO SAVE: Option 20

Save menu option 20 changes the default values for the GO SAVE command, menu
options 21, 22, and 23. You can change any of the default menu option parameters
for the GO SAVE command by using Save menu option 20. This option simplifies
the task of setting your save parameters.

| In order to change the defaults, you must have *CHANGE authority for both the
| QUSRSYS library and the QSRDFLTS data area in the QUSRSYS library.

When you enter the GO SAVE command, then select menu option 20, the server
displays the default parameter values for menu options 21, 22, and 23. If this is the
first time you have used option 20 from the Save menu, the server displays the
IBM-supplied default parameter values. You can change any or all of the parameter
values to suit your needs. For example, you can specify additional tape devices or
change the message queue delivery default. The server saves the new default
values in data area QSRDFLTS in library QUSRSYS. The server creates the
QSRDFLTS data area only after you change the IBM-supplied default values.

Once you define new values, you no longer need to worry about which, if any,
options to change on subsequent save operations. You can simply review your new
default options and then press Enter to start the save with the new default
parameters.

If you have multiple, distributed servers with the same save parameters on each
server, this option provides an additional benefit. You can simply define the
parameters from the Save menu, using option 20 on one server. Then, save the
QSRDFLTS data area, distribute the saved data area to the other servers, and
restore it.

Save your whole server with GO SAVE: Option 21

Option 21 saves everything on your server and allows you to perform the save
| while you are not there. Option 21 does not save spooled files.

18 OS/400 Backup and Recovery V5R1

| Option 21 saves all of your data for additional licensed programs, such as Domino
| or Integration for Windows Server when you select to vary off your network
| servers. Also, if you have Linux installed on a secondary logical partition, you can
| back up that partition when you select to vary off your network servers.

Option 21 puts your server into a restricted state. This means that no users can
access your server and the backup is the only thing that is running on your server.
It is best to run this option overnight for a small server or during the weekend for
larger servers.

| Note: If you are saving information on independent disk pools, make sure that
| you have varied on the independent disk pools before using Option 21.

Option Description Commands

Number
| 21 Entire server ENDSBS SBS(*ALL) OPTION(*IMMED)
| (QMNSAVE) CHGMSGQ MSGQ(QSYSOPR) DLVRY(*BREAK or *NOTIFY)
| SAVSYS
| SAVLIB LIB(*NONSYS) ACCPTH(*YES)
| SAVDLO DLO(*ALL) FLR(*ANY)
| SAV DEV('/QSYS.LIB/media-device-name.DEVD') +
| OBJ(('/*') ('/QSYS.LIB' *OMIT) +
| ('/QDLS' *OMIT))1 UPDHST(*YES)
| STRSBS SBSD(controlling-subsystem)
1
The command omits QSYS.LIB file system because the SAVSYS command and the SAVLIB
LIB(*NONSYS) command both save it. The command omits the QDLS file system because
the SAVDLO command saves it.

“Using GO SAVE: Options 21, 22, and 23” on page 20 provides you with
step-by-step instructions on how to save your entire server with menu option 21 of
the GO SAVE command.

Save system data with GO SAVE: Option 22

Option 22 saves only your system data. It does not save any user data. Option 22
puts your server into a restricted state. This means that no users can access your
server, and the backup is the only thing that is running on your server.

Option Description Commands

Number
22 System data only ENDSBS SBS(*ALL) OPTION(*IMMED)
(QSRSAVI) CHGMSGQ MSGQ(QSYSOPR) DLVRY(*BREAK or *NOTIFY)
SAVSYS
SAVLIB LIB(*IBM) ACCPTH(*YES)
SAV DEV('/QSYS.LIB/media-device-name.DEVD') +
OBJ(('/QIBM/ProdData') +
('/QOpenSys/QIBM/ProdData')) +
UPDHST(*YES)
STRSBS SBSD(controlling-subsystem)

“Using GO SAVE: Options 21, 22, and 23” on page 20 provides you with
step-by-step instructions on how to save your system data with menu option 22 of
the GO SAVE command.

Save user data with GO SAVE: Option 23

Option 23 saves all user data. This information includes files, records, and other
data that your users supply into your server. Option 23 puts your server into a

Chapter 2. Saving your Server 19

 restricted state. This means that no users can access your server, and the backup is
the only thing that is running on your server.

Option Description Commands

Number
23 All user data ENDSBS SBS(*ALL) OPTION(*IMMED)
(QSRSAVU) CHGMSGQ MSGQ(QSYSOPR) DLVRY(*BREAK or *NOTIFY)
SAVSECDTA
SAVCFG
SAVLIB LIB(*ALLUSR) ACCPTH(*YES)
SAVDLO DLO(*ALL) FLR(*ANY)
SAV DEV('/QSYS.LIB/media-device-name.DEVD') +
OBJ(('/*') ('/QSYS.LIB' *OMIT) +
('/QDLS' *OMIT) +
('/QIBM/ProdData' *OMIT) +
('/QOpenSys/QIBM/ProdData' *OMIT))1 +
UPDHST(*YES)
STRSBS SBSD(controlling-subsystem)
1
Menu option 23 omits the QSYS.LIB file system because the SAVSYS command, the
SAVSECDTA command, the SAVCFG command, and the SAVLIB LIB(*NONSYS) command
save it. The command omits the QDLS file system because the SAVDLO command saves it.
Menu option 23 also omits the /QIBM and /QOpenSys/QIBM directories because these
directories contain IBM supplied objects.

“Using GO SAVE: Options 21, 22, and 23” provides you with step-by-step
instructions on how to save your user data with menu option 23 of the GO SAVE
command.

Saving parts of your server with other GO SAVE command

menu options
You may perform the following GO SAVE command menu options.

Option Description Commands

Number
40 All libraries other ENDSBS SBS(*ALL) OPTION(*IMMED)
than the system CHGMSGQ MSGQ(QSYSOPR) DLVRY(*BREAK)
library (QMNSAVN) SAVLIB LIB(*NONSYS) ACCPTH(*YES)
STRSBS SBSD(controlling-subsystem)
41 All IBM libraries SAVLIB LIB(*IBM)
other than the system
library
42 All user libraries SAVLIB LIB(*ALLUSR)
43 All changed objects SAVCHGOBJ LIB(*ALLUSR)
in user libraries

Using GO SAVE: Options 21, 22, and 23

Use the following checklist for menu options 21, 22, and 23 of the GO SAVE
command. When appropriate, select the option that you require. If you choose to,
you can print system information during the procedure. Otherwise, “Printing
system information” on page 26 contains detailed instructions on how to print
system information if you do not want the Save menu option command to print
your system information automatically.
1. Sign on with a user profile that has *SAVSYS and *JOBCTL special authorities,
and also has sufficient authority to list different types of server resources. (The

20 OS/400 Backup and Recovery V5R1

 QSECOFR user profile contains all of these authorities.) This ensures that you
have the authority that you need to place the server in the necessary state and
to save everything.
2. If you have OptiConnect controllers, vary them off prior to the save operation.
You must vary off OptiConnect controllers before ending subsystems and
performing a save of the entire server, or before any save that ends the QSOC
subsystem. If you do not vary off OptiConnect controllers before ending
subsystems, they go into a failed status, the server marks them as damaged,
and the server does not save them. For more information, see OptiConnect for
OS/400, SC41-5414-02.
3. Make sure that Client Access is not active at your workstation. To deactivate
Client Access:
a. From the PC workstation, double-click the iSeries Workstation icon.
b. Double-click the Connections icon.
c. Click Disconnect.
d. If you have MQSeries (5769-MQ1), you need to quiesce MQSeries for
iSeries before you save the server. The MQSeries for OS/400 Administration,
GC33–1356 book has instructions for quiescing MQSeries for iSeries.
4. If you plan to run the save procedure immediately, make sure that no jobs are
running on the server: type WRKACTJOB.
If you plan to schedule the save procedure to run later, send a message to all
users informing them when the server will be unavailable.
5. Type GO SAVE at a command prompt to display the Save menu.
6. To perform an attended save of your server, go to step 8.
7. To perform an unattended save operation, continue with the following steps.
An unattended save operation prevents your save operation from stopping
because of unanswered messages:
a. Display the reply list sequence numbers to find what numbers are
available for use:
WRKRPYLE
b. If MSGID(CPA3708) is not already in your reply list, add it. For xxxx,
substitute an unused sequence number from 1 through 9999:
ADDRPYLE SEQNBR(xxxx) +
MSGID(CPA3708) +
RPY('G')
| c. Change your job to use the reply list and to notify you of any break
| messages that are sent:
| CHGJOB INQMSGRPY(*SYSRPYL) BRKMSG(*NOTIFY)

| Note: You can also set up a default so that whenever you select menu options
21, 22, or 23, the server will always use the reply list. To set up the
default, select menu option 20 from the Save menu. Specify Yes on the
Use system reply list option.
8. Select the option (21, 22, or 23) from the Save menu and press the Enter key.
A prompt display describes the function of the menu option that you selected.
9. After reading the prompt display, press the Enter key to continue. You are
shown the Specify Command Defaults display:

Chapter 2. Saving your Server 21

 Specify Command Defaults

Type choices, press Enter.

Devices . . . . . . . . . . . TAP01 Names

__________
__________
__________

Prompt for commands . . . . . Y Y=Yes, N=No

Check for active files . . . . Y Y=Yes, N=No

Message queue delivery . . . . *BREAK *BREAK, *NOTIFY

Start time . . . . . . . . . . *CURRENT *CURRENT, time

Vary off network servers . . . *NONE *NONE, *ALL,

__________ *WINDOWSNT

Unmount file systems . . . . . N Y=Yes, N=No

F3=Exit F12=Cancel More...

Specify Command Defaults

Type choices, press Enter.

Print system information . . . N Y=Yes, N=No

Use system reply list. . . . . N Y=Yes, N=No

F3=Exit F12=Cancel Bottom

10. Type your choices for the Devices prompt. You can specify as many as four
tape media device names. If you specify more than one device, the server
automatically switches to the next tape device when the current tape is full.
You may select only one DVD-RAM optical media device.
The first device for options 21 and 22 should be your alternate IPL device. If
you are creating media to install on another server, the device must be
compatible with the alternate IPL device for that server. This ensures that the
server can read the SAVSYS media if you need to restore your Licensed
Internal Code and the operating system.
11. Type your choice for the Prompt for commands prompt. Specify N (No) if you
want to run an unattended save. Specify Y (Yes) if you want to change the
defaults on the SAVxxx commands.

Note: If Y is specified to change the LABEL parameter for save commands, Y

must be specified if you use this media to restore the server.
12. Type your choice for the Check for active files prompt. Specify Y (Yes) if you
want the server to warn you if active files exist on the save media. The
warning you receive gives the following choices:
v Cancel the save operation.
v Insert new media and try the command again.
v Initialize the current media and try the command again.

Note: If you use DVD-RAM optical media for your save, the server sends
inquiry messages to the QSYSOPR message queue when it encounters
identical active files. The server sends the inquiry message for each
identical active file that it finds. See How optical media is different

22 OS/400 Backup and Recovery V5R1

 from tape media online in the Information Center, or the Optical Support
book for more information on optical media.
Specify N (No) if you want the server to write over any active files on the save
media without warning you.
13. Type your choice for the Message queue delivery prompt. Specify *NOTIFY if
you want to do an unattended save. This prevents communications messages
from stopping the save operation. If you specify *NOTIFY, severity 99
messages that are not associated with the save operation are sent to the
QSYSOPR message queue without interrupting the save process. For example,
messages that request a new volume be loaded interrupt the save operation
because they are associated with the job. You cannot continue until you reply
to these messages.
Specify *BREAK if you want to be interrupted for severity 99 messages that
require a reply.
14. Type your choice for the Start time prompt. You may schedule the start of the
save operation up to 24 hours later. For example, assume that the current time
is 4:30 p.m. on Friday. If you specify 2:30 for the start time, the save operation
begins at 2:30 a.m. on Saturday.
Notes:
a. The server uses the Delay Job (DLYJOB) command to schedule the save
operation. Your workstation will be unavailable from the time you request
the menu option until the save operation completes.
b. Make sure that your workstation is in a secure location. Your workstation
remains signed on, waiting for the job to start. If the server request
function is used to cancel the job, your workstation displays the Save
menu. The workstation remains signed on with your user profile and your
authority.
c. Make sure that the value for the QINACTITV system value is *NONE. If
the value for QINACTITV is other than *NONE, the workstation will vary
off in the amount of time specified. If you changed the value to *NONE,
write the old value down.
d. If you specify a delayed start and want your save operation to run
unattended, be sure you have done the following:
v Set up the system reply list.
v Specified *NONE on QINACTITV system value.
v Specified *NOTIFY on message queue delivery.
| v Specify *NOTIFY for any break messages.
v Responded N to the Prompt for commands prompt.
v Responded N to Check for active files.
15. Type your choice for the Vary off network servers prompt. If you use Integration
for Windows Server, you may vary off the network server descriptions before
beginning the save procedure.
| If have Linux installed on a secondary partition, you must select *ALL to save
| the data on that partition.
Select one of the following options to specify which network servers should
be varied off before the save operation is performed:
The Information Center provides additional information about the effects of
varying off the network servers. Select one of the following options to specify
which network servers should be varied off before the save operation is
performed:

Chapter 2. Saving your Server 23

 *NONE
Does not vary off network servers. The save operation will take longer
since the network server data will be saved in a format that allows
restoration of individual objects.
*ALL Varies off all network servers. The save operation will take less time
but the network server data will not be saved in a format that allows
restoration of individual objects. You will only be able to restore all of
the data from the network servers.
| You must select this option if you are saving data on a secondary
| logical partition with Linux installed on it.
*WINDOWSNT
Varies off all network servers of type *WINDOWSNT prior to the start
of the save. This allows the save of the network server storage spaces.
16. Type your choice for the Unmount file system prompt. If you use user-defined
file systems (UDFSs), you should unmount the UDFSs before beginning the
save procedure. Specify Y (Yes) if you want to allow all dynamically mounted
file systems to be unmounted. This allows you to save UDFSs and their
associated objects. IBM recommends that you unmount your UDFSs for
recovery purposes. For more information on UDFSs, refer to OS/400 Network
File System Support, SC41-5714-01.

Note: After the save operation completes, the server will not attempt to
remount the file systems.

Specify N (No) if you do not want all dynamically mounted file systems to be
unmounted. If you specify N, and you have mounted UDFSs, you will receive
a CPFA09E message for each mounted UDFS. The objects in the mounted UDFS
will be saved as if they belong to the mounted over file system.
17. Type your choice for the Print system information prompt. Specify Y (Yes) if you
want to print the system information. The system information may be useful
for disaster recovery. “Printing system information” on page 26 explains how
to print your system information manually without using the automatic GO
SAVE command menu option function.
18. Type your choice for the Use system reply list prompt. Specify Y (Yes) if you
want to use the system reply list when the server sends an inquiry message.
19. Press the Enter key. If you chose a later start time, your display shows
message CPI3716. The message tells when the save operation was requested
and when it will start. You cannot use the display until the save operation
completes. The input-inhibited indicator should appear. You have completed
the steps for setting up the save operation.
If you did not choose a later start time, continue with step 20. If the value for
QSYSOPR message queue delivery is *BREAK with a severity level of 60 or
lower, you must respond to the ENDSBS messages. This is true even if you
plan to run an unattended save operation specifying a start time of
*CURRENT.
20. If you responded Y to the system prompt, Prompt for commands, the End
Subsystem display appears. Type any changes and press the Enter key. While
the server is ending subsystems, you see the following messages. You must
respond to them if the QSYSOPR message queue is set to *BREAK with a
severity level of 60 or lower. Each message appears at least twice. Press the
Enter key to respond to each message.
a. CPF0994 ENDSBS SBS(*ALL) command being processed

24 OS/400 Backup and Recovery V5R1

 b. CPF0968 System ended to restricted condition

If you responded N to the Prompt for commands prompt, skip to step 22.
21. When the server is ready to perform each major step in the save operation,
you are shown the prompt display for that step. The time between prompt
displays may be quite long.
For option 21 (Entire system) these prompt displays appear:
ENDSBS SBS(*ALL) OPTION(*IMMED)
SAVSYS
SAVLIB LIB(*NONSYS) ACCPTH(*YES)
SAVDLO DLO(*ALL) FLR(*ANY)
SAV DEV('/QSYS.LIB/media-device-name.DEVD') +
OBJ(('/*') ('/QSYS.LIB' *OMIT) +
('/QDLS' *OMIT)) +
UPDHST(*YES)

STRSBS SBSD(controlling-subsystem)

For option 22 (System data only) these prompt displays appear:

ENDSBS SBS(*ALL) OPTION(*IMMED)
SAVSYS
SAVLIB LIB(*IBM) ACCPTH(*YES)
SAV DEV('/QSYS.LIB/media-device-name.DEVD') +
OBJ(('/QIBM/ProdData') +
('/QOpenSys/QIBM/ProdData')) +
UPDHST(*YES)
STRSBS SBSD(controlling-subsystem)

For option 23 (All user data) these prompt displays appear:

ENDSBS SBS(*ALL) OPTION(*IMMED)
SAVSECDTA
SAVCFG
SAVLIB LIB(*ALLUSR) ACCPTH(*YES)
SAVDLO DLO(*ALL) FLR(*ANY)
SAV DEV('/QSYS.LIB/media-device-name.DEVD') +
OBJ(('/*') ('/QSYS.LIB' *OMIT) +
('/QDLS' *OMIT) +
('/QIBM/ProdData' *OMIT) +
('/QOpenSys/QIBM/ProdData' *OMIT)) +
UPDHST(*YES) STRSBS SBSD(controlling-subsystem)

Type your changes at each prompt display and press the Enter key.
22. When the server sends a message that asks you to load the next volume, load
the next media and respond to the message. For example, if the message is the
following, load the next volume and then enter R (C cancels the operation):
Device was not ready or next volume was
not loaded (C R)

If a media error occurs

If an unrecoverable media error occurs during the SAVLIB procedure, see
How to recover from a media error during a SAVLIB operation. You can
find this subject under the Backing up your server topic in the
Information Center.

23. After the save completes, you should unmount user-defined file systems at
this point if you mounted them for the save operations.

Chapter 2. Saving your Server 25

 24. Change the QINACTITV system value back to its original value. You wrote
this value down in step 14c on page 23.
25. When the save operation completes, print the job log. It contains information
about the save operation. Use it to verify that the operation saved all objects.
Type one of the following:
DSPJOBLOG * *PRINT

Or
SIGNOFF *LIST

You have completed the save operation. Make sure that you mark all of your
media and store it in a safe, accessible place.

Printing system information

Printing the system information provides valuable information about your server
that will be useful during a system recovery. It is especially useful if you cannot
use your SAVSYS media to recover and must use your distribution media. Printing
this information requires *ALLOBJ, *IOSYSCFG, and *JOBCTL authority and
produces many spooled file listings. You may not need to print this information
every time you perform a backup. However, you should print it whenever
important information about your server changes.
1. Print your current disk configuration. This is essential if you plan to do a
model upgrade and you are using mirrored protection. Do the following:
a. Sign on with a user profile that has *SERVICE special authority.
b. Type STRSST on a command line and press the Enter key.
c. Select option 3 (Work with disk units) on the System Service Tools (SST)
display.
d. Select option 1 (Display disk configuration) on the Work with Disk Units
display.
e. Select option 3 (Display disk configuration protection) on the Display Disk
Configuration display.
f. Print the displays (there may be several) using the PRINT key for each
display.
g. Press F3 until you see the Exit System Service Tools display.
h. On the Exit System Service Tools display, press the Enter key.
2. If you are using logical partitions, print the logical partition configuration
information.
a. From the primary partition, type STRSST on a command line and press
Enter.
b. If you are using SST, select option 5 (Work with system partitions), and
press Enter. If you are using DST, select option 11 (Work with system
partitions), and press Enter.
c. From the Work With System Partitions menu, select option 1 (Display
partition information).
d. To display all system I/O resources from the Display Partition Information
menu, select option 5. To display only the allocated system I/O resources
from the Display Partition Information menu, select option 3.
e. At the Level of detail to display field, type *ALL to set the level of detail to
ALL.
f. Press F6 to print the system I/O configuration.
g. Select option 1 (132 characters wide) and press Enter.

26 OS/400 Backup and Recovery V5R1

 h. Press F12 to return to the Display Partition Information menu.
i. Select option 2 (Display partition processing configuration).
j. From the Display Partition Processing Configuration display, press the Print
Screen key to print the configuration.
k. Press F10 to display the Main Storage Size (MB) field.
l. Press the Print Screen key to print the configuration.
m. Press F10 to display the Interactive Performance field.
n. Press the Print Screen key to print the configuration.
o. Press F3 until you see the Exit System Service Tools display.
p. On the Exit System Service Tools display, press the Enter key.
3. Sign on with a user profile that has *ALLOBJ special authority, such as the
security officer. The server lists information only if you have the proper
authority. If you sign on as a user with less than *ALLOBJ authority, some of
the listings in these steps may not be complete. You must also be enrolled in
the system directory before you can print a list of all the folders on the server.
4. If you use the history log or if you have a requirement to keep it, do the
following:
a. Display the system log QHST. This automatically brings it up to date. Type:
DSPLOG LOG(QHST) OUTPUT(*PRINT)
b. Display all copies of the system log:
WRKF FILE(QSYS/QHST*)

Look at the list to verify that you saved all copies of the log that you may
need later.

Note: The history (QHST) log contains information such as date created,
and the last change date and time. To get more information about the
history (QHST) log, select option 8 (Display file description) on the
Work with Files display.
c. To prevent confusion about the date of the log, select the Delete option on
the Work with Files display. Delete all but the current copies of the system
log. This step improves the performance of the SAVSYS command.
5. Print the system information. You can do this by two different methods:
a. Using the GO SAVE command, on the Specify Command Defaults display,
select Y at the Print system information prompt.
b. Use the PRTSYSINF command.

The following table describes the spooled files that the server creates. The
PRTSYSINF command does not create empty spooled files. If some objects or
types of information do not exist on your server, you may not have all of the
files listed below.
Table 2. Spooled Files Created by the server
Spooled File User Data Description of Contents
Name
QPEZBCKUP DSPBCKUPL List of all user libraries
QPEZBCKUP DSPBCKUPL List of all folders
QSYSPRT DSPSYSVAL Current settings for all system values
QDSPNET DSPNETA Current settings for all network attributes
QSYSPRT DSPCFGL Configuration lists

Chapter 2. Saving your Server 27

 Table 2. Spooled Files Created by the server (continued)
Spooled File User Data Description of Contents
Name
QSYSPRT DSPEDTD User-defined edit descriptions ( a separate spooled
file for each)
QSYSPRT DSPPTF Details of all fixes that are installed on your server
QPRTRPYL WRKRYPLE All reply list entries
QSYSPRT DSPRCYAP Settings for access path recovery times
QSYSPRT DSPSRVA Settings for service attributes
QSYSPRT DSPNWSSTG Network server storage spaces information
QSYSPRT DSPPWRSCD Power on/off schedule
QSYSPRT DSPHDWRSC Hardware configuration reports (a separate spooled
file for each resource type, such as *CMN or *LWS)
QSYSPRT WRKOPTCFG Optical device descriptions (if your server has an
optical device and optical support is started when
you run the command)
QSYSPRT DSPRJECFG Remote job entry configurations
QPDSTSRV DSPDSTSRV SNADS configuration
QPRTSBSD DSPSBSD Subsystem descriptions (a separate spooled file for
each subsystem description on your server)
QSYSPRT DSPSFWRSC Installed licensed programs (Software Resources
List)
QPRTOBJD DSPOBJD A list of all the journals on your server
QPDSPJNA WRKJRNA The journal attributes for each journal that is not in
the QUSRSYS library (a separate file for each
journal). Typically, journals in the QUSRSYS library
are IBM-supplied journals. If you have your own
journals in the QUSRSYS library, you need to
manually print information about those journals.
QSYSPRT CHGCLNUP Settings for automatic cleanup
QPUSRPRF DSPUSRPRF Current values for the QSECOFR user profile
QPRTJOBD DSPJOBD Current values for the QDFTJOBD job description
QPJOBLOG PRTSYSINF The job log for this job1

1
On your server, this spooled file might be in the QEZJOBLOG output queue.

6. Print a list of directories in the root directory.

DSPLNK OBJ('/*') OUTPUT(*PRINT)
7. Print any IBM-supplied objects that you have modified, such as the QSYSPRT
print file.
8. If you maintain a CL program that contains your configuration information, use
the Retrieve Configuration Source (RTVCFGSRC) command to ensure that the
CL program is current.
RTVCFGSRC CFGD(*ALL) CFGTYPE(*ALL) +
SRCFILE(QGPL/QCLSRC) +
SRCMBR(SYSCFG)
9. Print these spooled files. Keep this information with your backup log or your
save system media for future reference. If you choose not to print the lists, use
the Copy Spooled File (CPYSPLF) command to copy them to database files. See

28 OS/400 Backup and Recovery V5R1

 the Information Center for information on how to do this. Make sure that the
database files are in a library that is saved when you perform the Save menu
option.

Go to “Using GO SAVE: Options 21, 22, and 23” on page 20.

Using the ObjectConnect/400 Function

ObjectConnect/400 is a set of CL commands for moving objects between iSeries
servers simply and efficiently. ObjectConnect/400 is included with the operating
system. You install it by selecting it on the Install Licensed Program display.

When you use an ObjectConnect command, the system moves the object directly to
the target system without using save files or distribution queues. ObjectConnect
provides better performance than other methods for moving objects between
systems, and ObjectConnect does not require additional disk space to store an
intermediate copy of the object that is being moved.

ObjectConnect commands are closely related to the SAVxxx and RSTxxx

commands. The ObjectConnect commands support most of the same parameters.
Table 3 shows a list of the ObjectConnect commands and the associated iSeries
save and restore commands. “Using the ObjectConnect Commands” on page 32
describes the functions that are performed by each command. The online command
help describes the parameters for each command.
Table 3. ObjectConnect/400 and Associated iSeries Save and Restore Commands
ObjectConnect Commands iSeries Save and Restore Commands

Save/Restore Integrated File System (SAVRST) Save (SAV), Restore (RST)

Save/Restore Object (SAVRSTOBJ) Save Object (SAVOBJ), Restore Object
(RSTOBJ)
Save/Restore Changed Object (SAVRSTCHG) Save Changed Object (SAVCHGOBJ),
Restore Object (RSTOBJ)
Save/Restore Library (SAVRSTLIB) Save Library (SAVLIB), Restore Library
(RSTLIB)
Save/Restore Document Library Object Save Document Library Object
(SAVRSTDLO) (SAVDLO), Restore Document Library
Object (RSTDLO)
Save/Restore Configuration (SAVRSTCFG) Save Configuration (SAVCFG), Restore
Configuration (RSTCFG)

To use the ObjectConnect functions, you must have ObjectConnect installed on

both the source and target systems. The systems must be connected with one of
the following methods:
v Local area network (LAN) or remote communications linewith APPC and
APPN*.
v Local area network (LAN) or remote communications line with TCP/IP with
AnyNet* support.
v Fiber optic bus with OptiConnect/400.

Chapter 2. Saving your Server 29

 Components of ObjectConnect/400
The basic components of ObjectConnect/400 are outlined below:

Component Description

QSR library This library contains all ObjectConnect objects.

QCMN subsystem If the source and target systems are connected with a
communications line or a LAN, ObjectConnect jobs run in the
QCMN subsystem.
QSOC subsystem If the source and target systems are connected with
OptiConnect/400, ObjectConnect jobs run in the QSOC
subsystem.
QSOCCT mode description ObjectConnect uses the IBM-supplied default mode description,
QSOCCT. You must start this mode description before you use
ObjectConnect commands by specifying the following:
STRMOD RMTLOCNAME(target) MODE(QSOCCT)
LCLLOCNAME(*NETATR) RMTNETID(*NETATR)
QUSER user profile This IBM-supplied user profile is used by ObjectConnect jobs.

Setting Up Your System to Use ObjectConnect/400

After you have installed ObjectConnect, you must set up your systems to run
ObjectConnect. You perform some tasks only once. You perform other tasks
regularly to prepare for ObjectConnect commands.

Do These Things Initially: If your systems are connected with a communications

line or a LAN, add a communications entry to the QCMN subsystem. Type the
following on both systems:
ADDCMNE SBSD(QCMN) DEV(*ALL) DFTUSR(QUSER)
MODE(QSOCCT)

If you are using a fiber-optic bus, see the OptiConnect for OS/400 book.

Do These Things Before You Run ObjectConnect Commands: Whenever you start
your system, you must also start the ObjectConnect environment. You can include
these tasks in your startup procedures, or you can perform them manually.

If your systems are connected with a communications line or a LAN, do the

following:
v Ensure that the QCMN subsystem is started.
v Ensure that the connection is varied on and active.
v Start the mode description by typing the following:
STRMOD RMTLOCNAME(target) MODE(QSOCCT)
LCLLOCNAME(*NETATR) RMTNETID(*NETATR)

If your systems are connected with OptiConnect/400, continue with ″How the
System Runs an ObjectConnect Command″.

How the System Runs an ObjectConnect Command

When you issue an ObjectConnect command, the system starts an ObjectConnect
job and establishes a conversation with the target system. Figure 4 on page 31
shows the flow of the job:

30 OS/400 Backup and Recovery V5R1

┌───────SAVRSTxxx command issued
│ │
│ ┌─────────┼─────────┐ ┌───────────────┐
│ │ Source │ │ │ Target │
│ │ System │ │ │ System │
│ │ N │ │ │
│ │ Optical Bus │ │ │
│ │ Interface │ │ Optical Bus │
│ │ │ │ │ conversation │
│ │ N │ │ completed │
│ │ Optical Bus │ │ O │
│ │ support ────────┼────────────────┼─┐ │ │
│ │ starts │ Optical Bus │ │ │ │
│ │ conversation │ connection │ │ │ │
│ │ │ and dialog │ │ │ │
│ ├────Machine────────┤ established ├─┼──Machine────┤
│ │ Interface │ │ │ Interface │
│ │ │ │ │ O │
│ │ │ │ N │ │
└─┼──SSave data │ │ Restore data │
│ transferred │ │ accepted │
│ │ │ │ O │
└────────┼──────────┘ └───────┼───────┘
│ │
└───────────────────────────────────┘
Optical Bus connection

Figure 4. ObjectConnect Job Flow

You can view the ObjectConnect job by working with the subsystem. Type
WRKACTJOB SBS(QCMN) if your systems are linked with communications support.
Type WRKACTJOB SBS(QSOC) if your systems are linked with OptiConnect/400. You
are shown the Work with Active Jobs display:
Work with Active Jobs AS009
03/31/95
CPU % .0 Elapsed time: 00:00:00 Active Jobs 60

Type options, press Enter.

2=Change 3=Hold 4=End 5=Work with 6=Release 7=Display messages
8=Work with spooled files 13=Disconnect ...

Opt Subsystem/Job User Type CPU % Function Status

_ QCMN QSYS SBS .0 DEQW
_ ENDCTL1 QCMN BCH .0 DEQW
_ RCHCTL2 QCMN ASJ .0 PGM-QYYCMGR DEQW

You can use the Work with Configuration Status (WRKCFGSTS) command to see
the activity on the communications or LAN link:
Work with Configuration Status AS009
03/31/95
Position to . . . . . __________ Starting characters

Type options, press Enter.

1=Vary on 2=Vary off 5=Work with job 8=Work with description
9=Display mode status ...

Opt Description Status ------------Job------

__ WWGLOCAL ACTIVE
__ WWGLOC1 ACTIVE
__ QSOCCT ACTIVE/DETACHED QPADEV0023 GREEN
__ QSOCCT ACTIVE/SOURCE QPADEV0024 GREEN

Chapter 2. Saving your Server 31

 Using the ObjectConnect Commands
The following topic shows the specific functions that are performed by the
ObjectConnect commands. You can use the Remote Location Name
(RMTLOCNAME) parameter on these commands to specify where the saved
objects in directories are to be restored. The system determines the method
(communications line or optical connection) for transferring data to that location.
ObjectConnect cannot run in restricted state.

Save/Restore (SAVRST) Command

You can use the Save/Restore (SAVRST) command to save one or more objects in
directories, send them to another system, and restore them. It can also save entire
directories (not to be confused with entire systems). The SAVRST command
supports the same options as the SAV command.

Save/Restore Object (SAVRSTOBJ) Command

You can use the Save/Restore Object (SAVRSTOBJ) command to save one or more
objects, send them to another system, and restore them. The SAVRSTOBJ command
supports the same options as the SAVOBJ command, including the use of the
OMITOBJ parameter.

Save/Restore Change Objects (SAVRSTCHG) Command

You can use the Save/Restore Changed Objects (SAVRSTCHG) command to save
one or more changed objects, send them to another system, and restore them. An
example of this would be a situation where you wanted to maintain duplicate sets
of files on two different systems. The SAVRSTCHG command supports most of the
same options as the SAVCHGOBJ command, including use of the OMITOBJ
parameters. You may use the OMITLIB parameter with this command, and you
may also specify generic values for the LIB parameter on this command.

Save/Restore Library (SAVRSTLIB) Command

You can use the Save/Restore Library (SAVRSTLIB) command to save one or more
libraries, send them to another system, and restore them. The SAVRSTLIB
command supports the same options as the SAVLIB command, including the use
of the OMITLIB and OMITOBJ parameters. You may also use generic values for
the *LIB parameter on this command.

Save/Restore Document Library Object (SAVRSTDLO) Command

You can use the Save/Restore Document Library Object (SAVRSTDLO) command
to save one or more document library objects, send them to another system, and
restore them. The SAVRSTDLO command supports the same options as the
SAVDLO command.

Save/Restore Configuration (SAVRSTCFG) Command

You can use the Save/Restore Configuration (SAVRSTCFG) command to save one
or more configuration objects, send them to another system, and restore them. The
SAVRSTCFG command supports most of the options and parameters as the
SAVCFG and RSTCFG commands.

When you copy your configuration by using the SAVRSTCFG command, the
system saves and restores the following object types:

*CFGL *CNNL *CRGM *NTBD

*CIO *COSD *MODD *TRA

32 OS/400 Backup and Recovery V5R1

Investigating ObjectConnect Problems
If all ObjectConnect commands are failing, do the following:
v Ensure the correct subsystem is active.
v Ensure the connection between systems is active.
v Ensure the correct remote location name is specified.

If you suspect a more specific problem, do the following:

1. Locate the failing job or job log on both the source and target system. An
informational message may exist between the save completion message and the
restore completion message. This message ID is CPFAD87. If this message
exists, use F1 to display the detailed message to determine the name of the job
log on the target system.
2. Display the job log information on the target system and find the following
message:
Corresponding source information from location &1;
3. Use F1 to display the detailed message. The detailed information indicates the
job name and number for the source job.
4. Inspect the job log information on both systems to locate any messages. The
messages each include secondary text that describes the recommended course
of action if one is needed.

If problems are identified as OptiConnect/400 or communications problems, see

the OptiConnect/400 Guide or the Communications Configuration book.

CPFAD84 Error Codes

If you receive a CPFAD84 message on either the source or target system, refer to
the error codes below to identify the problem. You may also use the Analyze
Problem (ANZPRB) command to report the problem.

Source System-specific Error Codes from CPFAD84 Message

SRC1 Unknown message type was received on the source system, sent from
target system. ObjectConnect does not expect the target system to send
messages other than escape, completion, diagnostic, or information type
messages.
SRC2 Zero messages sent from target system. ObjectConnect expects to get a
minimum of one message that indicates success or failure. If the target
system does not send any messages, then this is an error.
SRC3 Receive response above MI sent an invalid code within the message. This
indicates that something on the target system failed and could not
continues. Check the spool file on the target system.
SRC4 Cannot start save operation. Save code has sent an escape message,
indicating its inability to begin the save operation. This may be a source
type problem or a sink type problem. Check for vlogs and try again. The
number of simultaneous save operations or restore operations may have
exceeded the allowable limit.

Target System-specific Error Codes from CPFAD84 Message

TGT1 Spool file is invalid. This indicates that the spool file had messages in an
order that are not expected. This error may also occur if ObjectConnect
information message CPFAD85 is not in the spool file.
TGT2 Received a ’terminate’ message from below MI on source system. This is

Chapter 2. Saving your Server 33

 running over bus only. This indicates that the source has ended for some
reason and that it has notified the target system that it will not send any
more data. Refer to the source system job log.
TGT3 Send Response failed after Receive Request worked. Target system received
a function check while running via the bus.
TGT4 Received a function check while running via the bus, and did not receive
any information from the source system.
TGT5 Cannot start restore operation. Restore code has sent an escape message,
indicating its inability to begin the restore operation. This may be a source
type problem or a sink type problem. Check for vlogs and try again. The
number of simultaneous save operations or restore operations operations
may have exceeded the allowable limit.

Source or Target system Error Codes from CPFAD84 Message

F4BE Sent from below MI. This indicates that a valid ending of the job has
occurred. For example, the source system begins the save operation by
using the SAVRSTOBJ command. If it finds that there is no data to save in
the library, it returns a message that indicates that no objects were saved.
The source system sends a message to the target system indicating that no
data is being transferred. The job on the target system ends instead of
waiting for the data.
FxBF Received invalid error message from below MI. This can be received in the
CPF389C error message. This is never an expected.error code. Check for
vlogs and try the request again.
0000 Although this would usually indicate a valid function or return code, in
this situation it indicates that something out of the ordinary has failed. If
running via the bus, the bus manager has completed its operation in a
valid state, but something else has failed. Try the request again.

34 OS/400 Backup and Recovery V5R1

Part 3. Recovering Information on Your System
Chapter 3. Restore Procedures–General Actions for a failure in a user ASP disk
Information . . . . . . . . . . . . . . 39 unit–Checklist 11 . . . . . . . . . . . 86
The Relationship Between Save and Restore Actions for a failure in a user ASP disk
Commands . . . . . . . . . . . . . . 41 unit–Checklist 12 . . . . . . . . . . . 86
What Happens When You Restore Objects . . . . 42 Actions for a failure in a user ASP disk
Sequence for Restoring Related Objects . . . . . 45 unit–Checklist 13 . . . . . . . . . . . 88
Putting Your System in a Restricted State . . . . 45 Actions for non-load source disk unit
Reclaiming Storage . . . . . . . . . . . . 46 failure–Checklist 14 . . . . . . . . . . . 90
How to Reclaim Storage . . . . . . . . . 46 Actions for non-load source disk unit
Controlling Restoration of Security-Sensitive Objects 50 failure–Checklist 15 . . . . . . . . . . . 91
How to Set the QMCHPOOL, QBASPOOL, and Actions for non-load source disk unit
QPFRADJ System Values for a Different Processor failure–Checklist 16 . . . . . . . . . . . 92
or Memory . . . . . . . . . . . . . . 53 Actions for independent ASP disk
Locked Objects While Restoring . . . . . . . 53 failure–Checklist 17 . . . . . . . . . . . 93
How to Verify That Objects Are Restored | Actions for a failure in an independent ASP disk
Successfully . . . . . . . . . . . . . . 54 | unit–Checklist 18 . . . . . . . . . . . 93
Recovery from an Unsuccessful Restore Operation 56 | Actions for a failure in an independent ASP disk
Recovering from an Error While Restoring | unit–Checklist 19 . . . . . . . . . . . 94
Libraries . . . . . . . . . . . . . . 56 Recovering your entire system after a complete
Recovering from an Error While Restoring DLOs 57 system loss–Checklist 20 . . . . . . . . . 96
Recovering Mail . . . . . . . . . . . 57 Recovering your entire system after a complete
Recovering Documents and Folders . . . . 58 system loss including independent
How to Perform a Normal IPL . . . . . . . . 58 ASPs–Checklist 21 . . . . . . . . . . . 98
Parallel Restore Operations . . . . . . . . . 59 | Recovering your entire system after a complete
Recovery Considerations for Cryptographic Access | system loss including independent ASPs, when
Provider . . . . . . . . . . . . . . . 60 | Service Tools Network Interface is
| configured–Checklist 22 . . . . . . . . . 102
Chapter 4. Selecting the Right Recovery Strategy 61 Restoring a Logical Partition to Another Logical
Some Common Recovery Terminology . . . . . 61 Partition—Checklist 23 . . . . . . . . . 105
Recovery Procedure for a Power Failure . . . . . 62 Choosing the Procedure to Recover User
Recovery Procedure for a System Failure . . . . 63 Information . . . . . . . . . . . . . . 107
Recovery Procedure for a Program Failure or Recovering User Information Using
Human Error . . . . . . . . . . . . . . 63 Commands–Checklist 24. . . . . . . . . 108
Choosing the Recovery Procedure for a Disk Failure Using Option 21 from the Restore
or Disk Errors . . . . . . . . . . . . . 63 Menu–Checklist 25 . . . . . . . . . . 112
Actions for load source disk unit Using Options 22 and 23 from the Restore
failure–Checklist 1 . . . . . . . . . . . 65 Menu–Checklist 26 . . . . . . . . . . 114
Actions for load source disk unit Recovering User Information Using Tapes from
failure–Checklist 2 . . . . . . . . . . . 66 Operational Assistant Backup–Checklist 27 . . 117
Actions for load source disk unit
failure–Checklist 3 . . . . . . . . . . . 67 Chapter 5. Recovering the Licensed Internal
Actions for load source disk unit Code . . . . . . . . . . . . . . . . 121
failure–Checklist 4 . . . . . . . . . . . 68 How to Prepare for Loading the Licensed Internal
Actions for load source disk unit Code . . . . . . . . . . . . . . . . 122
failure–Checklist 5 . . . . . . . . . . . 72 Task 1–Getting Ready to Load the Licensed
Actions for non-load source disk unit failure or Internal Code . . . . . . . . . . . . 122
disk units in user ASP disk failure–Checklist 6. . 76 Task 2–Powering Down the System . . . . . 123
Actions for non-load source disk unit Task 3a–Preparing the System to Perform an IPL
failure–Checklist 7 . . . . . . . . . . . 77 from an Alternate Device . . . . . . . . 124
Actions for non-load source disk unit Task 3b-Preparing a Logical Partition to Perform
failure–Checklist 8 . . . . . . . . . . . 78 an IPL from an Alternate Device . . . . . . 125
Actions for non-load source disk unit Task 4–Loading the Licensed Internal Code from
failure–Checklist 9 . . . . . . . . . . . 79 Media . . . . . . . . . . . . . . . 125
Actions for non-load source disk unit How to Load the Licensed Internal Code . . . . 130
failure–Checklist 10 . . . . . . . . . . . 82 How to Recover Your Logical Partition
Configuration . . . . . . . . . . . . . 134

© Copyright IBM Corp. 1997, 2000, 2001 35

 How to Recover Your Disk Configuration After Task 4–Recovering Journals and Journal
Installing the Licensed Internal Code and Receivers in the QRCL Library . . . . . . 185
Initializing the System . . . . . . . . . . 135 Task 5–Restoring Libraries to the System
| How to Recover Your Disk Configuration Using Auxiliary Storage Pool . . . . . . . . . 186
| Operations Navigator at DST . . . . . . . . 138 Task 6–Restoring Document Library Objects to
How to Recover Your Disk Configuration . . . . 142 the System Auxiliary Storage Pool . . . . . 187
How to Start Your System After Restoring Licensed Task 7–Restoring User-Defined File Systems to
Internal Code . . . . . . . . . . . . . 145 the User Auxiliary Storage Pool . . . . . . 187
Disabling and Enabling the High-Speed Feature on Recovery Steps for Unmounted User-Defined
the 2440 Tape Unit . . . . . . . . . . . 145 File Systems (UDFS) . . . . . . . . . 187
Recovery Steps for Mounted User-Defined
Chapter 6. Restoring the Operating System . . 147 File Systems (UDFS) if Data is not Restored . 187
Choosing the Right Procedure for Restoring the Recovery Steps for Mounted User-Defined
Operating System . . . . . . . . . . . . 148 File Systems (UDFS) if Data is Restored . . 188
How to Load the Operating System Using a Task 8–Reclaiming Document Library Objects 188
Manual IPL . . . . . . . . . . . . . . 149 Task 9–Recovering Save Files from the QRCL
How to Restore the OS/400 Licensed Program . . 149 Library . . . . . . . . . . . . . . 188
Task 1–Starting to Restore the Operating System 150 Task 10–Associating Journal Receivers with
Task 2–Selecting the Installation Options . . . 154 Journals . . . . . . . . . . . . . . 189
Task 3–Selecting IPL Options . . . . . . . 158 Task 11–Restoring Object Ownership . . . . 190
Task 4–Setting Major System Options . . . . 160 How to Recover An Overflowed User Auxiliary
Task 5–Defining or Changing the System at IPL 160 Storage Pool . . . . . . . . . . . . . . 191
Task 6–Completing the IPL . . . . . . . . 162 Resetting An Overflowed User Auxiliary
Recovering from SRC A900 2000 . . . . . . . 164 Storage Pool without an IPL . . . . . . . 192
Creating a Configuration for 34xx Tape Units 165 Resetting An Overflowed User Auxiliary
Creating a Configuration for Other Tape Units 166 Storage Pool during an IPL . . . . . . . . 193
How to Delete Overflowed Objects during
Chapter 7. Starting the System After It Ends Recovery . . . . . . . . . . . . . . . 196
Abnormally . . . . . . . . . . . . . 167 How to Recover a Damaged User Auxiliary
What Happens When Your System Stops . . . . 167 Storage Pool . . . . . . . . . . . . . . 196
Using the Disk Configuration Error Report Task 1–Restoring User Profiles. . . . . . . 196
Display . . . . . . . . . . . . . . 168 Task 2–Determining the Contents of the Lost
Using the Main Storage Dump Manager Auxiliary Storage Pool . . . . . . . . . 197
Occurred Display . . . . . . . . . . . 168 Task 3–Determining Tasks to Restore Objects 197
How to Restart Your System . . . . . . . . 169 Task 4–Restoring Libraries to the User Auxiliary
Task 1–Performing an Attended IPL . . . . . 169 Storage Pool . . . . . . . . . . . . . 198
Task 2–Using the Edit Rebuild of Access Paths Task 5–Restoring Journals to the User Auxiliary
Display . . . . . . . . . . . . . . 171 Storage Pool . . . . . . . . . . . . . 198
Task 3–Using the Edit Check Pending Task 6–Restoring Documents to the User
Constraints Display . . . . . . . . . . 173 Auxiliary Storage Pool . . . . . . . . . 200
Task 4–Recovering from Damaged Objects and Task 7–Restoring User-Defined File Systems to
Unreadable Sectors . . . . . . . . . . 174 the User Auxiliary Storage Pool . . . . . . 200
Recovering Damaged Database Files. . . . 176 Recovery Steps for Unmounted User-Defined
Recovering a Damaged Journal . . . . . 178 File Systems (UDFS) . . . . . . . . . 200
Recovering a Damaged Journal Receiver . . 179 Recovery Steps for Mounted User-Defined
Recovering a Journaled Object That Is File Systems (UDFS) if Data is not Restored . 200
Damaged or Not Synchronized . . . . . 179 Recovery Steps for Mounted User-Defined
Recovering Damaged Objects in the File Systems (UDFS) if Data is Restored . . 201
Integrated File System (IFS) . . . . . . 180 Task 8–Restoring Journal Receivers to the User
Recovering Other Types of Damaged Objects 180 Auxiliary Storage Pool . . . . . . . . . 201
Task 9–Restore Save Files to the User Auxiliary
Chapter 8. Recovering Information in a User Storage Pool . . . . . . . . . . . . . 201
Auxiliary Storage Pool. . . . . . . . . . 181 How to Remove a Failed Disk Unit from the
Describing the Contents of Your User Auxiliary System ASP . . . . . . . . . . . . . . 202
Storage Pools . . . . . . . . . . . . . 181 Task 1–Access Dedicated Service Tools . . . . 202
Choosing the Procedure to Recover User ASPs . . 182 Task 2–Delete the Auxiliary Storage Pool Data 203
How to Recover a User ASP After Recovering the Task 3–Remove the Disk Unit from the
System ASP . . . . . . . . . . . . . . 182 Auxiliary Storage Pool Configuration . . . . 204
Task 1–Reclaiming Storage . . . . . . . . 183
Task 2–Restoring User Profiles. . . . . . . 184 Chapter 9. The Restore Menu . . . . . . . 207
Task 3–Restoring the Configuration . . . . . 185 What the Restore Menu Options Do . . . . . . 208

36 OS/400 Backup and Recovery V5R1

 How to Use Restore Menu Options 21, 22, and 23 208 How the System Restores Files with Referential
Constraints . . . . . . . . . . . . . 245
Chapter 10. How to Restore Specific Types of Referential Constraint Network–Example . . 246
Information. . . . . . . . . . . . . . 213 How the System Restores Files with Triggers 247
Recovering System Information . . . . . . . 213 Steps Before Deleting a Physical File . . . . 248
Sequence for Restoring Security Information . . . 213 Restoring Journals and Journal Receivers . . . . 248
Restoring User Profiles . . . . . . . . . . 214 Restoring Journals . . . . . . . . . . . 248
What Happens When You Restore User Profiles 215 Steps before Deleting a Journal . . . . . . 249
What You Should Know About Restoring User Restoring Journal Receivers. . . . . . . . 250
Profiles . . . . . . . . . . . . . . 216 How to Resolve Name Conflicts When
How the System Establishes Ownership for Restoring Journal Receivers. . . . . . . 251
Restored Objects . . . . . . . . . . . 218 How to Correct the Journal Receiver
How the System Establishes the Authorization Directory . . . . . . . . . . . . . 251
List for a Restored Object . . . . . . . . 218 Steps before Deleting a Journal Receiver . . . 251
How the System Establishes the Primary Group How the System Restores Programs . . . . . . 252
for Restored Objects . . . . . . . . . . 218 Restoring Programs to a Different Release . . . 253
Restoring Object Authorities . . . . . . . . 219 Restoring OPM Program Objects . . . . . . . 254
Overview of Restoring Authorities . . . . . 219 Restoring Save File Data. . . . . . . . . . 255
Restoring Authority On a System in a Restoring Spooled Output Files . . . . . . . 255
Non-Restricted State . . . . . . . . . . 220 Restoring Licensed Programs . . . . . . . . 255
What You Should Know Before Running Restoring Documents and Folders . . . . . . 255
RSTAUT . . . . . . . . . . . . . 220 RSTDLO Command Options . . . . . . . 255
Job Log Considerations . . . . . . . . 221 Using Multiple Concurrent DLO commands . . 256
Restoring Authority On a System in a Restricted Output from the RSTDLO Command . . . . 256
State . . . . . . . . . . . . . . . 225 Considerations and Restrictions . . . . . . 256
What the System Does When You Restore Moving Documents . . . . . . . . . 256
Authority. . . . . . . . . . . . . . 225 Searching Tape Files . . . . . . . . . 257
How to Restore Configuration Objects . . . . . 227 Selecting files from DVD-RAM optical media 257
Correcting Problems with the System Resource Search Index Database Errors . . . . . . 257
Management Information . . . . . . . . 228 Authority Required to Restore DLOs . . . 257
Recovering Devices That Will Not Vary On 228 How the System Restores New DLOs . . . 257
Recovering When You Change the Console Type 230 How the System Restores Existing DLOs . . 257
Recovering the System/36 Environment Size Limitations When Restoring Document
Configuration . . . . . . . . . . . . 230 Library Objects . . . . . . . . . . . 257
Restoring Logical Partitions . . . . . . . . 231 Restoring Folders . . . . . . . . . . . 258
Restoring Libraries . . . . . . . . . . . 231 Renaming Documents When Restoring . . . . 258
Restoring a Library From a Previous Release 232 Restoring OfficeVision/400 Mail and
Restoring Multiple Libraries . . . . . . . 232 Distribution Objects . . . . . . . . . . 258
Considerations and Restrictions . . . . . . 233 How the System Restores Descriptive
How to Restore All Libraries from a Single Save Information for DLOs . . . . . . . . . 259
Operation . . . . . . . . . . . . . 233 How the System Restores Authority and
How to Restore All Libraries from Multiple Ownership for DLOs . . . . . . . . . . 259
Save Operations . . . . . . . . . . . 234 When to Run the Rename Directory
How to Restore Objects . . . . . . . . . . 234 (RNMDIRE) Command . . . . . . . . . 260
| Restoring Objects That Are Journaled . . . . . 235 When to Run the Rename Document Library
| What Happens When You Restore Journaled Object (RNMDLO) Command . . . . . . . 260
| Objects to a Different Library . . . . . . . 235 Recovery of Text Index Files for Text Search
Restoring Database Files. . . . . . . . . . 236 Services . . . . . . . . . . . . . . . 260
Comparing File Attributes during a Restore Restoring Objects in Directories . . . . . . . 261
Operation . . . . . . . . . . . . . 238 Completing Recovery for the iSeries Integration for
How the System Matches File Members during Windows Server Product . . . . . . . . . 263
a Restore Operation . . . . . . . . . . 241 Recovery for save performed with Integrated
Restoring Members to a File . . . . . . . 241 xSeries Server varied off . . . . . . . . . 263
Restrictions on the File Member Parameter 241 Recovery for save performed with Integrated
Restoring Logical Files . . . . . . . . . 242 xSeries Server varied on . . . . . . . . . 264
How the System Restores Access Paths . . . . 242 | Recovering Linux in a Partition . . . . . . 264
Restoring a File Network–Examples . . . . 243 Recovery Steps for OS/400 Enhanced
How to Prevent the System from Rebuilding Integration for Novell NetWare . . . . . . 264
a Large Access Path . . . . . . . . . 245 Recovering a Domino™ Server. . . . . . . . 265
How the System Restores Files with Shared Recovering an entire Domino server . . . . . 265
Formats . . . . . . . . . . . . . . 245 Recovering Domino mail . . . . . . . . 266

Part 3. Recovering Information on Your System 37

 Recovering specific Domino databases . . . . 267 Missing Disk Units . . . . . . . . . . 298
Restoring changed objects to a Domino server 267 Saving a Unit . . . . . . . . . . . . 299
Example: Restoring changed Domino objects Restoring a Unit . . . . . . . . . . . 299
from a cumulative backup . . . . . . . 268 Active Mirrored Load Source Failure . . . . 299
Example: Restoring changed Domino objects System Cannot Find Active Mirrored Load
from a nightly backup . . . . . . . . 268 Source for IPL . . . . . . . . . . . 299
Example: Restoring Domino databases from Active Mirrored Load Source Being Used for
an incremental backup . . . . . . . . 269 IPL Fails . . . . . . . . . . . . . 300
Example: Restoring changed objects from a Active Mirrored Load Source Fails Late in
specific Domino subdirectory . . . . . . 269 the IPL or at Runtime . . . . . . . . 301
Restoring a Windows server . . . . . . . . 270 Cannot Read System Configuration Data
Restoring the NWSD and disk drives for from Active Mirrored Load Source . . . . 301
Windows server on iSeries . . . . . . . . 270 Unknown Unit 1 Status . . . . . . . . . 301
Restoring predefined disk drives for To Recover the State of the Unknown Load
Windows servers created on V4R5 and later Source . . . . . . . . . . . . . . 302
systems . . . . . . . . . . . . . 271 Display Incorrect Licensed Internal Code Install 302
Restoring predefined disk drives for
Windows servers created on pre-V4R5 Chapter 13. How to Restore Your System Using
systems . . . . . . . . . . . . . 271 Operational Assistant Tapes. . . . . . . . 305
Restoring user-defined disk drives for How to Restore Your Libraries. . . . . . . . 306
Windows servers on iSeries. . . . . . . 272 How to Restore Libraries That You Saved by Using
Restoring NWSDs for Windows server on a Backup List . . . . . . . . . . . . . 307
iSeries . . . . . . . . . . . . . . 273 How to Restore Changed Objects That You Saved
Recovering Windows server files . . . . . . 273 by Using Operational Assistant . . . . . . . 308
Examples: How to address parts of Windows
server . . . . . . . . . . . . . . . 274 Chapter 14. How to Restore the System from
Restrictions When Using the Restore Command 275 the Save Storage Media . . . . . . . . . 311
How to Restore Program Temporary Fixes. . . . 278 Task 1–Powering Down the System and Loading
the Licensed Internal Code . . . . . . . . . 311
Chapter 11. How to Restore Changed Objects Task 2–Restoring the Save Storage Tapes . . . . 312
and Apply Journaled Changes . . . . . . . 279 Task 3–Responding to Messages . . . . . . . 314
Task 1–Restoring Changed Objects . . . . . . 280 Task 4–Completing the Restore Storage Operation 315
Restoring Changed Objects by Library . . . . 280 Task 5–Restoring Additional Information . . . . 318
Restoring Changed Objects Individually . . . 280 Task 6–Restoring Program Temporary Fixes (PTFs) 318
Task 2–Restoring Changed Objects in Directories 281 How to Resume the Restore Storage Operation . . 318
Task 3–Determining Whether You Need to Apply
Journaled Changes . . . . . . . . . . . 282
Task 4–Determining What Journal Receivers to Use 282
Task 5–Applying Journaled Changes for User
Journals . . . . . . . . . . . . . . . 284
Task 6–Applying Journaled Changes for the
QAOSDIAJRN Journal . . . . . . . . . . 286
Task 7–Restoring Changed Documents and Folders 287

Chapter 12. Mirrored Protection Recovery

Actions . . . . . . . . . . . . . . . 289
System Actions for Permanent Errors . . . . . 289
Suspending Mirrored Units. . . . . . . . . 290
Resuming Mirrored Units . . . . . . . . . 291
Replacing a Mirrored Unit . . . . . . . . . 292
Using Spare Nonconfigured Units for
Replacement. . . . . . . . . . . . . 294
Mirrored Protection Recovery Actions
Performed by the Service Representative . . . 296
Actions When Concurrent Maintenance is
Possible . . . . . . . . . . . . . 296
Actions When Concurrent Maintenance is
Not Possible . . . . . . . . . . . . 296
Other Recovery Considerations for Mirrored
Protection . . . . . . . . . . . . . 297
Mirrored Protection Disk-Error Handling . . . 297

38 OS/400 Backup and Recovery V5R1

Chapter 3. Restore Procedures–General Information
Figure 5 on page 40 shows the menu options and commands that are available for
restoring information. It also shows the normal sequence for restoring information,
working from top to bottom. Figure 6 on page 41 shows what restore commands
can be used for information in the different file systems.

Note: Look for comprehensive information about how to save your iSeries server
on the Information Center. “Prerequisite and related information” on
page xix explains how to access the Information Center.

Compare these figures with the save information on the Information Center to see
the relationship between how things are saved and how they are restored. Use
them to gain a general understanding of what you need to restore and how you
might do it. Use the information in Chapter 4. Selecting the Right Recovery
Strategy to plan the correct recovery strategy for your situation.

© Copyright IBM Corp. 1997, 2000, 2001 39

 Restore Parts of the System Procedure for Restoring
Menu
┌───────────────────────────────┐W──────────────────────────────┐
│ Licensed Internal Code │ Option on Install Licensed │
│ │ Internal Code (LIC) Screen │
└───────────────────────────────┴W──────────────────────────────┘

┌───────────────────────────────┐W──────────────────────────────┐
│ OS/400 Objects in QSYS │ IPL or Install the System Menu│
└───────────────────────────────┘W──────────────────────────────┘

┌─┬─┬─S┌───────────────────────────────┐W──────────────────────────────┐
│ │ │ │ User Profiles │ RSTUSRPRF │
│ │ │ └───────────────────────────────┘W──────────────────────────────┘
│ │ 23
│ │ │ ┌───────────────────────────────┐W──────────────────────────────┐
│ │ │ │ Configuration Objects │ RSTCFG │
│ │ └─S└───────────────────────────────┘W──────────────────────────────┘
│ 22
│ │ ┌───────────────────────────────┐W───────────┐
│ │ │ IBM-supplied directories │ RST │
│ │ └───────────────────────────────┘W───────────┘
│ │
│ │ ┌───────────────────────────────┐W───────────┐W─────────────────┐
│ │ │ OS/400 Optional Libraries │ │ │
│ │ │ QHLPSYS QUSRTOOL │ RSTLIB │ │
│ │ ├───────────────────────────────┤ *IBM │ │
│ │ │ Licensed Program Libraries │ │ │
│ │ │ QRPG QCBL Qxxxxx │ │ RSTLIB │
│ └───S└───────────────────────────────┘W───────────┘ *NONSYS │
21 │
│ ┌─S┌───────────────────────────────┐W───────────┐ │
│ │ │ IBM Libraries with User Data │ │ │
│ │ │ QGPL QUSRSYS QS36F #LIBRARY │ RSTLIB │ │
│ │ ├───────────────────────────────┤ *ALLUSR │ │
│ │ │ User Libraries │ │ │
│ │ │ LIBA LIBB LIBC LIBxxx │ │ │
│ │ └───────────────────────────────┘W───────────┘W─────────────────┘
│ 23
│ │ ┌───────────────────────────────┐W──────────────────────────────┐
│ │ │ Filed Documents and Folders │ RSTDLO │
│ │ │ Distribution Objects │ │
│ │ └───────────────────────────────┘W──────────────────────────────┘
│ │
│ │ ┌───────────────────────────────┐W──────────────────────────────┐
│ │ │ User Objects in Directories │ RST │
└───┴─S└───────────────────────────────┘W──────────────────────────────┘

┌───────────────────────────────┐W──────────────────────────────┐
│ Saved Changes in Libraries, │ RSTLIB, RSTOBJ, RSTDLO, RST │
│ Documents, and Directories │ │
└───────────────────────────────┘W──────────────────────────────┘

┌───────────────────────────────┐W──────────────────────────────┐
│ Journaled Changes │ APYJRNCHG │
└───────────────────────────────┘W──────────────────────────────┘

┌─────S┌───────────────────────────────┐W──────────────────────────────┐
│ All │ Private Authorities │ RSTAUT │
└─────S└───────────────────────────────┘W──────────────────────────────┘

Figure 5. Restore Procedures

Note: RSTOBJ can also be used where RSTLIB is shown to restore objects.

40 OS/400 Backup and Recovery V5R1

 Figure 6. Save procedures and restore procedures for file systems

The Relationship Between Save and Restore Commands

Table 4 on page 42 shows which restore commands can be used, based on how the
objects were saved.

Note: For comprehensive information on saving your server, refer to the Backing
up your system topic on the Information Center at the following Web site:
http://www.ibm.com/eserver/iseries/infocenter

Chapter 3. Restore Procedures–General Information 41

 This Web site includes general information on the save commands below.
Table 4. Relationship Between Save and Restore Commands
Save Command Used Possible Restore Command

SAVOBJ RSTOBJ
RST
SAV RST
SAVLIB LIB(*NONSYS) RSTLIB SAVLIB(*NONSYS)
RSTLIB SAVLIB(*IBM)
RSTLIB SAVLIB(*ALLUSR)
RSTLIB SAVLIB(library-name)
RST
SAVLIB LIB(*ALLUSR) RSTLIB SAVLIB(*ALLUSR)
RSTLIB SAVLIB(library-name)
RST
SAVLIB LIB(*IBM) RSTLIB SAVLIB(*IBM)
RSTLIB SAVLIB(library-name)
RST
SAVLIB LIB(library-name) RSTLIB SAVLIB(library-name)
RST
SAVSECDTA RSTUSRPRF
RSTAUT1
SAVCFG RSTCFG
SAVSYS Restore Licensed Internal Code. (See
Chapter 5.)
Restore operating system. (See Chapter 6.)
RSTUSRPRF
RSTCFG
RSTAUT1
SAVDLO RSTDLO
RST
1
The RSTUSRPRF command restores authority information to temporary tables. The
RSTAUT command regrants private authorities using tables that are built as a part
of the RSTUSRPRF command.

What Happens When You Restore Objects

An object on this system is like a container. The object has information about the
container itself, such as the owner of the object and the last time it was saved. This
is the information you see when you display the object description (DSPOBJD
command). The object also has contents, such as the records in a database file or
the instructions in a program.

When you restore an object, the system takes different actions depending on the
following:
v Whether the object to be restored already exists.
v The allow object differences (ALWOBJDIF) parameter on the restore command.
v Whether the object was saved on a different system (serial number of the
processor).

With a few exceptions that relate to security, the contents of the object are always
restored. If the object exists, the system compares the object description
information on the system copy and the media copy and then makes decisions. For

42 OS/400 Backup and Recovery V5R1

 most information, the media version of the information is restored. For
security-relevant information, such as the public authority and the object owner,
the system version is left unchanged. In a few cases, such as the size of the object
and the date it was restored, the system determines a value when the object is
restored.

The allow object differences (ALWOBJDIF) parameter on the restorecommands is

primarily for security protection and integrity protection. For example, if system
security is important to you, you may want to take special action if someone
attempts to restore an object whose owner has been changed. Or, if the member
information about a database file does not match, you may have problems with the
integrity of your data. You can use the ALWOBJDIF parameter to prevent this.

The default value for the ALWOBJDIF parameter is *NONE. This means that if
important differences exist between the media version and the system version of
an object, you want the system to take special action. Normally, you should use the
default value. However, when you are restoring your information to a different
system, such as during a disaster recovery, you should specify ALWOBJDIF(*ALL).
Table 5 shows examples of the effect of the ALWOBJDIF parameter:
Table 5. Restoring Objects with ALWOBJDIF. Effect of ALWOBJDIF parameter when the
value on the media and on the system are different.
Value for Object after Restore Operation

Object Characteristic ALWOBJDIF(*NONE) ALWOBJDIF(*ALL) ALWOBJDIF(*FILELVL)

That Differs Specified Specified Specified

| Object owner Object is not restored Existing value1 Object is not restored
| Object primary group Object is not restored Existing value3 Object is not restored
| Object auditing Existing value Existing value Existing value
Authorization list, restore over existing object:
| Object on media is Object is not restored Object is restored and Object is not restored
secured by an is secured by
authorization list and authorization list of
object on system is object on system2
not secured by an
authorization list
| Object on media is Object is restored and Object is restored and Object is restored and
| not secured by an is secured by is secured by is secured by
| authorization list and authorization list of authorization list of authorization list of
| object on system is object on system object on system2 object on system
secured by an
authorization list
| Object on media is Object is not restored Object is restored and Object is not restored
secured by an is secured by
authorization list and authorization list of
object on system is object on system;
secured by a different message is sent to
authorization list user2
Authorization list, new object restored:
| Object is restored Object is restored and Object is restored and Object is restored and
| onto different system is not secured by an is secured by the is not secured by an
| than it was saved authorization list same authorization authorization list
from list that secured the
object when it was
saved, if the
authorization list
exists2
Database files:

Chapter 3. Restore Procedures–General Information 43

 Table 5. Restoring Objects with ALWOBJDIF (continued). Effect of ALWOBJDIF parameter
when the value on the media and on the system are different.
Value for Object after Restore Operation

Object Characteristic ALWOBJDIF(*NONE) ALWOBJDIF(*ALL) ALWOBJDIF(*FILELVL)

That Differs Specified Specified Specified

| Creation date for file File is not restored File is renamed on Logical file is not
| the system; copy is restored. System
| restored from media attempts to restore
| with media creation physical file data4
date; message is sent
to user.
| Creation date for Member is not Member is renamed Logical member is
| member restored on the system; copy not restored. System
| is restored from attempts to restore
| media with media physical member
| creation date; data4
message is sent to
user.
| Physical file data
| Level identifier for Physical file data is Physical file data is System attempts to
| file not restored not restored restore physical file
| data4
| Level identifier for Physical file data is Physical file data is System attempts to
| member not restored not restored restore physical
| member data4
Validation values:
| Validation value, with Program object is Program object is Program object is
| system value restored with restored without any restored with
| QSECURITY at 40 or authority changes authority changes authority changes
higher
| Validation value, with Object is restored Object is restored Object is restored
system value
QSECURITY lower
than 40
Program object without validation value, with system value QSECURITY
at 40 or higher:
| Program object Program object Program object
| restored with restored without restored with
| retranslation retranslation retranslation
1
Also applies to RST command with ALWOBJDIF(*OWNER)
2
Also applies to RST command with ALWOBJDIF(*AUTL)
3
Also applies to RST command with ALWOBJDIF(*PGP)
4
| Only applies to RSTLIB and RSTOBJ commands with
| ALWOBJDIF(*FILELVL)

The following topics provide more information about the effect of the ALWOBJDIF
parameter:
v “How the System Establishes Ownership for Restored Objects” on page 218
v “How the System Establishes the Authorization List for a Restored Object” on
page 218
v “Comparing File Attributes during a Restore Operation” on page 238
v “How the System Restores Programs” on page 252

44 OS/400 Backup and Recovery V5R1

 Sequence for Restoring Related Objects
| Some objects depend on other objects. When related objects are in the same library
| or directory, the system restores them in the correct order. If related objects are in
| different libraries or directories, you must restore them in the correct order or
| perform additional recovery steps after they are restored.

If possible, restore objects in this sequence:

| v Journals before journaled objects. If you restore a journaled object when the
| journal is not on the system, you must start journaling again after the journal is
| restored. Use the STRJRNPF command, the STRJRNAP command, the
| STRJRNOBJ command, or the STRJRN command. See “Restoring Objects That
| Are Journaled” on page 235 for more information.
v Journals before journal receivers. If you restore a journal receiver when the
journal is not on the system, you must associate the journal receivers with the
journal after it is is restored. Use the WRKJRN command. See “Restoring
Journals and Journal Receivers” on page 248 for more information.
v Physical files before logical files. You cannot restore a logical file if the based-on
physical files are not on the system. “How the System Restores Access Paths” on
page 242 describes how to restore logical files and based-on physical files that
are in different libraries.

Putting Your System in a Restricted State

Many recovery procedures require that your system have no other activity on it.
When no subsystems except the controlling subsystem are active on your system, it
is in a restricted state.

Use the End Subsystem (ENDSBS) command to put your systemin a restricted
state. You specify how you want the subsystems to end:

Possible Values for the OPTION Parameter of the ENDSBS Command:

*CNTRLD Allow active jobs to end themselves (if they are checking
to see if the job is being ended). If you specify *CNTRLD,
you can use the delay parameter to set a time for the
system to wait before ending subsystems immediately.
*IMMED End the subsystem immediately. Use this option if there
are no users on the system and no batch jobs running.

Note: Even if you have no activity on the system, jobs may be running under a
few system-provided subsystems, such as the QSYSWRK (subsystem
monitor) subsystem and the QCALSRV (calendar server) subsystem. You can
endall subsystems immediately without first ending these jobs. You will
receive messages that these subsystems ended abnormally.

Do the following to put your system in a restricted state:

1. Before putting your system in a restricted state, ensure that all users are signed
off and all jobs are ended.
2. To receive notification that the subsystems have ended, type the following and
press the Enter key:
CHGMSGQ MSGQ(QSYSOPR) DLVRY(*BREAK)
SEV(60)
3. To end all subsystems, type the following:

Chapter 3. Restore Procedures–General Information 45

 ENDSBS SBS(*ALL) OPTION(*CNTRLD)
DELAY(600)

Note: For the delay parameter, specify a number of seconds that allows your
system time to bring most jobs to a normal end. On a large, busy system,
you may need a longer delay.

A message is sent that indicates that the procedure for ending subsystems is in
progress. A final message is sent when the system is in a restricted state.

Reclaiming Storage
Use the reclaim storage procedure (RCLSTG command) to recover the
addressability of lost or damaged objects. This allows you to identify and then
restore those objects that were damaged. If an authorization list is found damaged
during reclaim storage, the objects secured by the damaged authorization are
associated with the system authorization list QRCLAUTL. To find out how to
recover from damaged authorization lists, look under the Programming topic on
the Information Center at the following Web
site:http://www.ibm.com/eserver/iseries/infocenter.

| The RCLSTG command has three parameters, SELECT, OMIT, and ASP. These
| parameters allow you to perform reclaim functions in one of the following ways:
v All reclaim functions are performed
v The database cross-reference table reclaim function is performed
v All reclaim functions are performed, except for the database cross-reference table
reclaim function
| v Reclaim the system ASP and all basic ASPs.The system ASP has an ASP number
| of 1. Basic ASPs have ASP numbers of 2 through 32.
| v Relcaim a specific independent ASP. Independent ASPs have a number greater
| than 32.

Note: The RCLSTG procedure requires auxiliary storage. If you are already using a
very high percentage of auxiliary storage, the RCLSTG procedure may not
complete successfully.

How to Reclaim Storage

To reclaim storage, do the following:
1. Sign on the system with a user profile that is authorized to the RCLSTG
command. Either sign on at the console or use the Transfer Job (TFRJOB)
command to transfer your job to the controlling subsystem.
2. Type DSPSYSVAL QALWUSRDMN. If the current value does not include the QRCL
(Reclaim Storage) library or does not specify *ALL, use the CHGSYSVAL
command to add QRCL to the list of libraries for this system value. Write the
current value here: __________________
3. Type DSPSYSVAL QCTLSBSD to display the name of your controlling subsystem.
Write the current value here: _________________
4. Ensure your system is in a restricted state. If it is not, follow the procedure in
“Putting Your System in a Restricted State” on page 45.
5. Start the reclaim storage process by typing one of the following:

RCLSTG Reclaim storage of the entire system.

46 OS/400 Backup and Recovery V5R1

 RCLSTG Reclaim storage of the database cross-reference table.
SELECT(*DBXREF)
RCLSTG OMIT(*DBXREF) Reclaim storage of the entire system except the database
cross-reference table.
| RCLSTG Reclaim the system ASP and all basic ASPs
| ASPDEV(*SYSBAS)
| RCLSTG Reclaim a specified independent ASP. Enter the name of the
| ASPDEV(device-description- device description that corresponds with the independent ASP.
| name)

6. Use the CHGSYSVAL command to set the QALWUSRDMN system value back
to its original setting. (You wrote the setting in step 2.)
7. When the reclaim storage procedure finishes, start the controlling subsystem by
typing the following:
STRSBS SBSD(controlling-subsystem)

(You wrote the name of the controlling subsystem in step 3.)

What Happens When You Reclaim Storage: The purpose of the RCLSTG command
is to ensure the following:
v Objects that reside permanently in auxiliary storage can be accessed.
v All auxiliary storage either is used properly or is available for use.

The system checks every object that resides permanently in auxiliary storage for
loss or damage. It does the following:
v If an object does not address a library or directory,it is placed in an
IBM-supplied library or directory based on the object type. The system may not
be able to retrieve description information for the object, such as:
– Program temporary fix (PTF) status.
– Save and restore information.
– Object attributes and text description.
v For objects that normally reside in libraries (the QSYS.LIB file system), the
system does the following:
– If a lost object with the same name and object type is alreadyin the Recovery
(QRCL) library, the system gives the object that it has just encountered a new
name. The name has the format QRCLnnnnnn, where nnnnnn is a unique
number. The original object name is placed in the text description for the
object in the QRCL library.

Note: You cannot rename journals and journal receivers. If the system
encounters two journals (or journal receivers) with the same name and
they both need to be placed in the QRCL library, the system renames
one of them. You cannot rename that journal or journal receiver back to
its original name. You must restore a previous version with the correct
name or re-create the journal or journal receiver. For this reason, you
should use a naming convention for journals and journal receivers that
is unique for the entire system, not just for a library.
– If data exists for a lost physical file, thesystem attempts to rebuild the file and
place it in the QRCL library. The text description for the object in the QRCL
library indicates that it has been rebuilt.

Chapter 3. Restore Procedures–General Information 47

 To use the physical file, create it again in the correct library with the correct
attributes. Then copy the data from the rebuilt file in the QRCL library to the
new physical file. The data in the file may not be complete.
– A user domain object can be placed in the QRCL libraryonly if the
QALWUSRDMN system value includes QRCL or specifies *ALL. Otherwise, a
lost user domain object is deleted. Most objects are system domain objects.
User domain objects have type *USRSPC, *USRIDX, or *USRQ.
– If an object does not have an owner, it is assigned toan IBM-supplied user
profile based on the object type. Most objects are assigned to the QDFTOWN
user profile.
– If descriptions for objects in a library cannot be accessed, the library is rebuilt.
– If an object is secured by a damaged authorization list or authority holder, the
system makes QRCLAUTL the authorization list for the object. You can use
the Display Authorization List Objects (DSPAUTLOBJ) command to determine
which objects are secured by the QRCLAUTL authorization list.
v If a lost object was in the Root file system, the object is placed in the /QReclaim
directory.
v If a lost object was in the QOpenSys file system, the object is placed in the
/QOpenSys/QReclaim directory.
v If an object in a directory is damaged to the extent that it is unusable, the system
deletes it. The RCLSTG command does not attempt to rebuild damaged objects.
v If a lost object was in a user-defined file system (UDFS), it is placed in the
QReclaim directory located in the root directory of the UDFS.
v If a lost object that was in a directory cannot be placed in the proper QReclaim
directory based on its original location, then it is placed into the root directory of
a special file system within the Auxiliary Storage Pool (ASP) in which the object
resides. This special file system is created by RCLSTG when needed. The file
system is named ’/dev/QASPxx/QReclaimFS.udfs’ where ’xx’ is the ASP
number.
v For objects in the Root, QOpenSys, or user-defined file systems, the system takes
actions for duplicate names or for unidentified object owners similar to the
actions taken for objects in the QSYS.LIB file system.

What To Do After Running the RCLSTG Procedure: Table 6 describes both where to
look for problems that the RCLSTG procedure detects and how to correct the
problems:
Table 6. Resolving Problems Detected by the RCLSTG Procedure
Where to Look for Problems How to Fix the Problem
Type DSPMSG QSYSOPR to display the QSYSOPR message Type DSPLOG QHST to display the history log. Look for
queue. Look for messages about damaged objects. messages about either damaged objects or rebuilt files.
1. Delete unusable objects by using the appropriate
DLTxxx command. Restore them by using the Restore
Object (RSTOBJ) command.
2. Copy data from rebuilt files to new files by using the
Copy File (CPYF) command.
Note: You may see a message indicating that objects
were deleted by the reclaim storage procedure. These are
internal system objects that are no longer needed.

48 OS/400 Backup and Recovery V5R1

Table 6. Resolving Problems Detected by the RCLSTG Procedure (continued)
Where to Look for Problems How to Fix the Problem
Type DSPLIB QRCL to display the QRCL library. Move objects from the QRCL library to the correct library
Note: If the reclaim storage procedure did not place any by using the Move Object (MOVOBJ) command.
objects in the QRCL library, you may receive a message
Notes:
that the library is not found. Ignore the message and
continue with the next step. 1. If IBM-supplied objects are in the QRCL library,
contact software support for assistance.
2. If you are unsure what to do with objects in the
QRCL library, use the SAVLIB command to save the
library to save media. Mark the media volume in case
you need the objects later.
Display the /QReclaim directory by using the Display Move objects from the /QReclaim directory to the correct
Link (DSPLNK) command. directory by using the Move (MOV) command.
Note: If the reclaim storage procedure did not place any
objects in the /QReclaim directory, you may receive a
message that the object is not found. Ignore the message
and continue with the next step.
Display the /QOpenSys/QReclaim directory by using Move objects from the /QOpenSys/QReclaim directory
the Display Link (DSPLNK) command. to the correct directory by using the Move (MOV)
Note: If the reclaim storage procedure did not place any command.
objects in the /QOpenSys/QReclaim directory, you may
receive a message that the object is not found. Ignore the
message and continue with the next step.
Type DSPMSG QSYSOPR to display the QSYSOPR message Use the Add Mounted File System (ADDMFS) command
queue. Look for CPFA0D7 messages. For each CPFA0D7 to mount the user-defined file system (UDFS) specified
message that contains a directory name beginning with in the CPFA0D7 message over a directory of your choice.
’/dev/QASPxx/’ (where ’xx’ is the number of an ASP), Then use the Display Link (DSPLNK) command to view
perform the action specified in the ″How to Fix the the contents of this UDFS. You may see objects with
Problem″ column. names beginning with ’QRCL’ or you may see a
directory named ’QReclaim’. If you see the ’QReclaim’
directory, look inside it to find the object names
beginning with ’QRCL’. These objects were previously
lost but have been relocated by the RCLSTG command.
Use the Move (MOV) command to move these objects
back to their original location. The original object names
may be specified in the CPFA0D7 message. If the original
names are not available, use the ″Display Attributes″
option in DSPLNK to view an object’s attributes in order
to attempt to identify it.
Type WRKOBJOWN QDFTOWN to display objects that are Use option 9 (Change owner) from the Work with
owned by the QDFTOWN user profile. Objects by owner display to transfer ownership to the
correct user profile.
Type DSPAUTLOBJ QRCLAUTL to display objects that are If necessary, assign the object to the correct authorization
secured by the QRCLAUTL authorization list. list by using the Edit Object Authority (EDTOBJAUT)
Note: If the reclaim storage procedure did not assign command.
any objects to the QRCLAUTL authorization list, you
may receive a message that the authorization list is not
found. Ignore the message.

Chapter 3. Restore Procedures–General Information 49

 Controlling Restoration of Security-Sensitive Objects
| You can use two different system values to control the restoration of
| security-sensitive objects:
| v Allow object restore operation (QALWOBJRST).
| v Verify object on restore (QVFYOBJRST).
| The QALWOBJRST system value determines whether objects that are
| security-sensitive may be restored to your system.The QVFYOBJRST system value
| is new for V5R1. It allows you to determine how the system restores objects with
| digital signatures.

QALWOBJRST system value

You can use QALWOBJRST to prevent anyone from restoring a system-state object
or an object that adopts authority. The QALWOBJRST system value affects
programs, service programs, modules, and SQL packages.

When your system is shipped, the QALWOBJRST system value is *ALL. This value
is necessary to install your system successfully.

Attention!
It is important to set the QALWOBJRST value to *ALL before performing
some system activities, such as:
v Installing a new release of the OS/400 licensed program.
v Installing new licensed programs.
v Recovering your system.
These activities may fail if the QALWOBJRST value is not *ALL.

If you are applying PTFs, set the QALWOBJRST value to *ALWPTF.

To ensure system security, return the QALWOBJRST value to your normal setting
after completing the system activity. Make sure the entire restore operation has
completed before changing the QALWOBJRST system value, or some objects may
not restore successfully.

You may specify multiple values for the QALWOBJRST system value, unless you
specify *ALL or *NONE.

50 OS/400 Backup and Recovery V5R1

 Possible Values for the QALWOBJRST System Value:

*ALL Any object may be restored to your system by a user with

the proper authority.
*NONE Security-sensitive objects, such as system state programs or
programs that adopt authority, may not be restored to the
system.
*ALWSYSST System state objects may be restored to the system.
*ALWPGMADP Objects that adopt authority may be restored to the
system.
*ALWPTF Security-sensitive objects may be restored only when the
restore is part of a Program Temporary Fix (PTF)
operation.
*ALWSETGID Allows the system to restore files with the S_ISGID
attribute enabled
*ALWSETUID Allows the system to restore files with the S_ISUID
attribute enabled
| *ALWVLDERR Allows the system to restore objects with validation (CRC)
| errors.

How to Set the QALWOBJRST System Value to Allow Complete Recovery

1. Type WRKSYSVAL QALWOBJRST and press the Enter key.
2. You are shown the Work with System Values display. Type 5 (Display) in the
Opt column next to QALWOBJRST and press the Enter key.
3. You are shown the Display System Value display. Write down the current
setting for use after you complete your recovery. If the value is *ALL, you do
not need to change it for your recovery. Skip to step 6.
4. Press F12 to return to the Work with System Values display. Type 2 (Change) in
the Opt column next to QALWOBJRST and press the Enter key.
5. You are shown the Change System Value display. Type *ALL for the value and
press the Enter key.
6. Press F12 to cancel the Work with System Values display.

How to Set the QALWOBJRST System Value to Restrict Restore Operations

1. Type WRKSYSVAL QALWOBJRST and press the Enter key.
2. You are shown the Work with System Values display. Type 2 (Change) in the
Opt column next to QALWOBJRST and press the Enter key.
3. You are shown the Change System Value display. Type the value you wrote
down in step 3 of 51. Press the Enter key.
4. Press F12 to cancel the Work with System Values display.

| QVFYOBJRST system value

| You can add digital signatures to objects so that users can verify the object’s
| integrity and origin. The objects affected by the QVFYOBJRST system value are as
| follows:
| v *PGM
| v *SRVPGM
| v *SQLPKG
| v *MODULE
| v *STMF objects with attached Java programs

Chapter 3. Restore Procedures–General Information 51

| You need to install Digital Certificate Manager (OS/400 option 34) before you can
| use the QVFYOBJRST system value to verify objects not signed by IBM. If Digital
| Certificate Manager is not installed, the system treats user-state objects being
| restored as unsigned objects even if they contain digital signatures. You do not
| need to restart your system for changes to this value to take effect. You can find
| more information about digital signatures in the Information Center at the
| following Web site:
| http://www.ibm.com/eserver/iseries/infocenter

| The QVFYOBJRST system value allows you to control signature verification of

| objects during a restore operation. The QVFYOBJRST system value has five options
| (option 3 is the default):
| 1. Do not verify signatures on restore.
| This is the only option that restores system-state or inherit-state objects without
| valid IBM-generated signatures.
| This option should not be used unless you have a large number of signed
| objects to restore which will fail their signature verification for some acceptable
| reason. Allowing a system-state or inherit-state object without a valid signature
| to restore represents an integrity risk to your system. If you choose to restore
| such an object onto your system by selecting this option, be sure to change it
| back to its previous value after the object has been restored.
| 2. Verify: Restore unsigned objects; Restore signed object, even if signatures are
| not valid.
| Restores unsigned user-state objects. Restores signed user-state objects, even if
| signatures are not valid. Does not restore system-state or inherit-state objects
| without valid IBM-generated signatures. This option should be used only if
| there are specific objects with signatures that are not valid which you want to
| restore. In general, it is dangerous to restore objects with signatures that are not
| valid on your system.
| 3. Verify: Restore unsigned objects; Restore signed objects only if signatures are
| valid.
| Restores unsigned user-state objects. Restores signed user-state objects only if
| signatures are valid. Does not restore system-state or inherit-state objects
| without valid IBM-generated signatures.
| 4. Verify: Do not restore unsigned objects; Restore signed objects, even if
| signatures are not valid.
| Does not restore unsigned user-state objects. Restores signed user-state objects,
| even if signatures are not valid. Does not restore system-state or inherit-state
| objects without valid IBM-generated signatures.
| 5. Verify: Do not restore unsigned objects; Restore signed objects only if
| signatures are valid.
| Does not restore unsigned user-state objects. Restores signed user-state objects
| only if signatures are valid. Does not restore system-state or inherit-state objects
| without valid IBM-generated signatures. This option is the most restrictive
| option and should be used when the only objects you want to allow to be
| restored are those which have been signed by trusted sources.

| How to Set the QVFYOBJRST System Value to Allow Complete Recovery

| 1. Type WRKSYSVAL QVFYOBJRST and press the Enter key.
| 2. You are shown the Work with System Values display. Type 5 (Display) in the
| Opt column next to QVFYOBJRST and press the Enter key.
| 3. You are shown the Display System Value display. Write down the current
| setting for use after you complete your recovery. If the value is 1, you do not
| need to change it for your recovery. Skip to step 6 on page 53.

52 OS/400 Backup and Recovery V5R1

| 4. Press F12 to return to the Work with System Values display. Type 2 (Change) in
| the Opt column next to QVFYOBJRST and press the Enter key.
| 5. You are shown the Change System Value display. Type 1 for the value and
| press the Enter key.
| 6. Press F12 to cancel the Work with System Values display.

How to Set the QMCHPOOL, QBASPOOL, and QPFRADJ System

Values for a Different Processor or Memory
1. Type WRKSYSVAL QMCHPOOL and press the Enter key.
2. You are shown the Work with System Values display. Type 5 (Display) in the
option column next to QMCHPOOL and press the Enter key.
3. You are shown the Display System Value display. As a general rule, if the
main storage size is 64M or higher, the system value should be 15 percent of
the main storage size. If the main storage size is less than 64M, the system
value should be 20 percent of the main storage size. If the value is correct,
you may proceed to step 7; if the value is incorrect, continue with the
following step.
4. Press F12 to return to the Work with System Values display. Type 2 (Change)
in the option column next to QMCHPOOL and press the Enter key.
5. You are shown the Change System Value display. As a general rule, if the
main storage size is 64M or higher, change the value to be 15 percent of the
main storage size. If the main storage size is less than 64M, change the value
to be 20 percent of the main storage size. (For a more precise setting of the
QMCHPOOL system value, refer to the Work Management book.) Press the
Enter key.
6. Press F12 to cancel the Work with System Values display.
7. Type WRKSYSVAL QBASPOOL and press the Enter key.
8. You are shown the Work with System Values display. Type 5 (Display) in the
option column next to QBASPOOL and press the Enter key.
9. You are shown the Display System Value display. The value should be equal
to the size of all the unallocated storage. If the value is correct, you may
proceed to step 12; if the value is incorrect, continue with the following step.
10. Press F12 to return to the Work with System Values display. Type 2 (Change)
in the option column next to QBASPOOL and press the Enter key.
11. You are shown the Change System Value display. Change the value to be the
size of all the unallocated storage and press the Enter key.
12. Press F12 to return to the Work with System Values display. Type 2 (Change)
in the option column next to QPFRADJ and press the Enter key.
13. You are shown the Change System Value display. Change the value to 2
(Adjustment at IPL and automatic adjustment) and press the Enter key.
14. Press F12 to cancel the Work with System Values display.

Locked Objects While Restoring

In general, an object is locked to prevent a read operation while it is being
restored. If the system cannot obtain a lock on an object within the specified time,
that object is not restored and a message is sent to the job log.

Chapter 3. Restore Procedures–General Information 53

 Table 7 shows the type of lock that is required to restore objects:
Table 7. Lock Type Needed for Restore Operation
Object Type Lock Type Needed for Restore Operation

Most object types *EXCL

*M36 object types *SHRRD
Job queue *SHRRD
Output queue *SHRRD
Message queue *EXCLRD
Library, during RSTLIB command *SHRUPD
Library, when object is being restored to it *SHRUPD

If you restore an Original Program Model (OPM) program that is running, the
program may end abnormally.

How to Verify That Objects Are Restored Successfully

You can use the joblog or an output file to determine which objects are restored
successfully.

Note: The system does not restore files to libraries QGPL and QUSRSYS if the file
names begin with QAPZ.No diagnostic message is sent indicating that these
files are not restored.

Using the Job Log: The restore commands send these messages:
CPC3703
Sent for each library that is restored.
CPF3773
Tells the number of objects that are restored and are not restored.
CPF3839
Completion message for RST command from media.
CPF383E
Completion message for RST command from a save file.
CPF9003
Completion message for RSTDLO command from media.
CPF909B
Completion message for RSTDLO command from a save file.

These messages tell the number of objects that are restored and the number of
objects that are not restored. An object is counted only if it fits the selection values
you specified. For example, assume that library LIB1 has 75 objects. The names of
74 of those objects begin with the characters ORD. You specify RSTOBJ OBJ(ORD*)
OBJTYPE(*ALL) SAVLIB(LIB1). If all objects restored successfully, the completion
message would say that 74 objects were restored to library LIB1. You would not be
notified that 1 object was not restored.

A diagnostic message is sent if:

An object could not be restored.
When the system cannot restore an object successfully, it is usually
because:

54 OS/400 Backup and Recovery V5R1

 v The object exists on the system and is being used. Restoring an object
requires an exclusive lock for most object types.
v The object is being saved or restored by another job.
v The object on the media is damaged.
v The user does not have the necessary authority to restore the object.
| v The object does not have a valid signature.
| v The object type is not supported in an independent ASP.
| v The user profile does not exist on the system.
| v The system found a validation error.
Security information was changed.
Under some conditions, the system may:
v Revoke public and private authority
v Change object ownership.
v Change the object’s primary group.
v Not link to the authorization list.
See “Sequence for Restoring Security Information” on page 213 for more
information.
An integrity change occurred.
v Journaling could not be started for an object that was being journaled at
the time of the save operation.
v A logical file is restored over a deleted and re-created physical file.
v The QAUDJRN (audit) journal was created by the system.You receive
message CPF7088. If you restore the operating system and the
QAUDLVL system value is not *NONE, the system creates the
QAUDJRN if it does not exist. This ensures that security auditing is
restarted for your system.

Using An Output File: Most restore commands create output that shows what was
restored. You can direct this output to a printer (OUTPUT(*PRINT)), a database file
(OUTPUT(*OUTFILE)), a stream file, or a user space. The default for restore
commands is not to create output. You must request it each time you run the
restore command. Or you can change the default for the OUTPUT parameter for
restore commands by using the Change Command Default (CHGCMDDFT)
command.

You can print the output and save it. Or you can create a program to analyze and
report on the information in the output file.

You can use the OUTPUT parameter with these commands:

RST RSTDLO RSTOBJ

RSTCFG RSTLIB RSTUSRPRF

The RST command places output in a stream file or a user space, rather than an
output file. “Appendix E. How to Create and Use Output from the Save and
Restore Commands” on page 857 provides the layouts. The RSTLIB, RSTOBJ, and
RST commands have an information type (INFTYPE) parameter to specify how
much detail you want in the output file.

Chapter 3. Restore Procedures–General Information 55

 The online information for the restore commands tells the names of the model
database outfiles they use for output.

Note: The output file you specify is in use throughout the restore operation.
Therefore, the system cannot restore it as part of the operation. Depending
on how you perform your restore operation, you may see a CPF379D
message in the joblog for the output file. If you want to restore the output
file after your restore operation completes, use the RSTOBJ command.

Recovery from an Unsuccessful Restore Operation

A restore operation can be unsuccessful either because an error was encountered
when trying to restore an object or because the operation was interrupted. If the
object existed on the system before the restore operation, it may be damaged by
the unsuccessful restore operation.

An object is not restored when an error is encountered. The error is either

recoverable or not.

Restore Operation Error Is Recoverable: If an object cannot be restored and the

error is recoverable, the following occurs:
v A diagnostic message is sent to the job log for each object that is not restored.
The message ID can vary, depending on why the object was not restored.
v Each object that is associated with the errors is not restored. However, other
objects not associated with the errors but involved in the same restore operation
are restored.
v Only the save and restore status information for the objects that were
successfully restored is updated.
v A count of the number of objects successfully restored and a count of the
number of objects not restored are sent to the user in a diagnostic message.

Restore Operation Error Is Not Recoverable: If the error is not recoverable, the
following occurs:
v Diagnostic messages are sent to the job log for each object.
v The save and restore status information for each object is not updated.
v A diagnostic message that identifies the error condition is sent to the user.
v The restore command ends immediately. No other objects are restored.

Recovering from an Error While Restoring Libraries

Some errors that occur during a restore operation stop the operation. Following are
a few examples of this type of error:
v An unrecoverable media error.
v Exceeding the maximum storage specified in the user profile of the user by
running the restore operation or in a user profile that owns objects that are being
restored.

If an error stops the restore operation, you can correct the error condition and then
start the restore operation where it ended. For example, if the maximum storage is
exceeded, you can increase the MAXSTG parameter in the user profile.

You can use the STRLIB parameter on the RSTLIB command to restart the restore
operation. The STRLIB parameter is valid only when *NONSYS, *ALLUSR, or
*IBM is specified for the restore operation,

56 OS/400 Backup and Recovery V5R1

 The basic recovery steps for a restore operation are:
1. Check the job log to determine the library where the previous RSTLIB
SAVLIB(*NONSYS, *IBM, or *ALLUSR) failed. Find the last library that is
restored which is indicated by a successful restore completion message.
2. Load the media volume of the SAVLIB LIB(*NONSYS, *ALLUSR, or *IBM)
media.
3. Type the following and press the Enter key:
RSTLIB SAVLIB(*NONSYS, *IBM or *ALLUSR)
DEV(media-device-name) ENDOPT(*LEAVE)
STRLIB(library-name)
OMITLIB(library-name)
v If the restore operation stopped because of a media error that you cannot
correct, the library-name for the STRLIB and the OMITLIB parameters should
be the library where the restore operation failed. This causes the operation to
start with the library after that library.
v If the failure was not related to a media error, the library-name for the STRLIB
and the OMITLIB parameters should be the name of the last library that was
successfully restored. This causes the operation to start with the library that
caused the error.
4. You will be asked to load the volume that contains the starting library.
5. After the restore operation is complete, restore the library that failed using the
media from a previous save operation.

Note: Consider eliminating the media volume with the media error from the
next save.

Recovering from an Error While Restoring DLOs

Some errors that occur during a restore operation stop the operation. Following are
a few examples of this type of error:
v An unrecoverable media error.
v Exceeding the maximum storage specified in the user profile of the user by
running the restore operation or in a user profile that owns objects that are being
restored.

If an error occurs that stops the restore operation, you can correct the error
condition and then start the restore operation where it ended. For example, if the
maximum storage is exceeded, you can increase the MAXSTG parameter in the
user profile.

If an unrecoverable error occurs when running the RSTDLO DLO(*ALL)

SAVFLR(*ANY) command, you must determine where the failure occurred and
continue the restore operation step-by-step. Do the following:
1. Check the job log to determine if the failure occurred on a distribution object or
a folder. The job log may identify where the failure occurred.
2. If the failure occurred on a distribution object, the restore operation failed when
the system was restoring mail. Go to “Recovering Mail”.
3. If the failure occurred on a folder, go to “Recovering Documents and Folders”
on page 58.

Recovering Mail
To recover, do one of the following:
v If you have daily save (SAVDLO DLO(*CHG or *MAIL)) media to restore later,
the system restores your mail during the restore process from this save media.

Chapter 3. Restore Procedures–General Information 57

 v Restore the mail from the next most current SAVDLO DLO(*ALL, *CHG, or
*MAIL) FLR(*ANY) media volumes. Type the following to restore the mail:
RSTDLO DLO(*MAIL) DEV(media-device-name)
v If you do not have any media volumes from SAVDLO DLO(*ALL, *CHG, or
*MAIL) FLR(*ANY), run the following program:
CALL PGM(QSYS/QOHFIXIX) PARM(Y)

Run this command so that the mail that was restored is usable. The system may
not have restored some of your mail.

If you need to restore the documents and folders from this set of save media,
continue with “Recovering Documents and Folders”.

Recovering Documents and Folders

If an unrecoverable error occurs during the RSTDLO procedure, you can restart the
procedure by using the SAVFLRparameter on the RSTDLO command.

The basic recovery steps for a restore operation are:

1. Check the job log to determine where the previous RSTDLO DLO(*ALL)
command failed. The job log identifies which folder failed to restore.

Note: If the failure occurred during the restore of mail, you need to restore all
documents and folders.
2. Find the first folder after the folder that failed to restore. Use the list that was
created during the last SAVDLO OUTPUT(*PRINT or *OUTFILE) operation or
use the DSPTAP DATA(*SAVRST) command to determine which first-level
folder is next. To find the first-level folders, find the object type *FLR. Look at
the Document or Folder Information column. The name of a first-level folder does
not contain a forward slash (/).
3. Load the first media volume of the SAVDLO DLO(*ALL) save media.

Note: You must always start with the first volume of the SAVDLO media for
each set of 300 first-level folders. You must load each volume in the set
of SAVDLO save media in sequence.
4. For each first-level folder, type the following and press the Enter key:
RSTDLO DLO(*ALL) SAVFLR(folder-name-list)
DEV(media-device-name)

Where the folder-name-list has the names of the first-level folders identified from
the list described in step 2. You can specify a limit of 300 first-level folders.

Repeat this step for each set of 300 first-level folders.

How to Perform a Normal IPL

You should perform a normal IPL at the end of any recovery before allowing users
to resume normal activity. Do the following:
1. Place the system in Normal mode.
Using Logical Partitions?:
If you are using logical partitions, perform these steps on the console of the
logical partition on which you want to perform an IPL:
a. Type STRSST on the command line and press Enter.

58 OS/400 Backup and Recovery V5R1

 b. On the System Service Tools display, select option 5 to work with system
partitions, and press Enter.
c. On the Work with System Partitions display, select option 2 to work with
system partition status, and press Enter.
d. On the Work with System Partition Status display, select normal mode by
typing a 9 in the Option field. Press Enter.
e. Press F3 until you see the Exit System Service Tools display. On the Exit
System Service Tools display, press Enter.
2. Ensure that no users are signed on and no jobs are active.
3. If you are not using logical partitions, continue with the next step. Otherwise, if
you are performing this operation from the primary partition, be sure to power
down all secondary partitions.
4. Type the following on a command line and press the Enter key:
PWRDWNSYS OPTION(*IMMED) RESTART(*YES)
5. When the IPL is complete, sign on the system.
6. Start any other subsystems that need to be started, such as QTCP or QSNADS.
STRSBS SBSD(subsystem-name)

Parallel Restore Operations

| Beginning with V4R4M0, you can perform restore operations while using more
| than one device simultaneously. The data that you restore in this manner must
| have been saved in parallel format. You can use the Restore Library (RSTLIB) or
| Restore Object (RSTOBJ) commands in conjunction with a media definition to
| perform a parallel restore. You can use a media definition with the RSTLIB
| command to restore *ALLUSR, *IBM, and *NONSYS libraries that were saved with
| a media definition.

It is possible to restore from a parallel save if you are using fewer devices than the
save operation used. However, IBM does not recommend this, due to the amount
of volume switching that you will need to do. IBM also does not recommend this
due to performance reasons. Restore operations that use fewer drives should only
be used occasionally to restore individual objects. Restore operations that use fewer
drives should never be used as part of a system recovery strategy, or to restore
large amounts of data. Whenever possible, the same number of devices that were
used during the save operation should be used during a restore operation.

A Display Tape (DSPTAP) command displays the list of objects that the system
saves across all media files. You only need one media file to display all of the
objects that the system saved during a parallel save operation. This list also
displays the number of media files that you need to restore data. However, you
need all of the media files to restore any of the objects that the system saved. This
may include multiple volumes.

IBM recommends that you use the same media definition object when saving and
restoring the same objects. If you use a different media definition object when
restoring, make sure that the same number of media files are defined within that
media definition object. If the number of media file definitions is different from the
number that exists on the storage media, you will receive an error message.

Chapter 3. Restore Procedures–General Information 59

Recovery Considerations for Cryptographic Access Provider
When you recover a system that includes Cryptographic Access Provider
(5769AC1), (5769AC2), or (5769AC3); the Cryptograhic Access Provider product
fails when you use it.You must reinstall Cryptographic Access Provider after
completing the recovery process in order to use it.

60 OS/400 Backup and Recovery V5R1

Chapter 4. Selecting the Right Recovery Strategy
Use this chapter to determine the correct procedure for recovering your system.
Before beginning your recovery, you must do the following:
v Make sure you understand what has caused the problem. Understanding the
cause helps you choose the correct recovery steps.
v Plan your recovery. Use Table 8 on page 64 to find the appropriate recovery
checklist for your situation.
v Make a copy of the checklist and check off each step as you complete it.
v Keep the checklist for future reference.
v Keep a record of what you have already done and what you do for the rest of
your recovery. This record is important if you need help later.
v If your problem requires hardware or software service, make sure you
understand what the service representative did. Do not be afraid to ask
questions, such as:
– Was a disk unit replaced? If yes, which one?
– Was the Licensed Internal Code restored? If yes, what option from the Install
Licensed Internal Code (LIC) menu was used?
– Did the disk configuration need to be recovered? Was it successful?
– Could the failed disk unit be pumped? How successfully?

Some Common Recovery Terminology

You may need to understand these terms when discussing your situation with
your service representative or software support:

Term Definition

Abnormal end (abend) A system failure or operator action that causes the system to
end without being able to end all jobs and close all files. Your
system may end abnormally because of a power failure or a
problem with certain hardware or software components.
Auxiliary storage pool A group of units defined from all the disk units that make up
auxiliary storage. Auxiliary storage pools (ASPs) allow you to
isolate objects on one or more specific disk units. This may
reduce the loss of data due to a disk media failure. In most
cases, only data stored on disk units in the affected ASP is lost.
Basic ASP A user auxiliary storage pool created by grouping together a
physical set of disk units and assigning them a number
between 2 and 32..
Dedicated service tools A set of tools to work with the system when the operating
(DST) system is not available or not working.
Disk configuration An internal system table that tells how the physical disk units
are arranged on your system. The disk configuration is used to
assign units to an auxiliary storage pool. The disk configuration
is stored on the load source unit.
Disk pump A commonly used term for the procedure used by a service
representative to attempt to copy data from a disk unit that has
failed.

© Copyright IBM Corp. 1997, 2000, 2001 61

 Term Definition

| Independent ASP A user auxiliary storage pool that can be made available
| (varied on) and made unavailable (varied off) without
| restarting the system. An independent ASP can be either
| switchable among multiple systems in a clustering environment
| or privately connected to a single system.
| Library user ASP A user ASP that contains libraries, directories, and folders and
| all the objects associated with them.
Licensed Internal Code The layer of iSeries architecture just above the hardware. You
must have the Licensed Internal Code on your machine before
you can restore the operating system.
Load source unit The first unit (unit 1) in the system ASP. It contains the
Licensed Internal Code and the disk configuration for your
system.
Nonlibrary user ASP A user ASP that may contain journals, journal receivers, and
save files. The libraries associated with these objects are in the
system ASP. A nonlibrary user ASP is sometimes called an
old-style ASP, because it was the only type of user ASP
available before Version 1 Release 3 of the OS/400 licensed
program.
System ASP An auxiliary storage pool that is created by the system and
always configured. The system ASP (ASP 1) contains the
Licensed Internal Code, licensed programs, and system
libraries. The system ASP may also contain user libraries,
folders, and directories. The system ASP contains all configured
disk units that are not assigned to a user ASP.
System service tools (SST) A subset of the DST tools. The tools available through SST,
such as displaying the disk configuration, can be used while
the operating system is running and other users are on the
system.
| User ASP A basic or independent auxiliary storage pool created by
| grouping together a physical set of disk units. You can assign a
| basic ASP a number between 2 and 32. The system assigns an
| independent ASP a number between 33 and 99. ASP 1 is
| always reserved as the system ASP.

Recovery Procedure for a Power Failure

If your system stops because power is lost, you need to follow special procedures
when you start the system again. Chapter 7. Starting the System After It Ends
Abnormally describes this procedure.

If you experience frequent power outages, consider using an uninterruptible power

supply for your system.

If power loss to workstations causes your system to perform constant error

recovery, you should modify your applications to handle losing communications to
workstations. “Chapter 31. Techniques and Programming Examples for Backup and
Recovery” on page 759 describes how to do this.

62 OS/400 Backup and Recovery V5R1

Recovery Procedure for a System Failure
A system failure is a problem either with hardware (other than DASD) or with
operating system software that causes your system to end abnormally. After your
service representative has corrected the problem, follow the procedure to start your
system after an abnormal end. Chapter 7. Starting the System After It Ends
Abnormally describes the procedure.

If the service representative replaced a disk unit, use the information in “Choosing
the Recovery Procedure for a Disk Failure or Disk Errors” to determine the correct
recovery procedure.

Recovery Procedure for a Program Failure or Human Error

You may need to recover objects because a program updated them incorrectly or
because a user deleted them. Review the information in “Chapter 10. How to
Restore Specific Types of Information” on page 213 for the kind of objects you are
restoring. Some objects have special considerations or need to be restored in a
particular sequence.

If you are restoring an object that does not exist on the system, private authorities
for the object are not restored with it. You can do one of the following:
v Reconstruct the private authorities manually, using the Edit Object Authority
(EDTOBJAUT) display.
v Restore private authorities by using this procedure:
1. Restore all user profiles from your most recent SAVSYS or SAVSECDTA tape.
Type: RSTUSRPRF. Restoring user profiles requires a restricted state.
2. Restore the objects you need to recover.
3. Restore authorities. Type: RSTAUT. Only one RSTAUT command can run on
the system at any given time.

Choosing the Recovery Procedure for a Disk Failure or Disk Errors

Attention
When a system reference code (SRC) appears on your system unit, use the
iSeries Licensed Internal Code Diagnostic Aids - Volume 1 to determine what the
code means. If you receive an SRC code that indicates a DASD problem, do
not perform an IPL before your service representative arrives. If you perform
an IPL, your service representative may not be able to recover data from the
damaged disk unit.

This topic describes the actions that you take if you are recovering because a disk
unit failed or was damaged. The steps you follow to recover from a disk failure
depend on:
v What unit failed.
v Whether disk protection, such as device parity protection or mirrored protection
is active.
v Whether you have user ASPs configured.

Chapter 4. Selecting the Right Recovery Strategy 63

 v Whether some or all of the sectors on the disk are damaged. If a disk unit must
be replaced, a service representative normally tries to copy the information from
the disk unit when it is replaced. This procedure is sometimes referred to as a
disk pump.

Use Table 8 to determine what recovery procedure you should follow, based on the
failure that has occurred on your system. To find your situation on the chart, ask
your service representative whether data was copied successfully (the results of the
disk pump):

Service Representative Terminology Terminology in Recovery Charts

Full pump None of the data is lost

Partial pump Some of the data is lost
Could not pump All of the data is lost

Recovery for Disk Errors That Do Not Require Disk Replacement: Some types of
disk units automatically recover from errors without needing to be replaced. In
some cases, however, sectors are damaged before the disk unit reassigns them and
some object damage occurs. If you receive a message indicating that object damage
has occurred and disk sectors have been reassigned, consider this to be the value
Some for the column Data Loss on the Failed Unit in Table 8.

If you are recovering from disk errors but you did not need a service
representative to replace the disk unit, you may need to perform tasks that are
normally performed by a service representative. Make a copy of the appropriate
checklist and mark it as follows:
1. Begin at the task immediately following “Attach the new disk unit”.
2. If the checklist contains a task called, “Restore the disk unit data”, skip that
task.
Table 8. Choosing the Correct Recovery Procedure for Disk Media Failure
Availability
Type of Unit That Data Loss on Protection on Failed User ASPs
Failed Failed Unit Unit Configured? Procedure to Follow
1
Load source unit None None N/A Checklist 1 on page 65
2 1
Load source unit Some None N/A Checklist 2 on page 66
Load source unit All None No Checklist 3 on page 67
Load source unit. No All None Yes Checklist 4 on page 68
user ASPs in
overflowed status3
Load source unit. One All None Yes Checklist 5 on page 72
or more user ASPs in
overflowed status3.
Non-load source unit None None N/A 1 Checklist 6 on page 76
in system ASP4
Non-load source unit Some2 None N/A 1 Checklist 7 on page 77
in system ASP4
Non-load source unit All None No Checklist 8 on page 78
in system ASP4

64 OS/400 Backup and Recovery V5R1

 Table 8. Choosing the Correct Recovery Procedure for Disk Media Failure (continued)
Availability
Type of Unit That Data Loss on Protection on Failed User ASPs
Failed Failed Unit Unit Configured? Procedure to Follow
Non-load source unit All None Yes Checklist 9 on page 79
in system ASP4. No
user ASPs in
overflowed status3.
Non-load source unit All None Yes Checklist 10 on page 82
in system ASP4. One
or more user ASPs in
overflowed status3.
Disk unit in user ASP None None Yes Checklist 6 on page 76
2
Disk unit in user ASP Some None Yes Checklist 11 on page 86
Disk unit in user ASP. All None Yes Checklist 12 on page 86
Failed unit not in
overflowed status3.
Disk unit in user ASP. All None Yes Checklist 13 on page 88
Failed unit in
overflowed status3.
| Disk unit in None None Yes Checklist 17 on page 93
| independent ASP
| Disk unit in Some2 None Yes Checklist 18 on page 93
| independent ASP
| Disk unit in All None Yes Checklist 19 on page 94
| independent ASP
Any N/A Mirrored protection N/A 1 Checklist 14 on page 90
1
Any N/A Device parity N/A Checklist 15 on page 91
protection

1
The recovery procedure is the same whether or not user ASPs are configured.
2
If the service representative was partially successful in saving data from a failed disk unit, you should
consider treating the situation as a complete data loss on the failed unit.
3
If a unit in your system ASP fails and a replacement is not immediately available, you can use the procedure
in Checklist 16 on page 92. This procedure allows you to return your system to operation. You will have less
disk storage and you will need to recover all the data in the system ASP.
4
Step 4 on page 192 describes how to determine whether a user ASP is in overflowed status.

Actions for load source disk unit failure–Checklist 1

This checklist should be used for the following problem situation:
Failed Unit:
Load source unit
Data Loss:
None
User ASP Configured:
N/A

Before you begin your recovery, make a copy of this checklist. Fill in the
appropriate areas as you and the service representative perform the recovery steps.

Chapter 4. Selecting the Right Recovery Strategy 65

 This checklist provides an important record of your recovery actions. It may help
you to diagnose any problems that occur after the recovery. It may also be useful
in evaluating your backup strategy.

Most steps in the checklist include references to other topics in this book. Refer to
these topics if you need more information about how to perform a particular step.
You may not need to perform some steps, such as restoring changed objects, if they
do not apply in your situation.
Table 9. Recovery Checklist for Disk Failure–Checklist 1
Task What To Do Where To Read More About It
Actions to Be Performed by the Service Representative
___ Task 1 Save the disk unit data.
___ Task 2 Attach the new disk unit.
___ Task 3 Install the Licensed Internal Code using “How to Prepare for Loading the
option 4 (Install Licensed Internal Code and Licensed Internal Code” on page 122
Restore Disk Unit Data)1. and “How to Load the Licensed
Internal Code” on page 130.
___ Task 4 Restore the disk unit data.
Actions to Be Performed by the User
___ Task 5 You must perform an IPL at this time. Chapter 7. Starting the System After It
Follow the procedure for starting the system Ends Abnormally, task 1 through task
after it ends abnormally. 4.

1
If you have a 2440 Tape Unit with the high-speed feature enabled, you must disable it before restoring the
Licensed Internal Code. See “Disabling and Enabling the High-Speed Feature on the 2440 Tape Unit” on
page 145 for instructions. You may enable the feature again after you complete all the steps in Task 3.

Actions for load source disk unit failure–Checklist 2

This checklist should be used for the following problem situation:
Failed Unit:
Load source unit
Data Loss:
Some
User ASP Configured:
N/A

Most steps in the checklist include references to other topics in this book. Refer to
these topics if you need more information about how to perform a particular step.
You may not need to perform some steps, such as restoring changed objects, if they
do not apply in your situation.
Table 10. Recovery Checklist for Disk Failure–Checklist 2
Task What To Do Where To Read More About It
Actions to Be Performed by the Service Representative
Task 1 Save the disk unit data.
Task 2 Attach the new disk unit.

66 OS/400 Backup and Recovery V5R1

Table 10. Recovery Checklist for Disk Failure–Checklist 2 (continued)
Task What To Do Where To Read More About It
Task 3 Install the Licensed Internal Code using “How to Prepare for Loading the
option 4 (Install Licensed Internal Code and Licensed Internal Code” on page 122
Restore Disk Unit Data)1. and “How to Load the Licensed
Internal Code” on page 130
Task 4 Restore the disk unit data.
Actions to Be Performed by the User
Task 5 You must perform an IPL at this time. Chapter 7. Starting the System After It
Follow the procedure for starting the system Ends Abnormally, task 1 through task
after it ends abnormally. 4.
Task 6 Restore the operating system. You are Chapter 6. Restoring the Operating
performing an abbreviated install operation. System, task 1 through task 6.
Note: You may have some objects that are
damaged. You may be required to perform a
complete restore of the operating system.
Task 7 If you restored the operating system using “Recovering System Information” on
distribution tapes, some system information, page 213.
such as access path recovery times and the
system reply list, may have been reset to
default values. Verify these values and correct
them if necessary.
Task 8 Reclaim storage. “Reclaiming Storage” on page 46.
Task 9 Evaluate the extent of the damage. Determine “Task 4–Recovering from Damaged
whether you will attempt to recover damaged Objects and Unreadable Sectors” on
objects or restore the entire system. Do not page 174.
skip this step.
Task 10 If you have decided to do a complete restore
operation, use Table 31 on page 107 to
determine the correct procedure for restoring
user information.
Task 11 If you have decided to attempt to recover
damaged objects, perform the tasks described
in “Task 4–Recovering from Damaged Objects
and Unreadable Sectors” on page 174.

1
If you have a 2440 Tape Unit with the high-speed feature enabled, you must disable it before restoring the
Licensed Internal Code. See “Disabling and Enabling the High-Speed Feature on the 2440 Tape Unit” on
page 145 for instructions. You may enable the feature again after you complete all the steps in Task 3.

Actions for load source disk unit failure–Checklist 3

This checklist should be used for the following problem situation:
Failed Unit:
Load source unit
Data Loss:
All
User ASP Configured:
No

Before you begin your recovery, make a copy of this checklist. Fill in the
appropriate areas as you and the service representative perform the recovery steps.

Chapter 4. Selecting the Right Recovery Strategy 67

 This checklist provides an important record of your recovery actions. It may help
you to diagnose any problems that occur after the recovery. It may also be useful
in evaluating your backup strategy.

Most steps in the checklist include references to other topics in this book. Refer to
these topics if you need more information about how to perform a particular step.
You may not need to perform some steps, such as restoring changed objects, if they
do not apply in your situation.

Task What to Do Where To Read More About

It
Actions to Be Performed by the Service Representative
___ Task 1 Attach the new disk unit.
___ Task 2 Prepare to load the Licensed “How to Prepare for Loading
Internal Code. the Licensed Internal Code” on
page 122
___ Task 3 Install the Licensed Internal Code “How to Load the Licensed
using option 3 (Install Licensed Internal Code” on page 130
Internal Code and Recover
Configuration)1.
___ Task 4 Recover the disk configuration “How to Recover Your Disk
(assignment of disks to ASPs and Configuration” on page 142
protection).
Actions to Be Performed by the User
___ Task 5 Restore the operating system, Chapter 16, ″Restoring the
beginning with “Task 1–Starting to Operating System″, task 1
Restore the Operating System” on through task 6.
page 150. You are performing a
complete restore operation.
___ Task 6 If you restored the operating “Recovering System
system using distribution media, Information” on page 213
some system information, such as
access path recovery times and the
system reply list, was returned to
default values. Set these values
correctly.
___ Task 7 Use Table 31 on page 107 to
determine the correct procedure
for recovering user information.
You will need to recover all user
data.

1
If you have a 2440 Tape Unit with the high-speed feature enabled, you must
disable it before restoring the Licensed Internal Code. See “Disabling and
Enabling the High-Speed Feature on the 2440 Tape Unit” on page 145 for
instructions. You may enable the feature again after you complete all the steps in
Task 3.

Actions for load source disk unit failure–Checklist 4

This checklist should be used for the following problem situation:
Failed Unit:
Load source unit

68 OS/400 Backup and Recovery V5R1

 Data Loss:
All
User ASP Configured:
Yes
User ASP Overflowed:
No

Attention!
When you replace a disk unit in your system ASP, the system loses
addressability to the objects in your user ASPs. Recovering object ownership
for objects other than DLOs will require manually assigning ownership for
every object in every user ASP. You may want to treat this situation as a total
recovery and restore all your information from your save media if the
following conditions are true:
1. You have a large number of objects in your user ASPs
2. You have thoroughly backed up your system
If you choose to do this, perform the steps that are described in “Recovering
your entire system after a complete system loss–Checklist 20” on page 96 to
recover your system.

Before you begin your recovery, make a copy of this checklist. Fill in the
appropriate areas as you and the service representative perform the recovery steps.
This checklist provides an important record of your recovery actions. It may help
you to diagnose any problems that occur after the recovery. It may also be useful
in evaluating your backup strategy.

Most steps in the checklist include references to other topics in this book. Refer to
these topics if you need more information about how to perform a particular step.
You may not need to perform some steps, such as restoring changed objects, if they
do not apply in your situation.
Table 11. Recovery Checklist for Disk Failure–Checklist 4
Task What To Do Where To Read More About It
Actions to Be Performed by the Service Representative
Task 1 Attach the new disk unit.
Task 2 Prepare to load the Licensed Internal Code. “How to Prepare for Loading the
Licensed Internal Code” on page 122.
Task 3 Install the Licensed Internal Code using “How to Load the Licensed Internal
option 3 (Install Licensed Internal Code and Code” on page 130.
Recover Configuration)1.
Task 4 Recover the disk configuration (assignment “How to Recover Your Disk
of disks to ASPs and protection). Configuration” on page 142.
Actions to Be Performed by the User
Task 5 Restore the operating system, beginning with Chapter 6. Restoring the Operating
“Task 1–Starting to Restore the Operating System, task 1 through task 6.
System” on page 150. You are performing a
complete restore operation.

Chapter 4. Selecting the Right Recovery Strategy 69

Table 11. Recovery Checklist for Disk Failure–Checklist 4 (continued)
Task What To Do Where To Read More About It
Task 6 If you restored the operating system using “Recovering System Information” on
distribution media, some system information, page 213.
such as access path recovery times and the
system reply list, was returned to default
values. Set these values correctly.
Task 7 If necessary, change the QALWOBJRST “Controlling Restoration of
system value by using the WRKSYSVAL Security-Sensitive Objects” on page 50.
command. Write the old value here:
______________
Task 8 If necessary, change the QVFYOBJRST “Controlling Restoration of
system value by using the WRKSYSVAL Security-Sensitive Objects” on page 50.
command. Write the old value here:
______________
Task 9 If necessary, change the system value that The System Values subtopic of the
controls whether the job log wraps when it is System Management topic in the
full. Use the Work with System Values iSeries Information Center.
command: WRKSYSVAL QJOBMSGQFL.
Write down the current value here:
______________. Then change the value to
*PRTWRAP.
Task 10 After changing the system values, sign off by “Describing the Contents of Your User
using the command SIGNOFF *LIST. Then, Auxiliary Storage Pools” on page 181.
using a newly created password, sign back
on as QSECOFR for the new values to take
effect.
Task 11 Describe or diagram, as much as possible, “Describing the Contents of Your User
the contents of your user ASPs before the Auxiliary Storage Pools” on page 181.
failure.
Task 12 Recover user profiles, configuration, libraries “How to Recover a User ASP After
in the system ASP, and the contents of your Recovering the System ASP” on
user ASPs. If you choose not to restore all of page 182, task 1 through task 11.
your libraries at this time, ensure that you
restore the QGPL and QUSRSYS libraries
along with the libraries that you are
restoring.
Task 13 Restore document library objects. “Restoring Documents and Folders” on
page 255.
Task 14 Restore your last complete save of “Restoring Objects in Directories” on
directories.2 page 261.
Task 15 If you have User-Defined File Systems in “Task 7–Restoring User-Defined File
User ASPs that do not restore correctly, you Systems to the User Auxiliary Storage
may need to perform additional recovery Pool” on page 187.
steps.
Task 16 Restore changed objects and apply journaled Chapter 11. How to Restore Changed
changes. Objects and Apply Journaled Changes,
task 1 through task 7.
Task 17 If you use Cryptographic Access Provider, “Recovery Considerations for
install the Cryptographic Access Provider Cryptographic Access Provider” on
licensed program (5769AC1), (5769AC2), or page 60
(5769AC3) using option 11 on the Work with
Licensed Programs menu.

70 OS/400 Backup and Recovery V5R1

 Table 11. Recovery Checklist for Disk Failure–Checklist 4 (continued)
Task What To Do Where To Read More About It
Task 18 Restore authority. Type: RSTAUT “Restoring Object Authorities” on
page 219.
Task 19 Reapply any PTFs that were applied since “How to Restore Program Temporary
your last SAVSYS operation. Fixes” on page 278.
Task 20 If you use Windows server on iSeries and “Completing Recovery for the iSeries
saved with the Integrated xSeries Server Integration for Windows Server
(NWSD) varied off, complete recovery for Product” on page 263.
the Windows server on iSeries.
| Task 21 If you run Linux and saved Linux by “Recovering Linux in a Partition” on
| varying off the network server description page 264
| (NWSD) for Linux, complete the recovery for
| Linux.
Task 22 If necessary, change the QALWOBJRST “Controlling Restoration of
system value back to its original value by Security-Sensitive Objects” on page 50.
using the WRKSYSVAL command.
Task 23 If necessary, change the QVFYOBJRST “Controlling Restoration of
system value back to its original value by Security-Sensitive Objects” on page 50.
using the WRKSYSVAL command.
Task 24 If necessary, change the QJOBMSGQFL The System Values subtopic of the
system value back to its original value by System Management topic in the
using the WRKSYSVAL command. iSeries Information Center.
Task 25 Perform either a:
SIGNOFF *LIST

or a
DSPJOBLOG * *PRINT

Check the job log to ensure all objects were

restored. The job log contains information
about the restore operation. To verify that all
objects were restored, spool the job log for
printing along with the job’s remaining
spooled output, if any.

Message CPC3703 is sent to the job log for

each library that was successfully restored.
Message CPF3773 is sent to tell you how
many objects were restored. Objects are not
restored for various reasons. Check for any
error messages, correct the errors, and then
restore those objects from the media.
Task 26 You must perform an IPL at this time. “How to Perform a Normal IPL” on
page 58.
Task 27 If you use Windows server on iSeries and “Completing Recovery for the iSeries
saved with the Integrated xSeries Server Integration for Windows Server
(NWSD) varied on, complete recovery for the Product” on page 263.
Windows server on iSeries product.

Chapter 4. Selecting the Right Recovery Strategy 71

Table 11. Recovery Checklist for Disk Failure–Checklist 4 (continued)
Task What To Do Where To Read More About It
Task 28 If IBM Content Manager OnDemand for
iSeries is installed, complete journaling for
IBM Content Manager OnDemand for iSeries
by typing the following commands:
CALL QRDARS/QRLCSTRJ PARM('RLC')
CALL QRDARS/QRLCSTRJ PARM('RLR')
CALL QRDARS/QRLCSTRJ PARM('RLO')
Task 29 Review job logs or output from your restore “How to Verify That Objects Are
operations to ensure that all objects were Restored Successfully” on page 54.
restored successfully.

1
If you have a 2440 Tape Unit with the high-speed feature enabled, you must disable it before restoring the
Licensed Internal Code. See “Disabling and Enabling the High-Speed Feature on the 2440 Tape Unit” on
page 145 for instructions. You may enable the feature again after you complete all the steps in 69.
2
You may receive one of the following messages:
CPD377A: Object not restored, /QNTC.
CPD377A: Object not restored, /QNetWare.

These objects cannot be restored until their file systems have been mounted during the IPL. You may
ignore these messages. The additional recovery tasks will guide you through the steps to restore these
objects.
Note: Since the OS/400 Enhanced Integration for Novell NetWare software resides on a remote server, you
do not have to restore your Netware data when you restore your server. Previously, the OS/400 Integration
for Novell NetWare product ran on a Integrated xSeries Server and you had to restore the Novell product
if you completely restored your server.

Actions for load source disk unit failure–Checklist 5

This checklist should be used for the following problem situation:
Failed Unit:
Load source unit
Data Loss:
All
User ASP Configured:
Yes
User ASP Overflowed:
Yes

72 OS/400 Backup and Recovery V5R1

 Attention!
When you replace a disk unit in your system ASP, the system loses
addressability to the objects in your user ASPs. Recovering object ownership
for objects other than DLOs will require manually assigning ownership for
every object in every user ASP. You may want to treat this situation as a total
recovery and restore all your information from your save media if the
following conditions are true:
1. You have a large number of objects in your user ASPs
2. You have thoroughly backed up your system
If you choose to do this, perform the steps that are described in “Recovering
your entire system after a complete system loss–Checklist 20” on page 96 to
recover your system.

Table 12. Recovery Checklist for Disk Failure–Checklist 5

Task What To Do Where To Read More About It
Actions to Be Performed by the Service Representative
Task 1 Attach the new disk unit.
Task 2 Prepare to load the Licensed Internal Code. “How to Prepare for Loading the
Licensed Internal Code” on page 122
Task 3 Install the Licensed Internal Code using “How to Load the Licensed Internal
option 3 (Install Licensed Internal Code and Code” on page 130
Recover Configuration)1.
Task 4 Recover the disk configuration (assignment “How to Recover Your Disk
of disk to ASPs and protection). Configuration” on page 142.
Actions to Be Performed by the User
Task 5 Restore the operating system, beginning with Chapter 6. Restoring the Operating
“Task 1–Starting to Restore the Operating System, task 1 through task 6.
System” on page 150. You are performing a
complete restore operation.
Task 6 If you restored the operating system using “Recovering System Information” on
distribution media, some system information, page 213.
such as access path recovery times and the
system reply list, may have been reset to
default values. Verify these values and
correct them if necessary.
Task 7 If necessary, change the QALWOBJRST “Controlling Restoration of
system value. Write the old value here: Security-Sensitive Objects” on page 50.
______________
Task 8 If necessary, change the QVFYOBJRST “Controlling Restoration of
system value. Write the old value here: Security-Sensitive Objects” on page 50.
______________
Task 9 If necessary, change the system value that The System Values subtopic of the
controls whether the job log wraps when it is System Management topic in the
full. Use the Work with System Values iSeries Information Center.
command: WRKSYSVAL QJOBMSGQFL.
Write down the current value here:
______________. Then change the value to
*PRTWRAP.

Chapter 4. Selecting the Right Recovery Strategy 73

Table 12. Recovery Checklist for Disk Failure–Checklist 5 (continued)
Task What To Do Where To Read More About It
Task 10 After changing the system values, sign off by “Describing the Contents of Your User
using the command SIGNOFF *LIST. Then, Auxiliary Storage Pools” on page 181.
using a newly created password, sign back
on as QSECOFR for the new values to take
effect.
Task 11 Describe or diagram, as much as possible, “Describing the Contents of Your User
the contents of your user ASPs before the Auxiliary Storage Pools” on page 181.
failure.
Task 12 Recover user profiles, configuration, libraries “How to Recover a User ASP After
in the system ASP, and the contents of the Recovering the System ASP” on
user ASPs that were not in overflowed page 182, task 1 through task 11.
status. If you choose not to restore all of
your libraries at this time, ensure that you
restore the QGPL and QUSRSYS libraries
along with the libraries that you are
restoring.
Task 13 Recover the objects in the user ASPs that “How to Recover a Damaged User
were overflowed. Auxiliary Storage Pool” on page 196,
task 1 through task 9.
Task 14 Restore document library objects to the “Restoring Documents and Folders” on
system ASP and to any overflowed user page 255.
ASPs that had DLOs.
Task 15 Restore your last complete save of “Restoring Objects in Directories” on
directories.2 page 261.
Task 16 If you have User-Defined File Systems in “Task 7–Restoring User-Defined File
User ASPs that do not restore correctly, you Systems to the User Auxiliary Storage
may need to perform additional recovery Pool” on page 187.
steps.
Task 17 Restore changed objects and apply journaled “Chapter 11. How to Restore Changed
changes. Objects and Apply Journaled Changes”
on page 279.
Task 18 If you use Cryptographic Access Provider, “Recovery Considerations for
install the Cryptographic Access Provider Cryptographic Access Provider” on
licensed program (5769AC1), (5769AC2), or page 60
(5769AC3) using option 11 on the Work with
Licensed Programs menu.
Task 19 Restore authority. Type: RSTAUT “Restoring Object Authorities” on
page 219.
Task 20 Reapply any PTFs that were applied since “How to Restore Program Temporary
your last SAVSYS operation. Fixes” on page 278.
Task 21 If you use Windows server on iSeries and “Completing Recovery for the iSeries
saved with the Integrated xSeries Server Integration for Windows Server
(NWSD) varied off, complete recovery for Product” on page 263.
the Windows server on iSeries.
Task 22 If you run Linux and saved Linux by “Recovering Linux in a Partition” on
varying off the network server description page 264
(NWSD) for Linux, complete the recovery for
Linux.
Task 23 If necessary, use the WRKSYSVAL command “Controlling Restoration of
to change the QALWOBJRST system value Security-Sensitive Objects” on page 50.
back to its original value.

74 OS/400 Backup and Recovery V5R1

Table 12. Recovery Checklist for Disk Failure–Checklist 5 (continued)
Task What To Do Where To Read More About It
Task 24 If necessary, use the WRKSYSVAL command “Controlling Restoration of
to change the QVFYOBJRST system value Security-Sensitive Objects” on page 50.
back to its original value.
Task 25 If necessary, change the QJOBMSGQFL The System Values subtopic of the
system value back to its original value by System Management topic in the
using the WRKSYSVAL command. iSeries Information Center.
Task 26 Perform either a:
SIGNOFF *LIST

or a
DSPJOBLOG * *PRINT

Check the job log to ensure all objects were

restored. The job log contains information
about the restore operation. To verify that all
objects were restored, spool the job log for
printing along with the job’s remaining
spooled output, if any.

Message CPC3703 is sent to the job log for

each library that was successfully restored.
Message CPF3773 is sent to tell you how
many objects were restored. Objects are not
restored for various reasons. Check for any
error messages, correct the errors, and then
restore those objects from the media.
Task 27 You must perform a normal IPL at this “How to Perform a Normal IPL” on
time. page 58.
Task 28 If you use Windows server on iSeries and “Completing Recovery for the iSeries
saved with the Integrated xSeries Server Integration for Windows Server
(NWSD) varied on, complete recovery for the Product” on page 263.
Windows server on iSeries product.
Task 29 If IBM Content Manager OnDemand for
iSeries is installed, complete journaling for
IBM Content Manager OnDemand for iSeries
by typing the following commands:
CALL QRDARS/QRLCSTRJ PARM('RLC')
CALL QRDARS/QRLCSTRJ PARM('RLR')
CALL QRDARS/QRLCSTRJ PARM('RLO')
Task 30 Review job logs or output from your restore “How to Verify That Objects Are
operations to ensure that all objects were Restored Successfully” on page 54.
restored successfully.

Chapter 4. Selecting the Right Recovery Strategy 75

Table 12. Recovery Checklist for Disk Failure–Checklist 5 (continued)
Task What To Do Where To Read More About It

1
If you have a 2440 Tape Unit with the high-speed feature enabled, you must disable it before restoring the
Licensed Internal Code. See “Disabling and Enabling the High-Speed Feature on the 2440 Tape Unit” on
page 145 for instructions. You may enable the feature again after you complete all the steps in Task 3.
2
You may receive one of the following messages:
CPD377A: Object not restored, /QNTC.
CPD377A: Object not restored, /QNetWare.

These objects cannot be restored until their file systems have been mounted during the IPL. The additional
recovery tasks will guide you through the steps to restore these objects.
Note: Since the OS/400 Enhanced Integration for Novell NetWare software resides on a remote server, you
do not have to restore your Netware data when you restore your. Previously, the OS/400 Integration for
Novell NetWare product ran on a Integrated xSeries Server and you had to restore the Novell product if
you completely restored your.

Actions for non-load source disk unit failure or disk units in

user ASP disk failure–Checklist 6
This checklist should be used for the following problem situation:
Failed Unit:
Non-load source unit in system ASP or Disk unit in user ASP
Data Loss:
None
User ASP Configured:
N/A

Before you begin your recovery, make a copy of this checklist. Fill in the
appropriate areas as you and the service representative perform the recovery steps.
This checklist provides an important record of your recovery actions. It may help
you to diagnose any problems that occur after the recovery. It may also be useful
in evaluating your backup strategy.

Most steps in the checklist include references to other topics in this book. Refer to
these topics if you need more information about how to perform a particular step.
You may not need to perform some steps, such as restoring changed objects, if they
do not apply in your situation.
Table 13. Recovery Checklist for Disk Failure–Checklist 6
Task What To Do Where To Read More About It
Actions to Be Performed by the Service Representative
Task 1 Save the disk unit data.
Task 2 Attach a new disk unit.
Task 3 Restore data to the new disk unit.
Actions to Be Performed by the User
Task 4 Perform an IPL. Follow the procedure for Chapter 7. Starting the System After It
starting the system after it ends abnormally. Ends Abnormally, task 1 through task
4.

76 OS/400 Backup and Recovery V5R1

 Actions for non-load source disk unit failure–Checklist 7
This checklist should be used for the following problem situation:
Failed Unit:
Non-load source unit in system ASP
Data Loss:
Some
User ASP Configured:
N/A

Before you begin your recovery, make a copy of this checklist. Fill in the
appropriate areas as you and the service representative perform the recovery steps.
This checklist provides an important record of your recovery actions. It may help
you to diagnose any problems that occur after the recovery. It may also be useful
in evaluating your backup strategy.

Most steps in the checklist include references to other topics in this book. Refer to
these topics if you need more information about how to perform a particular step.
You may not need to perform some steps, such as restoring changed objects, if they
do not apply in your situation.
Table 14. Recovery Checklist for Disk Failure–Checklist 7
Task What To Do Where To Read More About It
Actions to Be Performed by the Service Representative
Task 1 Save the disk unit data.
Task 2 Attach the new disk unit.
Task 3 Restore the disk unit data.
Actions to Be Performed by the User
Task 4 Restore the operating system, beginning with Chapter 6. Restoring the Operating
“Task 1–Starting to Restore the Operating System, task 1 through task 6.
System” on page 150. You are performing a
complete restore operation.
Task 5 If you restored the operating system using “Recovering System Information” on
distribution media, some system information, page 213.
such as access path recovery times and the
system reply list, may have been reset to
default values. Verify these values and
correct them if necessary.
Task 6 Reclaim storage. “Reclaiming Storage” on page 46.
Task 7 Evaluate the extent of the damage. “Task 4–Recovering from Damaged
Determine whether you will attempt to Objects and Unreadable Sectors” on
recover damaged objects or restore the entire page 174.
system. Do not skip this step.
Task 8 If you have decided to do a complete restore
operation, use Table 31 on page 107 to
determine the correct procedure for
recovering user information.
Task 9 If you have decided to attempt to recover
damaged objects, perform the tasks in “Task
4–Recovering from Damaged Objects and
Unreadable Sectors” on page 174.

Chapter 4. Selecting the Right Recovery Strategy 77

Table 14. Recovery Checklist for Disk Failure–Checklist 7 (continued)
Task What To Do Where To Read More About It

1
If you have a 2440 Tape Unit with the high-speed feature enabled, you must disable it before restoring the
Licensed Internal Code. See “Disabling and Enabling the High-Speed Feature on the 2440 Tape Unit” on
page 145 for instructions.

Actions for non-load source disk unit failure–Checklist 8

This checklist should be used for the following problem situation:
Failed Unit:
Non-load source unit in system ASP
Data Loss:
All
User ASP Configured:
No

Most steps in the checklist include references to other topics in this book. Refer to
these topics if you need more information about how to perform a particular step.
You may not need to perform some steps, such as restoring changed objects, if they
do not apply in your situation.
Table 15. Recovery Checklist for Disk Failure–Checklist 8
Task What To Do Where To Read More About It
Actions to Be Performed by the Service Representative
Task 1 Attach the new disk unit.
Task 2 Delete the ASP data.
Task 3 Restore the Licensed Internal Code using “How to Prepare for Loading the
option 1 (Restore Licensed Internal Code)1. If Licensed Internal Code” on page 122
user ASPs are configured, they remain intact. and “How to Load the Licensed
Internal Code” on page 130.
Actions to Be Performed by the User
Task 4 Restore the operating system, beginning with Chapter 6. Restoring the Operating
“Task 1–Starting to Restore the Operating System, task 1 through task 6.
System” on page 150. You are performing a
complete restore operation.
Task 5 If you restored the operating system using “Recovering System Information” on
distribution media, some system information, page 213.
such as access path recovery times and the
system reply list, may have been reset to
default values. Verify these values and
correct them if necessary.
Task 6 Reclaim storage. “Reclaiming Storage” on page 46.
Task 7 Use Table 31 on page 107 to determine the
correct procedure for recovering user
information.

1
If you have a 2440 Tape Unit with the high-speed feature enabled, you must disable it before restoring the
Licensed Internal Code. See “Disabling and Enabling the High-Speed Feature on the 2440 Tape Unit” on
page 145 for instructions. You may enable the feature again after you complete all the steps in Task 3.

78 OS/400 Backup and Recovery V5R1

 Actions for non-load source disk unit failure–Checklist 9
This checklist should be used for the following problem situation:
Failed Unit:
Non-load source unit in system ASP
Data Loss:
All
User ASP Configured:
Yes
User ASP Overflowed:
No

Attention!
When you replace a disk unit in your system ASP, the system loses
addressability to the objects in your user ASPs. Recovering object ownership
for objects other than DLOs will require manually assigning ownership for
every object in every user ASP. You may want to treat this situation as a total
recovery and restore all your information from your save media if the
following conditions are true:
1. You have a large number of objects in your user ASPs
2. You have thoroughly backed up your system
If you choose to do this, perform the steps that are described in “Recovering
your entire system after a complete system loss–Checklist 20” on page 96 to
recover your system.

Table 16. Recovery Checklist for Disk Failure–Checklist 9

Task What To Do Where To Read More About It
Actions to Be Performed by the Service Representative
Task 1 Delete the data in the ASP that contains the
failed unit.
Task 2 Replace the failed disk unit.
Task 3 Configure the replacement disk unit by
adding it to the correct ASP.
Task 4 Restore the Licensed Internal Code using “How to Prepare for Loading the
option 1 (Restore Licensed Internal Code)1. If Licensed Internal Code” on page 122
user ASPs are configured, they remain intact. and “How to Load the Licensed
Internal Code” on page 130.
Actions to Be Performed by the User
Task 5 Restore the operating system, beginning with Chapter 6. Restoring the Operating
“Task 1–Starting to Restore the Operating System, task 1 through task 6.
System” on page 150. You are performing a
complete restore operation.
Task 6 If you restored the operating system using “Recovering System Information” on
distribution media, some system information, page 213.
such as access path recovery times and the
system reply list, may have been reset to
default values. Verify these values and
correct them if necessary.

Chapter 4. Selecting the Right Recovery Strategy 79

 Table 16. Recovery Checklist for Disk Failure–Checklist 9 (continued)
Task What To Do Where To Read More About It
Task 7 If necessary, change the QALWOBJRST “Controlling Restoration of
system value. Write the old value here: Security-Sensitive Objects” on page 50.
______________
| Task 8 If necessary, change the QVFYOBJRST “Controlling Restoration of
system value. Write the old value here: Security-Sensitive Objects” on page 50.
______________
Task 9 If necessary, change the system value that The System Values subtopic of the
controls whether the job log wraps when it is System Management topic in the
full. Use the Work with System Values iSeries Information Center.
command: WRKSYSVAL QJOBMSGQFL.
Write down the current value here:
______________. Then change the value to
*PRTWRAP.
Task 10 After changing the system values, sign off by “Describing the Contents of Your User
using the command SIGNOFF *LIST. Then, Auxiliary Storage Pools” on page 181.
using a newly created password, sign back
on as QSECOFR for the new values to take
effect.
Task 11 Describe or diagram, as much as possible, “Describing the Contents of Your User
the contents of your user ASPs before the Auxiliary Storage Pools” on page 181.
failure.
Task 12 Recover user profiles, configuration, libraries “How to Recover a User ASP After
in the system ASP, and the contents of your Recovering the System ASP” on
user ASPs. If you choose not to restore all of page 182, task 1 through task 11.
your libraries at this time, ensure that you
restore the QGPL and QUSRSYS libraries
along with the libraries that you are
restoring.
Task 13 Restore document library objects to the “Restoring Documents and Folders” on
system ASP. page 255.
Task 14 Restore your last complete save of “Restoring Objects in Directories” on
directories.2 page 261.
Task 15 If you have User-Defined File Systems in “Task 7–Restoring User-Defined File
User ASPs that do not restore correctly, you Systems to the User Auxiliary Storage
may need to perform additional recovery Pool” on page 187.
steps.
Task 16 Restore changed objects and apply journaled Chapter 11. How to Restore Changed
changes. Objects and Apply Journaled Changes,
task 1 through task 7.
Task 17 If you use Cryptographic Access Provider, “Recovery Considerations for
install the Cryptographic Access Provider Cryptographic Access Provider” on
licensed program (5769AC1), (5769AC2), or page 60
(5769AC3) using option 11 on the Work with
Licensed Programs menu.
Task 18 Restore authority. Type: RSTAUT “Restoring Object Authorities” on
page 219.
Task 19 Reapply any PTFs that were applied since “How to Restore Program Temporary
your last SAVSYS operation. Fixes” on page 278.
Task 20 If you use Windows server on iSeries and “Completing Recovery for the iSeries
saved with the Integrated xSeries Server Integration for Windows Server
(NWSD) varied off, complete recovery for Product” on page 263.
the Windows server on iSeries product.

80 OS/400 Backup and Recovery V5R1

 Table 16. Recovery Checklist for Disk Failure–Checklist 9 (continued)
Task What To Do Where To Read More About It
Task 21 If you run Linux and saved Linux by “Recovering Linux in a Partition” on
varying off the network server description page 264
(NWSD) for Linux, complete the recovery for
Linux.
Task 22 If necessary, change the QALWOBJRST “Controlling Restoration of
system value back to its original value by Security-Sensitive Objects” on page 50.
using the WRKSYSVAL command.
| Task 23 If necessary, change the QVFYOBJRST “Controlling Restoration of
system value back to its original value by Security-Sensitive Objects” on page 50.
using the WRKSYSVAL command.
Task 24 If necessary, change the QJOBMSGQFL The System Values subtopic of the
system value back to its original value by System Management topic in the
using the WRKSYSVAL command. iSeries Information Center.
Task 25 Perform either a:
SIGNOFF *LIST

or a
DSPJOBLOG * *PRINT

Check the job log to ensure all objects were

restored. The job log contains information
about the restore operation. To verify that all
objects were restored, spool the job log for
printing along with the job’s remaining
spooled output, if any.

Message CPC3703 is sent to the job log for

each library that was successfully restored.
Message CPF3773 is sent to tell you how
many objects were restored. Objects are not
restored for various reasons. Check for any
error messages, correct the errors, and then
restore those objects from the media.
Task 26 You must perform a normal IPL at this “How to Perform a Normal IPL” on
time. page 58.
Task 27 If you use Windows server on iSeries and “Completing Recovery for the iSeries
saved with the Integrated xSeries Server Integration for Windows Server
(NWSD) varied on, complete recovery for the Product” on page 263.
Windows server on iSeries product.
Task 28 If IBM Content Manager OnDemand for
iSeries is installed, complete journaling for
IBM Content Manager OnDemand for iSeries
by typing the following commands:
CALL QRDARS/QRLCSTRJ PARM('RLC')
CALL QRDARS/QRLCSTRJ PARM('RLR')
CALL QRDARS/QRLCSTRJ PARM('RLO')
Task 29 Review job logs or output from your restore “How to Verify That Objects Are
operations to ensure that all objects were Restored Successfully” on page 54.
restored successfully.

Chapter 4. Selecting the Right Recovery Strategy 81

Table 16. Recovery Checklist for Disk Failure–Checklist 9 (continued)
Task What To Do Where To Read More About It

1
If you have a 2440 Tape Unit with the high-speed feature enabled, you must disable it before restoring the
Licensed Internal Code. See “Disabling and Enabling the High-Speed Feature on the 2440 Tape Unit” on
page 145 for instructions. You may enable the feature again after you complete all the steps in Task 4.
2
You may receive one of the following messages:
CPD377A: Object not restored, /QNTC.
CPD377A: Object not restored, /QNetWare.

These objects cannot be restored until their file systems have been mounted during the IPL. You may
ignore these messages. The additional recovery tasks will guide you through the steps to restore these
objects.
Note: Since the OS/400 Enhanced Integration for Novell NetWare software resides on a remote server, you
do not have to restore your Netware data when you restore your. Previously, the OS/400 Integration for
Novell NetWare product ran on a Integrated xSeries Server and you had to restore the Novell product if
you completely restored your.

Actions for non-load source disk unit failure–Checklist 10

This checklist should be used for the following problem situation:
Failed Unit:
Non-load source unit in system ASP
Data Loss:
All
User ASP Configured:
Yes
User ASP Overflowed:
Yes

Attention!
When you replace a disk unit in your system ASP, the system loses
addressability to the objects in your user ASPs. Recovering object ownership
for objects other than DLOs will require manually assigning ownership for
every object in every user ASP. You may want to treat this situation as a total
recovery and restore all your information from your save media if the
following conditions are true:
1. You have a large number of objects in your user ASPs
2. You have thoroughly backed up your system
If you choose to do this, perform the steps that are described in “Recovering
your entire system after a complete system loss–Checklist 20” on page 96 to
recover your system.

Table 17. Recovery Checklist for Disk Failure–Checklist 10

Task What To Do Where To Read More About It
Actions to Be Performed by the Service Representative
Task 1 Physically remove the failed disk unit from
the system.

82 OS/400 Backup and Recovery V5R1

 Table 17. Recovery Checklist for Disk Failure–Checklist 10 (continued)
Task What To Do Where To Read More About It
Task 2 Delete the data in the ASP that contains the
failed unit. When you delete the data in the
system ASP, the system also deletes data in
any user ASPs that have a status of
overflowed.
Task 3 Install the replacement disk unit.
Task 4 Configure the replacement disk unit by
selecting the ’Replace configured unit’
function on the Work with Disk Units
display.
Task 5 Restore the Licensed Internal Code using “How to Prepare for Loading the
option 1 (Restore Licensed Internal Code)1. If Licensed Internal Code” on page 122
user ASPs are configured and are not and “How to Load the Licensed
overflowed, they remain intact. Internal Code” on page 130.
Actions to Be Performed by the User
Task 6 Restore the operating system, beginning with Chapter 6. Restoring the Operating
“Task 1–Starting to Restore the Operating System, task 1 through task 6.
System” on page 150. You are performing a
complete restore operation.
Task 7 If you restored the operating system using “Recovering System Information” on
distribution media, some system information page 213.
such as access path recovery times and the
system reply list may have been reset to
default values. Verify these values and
correct them if necessary.
Task 8 If necessary, change the QALWOBJRST “Controlling Restoration of
system value. Write the old value here: Security-Sensitive Objects” on page 50.
______________
| Task 9 If necessary, change the QVFYOBJRST “Controlling Restoration of
system value. Write the old value here: Security-Sensitive Objects” on page 50.
______________
Task 10 If necessary, change the system value that The System Values subtopic of the
controls whether the job log wraps when it is System Management topic in the
full. Use the Work with System Values iSeries Information Center.
command: WRKSYSVAL QJOBMSGQFL.
Write down the current value here:
______________. Then change the value to
*PRTWRAP.
Task 11 After changing the system values, sign off by
using the command SIGNOFF *LIST. Then,
using a newly created password, sign back
on as QSECOFR for the new values to take
effect.
Task 12 Describe or diagram, as much as possible, “Describing the Contents of Your User
the contents of your user ASPs before the Auxiliary Storage Pools” on page 181.
failure.

Chapter 4. Selecting the Right Recovery Strategy 83

 Table 17. Recovery Checklist for Disk Failure–Checklist 10 (continued)
Task What To Do Where To Read More About It
Task 13 Recover user profiles, configuration, libraries “How to Recover a User ASP After
in the system ASP, and the contents of any Recovering the System ASP” on
user ASPs that were not in overflowed page 182, task 1 through task 11.
status. If you choose not to restore all of
your libraries at this time, ensure that you
restore the QGPL and QUSRSYS libraries
along with the libraries that you are
restoring.
Task 14 Recover the objects in the user ASPs that “How to Recover a Damaged User
were overflowed. Auxiliary Storage Pool” on page 196,
task 1 through task 9.
Task 15 Restore document library objects to the “Restoring Documents and Folders” on
system ASP and to any overflowed user page 255.
ASPs that had DLOs.
Task 16 Restore your last complete save of “Restoring Objects in Directories” on
directories.2 page 261.
Task 17 If you have User-Defined File Systems in “Task 7–Restoring User-Defined File
User ASPs that do not restore correctly, you Systems to the User Auxiliary Storage
may need to perform additional recovery Pool” on page 187.
steps.
Task 18 Restore changed objects and apply journaled Chapter 11. How to Restore Changed
changes. Objects and Apply Journaled Changes,
task 1 through task 7.
Task 19 If you use Cryptographic Access Provider, “Recovery Considerations for
install the Cryptographic Access Provider Cryptographic Access Provider” on
licensed program (5769AC1), (5769AC2), or page 60
(5769AC3) using option 11 on the Work with
Licensed Programs menu.
Task 20 Restore authority. Type: RSTAUT “Restoring Object Authorities” on
page 219.
Task 21 Reapply any PTFs that were applied since “How to Restore Program Temporary
your last SAVSYS operation. Fixes” on page 278.
Task 22 If you use Windows server on iSeries and “Completing Recovery for the iSeries
saved with the Integrated xSeries Server Integration for Windows Server
(NWSD) varied off, complete recovery for Product” on page 263.
the Windows server on iSeries product.
Task 23 If you run Linux and saved Linux by “Recovering Linux in a Partition” on
varying off the network server description page 264.
(NWSD) for Linux, complete the recovery for
Linux.
Task 24 If necessary, change the QALWOBJRST “Controlling Restoration of
system value back to its original value by Security-Sensitive Objects” on page 50.
using the WRKSYSVAL command.
| Task 25 If necessary, change the QVFYOBJRST “Controlling Restoration of
system value back to its original value by Security-Sensitive Objects” on page 50.
using the WRKSYSVAL command.
Task 26 If necessary, change the QJOBMSGQFL The System Values subtopic of the
system value back to its original value by System Management topic in the
using the WRKSYSVAL command. iSeries Information Center.

84 OS/400 Backup and Recovery V5R1

Table 17. Recovery Checklist for Disk Failure–Checklist 10 (continued)
Task What To Do Where To Read More About It
Task 27 Perform either a:
SIGNOFF *LIST

or a
DSPJOBLOG * *PRINT

Check the job log to ensure all objects were

restored. The job log contains information
about the restore operation. To verify that all
objects were restored, spool the job log for
printing along with the job’s remaining
spooled output, if any.

Message CPC3703 is sent to the job log for

each library that was successfully restored.
Message CPF3773 is sent to tell you how
many objects were restored. Objects are not
restored for various reasons. Check for any
error messages, correct the errors, and then
restore those objects from the media.
Task 28 You must perform a normal IPL at this “How to Perform a Normal IPL” on
time. page 58.
Task 29 If you use Windows server on iSeries and “Completing Recovery for the iSeries
saved with the Integrated xSeries Server Integration for Windows Server
(NWSD) varied on, complete recovery for the Product” on page 263.
Windows server on iSeries product.
Task 30 If IBM Content Manager OnDemand for
iSeries is installed, complete journaling for
IBM Content Manager OnDemand for iSeries
by typing the following commands:
CALL QRDARS/QRLCSTRJ PARM('RLC')
CALL QRDARS/QRLCSTRJ PARM('RLR')
CALL QRDARS/QRLCSTRJ PARM('RLO')
Task 31 Review job logs or output from your restore “How to Verify That Objects Are
operations to ensure that all objects were Restored Successfully” on page 54.
restored successfully.

1
If you have a 2440 Tape Unit with the high-speed feature enabled, you must disable it before restoring the
Licensed Internal Code. See “Disabling and Enabling the High-Speed Feature on the 2440 Tape Unit” on
page 145 for instructions. You may enable the feature again after you complete all the steps in Task 4.
2
You may receive one of the following messages:
CPD377A: Object not restored, /QNTC.
CPD377A: Object not restored, /QNetWare.

These objects cannot be restored until their file systems have been mounted during the IPL. You may
ignore these messages. The additional recovery tasks will guide you through the steps to restore these
objects.
Note: Since the OS/400 Enhanced Integration for Novell NetWare software resides on a remote server, you
do not have to restore your Netware data when you restore your server. Previously, the OS/400 Integration
for Novell NetWare product ran on a Integrated xSeries Server and you had to restore the Novell product
if you completely restored your server.

Chapter 4. Selecting the Right Recovery Strategy 85

 Actions for a failure in a user ASP disk unit–Checklist 11
This checklist should be used for the following problem situation:
Failed Unit:
User ASP
Data Loss:
Some

Most steps in the checklist include references to other topics in this book. Refer to
these topics if you need more information about how to perform a particular step.
You may not need to perform some steps, such as restoring changed objects, if they
do not apply in your situation.
Table 18. Recovery Checklist for Disk Failure–Checklist 11
Task What To Do Where To Read More About It
Actions to Be Performed by the Service Representative
Task 1 Save the disk unit.
Task 2 Attach the new disk unit.
Task 3 Restore the disk unit data.
Actions to Be Performed by the User
Task 4 You must perform an IPL at this time. Chapter 7. Starting the System After It
Follow the procedure for starting the system Ends Abnormally, task 1 through task
after it ends abnormally. 4.
Task 5 Reclaim storage. “Reclaiming Storage” on page 46.
Task 6 Evaluate the extent of the damage. “Task 4–Recovering from Damaged
Determine whether you will attempt to Objects and Unreadable Sectors” on
recover damaged objects or restore the entire page 174.
system. Do not skip this step.
Task 7 If you have decided to do a complete restore
operation, use Table 31 on page 107 to
determine the correct procedure for
recovering user information.
Task 8 If you have decided to attempt to recover
damaged objects, perform the tasks in “Task
4–Recovering from Damaged Objects and
Unreadable Sectors” on page 174.

Actions for a failure in a user ASP disk unit–Checklist 12

This checklist should be used for the following problem situation:
Failed Unit:
User ASP Not in overflow status
Data Loss:
All

Before you begin your recovery, make a copy of this checklist. Fill in the
appropriate areas as you and the service representative perform the recovery steps.
This checklist provides an important record of your recovery actions. It may help
you to diagnose any problems that occur after the recovery. It may also be useful
in evaluating your backup strategy.

86 OS/400 Backup and Recovery V5R1

 Most steps in the checklist include references to other topics in this book. Refer to
these topics if you need more information about how to perform a particular step.
You may not need to perform some steps, such as restoring changed objects, if they
do not apply in your situation.
Table 19. Recovery Checklist for Disk Failure–Checklist 12
Task What To Do Where To Read More About It
Actions to Be Performed by the Service Representative
Task 1 Physically remove the failed disk unit from
the system.
Task 2 Delete the data in the ASP that contains the
failed unit.
Task 3 Install the replacement disk unit.
Task 4 Configure the replacement disk unit by
selecting the ’Replace configured unit’
function on the Work with Disk Units
display.
Actions to Be Performed by the User
Task 5 You must perform an IPL at this time. Chapter 7. Starting the System After It
Follow the procedure for starting the system Ends Abnormally, task 1 through task
after it ends abnormally. 4.
Task 6 Restore user profiles: RSTUSRPRF “Restoring User Profiles” on page 214.
USRPRF(*ALL) DEV(TAP01)
Task 7 If necessary, change the QALWOBJRST “Controlling Restoration of
system value. Write the old value here: Security-Sensitive Objects” on page 50.
______________
| Task 8 If necessary, change the QVFYOBJRST “Controlling Restoration of
system value. Write the old value here: Security-Sensitive Objects” on page 50.
______________
Task 9 If necessary, change the system value that The System Values subtopic of the
controls whether the job log wraps when it is System Management topic in the
full. Use the Work with System Values iSeries Information Center.
command: WRKSYSVAL QJOBMSGQFL.
Write down the current value here:
______________
Task 10 After changing the system values, sign off by
using the command SIGNOFF *LIST. Then,
using a newly created password, sign back
on as QSECOFR for the new values to take
effect.
Task 11 Recover the objects in the user ASP. “How to Recover a Damaged User
Auxiliary Storage Pool” on page 196,
task 1 through task 9.
Task 12 If you have User-Defined File Systems in “Task 7–Restoring User-Defined File
User ASPs that do not restore correctly, you Systems to the User Auxiliary Storage
may need to perform additional recovery Pool” on page 187.
steps.

Chapter 4. Selecting the Right Recovery Strategy 87

 Table 19. Recovery Checklist for Disk Failure–Checklist 12 (continued)
Task What To Do Where To Read More About It
Task 13 Perform either a:
SIGNOFF *LIST

or a
DSPJOBLOG * *PRINT

Check the job log to ensure all objects were

restored. The job log contains information
about the restore operation. To verify that all
objects were restored, spool the job log for
printing along with the job’s remaining
spooled output, if any.

Message CPC3703 is sent to the job log for

each library that was successfully restored.
Message CPF3773 is sent to tell you how
many objects were restored. Objects are not
restored for various reasons. Check for any
error messages, correct the errors, and then
restore those objects from the media.
Task 14 Restore changed objects to the user ASP. Chapter 11. How to Restore Changed
Apply journaled changes to objects in the Objects and Apply Journaled Changes,
user ASP. task 1 through task 7.
Task 15 If you use Cryptographic Access Provider, “Recovery Considerations for
install the Cryptographic Access Provider Cryptographic Access Provider” on
licensed program (5769AC1), (5769AC2), or page 60
(5769AC3) using option 11 on the Work with
Licensed Programs menu.
Task 16 Restore authority. Type: RSTAUT “Restoring Object Authorities” on
page 219.
Task 17 If necessary, change the QALWOBJRST “Controlling Restoration of
system value back to its original value by Security-Sensitive Objects” on page 50.
using the WRKSYSVAL command.
| Task 18 If necessary, change the QVFYOBJRST “Controlling Restoration of
system value back to its original value by Security-Sensitive Objects” on page 50.
using the WRKSYSVAL command.
Task 19 If necessary, change the QJOBMSGQFL The System Values subtopic of the
system value back to its original value by System Management topic in the
using the WRKSYSVAL command. iSeries Information Center.
Task 20 You must perform an IPL at this time. “How to Perform a Normal IPL” on
page 58.
Task 21 Review job logs or output from your restore “How to Verify That Objects Are
operations to ensure that all objects were Restored Successfully” on page 54.
restored successfully.

Actions for a failure in a user ASP disk unit–Checklist 13

This checklist should be used for the following problem situation:
Failed Unit:
User ASP In overflow status
Data Loss:
All

88 OS/400 Backup and Recovery V5R1

 Before you begin your recovery, make a copy of this checklist. Fill in the
appropriate areas as you and the service representative perform the recovery steps.
This checklist provides an important record of your recovery actions. It may help
you to diagnose any problems that occur after the recovery. It may also be useful
in evaluating your backup strategy.

Most steps in the checklist include references to other topics in this book. Refer to
these topics if you need more information about how to perform a particular step.
You may not need to perform some steps, such as restoring changed objects, if they
do not apply in your situation.
Table 20. Recovery Checklist for Disk Failure–Checklist 13
Task What To Do Where To Read More About It
Actions to Be Performed by the Service Representative
Task 1 Physically remove the failed disk unit from
the system.
Task 2 Delete the data in the ASP that contains the
failed unit.
Task 3 Install the replacement disk unit.
Task 4 Configure the replacement disk unit by
selecting the ’Replace configured unit’
function on the Work with Disk Units
display.
Actions to Be Performed by the User
Task 5 You must perform an IPL at this time. “Chapter 7. Starting the System After It
Follow the procedure for starting the system Ends Abnormally” on page 167.
after it ends abnormally.
Task 6 Reclaim storage. “Reclaiming Storage” on page 46.
Task 7 Delete the overflowed objects. “How to Delete Overflowed Objects
during Recovery” on page 196.
Task 8 If necessary, change the QALWOBJRST “Controlling Restoration of
system value. Write the old value here: Security-Sensitive Objects” on page 50.
______________
| Task 9 If necessary, change the QVFYOBJRST “Controlling Restoration of
| system value. Write the old value here: Security-Sensitive Objects” on page 50.
| ______________
Task 10 If necessary, change the system value that The System Values subtopic of the
controls whether the job log wraps when it is System Management topic in the
full. Use the Work with System Values iSeries Information Center.
command: WRKSYSVAL QJOBMSGQFL.
Write down the current value here:
______________. Then change the value to
*PRTWRAP.
Task 11 After changing the system values, sign off by
using the command SIGNOFF *LIST. Then,
using a newly created password, sign back
on as QSECOFR for the new values to take
effect.
Task 12 Recover the objects in the user ASP. “How to Recover a Damaged User
Auxiliary Storage Pool” on page 196,
task 1 through task 9.

Chapter 4. Selecting the Right Recovery Strategy 89

 Table 20. Recovery Checklist for Disk Failure–Checklist 13 (continued)
Task What To Do Where To Read More About It
Task 13 If you have User-Defined File Systems in “Task 7–Restoring User-Defined File
User ASPs that do not restore correctly, you Systems to the User Auxiliary Storage
may need to perform additional recovery Pool” on page 187.
steps.
Task 14 Restore changed objects to the user ASP. Chapter 11. How to Restore Changed
Apply journaled changes to objects in the Objects and Apply Journaled Changes,
user ASP. task 1 through task 7.
Task 15 If you use Cryptographic Access Provider, “Recovery Considerations for
install the Cryptographic Access Provider Cryptographic Access Provider” on
licensed program (5769AC1), (5769AC2), or page 60
(5769AC3) using option 11 on the Work with
Licensed Programs menu.
Task 16 Restore authority. Type: RSTAUT “Restoring Object Authorities” on
page 219.
Task 17 If necessary, change the QALWOBJRST “Controlling Restoration of
system value. Security-Sensitive Objects” on page 50.
| Task 18 If necessary, change the QVFYOBJRST “Controlling Restoration of
| system value. Security-Sensitive Objects” on page 50.
Task 19 If necessary, change the QJOBMSGQFL The System Values subtopic of the
system value back to its original value by System Management topic in the
using the WRKSYSVAL command. iSeries Information Center.
Task 20 Perform either a:
SIGNOFF *LIST

or a
DSPJOBLOG * *PRINT

Check the job log to ensure all objects were

restored. The job log contains information
about the restore operation. To verify that all
objects were restored, spool the job log for
printing along with the job’s remaining
spooled output, if any.

Message CPC3703 is sent to the job log for

each library that was successfully restored.
Message CPF3773 is sent to tell you how
many objects were restored. Objects are not
restored for various reasons. Check for any
error messages, correct the errors, and then
restore those objects from the media.
Task 21 You must perform a normal IPL at this “How to Perform a Normal IPL” on
time. page 58.
Task 22 Review job logs or output from your restore “How to Verify That Objects Are
operations to ensure that all objects were Restored Successfully” on page 54.
restored successfully.

Actions for non-load source disk unit failure–Checklist 14

This checklist should be used for the following problem situation:
Failed Unit:
Any

90 OS/400 Backup and Recovery V5R1

 Mirrored Protection:
Yes

Before you begin your recovery, make a copy of this checklist. Fill in the
appropriate areas as you and the service representative perform the recovery steps.
This checklist provides an important record of your recovery actions. It may help
you to diagnose any problems that occur after the recovery. It may also be useful
in evaluating your backup strategy.

Most steps in the checklist include references to other topics in this book. Refer to
these topics if you need more information about how to perform a particular step.
You may not need to perform some steps, such as restoring changed objects, if they
do not apply in your situation.

Note: For many failures, the system does not have to stopped and started again.
The service representative can repair the failed component while the system
continues to run. See “Chapter 12. Mirrored Protection Recovery Actions” on
page 289.
Table 21. Recovery Checklist for Disk Failure–Checklist 14
Task What To Do Where To Read More About It
Actions to Be Performed by the Service Representative
Task 1 Replace the failed disk unit.
Task 2 Resume mirrored protection.
Actions to Be Performed by the User
Task 3 Ensure that disk configuration is correct. Chapter 27, “Working with Mirrored
Protection”.

Actions for non-load source disk unit failure–Checklist 15

This checklist should be used for the following problem situation:
Failed Unit:
Any
Device Parity Protection:
Yes

Most steps in the checklist include references to other topics in this book. Refer to
these topics if you need more information about how to perform a particular step.
You may not need to perform some steps, such as restoring changed objects, if they
do not apply in your situation.

Note: For many failures, the system does not have to stopped and started again.
The service representative can repair the failed component while the system
continues to run. See “Chapter 26. Working with Device Parity Protection”
on page 695.
Table 22. Recovery Checklist for Disk Failure–Checklist 15
Task What To Do Where To Read More About It
Actions to Be Performed by the Service Representative
Task 1 Attach the new disk unit.
Task 2 Rebuild the failed device parity disk unit
data.

Chapter 4. Selecting the Right Recovery Strategy 91

Table 22. Recovery Checklist for Disk Failure–Checklist 15 (continued)
Task What To Do Where To Read More About It
Actions to Be Performed by the User
Task 3 Ensure that disk configuration is correct. Chapter 26, “Working with Device
Parity Protection”.

Actions for non-load source disk unit failure–Checklist 16

This checklist should be used for the following problem situation:
Failed Unit:
Non-Load source unit in system ASP
Data Loss:
N/A Disk unit cannot be replaced immediately

Before you begin your recovery, make a copy of this checklist. Fill in the
appropriate areas as you and the service representative perform the recovery steps.
This checklist provides an important record of your recovery actions. It may help
you to diagnose any problems that occur after the recovery. It may also be useful
in evaluating your backup strategy.

Most steps in the checklist include references to other topics in this book. Refer to
these topics if you need more information about how to perform a particular step.
You may not need to perform some steps, such as restoring changed objects, if they
do not apply in your situation.
Table 23. Recovery Checklist for Disk Failure–Checklist 16
Task What To Do Where To Read More About It
Actions to Be Performed by the User
Task 1 Remove the failed disk unit from the “How to Remove a Disk Unit from
configuration. an Auxiliary Storage Pool” on
page 678.
Task 2 Restore the Licensed Internal Code “How to Prepare for Loading the
using option 1 (Restore Licensed Licensed Internal Code” on page 122
Internal Code)1. and “How to Load the Licensed
Internal Code” on page 130
Task 3 Restore the operating system, Chapter 6. Restoring the Operating
beginning with “Task 1–Starting to System, task 1 through task 6.
Restore the Operating System” on
page 150. You are performing a
complete restore operation.
Task 4 If you restored the operating system “Recovering System Information” on
using distribution media, some page 213.
system information, such as access
path recovery times and the system
reply list, may have been reset to
default values. Verify these values
and correct them if necessary.
Task 5 Use Table 31 on page 107 to
determine the correct procedure for
recovering user information.

92 OS/400 Backup and Recovery V5R1

 Table 23. Recovery Checklist for Disk Failure–Checklist 16 (continued)
Task What To Do Where To Read More About It

1
If you have a 2440 Tape Unit with the high-speed feature enabled, you must disable it before restoring the
Licensed Internal Code. See “Disabling and Enabling the High-Speed Feature on the 2440 Tape Unit” on
page 145 for instructions. You may enable the feature again after you complete all the steps in Task 2.

Actions for independent ASP disk failure–Checklist 17

This checklist should be used for the following problem situation:
Failed Unit:
Disk unit in independent ASP
Data Loss:
None
User ASP Configured:
N/A

Before you begin your recovery, make a copy of this checklist. Fill in the
appropriate areas as you and the service representative perform the recovery steps.
This checklist provides an important record of your recovery actions. It may help
you to diagnose any problems that occur after the recovery. It may also be useful
in evaluating your backup strategy.

Most steps in the checklist include references to other topics in this book. Refer to
these topics if you need more information about how to perform a particular step.
You may not need to perform some steps, such as restoring changed objects, if they
do not apply in your situation.
Table 24. Recovery Checklist for Disk Failure–Checklist 17
Task What To Do Where To Read More About It
Actions to Be Performed by the Service Representative
Task 1 Save the disk unit data.
Task 2 Attach a new disk unit.
Task 3 Restore data to the new disk unit.
Actions to Be Performed by the User
Task 4 Vary on the independent ASP. Use the VRYCFG command or the
Operations Navigator interface to vary
on the independent ASP.

| Actions for a failure in an independent ASP disk

| unit–Checklist 18
| This checklist should be used for the following problem situation:
| Failed Unit:
| Independent ASP
| Data Loss:
| Some

Chapter 4. Selecting the Right Recovery Strategy 93

| Most steps in the checklist include references to other topics in this book. Refer to
| these topics if you need more information about how to perform a particular step.
| You may not need to perform some steps, such as restoring changed objects, if they
| do not apply in your situation.
| Table 25. Recovery Checklist for Disk Failure–Checklist 18
| Task What To Do Where To Read More About It
| Actions to Be Performed by the Service Representative
| Task 1 Save the disk unit.
| Task 2 Attach the new disk unit.
| Task 3 Restore the disk unit data.
| Actions to Be Performed by the User
| Task 4 You must vary on the independent ASP Use the VRYCFG command or the
| now. Operations Navigator interface to vary
| on the independent ASP.
| Task 5 Reclaim storage. “Reclaiming Storage” on page 46.
| Task 6 Evaluate the extent of the damage. “Task 4–Recovering from Damaged
| Determine whether you will attempt to Objects and Unreadable Sectors” on
| recover damaged objects or restore the entire page 174.
| system. Do not skip this step.
| Task 7 If you have decided to proceed, continue “How to Recover a Damaged User
| with the restore operation for independent Auxiliary Storage Pool” on page 196
| ASP data.
| Task 8 If you have decided to attempt to recover
| damaged objects, perform the tasks in “Task
| 4–Recovering from Damaged Objects and
| Unreadable Sectors” on page 174.
|

| Actions for a failure in an independent ASP disk

| unit–Checklist 19
| This checklist should be used for the following problem situation:
| Failed Unit:
| Independent ASP
| Data Loss:
| All

| Before you begin your recovery, make a copy of this checklist. Fill in the
| appropriate areas as you and the service representative perform the recovery steps.
| This checklist provides an important record of your recovery actions. It may help
| you to diagnose any problems that occur after the recovery. It may also be useful
| in evaluating your backup strategy.

| Most steps in the checklist include references to other topics in this book. Refer to
| these topics if you need more information about how to perform a particular step.
| You may not need to perform some steps, such as restoring changed objects, if they
| do not apply in your situation.
| Table 26. Recovery Checklist for Disk Failure–Checklist 19
| Task What To Do Where To Read More About It
| Actions to Be Performed by the Service Representative

94 OS/400 Backup and Recovery V5R1

| Table 26. Recovery Checklist for Disk Failure–Checklist 19 (continued)
| Task What To Do Where To Read More About It
| Task 1 Physically remove the failed disk unit from
| the system.
| Task 2 Delete the data in the ASP that contains the
| failed unit.
| Task 3 Install the replacement disk unit.
| Task 4 Configure the replacement disk unit by
| selecting the ’Replace configured unit’
| function on the Work with Disk Units
| display.
| Actions to Be Performed by the User
| Task 5 Vary on the IASP at this time. Use the VRYCFG command or the
| Operations Navigator interface to vary
| on the independent ASP.
| Task 6 Unmount ’/dev/iasp- You can display where UDFS are
| name/QDEFAULT.UDFS’ and any other mounted with the DSPUDFS
| UDFS associated with the independent ASP. command. Then use the UNMOUNT
| command.
| Task 7 If necessary, change the QALWOBJRST “Controlling Restoration of
| system value. Write the old value here: Security-Sensitive Objects” on page 50.
| ______________
| Task 8 If necessary, change the QVFYOBJRST “Controlling Restoration of
| system value. Write the old value here: Security-Sensitive Objects” on page 50.
| ______________
| Task 9 If necessary, change the system value that The System Values subtopic of the
| controls whether the job log wraps when it is System Management topic in the
| full. Use the Work with System Values iSeries Information Center.
| command: WRKSYSVAL QJOBMSGQFL.
| Write down the current value here:
| ______________
| Task 10 After changing the system values, sign off by
| using the command SIGNOFF *LIST. Then,
| using a newly created password, sign back
| on as QSECOFR for the new values to take
| effect.
| Task 11 Recover the objects in the independent user “How to Recover a Damaged User
| ASP. Auxiliary Storage Pool” on page 196.
| Note: If you know which user profiles own
| objects in the independent ASP, you can
| specify the individual profiles and avoid
| restricting your system to process USRPRF
| *ALL.
| Task 12 Restore changed objects to the independent Task 2–Restoring Changed Objects in
| user ASP. Directories.
| Task 13 Restore authority. Type: RSTAUT “Restoring Object Authorities” on
| page 219.

Chapter 4. Selecting the Right Recovery Strategy 95

| Table 26. Recovery Checklist for Disk Failure–Checklist 19 (continued)
| Task What To Do Where To Read More About It
| Task 14 Perform either a:
| SIGNOFF *LIST

| or a
| DSPJOBLOG * *PRINT

| Check the job log to ensure all objects were

| restored. The job log contains information
| about the restore operation. To verify that all
| objects were restored, spool the job log for
| printing along with the job’s remaining
| spooled output, if any.

| Message CPF3773 is sent to tell you how

| many objects were restored. Objects are not
| restored for various reasons. Check for any
| error messages, correct the errors, and then
| restore those objects from the media.
| Task 15 If necessary, change the QALWOBJRST “Controlling Restoration of
| system value back to its original value by Security-Sensitive Objects” on page 50.
| using the WRKSYSVAL command.
| Task 16 If necessary, change the QVFYOBJRST “Controlling Restoration of
| system value back to its original value by Security-Sensitive Objects” on page 50.
| using the WRKSYSVAL command.
| Task 17 If necessary, change the QJOBMSGQFL The System Values subtopic of the
| system value back to its original value by System Management topic in the
| using the WRKSYSVAL command. iSeries Information Center.
| Task 18 Mount the UDFS that was unmounted in Use the MOUNT command to mount
| Task 6. ’/dev/iasp-name/QDEFAULT.UDFS’.
| Task 19 Review job logs or output from your restore “How to Verify That Objects Are
| operations to ensure that all objects were Restored Successfully” on page 54.
| restored successfully.
|

| Recovering your entire system after a complete system

loss–Checklist 20
| This checklist should be used if you need to restore your entire system to a system
| that is running the same version of the OS/400 licensed program. Do not use this
| checklist if you are performing a RISC to RISC upgrade. For information on
| performing a RISC to RISC upgrade, refer to iSeries 940x RISC-to-RISC Road Map,
| SA41-5155-05.

Note: If the system you must recover contains an independent ASP, refer to
“Recovering your entire system after a complete system loss including
independent ASPs–Checklist 21” on page 98 or “Recovering your entire
system after a complete system loss including independent ASPs, when
Service Tools Network Interface is configured–Checklist 22” on page 102.

Before you begin your recovery, make a copy of this checklist. Fill in the
appropriate areas as you and the service representative perform the recovery steps.
This checklist provides an important record of your recovery actions. It may help
you to diagnose any problems that occur after the recovery. It may also be useful
in evaluating your backup strategy.

96 OS/400 Backup and Recovery V5R1

 Most steps in the checklist include references to other topics in this book. Refer to
these topics if you need more information about how to perform a particular step.
You may not need to perform some steps, such as restoring changed objects, if they
do not apply in your situation.
Table 27. Recovery Checklist for Complete System Loss–Checklist 20
Task What To Do Where To Read More About It
Actions to Be Performed by the User
Task 1 Prepare to load the Licensed Internal “How to Prepare for Loading the
Code1. Licensed Internal Code” on page 122.
Task 2 Install the Licensed Internal Code “How to Prepare for Loading the
using option 2 (Install Licensed Licensed Internal Code” on page 122
Internal Code and Initialize System)1. and “How to Load the Licensed
Internal Code” on page 130
Task 3 Configure the disk units (assign to “Chapter 24. Procedures for
ASP and set up disk protection). If Configuring Disks and Disk
you saved any User-Defined File Protection” on page 643 and
Systems (UDFS), you must configure “Chapter 25. Working with Auxiliary
your user ASPs or the UDFS will not Storage Pools” on page 669.
restore.
Task 4 Restore the operating system, “How to Restore the OS/400
beginning with “Task 1–Starting to Licensed Program” on page 149.
Restore the Operating System” on
page 150. You are performing a
complete restore operation.
Task 5 If you restored the operating system “Recovering System Information” on
using distribution media, some page 213.
system information, such as access
path recovery times and the system
reply list may have been reset to
default values. Verify these values
and correct them if necessary.
Task 6 Recover user information from your “Choosing the Procedure to Recover
save media. Restore changed objects User Information” on page 107.
and apply journal entries. If you are
restoring to a different system, or a
different logical partition, you must
specify ALWOBJDIF(*ALL) on the
RSTxxx commands.
Note: If you use Backup Recovery
and Media Services, refer to your
Backup Recovery and Media Services
recovery report to recover your user
information.
Task 7 If you are not sure what the “What Happens When You Restore
password is for the QSECOFR profile User Profiles” on page 215.
that was restored from tape, change
it before signing off: CHGUSRPRF
USRPRF(QSECOFR)
PASSWORD(new-password)
Task 8 If you restored from distribution “Recovering System Information” on
media, restore your system page 213.
information to the correct settings.

Chapter 4. Selecting the Right Recovery Strategy 97

 Table 27. Recovery Checklist for Complete System Loss–Checklist 20 (continued)
Task What To Do Where To Read More About It
Task 9 Perform either a
SIGNOFF *LIST

or a
DSPJOBLOG * *PRINT

Check the job log to ensure that all

objects were restored. The job log
contains information about the
restore operation. To verify that all
objects were restored, spool the job
log for printing along with the job’s
remaining spooled output, if any.

Message CPC3703 is sent to the job

log for each library that was
successfully restored. Message
CPF3773 is sent to tell you how
many objects were restored. Objects
are not restored for various reasons.
Check for any error messages, correct
the error, and then restore those
objects from the media.
Task 10 You must perform a normal IPL at “How to Perform a Normal IPL” on
this time. page 58.
Task 11 If you use Windows server on iSeries “Completing Recovery for the iSeries
and saved with the Integrated xSeries Integration for Windows Server
Server (NWSD) varied on, complete Product” on page 263.
recovery for the Windows server on
iSeries.
Task 12 If IBM Content Manager OnDemand
for iSeries is installed, complete
journaling for IBM Content Manager
OnDemand for iSeries by typing the
following commands:
CALL QRDARS/QRLCSTRJ PARM('RLC')
CALL QRDARS/QRLCSTRJ PARM('RLR')
CALL QRDARS/QRLCSTRJ PARM('RLO')
Task 13 Review job logs or output from your “How to Verify That Objects Are
restore operations to ensure that all Restored Successfully” on page 54.
objects were restored successfully.

1
If you have a 2440 Tape Unit with the high-speed feature enabled, you must disable it before restoring the
Licensed Internal Code. See “Disabling and Enabling the High-Speed Feature on the 2440 Tape Unit” on
page 145 for instructions. You may enable the feature again after you complete all the steps in Task 2.

Recovering your entire system after a complete system loss

including independent ASPs–Checklist 21
| This checklist should be used if you need to restore your entire system that
| includes an independent disk pool to a system that is running the same version of
| the OS/400 licensed program. If you have the Service Tools Network Interface
| configured to access DST disk management function with Operations Navigator,

98 OS/400 Backup and Recovery V5R1

| refer to “Recovering your entire system after a complete system loss including
| independent ASPs, when Service Tools Network Interface is configured–Checklist
| 22” on page 102 . Do not use this checklist if you are performing a RISC to RISC
| upgrade. For information on performing a RISC to RISC upgrade, refer to iSeries
| 940x RISC-to-RISC Road Map, SA41-5155-05.

| Note: If you are restoring a clustered system with independent ASPs, refer to the
| Clustering topic in the Information Center at
| www.ibm.com/eserver/iseries/infocenter, along with this checklist.

| Before you begin your recovery, make a copy of this checklist. Fill in the
appropriate areas as you and the service representative perform the recovery steps.
This checklist provides an important record of your recovery actions. It may help
you to diagnose any problems that occur after the recovery. It may also be useful
in evaluating your backup strategy.

Most steps in the checklist include references to other topics in this book. Refer to
these topics if you need more information about how to perform a particular step.
You may not need to perform some steps, such as restoring changed objects, if they
do not apply in your situation.
Table 28. Recovery Checklist for Complete System Loss–Checklist 21
Task What To Do Where To Read More About It
Actions to Be Performed by the User
Task 1 Prepare to load the Licensed Internal “How to Prepare for Loading the
Code1. Licensed Internal Code” on page 122.
Task 2 Install the Licensed Internal Code “How to Prepare for Loading the
using option 2 (Install Licensed Licensed Internal Code” on page 122
Internal Code and Initialize System)1. and “How to Load the Licensed
Internal Code” on page 130
Task 3 Configure the disk units (assign to “Chapter 24. Procedures for
ASP and set up disk protection). If Configuring Disks and Disk
you saved any User-Defined File Protection” on page 643 and
Systems (UDFS), you must configure “Chapter 25. Working with Auxiliary
your user ASPs or the UDFS will not Storage Pools” on page 669.
restore.
Note: You will configure and restore
independent ASPs in a later step.
Task 4 Restore the operating system, “How to Restore the OS/400
beginning with “Task 1–Starting to Licensed Program” on page 149.
Restore the Operating System” on
page 150. You are performing a
complete restore operation.
Task 5 If you restored the operating system “Recovering System Information” on
using distribution media, some page 213.
system information, such as access
path recovery times and the system
reply list may have been reset to
default values. Verify these values
and correct them if necessary.

Chapter 4. Selecting the Right Recovery Strategy 99

Table 28. Recovery Checklist for Complete System Loss–Checklist 21 (continued)
Task What To Do Where To Read More About It
Task 6 Recover user information from your “Choosing the Procedure to Recover
save media. Restore changed objects User Information” on page 107.
and apply journal entries. If you are
restoring to a different system, or a
different logical partition, you must
specify ALWOBJDIF(*ALL) on the
RSTxxx commands.
Note: Because you have not
configured independent ASPs, you
will receive messages that some
objects are not restored.
Note: You may want to wait to
restore authorities until after you
have configured the independent
ASP to eliminate doing the procedure
twice.
Note: If you use Backup Recovery
and Media Services, refer to your
Backup Recovery and Media Services
recovery report to recover your user
information.
Task 7 If you are not sure what the “What Happens When You Restore
password is for the QSECOFR profile User Profiles” on page 215.
that was restored from tape, change
it before signing off: CHGUSRPRF
USRPRF(QSECOFR)
PASSWORD(new-password)
Task 8 If you restored from distribution “Recovering System Information” on
media, restore your system page 213.
information to the correct settings.
Task 9 Perform either a
SIGNOFF *LIST

or a
DSPJOBLOG * *PRINT

Check the job log to ensure that all

objects were restored. The job log
contains information about the
restore operation. To verify that all
objects were restored, spool the job
log for printing along with the job’s
remaining spooled output, if any.

Message CPC3703 is sent to the job

log for each library that was
successfully restored. Message
CPF3773 is sent to tell you how
many objects were restored. Objects
are not restored for various reasons.
Check for any error messages, correct
the error, and then restore those
objects from the media.
Task 10 You must perform a normal IPL at “How to Perform a Normal IPL” on
this time. page 58.

100 OS/400 Backup and Recovery V5R1

 Table 28. Recovery Checklist for Complete System Loss–Checklist 21 (continued)
Task What To Do Where To Read More About It
Task 11 If you use Windows server on iSeries “Completing Recovery for the iSeries
and saved with the Integrated xSeries Integration for Windows Server
Server (NWSD) varied on, complete Product” on page 263.
recovery for the Windows server on
iSeries.
Task 12 Delete the device description that is You can view the restored
associated with the independent ASP configurations with the WRKCFGSTS
as a result of the restore *DEV *ASP command. Then use the
configuration command that you DLTDEVD DEVD(iasp-name)
performed earlier in this checklist. command to delete the device
You will recreate the device description associated with the
description in the next step. independent ASP.
Task 13 You may need to configure the Tips for Managing and Monitoring
Service Tools Server for OS/400 in Authority chapter of Tips and Tools for
order to access disk management Securing Your iSeries.
functions.
Task 14 Configure independent ASPs through Operations Navigator online help for
Operations Navigator. disk units.
Task 15 Verify the RESOURCE and vary on Use the vary configuration (VRYCFG)
the independent ASP now. This will command.
create a directory for the independent
ASP and automatically mount the
UDFS to that directory.
Task 16 Unmount ’/dev/iasp- You can display where UDFS are
name/QDEFAULT.UDFS’ and any mounted with the DSPUDFS
other UDFS associated with the command. Then use the UNMOUNT
independent ASP. command.
| Task 17 Restore user profiles. “Task 2–Restoring User Profiles” on
| Note: If you chose not to run the page 184.
| RSTAUT command before the IPL, you
| do not need to restore user profiles
| again.
| Task 18 Restore independent ASP data. “Recovery Steps for Unmounted
User-Defined File Systems (UDFS)”
on page 187
Task 19 Restore authority. Type: RSTAUT “Restoring Object Authorities” on
page 219.
Task 20 If you want to continue operations Mount ’/dev/iasp-
with the independent ASP varied on, name/QDEFAULT.UDFS’ and any
you must mount the UDFS that you other UDFS associated with the
unmounted in a previous step. independent ASP.
Task 21 If IBM Content Manager OnDemand
for iSeries is installed, complete
journaling for IBM Content Manager
OnDemand for iSeries by typing the
following commands:
CALL QRDARS/QRLCSTRJ PARM('RLC')
CALL QRDARS/QRLCSTRJ PARM('RLR')
CALL QRDARS/QRLCSTRJ PARM('RLO')
Task 22 Review job logs or output from your “How to Verify That Objects Are
restore operations to ensure that all Restored Successfully” on page 54.
objects were restored successfully.

Chapter 4. Selecting the Right Recovery Strategy 101

 Table 28. Recovery Checklist for Complete System Loss–Checklist 21 (continued)
Task What To Do Where To Read More About It

1
If you have a 2440 Tape Unit with the high-speed feature enabled, you must disable it before restoring the
Licensed Internal Code. See “Disabling and Enabling the High-Speed Feature on the 2440 Tape Unit” on
page 145 for instructions. You may enable the feature again after you complete all the steps in Task 2.

| Recovering your entire system after a complete system loss

| including independent ASPs, when Service Tools Network
| Interface is configured–Checklist 22
| This checklist should be used if you need to restore your entire system to a system
| that is running the same version of the OS/400 licensed program. Use this
| checklist if the system you are recovering contained an independent auxiliary
| storage pool (ASP) and has the service tools network interface configured for DST
| access to disk management functions through Operations Navigator. Do not use
| this checklist if you are performing a RISC to RISC upgrade. For information on
| performing a RISC to RISC upgrade, refer to iSeries 940x RISC-to-RISC Road Map,
| SA41-5155-05.

| Note: If you are restoring a clustered system with independent ASPs, refer to the
| Clustering topic in the Information Center at
| www.ibm.com/eserver/iseries/infocenter, along with this checklist.

| Before you begin your recovery, make a copy of this checklist. Fill in the
| appropriate areas as you and the service representative perform the recovery steps.
| This checklist provides an important record of your recovery actions. It may help
| you to diagnose any problems that occur after the recovery. It may also be useful
| in evaluating your backup strategy.

| Most steps in the checklist include references to other topics in this book. Refer to
| these topics if you need more information about how to perform a particular step.
| You may not need to perform some steps, such as restoring changed objects, if they
| do not apply in your situation.
| Table 29. Recovery Checklist for Complete System Loss–Checklist 22
| Task What To Do Where To Read More About It
| Actions to Be Performed by the User
| Task 1 Prepare to load the Licensed Internal “How to Prepare for Loading the
| Code1. Licensed Internal Code” on page 122.
| Task 2 Install the Licensed Internal Code “How to Prepare for Loading the
| using option 2 (Install Licensed Licensed Internal Code” on page 122
| Internal Code and Initialize System)1. and “How to Load the Licensed
| Internal Code” on page 130
| Task 3 You may need to configure the Tips for Managing and Monitoring
| Service Tools Server in order to Authority chapter of Tips and Tools for
| access disk management functions at Securing Your iSeries.
| DST.

102 OS/400 Backup and Recovery V5R1

| Table 29. Recovery Checklist for Complete System Loss–Checklist 22 (continued)
| Task What To Do Where To Read More About It
| Task 4 Use dedicated service tools (DST) to “How to Recover Your Disk
| access Operations Navigator which is Configuration Using Operations
| needed to configure independent Navigator at DST” on page 138 and
| ASPs. Configure the disk units the online help for Disk Units in
| (assign to ASP and set up disk Operations Navigator. To configure
| protection). If you saved any disk units to system and basic ASPs
| User-Defined File Systems (UDFS), via the character-based interface,
| you must configure your basic and refer to “Chapter 24. Procedures for
| independent user ASPs or the UDFS Configuring Disks and Disk
| will not restore. You may configure Protection” on page 643 and
| system and basic ASPs via the “Chapter 25. Working with Auxiliary
| character-based interface. Storage Pools” on page 669.
| Task 5 Restore the operating system, “Task 2–Selecting the Installation
| beginning with ″Task 2–Selecting the Options” on page 154.
| Install Options. You are performing a
| complete restore operation.
| Task 6 If you restored the operating system “Recovering System Information” on
| using distribution media, some page 213.
| system information, such as access
| path recovery times and the system
| reply list may have been reset to
| default values. Verify these values
| and correct them if necessary.
| Task 7 Vary on the IASP at this time. Use the VRYCFG command or the
| Operations Navigator interface to
| vary on the independent ASP.
| Task 8 Unmount ’/dev/iasp- You can display where UDFS are
| name/QDEFAULT.UDFS’ and any mounted with the DSPUDFS
| other UDFS associated with the command. Then use the UNMOUNT
| independent ASP. command.
| Task 9 Recover user information from your “Recovering User Information Using
| save media. Restore changed objects Commands–Checklist 24” on
| and apply journal entries. If you are page 108.
| restoring to a different system, or a
| different logical partition, you must
| specify ALWOBJDIF(*ALL) on the
| RSTxxx commands.
| Note: If you use Backup Recovery
| and Media Services, refer to your
| Backup Recovery and Media Services
| recovery report to recover your user
| information.
| Task 10 If you are not sure what the “What Happens When You Restore
| password is for the QSECOFR profile User Profiles” on page 215.
| that was restored from tape, change
| it before signing off: CHGUSRPRF
| USRPRF(QSECOFR)
| PASSWORD(new-password)
| Task 11 If you restored from distribution “Recovering System Information” on
| media, restore your system page 213.
| information to the correct settings.

Chapter 4. Selecting the Right Recovery Strategy 103

| Table 29. Recovery Checklist for Complete System Loss–Checklist 22 (continued)
| Task What To Do Where To Read More About It
| Task 12 Perform either a
| SIGNOFF *LIST

| or a
| DSPJOBLOG * *PRINT

| Check the job log to ensure that all

| objects were restored. The job log
| contains information about the
| restore operation. To verify that all
| objects were restored, spool the job
| log for printing along with the job’s
| remaining spooled output, if any.

| Message CPC3703 is sent to the job

| log for each library that was
| successfully restored. Message
| CPF3773 is sent to tell you how
| many objects were restored. Objects
| are not restored for various reasons.
| Check for any error messages, correct
| the error, and then restore those
| objects from the media.
| Task 13 You must perform a normal IPL at “How to Perform a Normal IPL” on
| this time. page 58.
| Task 14 If you use Windows server on iSeries “Completing Recovery for the iSeries
| and saved with the Integrated xSeries Integration for Windows Server
| Server (NWSD) varied on, complete Product” on page 263.
| recovery for the Windows server on
| iSeries.
| Task 15 If IBM Content Manager OnDemand
| for iSeries is installed, complete
| journaling for IBM Content Manager
| OnDemand for iSeries by typing the
| following commands:
| CALL QRDARS/QRLCSTRJ PARM('RLC')
| CALL QRDARS/QRLCSTRJ PARM('RLR')
| CALL QRDARS/QRLCSTRJ PARM('RLO')
| Task 16 If you want the independent ASP Use the VRYCFG command or the
| available for system operations, you Operations Navigator interface to
| must vary it on now. vary on the independent ASP.
| Task 17 Review job logs or output from your “How to Verify That Objects Are
| restore operations to ensure that all Restored Successfully” on page 54.
| objects were restored successfully.

1
| If you have a 2440 Tape Unit with the high-speed feature enabled, you must disable it before restoring the
| Licensed Internal Code. See “Disabling and Enabling the High-Speed Feature on the 2440 Tape Unit” on
| page 145 for instructions. You may enable the feature again after you complete all the steps in Task 2.
|
|

104 OS/400 Backup and Recovery V5R1

| Restoring a Logical Partition to Another Logical
Partition—Checklist 23
This checklist should be used if you need to restore one logical partition to
another.

Before you begin your recovery, make a copy of this checklist. Fill in the
appropriate areas as you and the service representative perform the recovery steps.
This checklist provides an important record of your recovery actions. It may help
you to diagnose any problems that occur after the recovery. It may also be useful
in evaluating your backup strategy.

Most steps in the checklist include references to other topics in this book. Refer to
these topics if you need more information about how to perform a particular step.
You may not need to perform some steps, such as restoring changed objects, if they
do not apply in your situation.
Table 30. Recovery Checklist for Complete System Loss–Checklist 23
Task What To Do Where To Read More About It
Actions to Be Performed by the User
Task 1 Prepare to load the Licensed Internal “How to Prepare for Loading the
Code1. Licensed Internal Code” on page 122.
Task 2 Install the Licensed Internal Code “How to Prepare for Loading the
using option 3 (Install Licensed Licensed Internal Code” on page 122
Internal Code and Recover and “How to Load the Licensed
Configuration)1. Internal Code” on page 130
Task 3 Configure the disk units (assign to “Chapter 24. Procedures for
ASP and set up disk protection). If Configuring Disks and Disk
you saved any User-Defined File Protection” on page 643 and
Systems (UDFS), you must configure “Chapter 25. Working with Auxiliary
your user ASPs or the UDFS will not Storage Pools” on page 669.
restore.
Task 4 Restore the operating system, “How to Restore the OS/400
beginning with “Task 1–Starting to Licensed Program” on page 149.
Restore the Operating System” on
page 150. You are performing a
complete restore operation.
Task 5 If you restored the operating system “Recovering System Information” on
using distribution media, some page 213.
system information, such as access
path recovery times and the system
reply list may have been reset to
default values. Verify these values
and correct them if necessary.
Task 6 Recover user information from your “Choosing the Procedure to Recover
save media. Restore changed objects User Information” on page 107.
and apply journal entries. If you are
restoring to a different system, or a
different logical partition, you must
specify ALWOBJDIF(*ALL) on the
RSTxxx commands.

Chapter 4. Selecting the Right Recovery Strategy 105

Table 30. Recovery Checklist for Complete System Loss–Checklist 23 (continued)
Task What To Do Where To Read More About It
Task 7 If you are not sure what the “What Happens When You Restore
password is for the QSECOFR profile User Profiles” on page 215.
that was restored from tape, change
it before signing off: CHGUSRPRF
USRPRF(QSECOFR)
PASSWORD(new-password)
Task 8 If you restored from distribution “Recovering System Information” on
media, restore your system page 213.
information to the correct settings.
Task 9 Perform either a
SIGNOFF *LIST

or a
DSPJOBLOG * *PRINT

Check the job log to ensure that all

objects were restored. The job log
contains information about the
restore operation. To verify that all
objects were restored, spool the job
log for printing along with the job’s
remaining spooled output, if any.

Message CPC3703 is sent to the job

log for each library that was
successfully restored. Message
CPF3773 is sent to tell you how
many objects were restored. Objects
are not restored for various reasons.
Check for any error messages, correct
the error, and then restore those
objects from the media.
Task 10 You must perform a normal IPL at “How to Perform a Normal IPL” on
this time. page 58.
Task 11 If you use Windows server on iSeries “Completing Recovery for the iSeries
and saved with the Integrated xSeries Integration for Windows Server
Server (NWSD) varied on, complete Product” on page 263.
recovery for the Windows server on
iSeries product.
Task 12 If IBM Content Manager OnDemand
for iSeries is installed, complete
journaling for IBM Content Manager
OnDemand for iSeries by typing the
following commands:
CALL QRDARS/QRLCSTRJ PARM('RLC')
CALL QRDARS/QRLCSTRJ PARM('RLR')
CALL QRDARS/QRLCSTRJ PARM('RLO')
Task 13 Review job logs or output from your “How to Verify That Objects Are
restore operations to ensure that all Restored Successfully” on page 54.
objects were restored successfully.

106 OS/400 Backup and Recovery V5R1

Table 30. Recovery Checklist for Complete System Loss–Checklist 23 (continued)
Task What To Do Where To Read More About It

1
If you have a 2440 Tape Unit with the high-speed feature enabled, you must disable it before restoring the
Licensed Internal Code. See “Disabling and Enabling the High-Speed Feature on the 2440 Tape Unit” on
page 145 for instructions. You may enable the feature again after you complete all the steps in Task 2.

Choosing the Procedure to Recover User Information

Your first step in a recovery is to return your system to a normal operating
condition. This may require:
v Replacing hardware
v Restoring or installing the Licensed Internal Code
v Performing an IPL after the system ends abnormally

When your system is running normally, you are ready to recover user information.
Use Table 31 to determine the procedure you should follow. In the table, N/A in a
column means that the recovery procedure is the same, whether you respond yes
or no.
Table 31. Choosing the Correct Recovery Procedure for User Information
Do You Have Do You Want to
Are You Save SAVCHGOBJs or Use Menu
Recovering All Procedure Journals to Options to
ASPs? Used Apply? Recover? Recovery Procedure to Follow
Yes Commands N/A See note 1. “Recovering User Information Using
Commands–Checklist 24” on page 108
Yes Save menu No Yes “Using Option 21 from the Restore
option 21 Menu–Checklist 25” on page 112
Yes Save menu Yes N/A “Recovering User Information Using
option 21 Commands–Checklist 24” on page 108
Yes Save menu No No “Recovering User Information Using
option 21 Commands–Checklist 24” on page 108
Yes Save menu No Yes “Using Options 22 and 23 from the Restore
option 22 Menu–Checklist 26” on page 114

Save menu
option 23
Yes Save menu Yes N/A “Recovering User Information Using
option 22 Commands–Checklist 24” on page 108

Save menu
option 23
Yes Save menu No No “Recovering User Information Using
option 22 Commands–Checklist 24” on page 108

Save menu
option 23
Yes Save menu No Yes “Using Options 22 and 23 from the Restore
option 21 Menu–Checklist 26” on page 114

Save menu
option 23

Chapter 4. Selecting the Right Recovery Strategy 107

Table 31. Choosing the Correct Recovery Procedure for User Information (continued)
Do You Have Do You Want to
Are You Save SAVCHGOBJs or Use Menu
Recovering All Procedure Journals to Options to
ASPs? Used Apply? Recover? Recovery Procedure to Follow
Yes Save menu Yes N/A “Recovering User Information Using
option 21 Commands–Checklist 24”

Save menu
option 23
Yes Save menu No No “Recovering User Information Using
option 21 Commands–Checklist 24”

Save menu
option 23
Yes Operational N/A N/A “Recovering User Information Using Tapes
Assistant from Operational Assistant
Backup2 Backup–Checklist 27” on page 117
No Any N/A N/A “Recovering User Information Using
Commands–Checklist 24”

1
If you save using commands rather than menu options, you should recover using commands.
2
You have saved using either the RUNBCKUP command or the Run Backup menu.

Recovering User Information Using Commands–Checklist 24

This checklist shows the sequence of steps you should use to recover user
information using commands. You may need to perform some tasks more than
once. The correct steps for your situation depend on:
v How you saved your information.
v Whether you use journaling or whether applications you have purchased use
journaling.
v Whether you have document library objects.
v Whether you save changed objects.

Before you begin recovering user information, make a copy of this checklist. Fill
in the appropriate areas as you perform the recovery steps. This checklist provides
an important record of your recovery actions. It may help you diagnose any
problems that occur after the recovery. It may also be useful in evaluating your
backup strategy.

Most steps in the checklist include references to other topics in this book. Refer to
these topics if you need more information about how to perform a particular step.
You may not need to perform some steps, such as restoring changed objects and
applying journal changes, if they do not apply in your situation.

Restoring to a Different System or Different Logical Partition?

v You must specify ALWOBJDIF(*ALL) on the RSTxxx commands.
v You must specify SRM(*NONE) on the RSTCFG command.
v Network attributes may be reset to the IBM-supplied defaults.

108 OS/400 Backup and Recovery V5R1

 Table 32. Checklist for Recovering User Information Using Commands
Task What To Do Where To Read More About It
Task 1 If your system is not already in a restricted “Putting Your System in a Restricted
state, ensure that all users are off the system State” on page 45.
and that all jobs are ended. Then type ENDSBS
SBS(*ALL) OPTION(*CNTRLD) DELAY(600)1,2.
Task 2 If necessary, change the QALWOBJRST “Controlling Restoration of
system value. Write the old value here: Security-Sensitive Objects” on page 50.
______________
| Task 3 If necessary, change the QVFYOBJRST “Controlling Restoration of
system value. Write the old value here: Security-Sensitive Objects” on page 50.
______________
Task 4 If necessary, change the system value that The System Values subtopic of the
controls whether the job log wraps when it is System Management topic in the
full. Use the Work with System Values iSeries Information Center.
command: WRKSYSVAL QJOBMSGQFL.
Write down the current value here:
______________. Then change the value to
*PRTWRAP.
Task 5 After changing the system values, sign off by
using the command SIGNOFF *LIST. Then,
using a newly created password, sign back
on as QSECOFR for the new values to take
effect.
Task 6 If restoring to a system with a different “How to Set the QMCHPOOL,
processor or memory, ensure the QBASPOOL, and QPFRADJ System
QMCHPOOL, QBASPOOL, and QPFRADJ Values for a Different Processor or
system values are correct by using the Memory” on page 53.
WRKSYSVAL command.
Task 7 Prevent messages that are not related to the
recovery from interrupting by typing:
CHGMSGQ MSGQ(QSYSOPR) DLVRY(*NOTIFY)
SEV(99)
What ENDOPT?

When you are restoring from tape, you tell the system whether or not to rewind the tape. If you are using tape in
the tasks that follow, specify ENDOPT(*LEAVE) when you have additional steps. Specify ENDOPT(*REWIND) for your last
step.
Task 8 Restore user profiles: RSTUSRPRF DEV(TAP01) “Restoring User Profiles” on page 214.
USRPRF(*ALL)
Task 9 Restore device configuration: RSTCFG “How to Restore Configuration
OBJ(*ALL) OBJTYPE(*ALL) DEV(TAP01) Objects” on page 227.
Task 10 If you configured an independent ASP Use the vary configuration (VRYCFG)
through Operations Navigator at DST, you command or the Operations Navigator
need to verify the RESOURCE and vary on interface to vary off the independent
the independent ASP now. This will create a ASP.
directory for the independent ASP and
automatically mount the UDFS to that
directory.
Task 11 Once you vary on the independent ASP, you You can display where UDFS are
need to unmount ’/dev/iasp- mounted with the DSPUDFS
name/QDEFAULT.UDFS’ because you saved command. Then use the UNMOUNT
the independent ASP when it was varied off. command.

Chapter 4. Selecting the Right Recovery Strategy 109

 Table 32. Checklist for Recovering User Information Using Commands (continued)
Task What To Do Where To Read More About It
Task 12 Restore the libraries to each ASP that you are “Restoring Libraries” on page 231.
recovering. If you choose not to restore all of
your libraries at this time, ensure that you
restore the QGPL and QUSRSYS libraries
along with the libraries that you are
restoring. When recovering the entire system,
there is no need to restore QGPL and
QUSRSYS libraries first.
Task 13 Restore the ownership for DLOs in the user “Task 8–Reclaiming Document Library
ASPs you are restoring. Objects” on page 188.
Task 14 Restore your last complete save of document “Restoring Documents and Folders” on
library objects to each user ASP you are page 255.
recovering.
Task 15 Restore your last complete save of “Restoring Objects in Directories” on
directories.3 page 261.
Task 16 If you have User-Defined File Systems in “Task 7–Restoring User-Defined File
User ASPs that do not restore correctly, you Systems to the User Auxiliary Storage
may need to perform additional recovery Pool” on page 187.
steps.
Task 17 Restore changed objects and apply journaled Chapter 11. How to Restore Changed
changes. Objects and Apply Journaled Changes,
task 1 through task 7.
Task 18 If you use Cryptographic Access Provider, “Recovery Considerations for
install the Cryptographic Access Provider Cryptographic Access Provider” on
licensed program (5769AC1), (5769AC2), or page 60
(5769AC3) using option 11 on the Work with
Licensed Programs menu.
Task 19 Restore authority. Type: RSTAUT “Restoring Object Authorities” on
page 219.
Task 20 Reapply any PTFs that were applied since “How to Restore Program Temporary
your last SAVSYS operation. Fixes” on page 278.
Task 21 If you use Windows server on iSeries and “Completing Recovery for the iSeries
saved with the Integrated xSeries Server Integration for Windows Server
(NWSD) varied off, complete recovery for Product” on page 263.
the Windows server on iSeries.
Task 22 If you run Linux and saved Linux by “Recovering Linux in a Partition” on
varying off the network server description page 264
(NWSD) for Linux, complete the recovery for
Linux.
Task 23 If necessary, change the QALWOBJRST “Controlling Restoration of
system value back to its original value by Security-Sensitive Objects” on page 50.
using the WRKSYSVAL command.
| Task 24 If necessary, change the QVFYOBJRST “Controlling Restoration of
system value back to its original value by Security-Sensitive Objects” on page 50.
using the WRKSYSVAL command.
Task 25 If necessary, change the QJOBMSGQFL The System Values subtopic of the
system value back to its original value by System Management topic in the
using the WRKSYSVAL command. iSeries Information Center.
Task 26 If you are recovering from a complete system
loss, return to the appropriate checklist.
Continue with the tasks on that checklist.

110 OS/400 Backup and Recovery V5R1

Table 32. Checklist for Recovering User Information Using Commands (continued)
Task What To Do Where To Read More About It
Task 27 Perform either a:
SIGNOFF *LIST

or a
DSPJOBLOG * *PRINT

Check the job log to ensure all objects were

restored. The job log contains information
about the restore operation. To verify that all
objects were restored, spool the job log for
printing along with the job’s remaining
spooled output, if any.

Message CPC3703 is sent to the job log for

each library that was successfully restored.
Message CPF3773 is sent to tell you how
many objects were restored. Objects are not
restored for various reasons. Check for any
error messages, correct the errors, and then
restore those objects from the media.
Task 28 You must perform a normal IPL at this “How to Perform a Normal IPL” on
time. page 58.
Task 29 If you use Windows server on iSeries and “Completing Recovery for the iSeries
saved with the Integrated xSeries Server Integration for Windows Server
(NWSD) varied on, complete recovery for the Product” on page 263.
Windows server on iSeries.
Task 30 If IBM Content Manager OnDemand for
iSeries is installed, complete journaling for
IBM Content Manager OnDemand for iSeries
by typing the following commands:
CALL QRDARS/QRLCSTRJ PARM('RLC')
CALL QRDARS/QRLCSTRJ PARM('RLR')
CALL QRDARS/QRLCSTRJ PARM('RLO')
Task 31 Review job logs or output from your restore “How to Verify That Objects Are
operations to ensure that all objects were Restored Successfully” on page 54.
restored successfully.

1
Your system must be in a restricted state to restore user profiles. Other steps in the recovery may not
require a restricted state. However, to ensure the success of your recovery and better performance when
you are restoring information, a restricted state is recommended.
2
For the delay parameter, specify a number of seconds that allows your system time to bring most jobs to a
normal end. On a large, busy system, you may need a longer delay.
3
You may receive one of the following messages:
CPD377A: Object not restored, /QNTC.
CPD377A: Object not restored, /QNetWare.

These objects cannot be restored until their file systems have been mounted during the IPL. The additional
recovery tasks will guide you through the steps to restore these objects.
Note: Since the OS/400 Enhanced Integration for Novell NetWare resides on a remote server, you do not
have to restore your Netware data when you restore your server. Previously, the OS/400 Integration for
Novell NetWare product ran on a Integrated xSeries Server and you had to restore the Novell product if
you completely restored your server.

Chapter 4. Selecting the Right Recovery Strategy 111

 Using Option 21 from the Restore Menu–Checklist 25
This checklist shows the sequence of steps you should use to recover user
information using option 21 from the Restore menu. Option 21 restores your
system to your last complete save.

Before you begin recovering user information, make a copy of this checklist. Fill
in the appropriate areas as you perform the recovery steps. This checklist provides
an important record of your recovery actions. It may help you diagnose any
problems that occur after the recovery. It may also be useful in evaluating your
backup strategy.

Most steps in the checklist include references to other topics in this book. Refer to
these topics if you need more information about how to perform a particular step.
You may not need to perform some steps, such as restoring changed objects and
applying journal changes, if they do not apply in your situation.

Restoring to a Different System?

If you are restoring to a different system or to a different logical partition, be
aware of the following:
v You must specify ALWOBJDIF(*ALL) on the RSTxxx commands.
v You must specify SRM(*NONE) on the RSTCFG command.
v Network attributes may be reset to the IBM-supplied defaults.

Note: An option is available on the restore menu that indicates that you are
restoring to a different system. If you selected this option, the system
automatically specifies the first two items that are listed above. You
should also specify this option if you are restoring to a different logical
partition.

Table 33. Checklist for Recovering User Information Using Option 21

Task What To Do Where To Read More About It
Task 1 If necessary, change the QALWOBJRST “Controlling Restoration of
system value. Write the old value here: Security-Sensitive Objects” on page 50.
______________
| Task 2 If necessary, change the QVFYOBJRST “Controlling Restoration of
| system value. Write the old value here: Security-Sensitive Objects” on page 50.
| ______________
Task 3 If necessary, change the system value that The System Values subtopic of the
controls whether the job log wraps when it is System Management topic in the
full. Use the Work with System Values iSeries Information Center.
command: WRKSYSVAL QJOBMSGQFL.
Write down the current value here:
______________. Then change the value to
*PRTWRAP.
Task 4 After changing the system values, sign off by
using the command SIGNOFF *LIST. Then,
using a newly created password, sign back
on as QSECOFR for the new values to take
effect.

112 OS/400 Backup and Recovery V5R1

 Table 33. Checklist for Recovering User Information Using Option 21 (continued)
Task What To Do Where To Read More About It
Task 5 If restoring to a system with a different “How to Set the QMCHPOOL,
processor or memory, ensure the QBASPOOL, and QPFRADJ System
QMCHPOOL, QBASPOOL, and QPFRADJ Values for a Different Processor or
system values are correct by using the Memory” on page 53.
WRKSYSVAL command.
Task 6 Perform option 21 from the Restore menu. “How to Use Restore Menu Options 21,
Use your most recent tapes from performing 22, and 23” on page 208.
option 21 on the Save menu. If you are
recovering using “Recovering your entire
system after a complete system
loss–Checklist 20” on page 96 and restoring
to a different system, use the ″Restore to
different system″ option on the Specify
Command Defaults display. You should
also use this option if you restoring to a
different logical partition. This option will
automatically specify ALWOBJDIF(*ALL)
on the RSTxxx commands and
SRM(*NONE) on the RSTCFG command.1
Task 7 If you use Windows server on iSeries and “Completing Recovery for the iSeries
saved with the Integrated xSeries Server Integration for Windows Server
(NWSD) varied off, complete recovery for Product” on page 263.
the Windows server on iSeries product.
Task 8 If you run Linux and saved Linux by “Recovering Linux in a Partition” on
varying off the network server description page 264
(NWSD) for Linux, complete the recovery for
Linux.
Task 9 If you have User-Defined File Systems in “Task 7–Restoring User-Defined File
User ASPs that do not restore correctly, you Systems to the User Auxiliary Storage
may need to perform additional recovery Pool” on page 187.
steps.
Task 10 Reapply any PTFs that were applied since “How to Restore Program Temporary
your last SAVSYS operation. Fixes” on page 278.
Task 11 If necessary, change the QALWOBJRST “Controlling Restoration of
system value back to its original value by Security-Sensitive Objects” on page 50.
using the WRKSYSVAL command.
| Task 12 If necessary, change the QVFYOBJRST “Controlling Restoration of
system value back to its original value by Security-Sensitive Objects” on page 50.
using the WRKSYSVAL command.
Task 13 If necessary, change the QJOBMSGQFL The System Values subtopic of the
system value back to its original value by System Management topic in the
using the WRKSYSVAL command. iSeries Information Center.
Task 14 If you are recovering from a complete system
loss, return to “Recovering your entire
system after a complete system
loss–Checklist 20” on page 96. Continue with
task 7 on that checklist.
Task 15 You must perform a normal IPL at this “How to Perform a Normal IPL” on
time. page 58.
Task 16 If you use Windows server on iSeries and “Completing Recovery for the iSeries
saved with the Integrated xSeries Server Integration for Windows Server
(NWSD) varied on, complete recovery for the Product” on page 263.
Windows server on iSeries product.

Chapter 4. Selecting the Right Recovery Strategy 113

 Table 33. Checklist for Recovering User Information Using Option 21 (continued)
Task What To Do Where To Read More About It
Task 17 If IBM Content Manager OnDemand for
iSeries is installed, complete journaling for
IBM Content Manager OnDemand for iSeries
by typing the following commands:
CALL QRDARS/QRLCSTRJ PARM('RLC')
CALL QRDARS/QRLCSTRJ PARM('RLR')
CALL QRDARS/QRLCSTRJ PARM('RLO')
Task 18 Review job logs or output from your restore “How to Verify That Objects Are
operations to ensure that all objects were Restored Successfully” on page 54.
restored successfully.

1
You may receive one of the following messages:
CPD377A: Object not restored, /QNTC.
CPD377A: Object not restored, /QNetWare.

These objects cannot be restored until their file systems have been mounted during the IPL. The additional
recovery tasks will guide you through the steps to restore these objects.
Note: Since the OS/400 Enhanced Integration for Novell NetWare resides on a remote server, you do not
have to restore your Netware data when you restore your server. Previously, the OS/400 Integration for
Novell NetWare product ran on a Integrated xSeries Server and you had to restore the Novell product if
you completely restored your server.

Using Options 22 and 23 from the Restore Menu–Checklist 26

This checklist shows the sequence of steps you should use to recover user
information using option 22 and 23 from the restore menu. Option 22 restores your
IBM-supplied libraries to your last save. Option 23 restores your user libraries to
your last save.

Before you begin recovering user information, make a copy of this checklist. Fill
in the appropriate areas as you perform the recovery steps. This checklist provides
an important record of your recovery actions. It may help you diagnose any
problems that occur after the recovery. It may also be useful in evaluating your
backup strategy.

Most steps in the checklist include references to other topics in this book. Refer to
these topics if you need more information about how to perform a particular step.
You may not need to perform some steps, such as restoring changed objects and
applying journal changes, if they do not apply in your situation.
Table 34. Checklist for Recovering User Information Using Options 22 and 23
Task What To Do Where To Read More About It
Task 1 If necessary, change the QALWOBJRST “Controlling Restoration of
system value back to its original value by Security-Sensitive Objects” on page 50.
using the WRKSYSVAL command. Write the
old value here: ______________
| Task 2 If necessary, change the QVFYOBJRST “Controlling Restoration of
| system value back to its original value by Security-Sensitive Objects” on page 50.
| using the WRKSYSVAL command. Write the
| old value here: ______________

114 OS/400 Backup and Recovery V5R1

Table 34. Checklist for Recovering User Information Using Options 22 and 23 (continued)
Task What To Do Where To Read More About It
Task 3 If necessary, change the system value that The System Values subtopic of the
controls whether the job log wraps when it is System Management topic in the
full. Use the Work with System Values iSeries Information Center.
command: WRKSYSVAL QJOBMSGQFL.
Write down the current value here:
______________. Then change the value to
*PRTWRAP.
Task 4 After changing the system values, sign off by
using the command SIGNOFF *LIST. Then,
using a newly created password, sign back
on as QSECOFR for the new values to take
effect.
Task 5 If restoring to a system with a different “How to Set the QMCHPOOL,
processor or memory, ensure the QBASPOOL, and QPFRADJ System
QMCHPOOL, QBASPOOL, and QPFRADJ Values for a Different Processor or
system values are correct by using the Memory” on page 53.
WRKSYSVAL command.
Task 6 Perform option 22 from the Restore menu to “How to Use Restore Menu Options 21,
restore IBM-supplied libraries and 22, and 23” on page 208.
directories. Use your most recent tapes from
performing either option 21 or option 22 on
the Save menu. If you are recovering using
“Recovering your entire system after a
complete system loss–Checklist 20” on
page 96 and restoring to a different system,
use the ″Restore to different system″ option
on the Specify Command Defaults display.
You should also use this option if you
restoring to a different logical partition.
This option will automatically specify
ALWOBJDIF(*ALL) on the RSTxxx
commands and SRM(*NONE) on the
RSTCFG command.1
Task 7 Perform option 23 from the Restore menu to “How to Use Restore Menu Options 21,
restore user libraries and user directories. 22, and 23” on page 208.
Use your most recent tapes from performing
either option 21 or option 23 on the Save
menu. If you are recovering using
“Recovering your entire system after a
complete system loss–Checklist 20” on
page 96 and restoring to a different system,
use the ″Restore to different system″ option
on the Specify Command Defaults display.
You should also use this option if you
restoring to a different logical partition.
This option will automatically specify
ALWOBJDIF(*ALL) on the RSTxxx
commands and SRM(*NONE) on the
RSTCFG command.1
Task 8 If you use Windows server on iSeries and “Completing Recovery for the iSeries
saved with the Integrated xSeries Server Integration for Windows Server
(NWSD) varied off, complete recovery for Product” on page 263.
the Windows server on iSeries product.

Chapter 4. Selecting the Right Recovery Strategy 115

 Table 34. Checklist for Recovering User Information Using Options 22 and 23 (continued)
Task What To Do Where To Read More About It
Task 9 If you run Linux and saved Linux by “Recovering Linux in a Partition” on
varying off the network server description page 264
(NWSD) for Linux, complete the recovery for
Linux.
Task 10 If you have User-Defined File Systems in “Task 7–Restoring User-Defined File
User ASPs that do not restore correctly, you Systems to the User Auxiliary Storage
may need to perform additional recovery Pool” on page 187.
steps.
Task 11 Reapply any PTFs that were applied since “How to Restore Program Temporary
your last SAVSYS operation. Fixes” on page 278.
Task 12 If necessary, change the QALWOBJRST “Controlling Restoration of
system value. Security-Sensitive Objects” on page 50
on 51.
| Task 13 If necessary, change the QVFYOBJRST “Controlling Restoration of
system value. Security-Sensitive Objects” on page 50
on 51.
Task 14 If necessary, change the QJOBMSGQFL The System Values subtopic of the
system value back to its original value by System Management topic in the
using the WRKSYSVAL command. iSeries Information Center.
Task 15 Perform either a:
SIGNOFF *LIST

or a
DSPJOBLOG * *PRINT

Check the job log to ensure all objects were

restored. The job log contains information
about the restore operation. To verify that all
objects were restored, spool the job log for
printing along with the job’s remaining
spooled output, if any.

Message CPC3703 is sent to the job log for

each library that was successfully restored.
Message CPF3773 is sent to tell you how
many objects were restored. Objects are not
restored for various reasons. Check for any
error messages, correct the errors, and then
restore those objects from the media.
Task 16 You must perform a normal IPL at this “How to Perform a Normal IPL” on
time. page 58.
Task 17 If you use Windows server on iSeries and “Completing Recovery for the iSeries
saved with the Integrated xSeries Server Integration for Windows Server
(NWSD) varied on, complete recovery for the Product” on page 263.
Windows server on iSeries product.
Task 18 If IBM Content Manager OnDemand for
iSeries is installed, complete journaling for
IBM Content Manager OnDemand for iSeries
by typing the following commands:
CALL QRDARS/QRLCSTRJ PARM('RLC')
CALL QRDARS/QRLCSTRJ PARM('RLR')
CALL QRDARS/QRLCSTRJ PARM('RLO')

116 OS/400 Backup and Recovery V5R1

Table 34. Checklist for Recovering User Information Using Options 22 and 23 (continued)
Task What To Do Where To Read More About It
Task 19 Review job logs or output from your restore “How to Verify That Objects Are
operations to ensure that all objects were Restored Successfully” on page 54.
restored successfully.

1
You may receive one of the following messages:
CPD377A: Object not restored, /QNTC.
CPD377A: Object not restored, /QNetWare.

These objects cannot be restored until their file systems have been mounted during the IPL. The additional
recovery tasks will guide you through the steps to restore these objects.
Note: Since the OS/400 Enhanced Integration for Novell NetWare resides on a remote server, you do not
have to restore your Netware data when you restore your server. Previously, the OS/400 Integration for
Novell NetWare product ran on a Integrated xSeries Server and you had to restore the Novell product if
you completely restored your server.

Recovering User Information Using Tapes from Operational

Assistant Backup–Checklist 27
This checklist shows the sequence of steps you should use to recover user
information when you have saved using Operational Assistant backup. These
procedures assume that all of your backup is done using Operational Assistant.
You have not mixed Operational Assistant backup with other save methods.

Before you begin recovering user information, make a copy of this checklist. Fill
in the appropriate areas as you perform the recovery steps. This checklist provides
an important record of your recovery actions. It may help you diagnose any
problems that occur after the recovery. It may also be useful in evaluating your
backup strategy.

Most steps in the checklist include references to other topics in this book. Refer to
these topics if you need more information about how to perform a particular step.
You may not need to perform some steps, such as restoring changed objects and
applying journal changes, if they do not apply in your situation.

Restoring to a Different System?

If you are restoring to a different system or to a different logical partition, be
aware of the following:
v You must specify ALWOBJDIF(*ALL) on the RSTxxx commands.
v You must specify SRM(*NONE) on the RSTCFG command.
v Network attributes are reset to the IBM-supplied defaults.

Table 35. Checklist for Recovering User Information Using Operational Assistant Backup Tapes
Task What To Do Where To Read More About It
Task 1 If your system is operational and the
QUSRSYS library is on the system, print the
Backup Status and the Backup History by
typing: DSPBCKSTS OUTPUT(*PRINT).

Chapter 4. Selecting the Right Recovery Strategy 117

 Table 35. Checklist for Recovering User Information Using Operational Assistant Backup Tapes (continued)
Task What To Do Where To Read More About It
Task 2 If your system is operational and the
QUSRSYS library is on the system, print the
Backup List by typing: DSPBCKUPL
OUTPUT(*PRINT).
Task 3 If your system is not already in a restricted “Putting Your System in a Restricted
state, ensure all users are off the system. State” on page 45.
Then type ENDSBS SBS(*ALL)
OPTION(*CNTRLD) DELAY(600)1,2.
| Task 4 If necessary, change the QVFYOBJRST “Controlling Restoration of
| system value back to its original value by Security-Sensitive Objects” on page 50.
| using the WRKSYSVAL command. Write the
| old value here: ______________
Task 5 If necessary, change the QALWOBJRST “Controlling Restoration of
system value back to its original value by Security-Sensitive Objects” on page 50.
using the WRKSYSVAL command. Write the
old value here: ______________
Task 6 If necessary, change the system value that The System Values subtopic of the
controls whether the job log wraps when it is System Management topic in the
full. Use the Work with System Values iSeries Information Center.
command: WRKSYSVAL QJOBMSGQFL.
Write down the current value here:
______________ Then change the value to
*PRTWRAP.
Task 7 After changing the system values, sign off by
using the command SIGNOFF *LIST. Then,
using a newly created password, sign back
on as QSECOFR for the new values to take
effect.
Task 8 If restoring to a system with a different “How to Set the QMCHPOOL,
processor or memory, ensure the QBASPOOL, and QPFRADJ System
QMCHPOOL, QBASPOOL, and QPFRADJ Values for a Different Processor or
system values are correct by using the Memory” on page 53.
WRKSYSVAL command.
Task 9 Prevent messages that are not related to the
recovery from interrupting by typing:
CHGMSGQ MSGQ(QSYSOPR) DLVRY(*NOTIFY)
SEV(99)
Task 10 Restore user profiles: RSTUSRPRF DEV(TAP01) “Restoring User Profiles” on page 214.
USRPRF(*ALL).
Task 11 Restore device configuration: RSTCFG “How to Restore Configuration
OBJ(*ALL) OBJTYPE(*ALL) DEV(TAP01) Objects” on page 227.
Task 12 Restore the libraries to each ASP that you are “How to Restore Your Libraries” on
recovering. If you choose not to restore all of page 306
your libraries at this time, ensure that you
restore the QGPL and QUSRSYS libraries
along with the libraries that you are
restoring. When recovering the entire system,
there is no need to restore QGPL and
QUSRSYS libraries first.
Task 13 Restore the ownership for DLOs in the user “Task 8–Reclaiming Document Library
ASPs that you are restoring. Objects” on page 188.

118 OS/400 Backup and Recovery V5R1

 Table 35. Checklist for Recovering User Information Using Operational Assistant Backup Tapes (continued)
Task What To Do Where To Read More About It
Task 14 Restore your last complete save of document “Restoring Documents and Folders” on
library objects to each user ASP that you are page 255.
recovering.
Task 15 Restore your last complete save of “Restoring Objects in Directories” on
directories3. page 261.
Task 16 Restore incremental backups of libraries. “How to Restore Libraries That You
Saved by Using a Backup List” on
page 307.
Task 17 Restore changed objects. “How to Restore Changed Objects That
You Saved by Using Operational
Assistant” on page 308.
Task 18 If you use Cryptographic Access Provider, “Recovery Considerations for
install the Cryptographic Access Provider Cryptographic Access Provider” on
licensed program (5769AC1), (5769AC2), or page 60
(5769AC3) using option 11 on the Work with
Licensed Programs menu.
Task 19 Restore authority. Type: RSTAUT “Restoring Object Authorities” on
page 219.
Task 20 If you have User-Defined File Systems in “Task 7–Restoring User-Defined File
User ASPs that do not restore correctly, you Systems to the User Auxiliary Storage
may need to perform additional recovery Pool” on page 187.
steps.
Task 21 If you use Windows server on iSeries and “Completing Recovery for the iSeries
saved with the Integrated xSeries Server Integration for Windows Server
(NWSD) varied off, complete recovery for Product” on page 263.
the Windows server on iSeries product.
Task 22 If you run Linux and saved Linux by “Recovering Linux in a Partition” on
varying off the network server description page 264
(NWSD) for Linux, complete the recovery for
Linux.
Task 23 If necessary, change the QALWOBJRST “Controlling Restoration of
system value. Security-Sensitive Objects” on page 50.
| Task 24 If necessary, change the QVFYOBJRST “Controlling Restoration of
system value. Security-Sensitive Objects” on page 50.
Task 25 If necessary, change the QJOBMSGQFL The System Values subtopic of the
system value back to its original value by System Management topic in the
using the WRKSYSVAL command. iSeries Information Center.

Chapter 4. Selecting the Right Recovery Strategy 119

Table 35. Checklist for Recovering User Information Using Operational Assistant Backup Tapes (continued)
Task What To Do Where To Read More About It
Task 26 Perform either a:
SIGNOFF *LIST

or a
DSPJOBLOG * *PRINT

Check the job log to ensure all objects were

restored. The job log contains information
about the restore operation. To verify that all
objects were restored, spool the job log for
printing along with the job’s remaining
spooled output, if any.

Message CPC3703 is sent to the job log for

each library that was successfully restored.
Message CPF3773 is sent to tell you how
many objects were restored. Objects are not
restored for various reasons. Check for any
error messages, correct the errors, and then
restore those objects from the media.
Task 27 You must perform a normal IPL at this “How to Perform a Normal IPL” on
time. page 58.
Task 28 If you use Windows server on iSeries and “Completing Recovery for the iSeries
saved with the Integrated xSeries Server Integration for Windows Server
(NWSD) varied on, complete recovery for the Product” on page 263.
Windows server on iSeries product.
Task 29 If IBM Content Manager OnDemand for
iSeries is installed, complete journaling for
IBM Content Manager OnDemand for iSeries
by typing the following commands:
CALL QRDARS/QRLCSTRJ PARM('RLC')
CALL QRDARS/QRLCSTRJ PARM('RLR')
CALL QRDARS/QRLCSTRJ PARM('RLO')
Task 30 Review job logs or output from your restore “How to Verify That Objects Are
operations to ensure that all objects were Restored Successfully” on page 54.
restored successfully.

1
Your system must be in a restricted state to restore user profiles. Other steps in the recovery may not
require a restricted state. However, to ensure the success of your recovery and better performance when
you are restoring information, a restricted state is recommended.
2
For the delay parameter, specify a number of seconds that allows your system time to bring most jobs to a
normal end. On a large, busy system, you may need a longer delay.
3
You may receive one of the following messages:
CPD377A: Object not restored, /QNTC.
CPD377A: Object not restored, /QNetWare.

These objects cannot be restored until their file systems have been mounted during the IPL. The additional
recovery tasks will guide you through the steps to restore these objects.
Note: Since the OS/400 Enhanced Integration for Novell NetWare resides on a remote server, you do not
have to restore your Netware data when you restore your server. Previously, the OS/400 Integration for
Novell NetWare product ran on a Integrated xSeries Server and you had to restore the Novell product if
you completely restored your server.

120 OS/400 Backup and Recovery V5R1

Chapter 5. Recovering the Licensed Internal Code
Licensed Internal Code is the layer of iSeries server architecture just above the
hardware. You must have the Licensed Internal Code on your machine before you
can restore the operating system. You must use the control panel on your system
unit to start the recovery of the Licensed Internal Code.

The Install Licensed Internal Code (LIC) menu provides several methods for
loading the Licensed Internal Code to your system. Table 36 describes the options
and how they are used:
Table 36. Options from the Install the Licensed Internal Code (LIC) Menu
Option Number Description Purpose
1 Restore Licensed Internal Restores the Licensed Internal Code without removing other
Code information that is on the system. Option 1 is similar to Function
Code 23 on earlier versions of the iSeries or AS/400e server.
Option 1 is normally used in the following situations:
v You are encountering problems with the operating system, such
as damaged objects. You sometimes need to restore the
Licensed Internal Code before restoring the operating system.
v The software support center recommends it.
v You have replaced a failed disk unit other than unit 1 in the
system ASP.
v You are updating your system to a new release. See the Software
Installation book for the procedures to install a new release of
the iSeries server.
2 Install the Licensed Internal Installs the Licensed Internal Code and removes all data from all
Code and Initialize system disk units. Option 2 is similar to Function Code 24 on earlier
versions of the iSeries or AS/400e server. Option 2 is normally
used in the following situations:
v You are doing a restore operation using the SAVSTG media.
v You are restoring to another system to recover from a complete
system loss.
v You are recovering with SAVSYS media that is at a previous
release than what is currently installed on the system.

If a system is configured to use Operations Console, and that

system undergoes a backup and recovery cycle, you will have to
perform the following steps:
1. Perform an initial program load (IPL) in Manual mode.
2. Use dedicated service tools (DST) to reconfigure the system so
that it will detect the PC console when you perform an IPL in
Normal mode.
Detailed instructions on setting up Operations Console are in the
document Operations Console Setup.

© Copyright IBM Corp. 1997, 2000, 2001 121

Table 36. Options from the Install the Licensed Internal Code (LIC) Menu (continued)
Option Number Description Purpose
3 Install Licensed Internal Code Installs the Licensed Internal Code and prompts you to begin the
and Recover Configuration procedure to recover information about how the disks were
configured on your system (including ASP assignments and
protection). Option 3 is similar to Function Code 24 on earlier
version of the iSeries or AS/400e server. Option 3 is normally
used in the following situations:
v You have replaced the load source unit.
v The software support center recommends it.
4 Install Licensed Internal Code Installs the Licensed Internal Code and restores data to a
and Restore Disk Unit Data replacement disk unit. This option is used only by a service
representative after data was successfully saved (pumped) from a
failed load source disk unit.
5 Install Licensed Internal Code This option is used in certain cases when you are changing to a
and Upgrade Load Source PowerPC based system. See the AS/400 Road Map for Changing to
PowerPC Technology for more information.

The recovery checklists in Chapter 4 specify which procedures in this chapter are
required for your situation.

Attention!
Make sure you use the correct procedure for your situation. Some of the
procedures in this chapter will remove all data from your system.

How to Prepare for Loading the Licensed Internal Code

The tasks for starting to load the Licensed Internal Code include the following:
v Finding the right media and documentation.
v Stopping your system, if it is running.
v Performing an IPL from an alternate device, either tape or optical media.
Check off each step in these tasks as you complete it.

Task 1–Getting Ready to Load the Licensed Internal Code

Find These Things Before You Begin:
v Your most recent SAVSYS media. One of the following creates a SAVSYS media
image:
– Running the Save System (SAVSYS) command.
– Using option 21 from the Save menu.
– Using option 22 from the Save menu.
– Using option 11 from the Run Backup menu.

122 OS/400 Backup and Recovery V5R1

 Attention!
DO NOT use a media volume that you created through DST by using
option 5=Save Licensed Internal Code from the IPL or Install the System
menu unless you have been instructed to do so by Software Services. A
media volume that is created through this process does not contain the
Licensed Internal Code PTF Inventory information or the OS/400
Operating System. If you perform the recovery process using this media
volume, you will have to re-install the Licensed Internal Code from either a
SAVSYS media volume or from your distribution media before you can
load any PTFs onto the system.

v If you enabled your device as an alternate installation device, you will need the
Licensed Internal Code CD-ROM. (An alternate installation device is an alternate
IPL device that is connected to a bus other than the system bus (bus 1).) Refer to
“Chapter 23. Using an Alternate Installation Device” on page 635 for more
information.
v If you do not have current SAVSYS media or they are damaged, you need the
following:
– The distribution media (optical media or tape) that is supplied by IBM.
– All the optical media for program temporary fixes you have applied. Use the
distribution media only if you do not have SAVSYS media. If you use the
distribution media to restore the Licensed Internal Code, you will lose some
of your system information, such as the program temporary fixes you have
applied.
v The list of all the program temporary fixes (PTFs) applied to your system at the
time you saved the entire system. This list should be attached to your backup
log or found with your SAVSYS media.
v The keystick for the system if it is not already inserted in the control panel.
v The manual for the tape or optical device that is your alternate IPL device. It
describes other SRC codes you might see.

Do These Things Before You Begin:

v Clean the read and write heads of the tape unit if you are using a tape device.
v If your source system (the system that was saved and needs recovery) is
operational, print a list of all the program temporary fixes (PTFs) currently on
the system. Type the following and press the Enter key:
DSPPTF LICPGM(*ALL) OUTPUT(*PRINT)

Task 2–Powering Down the System

Using a 2440 Tape Unit?

Be sure to disable the high-speed feature before restoring the Licensed
Internal Code. See “Disabling and Enabling the High-Speed Feature on the
2440 Tape Unit” on page 145 for instructions.

Attention: If you are loading the Licensed Internal Code in a secondary partition,
you do not need to power down the system.

Chapter 5. Recovering the Licensed Internal Code 123

 If your system is already powered off or if you are recovering to a system at an
IBM Business Recovery Services Center, skip this task and begin with “Task
3a–Preparing the System to Perform an IPL from an Alternate Device”. It is not
necessary to power down a system that has no activity on the system.
1. Sign on the system as QSECOFR.
2. Change the QSYSOPR message queue:
CHGMSGQ MSGQ(QSYSOPR) DLVRY(*BREAK)
SEV(60)
3. Bring your system to a restricted state:
ENDSBS SBS(*ALL) OPTION(*CNTRLD)
DELAY(600)

Note: Specify a number of seconds for the delay parameter that is long enough
for your system to bring most jobs to a normal end. On a large, busy
system, you may need more time.

The system displays a message that subsystem ending is in progress. The

system displays another message when all subsystems have ended and the
system is in a restricted state. After the subsystems have ended, continue with
the next step.
4. Change the QSYSOPR message queue:
CHGMSGQ MSGQ(QSYSOPR) DLVRY(*BREAK)
SEV(99)

Note: Communications messages with a severity of 99 that require a reply can

interrupt your restore operation. You can either identify these messages
and add them to the system reply list, or you can change the delivery of
the QSYSOPR message queue to *NOTIFY. This prevents the
communications messages from interrupting the interactive restore
operation.
5. Power down the system:
PWRDWNSYS OPTION(*IMMED)

Attention Logical Partitioning users!

Before issuing this command, be certain that all secondary partitions have
been powered off.

Note: When the Power On light goes off, continue with the next task.

Task 3a–Preparing the System to Perform an IPL from an

Alternate Device
To perform an initial program load (IPL) from tape, optical media, or CD-ROM,
you must use the control panel on the system unit. The steps vary slightly based
| on the type of system unit that you have. Click on Getting started with iSeries
| under the System planning and installation topic in the Information Center if you
| are unsure of the procedures for your system. You can find the Information Center
| at the following Web site:
http://www.ibm.com/eserver/iseries/infocenter/

124 OS/400 Backup and Recovery V5R1

| Note: This task only applies to the primary partition. If you are installing to a
| secondary partition, use the Work with system partitions option in SST or
| DST on the primary partition. See the Information Center for more
| information about logical partitions.

Do the following:
1. If your system unit has a lock on the control panel, use the key or keystick to
unlock the control panel.
2. Place the system in Manual mode.
3. Press the Function Select switch (or buttons) to display 02 (IPL) in the Function
display.
4. Press the Enter button on the control panel.
5. Press the Function Select switch (or buttons) to display D (IPL from tape, optical
media, or CD-ROM) in the Data display.
6. Press the Enter button on the control panel.
7. Ensure that any switches for the alternate IPL device and for all disk units are
in the On position.

Task 3b-Preparing a Logical Partition to Perform an IPL from

an Alternate Device
To perform an IPL from tape, optical media, or CD-ROM, you must use the Work
with Partition Status display from the primary partition. Perform the following
steps on the primary partition:
1. Type STRSST on a command line, and press Enter.
2. On the System Service Tools (SST) display, select option 5 to work with system
partitions, and press Enter.
3. On the Work with System Partitions display, select option 2 to work with
system partition status, and press Enter.
4. On the Work with System Partition Status display, select manual mode by
typing a 10 in the option field next to the partition that you are performing the
alternate IPL on. Press Enter.
5. Select source D by typing a D in the option field next to the partition on which
you are performing the alternate IPL . Press Enter.

Task 4–Loading the Licensed Internal Code from Media

Note: If you are working in a secondary partition, keep the following
considerations in mind for this task:
v You may skip step 3 below (turn on system power), since you have not
powered off.
v In step 4, you are using the virtual control panel instead of the system
unit control panel.
v Instructions dealing with alternate installation do not apply to secondary
partitions. (You can install from any tape device or optical device in the
partition.)
1. Find the Licensed Internal Code tape or optical media. It is the first volume of
the most current set of SAVSYS media or the first volume of the distribution
optical media.

Chapter 5. Recovering the Licensed Internal Code 125

 Attention!
v DO NOT use save media that was created through DST by using
option 5=Save Licensed Internal Code from the IPL or Install the
System menu unless you have been instructed to do so by Software
Services. The save media that you create through this process does not
contain the Licensed Internal Code PTF Inventory information or the
OS/400 Operating System. If you perform the recovery process using
this save media, you will have to re-install the Licensed Internal Code
from either a SAVSYS media or from your distribution media before
you can load any PTFs onto the system.
v Use the distribution media only if no SAVSYS media volume exists. If
you use the distribution media, some system information will be lost.
This includes, but is not limited to, PTFs and PTF packages. If you use
the distribution media, you must reinstall all cumulative PTF packages
and individual PTFs applied after the initial installation of your
system.

2. Place the media volume in the device that you use for the IPL, or place the
optical media in the optical disk unit. When you start the IPL, the system
searches the alternate IPL devices for the correct media. For more information
on loading the tape or optical media, see the setup manual for the device.
Notes:
a. If you cannot load your alternate IPL device when the power is off,
continue with the next step. The system prompts you later with an SRC
code for the tape device or optical device.
b. If you use a tape device that you enabled as an alternate installation
device, you must load both the Licensed Internal Code CD-ROM and your
tape media. (An alternate installation device is an alternate IPL device that
is connected to a bus other than the system bus (bus 1).) Refer to
“Chapter 23. Using an Alternate Installation Device” on page 635 for more
information.
3. Turn on the power to the system.

Using Logical Partitions?

On the primary partition, go to the Work with System Partition Status
display. Turn the power on by typing a 1 in the option field next to the
partition on which you are performing the alternate IPL .

4. If you could not load your media volume in step 2, load the first media
volume into the device that you use for IPL. Make the device ready and then
continue with the next step.

Note: If you did not power down your system after ending the subsystems,
do the following:
a. Press the Function Select switch (or buttons) to display 03 (continue
the IPL) in the Function display on the control panel.
b. Press the Enter button on the control panel.

126 OS/400 Backup and Recovery V5R1

 Using Logical Partitions?
If you are performing an alternate IPL for a logical partition,
perform the following steps:
1) On the primary partition, go to the Work with System
Partition Status display. Select IPL restart by typing a 3 in
the option field next to the partition on which you are
performing the alternate IPL .
2) Press Enter.
3) You will be shown the Confirm Alternate IPL display. The
system denotes the selected alternate IPL device with a
percent (%) sign. If this is the correct alternate IPL device,
press Enter to continue with the IPL and continue with
step 5.
If there is no alternate IPL device defined, or if you want to
select a different alternate IPL device, press F11 (Select
alternate IPL resource). On the Select Alternate IPL
Resource display, type a 1 in the option field next to the
Storage IOP of the device that you want to select. Press the
Enter key. You will be shown the Confirm Alternate IPL
Resource display. Press Enter to confirm your choice. Press
F12 to return to the Confirm Alternate IPL display. Press
Enter to continue with the IPL.

5. Ensure that the tape device or optical device is online or ready. No action is
required for devices that perform this step automatically (such as the tape
cartridge unit).
6. Ensure that the console display is turned on. After a delay, you should see the
Install Licensed Internal Code menu. The length of the delay varies,
depending on your system configuration and the speed of your alternate IPL
device. The delay is usually between 5 minutes and 30 minutes. When you see
this menu, continue with step 7 on page 128.
If the system attention light appears and one of the SRC codes that are shown
in Table 37 is displayed in the Data display, complete the instructions for that
SRC code.

Note: If you are using logical partitions, the SRC codes will be shown from
the primary partition on the Work with Partition Status or the Monitor
Partition Status displays.
Table 37. SRC Codes When Loading the Licensed Internal Code
SRC Code Why It Appears What You Do
A1xx 1933 The device for the alternate IPL is not Make sure that you loaded the
A12x 1933 ready. correct media volume. Make the
(’x’ is any device unit ready. Wait for the
character) System Attention light to go off.
Then continue with the next step.
If the Sytem Attention light stays
on for more than 5 minutes, check
to see if you have the correct tape
loaded in the device for the
alternate IPL and make the device
ready. Then continue with the next
step.

Chapter 5. Recovering the Licensed Internal Code 127

 Table 37. SRC Codes When Loading the Licensed Internal Code (continued)
SRC Code Why It Appears What You Do
B1xx 1803 The device for the alternate IPL was not Make sure that you powered on
B1xx 1806 found or was not ready. the device, that you loaded the
B1xx 1938 correct media volume, and the
media volume is ready. Then
continue with the next step.
B1xx 1934 The wrong media volume is loaded. Or Load the correct media volume
the high-speed feature is enabled on the and make the device ready. Then
2440 Tape Unit. continue with the next step or
disable the high-speed feature on
the 2440 Tape Unit.
B608 1105 This SRC occurs when you exit from the Power down the system. If
automatic installation because an alternate necessary, fix the device. Power on
installation device attached to the system the system. Verify that the
is enabled but is otherwise not ready. The alternate installation device is
device may not be ready because the enabled. Load the media volume
media volume is not loaded or the device in the alternate installation device
may not be enabled as an alternate and start the installation procedure
installation device. Or, there may be a again.
problem with the device.
2507 0001 A media volume is not loaded in the Make sure the correct media
2642 0001 device for the alternate IPL. volume is loaded in the correct
2643 0001 device. Then continue with the
next step.
2644 3136 The device is assigned to another system. Go to the other system and vary
off the device. Make sure that you
loaded the correct media volume.
Then continue with the next step.
Any other The system encountered a problem
SRC loading the Licensed Internal Code.

If the System Attention light is lit and no SRC code appears on the control
panel, do the following:
a. Press the Function Select switch (or buttons) to display 03 (continue the
IPL) in the Function display on the control panel.
b. Press the Enter button on the control panel.
Then continue with the next step.
7. You are shown the Install Licensed Internal Code (LIC) display.
If you have an alternate installation device attached to the system, perform
Install Licensed Internal Code

Select one of the following:

1. Install Licensed Internal Code

2. Work with Dedicated Service Tools (DST)
3. Define alternate installation device

steps 8 through 10. If you do not have an alternate installation device attached
to the system, type a 1 and press the Enter key.

128 OS/400 Backup and Recovery V5R1

 Stop!
You are now ready to recover your Licensed Internal Code. Consult your
recovery checklist before continuing. The checklist tells you the correct
option to select from the Install Licensed Internal Code (LIC) display.

8. If you have an alternate installation device attached to the system, type a 3 to

verify its address and determine whether it is enabled or disabled.
9. The Select Alternate Installation Device Bus screen appears.
The Selected column shows the bus where the alternate load source is
Select Alternate Installation Device Bus
System: YOURSYS
Type Option, press Enter.
1=Select

Option Bus Number Selected

_ 2
_ 3 *
_ 4
_ 5
_ 6
_ 7
_ 8
_ 9
_ A
_ B
_ C
_ D
More.....
F2=Deselect device F3=Exit 12=Cancel

currently defined. You can use the F2 key to deselect the current bus and then
use option 1 to select another. All buses that exist on the system are displayed.

After pressing the Enter key, there will be a brief delay (up to 10 minutes)
while the bus is initialized. Following this delay, the Select Alternate
Installation Device screen appears.
Select Alternate Installation Device
System: YOURSYS
Type option, press Enter.
1=Select 5=Details

Resource Serial
Option Name Type Model Number Selected
_ TAP01 6380 001 00-1017187
_ TAP08 3287 030 32-234333
_ TAP02 6380 001 00-2017187
_ TAP05 3287 030 72-234333 *
_ TAP09 6380 001 00-1015187
_ TAP16 3287 030 22-234633

F2=Deselect device F3=Exit F5=Refresh F12=Cancel

Type a 1 in the Option field to select the device you wish to use, and press the
Enter key.

Note: When installing from an alternate installation device, be sure that only
one device contains valid install media. This will prevent the wrong
version of the Licensed Internal Code from being installed.

Chapter 5. Recovering the Licensed Internal Code 129

 10. The Install Licensed Internal Code display appears. Type a 1 and press the
Enter key.

Stop!
You are now ready to recover your Licensed Internal Code. Consult your
recovery checklist before continuing. The checklist tells you the correct option
to select from the Install Licensed Internal Code (LIC) display.

How to Load the Licensed Internal Code

If you receive an error screen:

See “Appendix A. Licensed Internal Code Installation Error Screens” on
page 775 for information on the possible error screens that can be displayed
during LIC installation.

If you are using an alternate installation device and you receive an error
screen, it may be due to one of the following conditions:
v You are trying to install from CD-ROM when an alternate installation
device is enabled.
v You are trying to use an alternate installation device which is not enabled.
Review “How to Set Up an Alternate Installation Device” on page 635 and
“How to Disable an Alternate Installation Device” on page 638 and perform
the appropriate procedure.

Note: You may find that the address information is not available, or that the
system configuration has changed so that the address information is
not correct. If this occurs, you must determine the address information
by a physical inspection of your system configuration. This inspection
can be difficult and may vary depending on your system model and
the specific configuration of your IO buses. For this reason, IBM
recommends that you call your next level of support for assistance in
determining the addresses you need for the alternate installation
device. A service agreement may be required for this type of assistance.

To complete the procedure for loading the Licensed Internal Code to your system
during a recovery, do the following:
1. You should see the Install Licensed Internal Code (LIC) display:

130 OS/400 Backup and Recovery V5R1

 Install Licensed Internal Code (LIC)

Disk selected to write the Licensed Internal Code to:

Serial Number Type Model I/O Bus Controller Device
xx-xxxxxxx xxxx xxx x x x

Select one of the following:

1. Restore Licensed Internal Code

2. Install Licensed Internal Code and Initialize system
3. Install Licensed Internal Code and Recover Configuration
4. Install Licensed Internal Code and Restore Disk Unit Data
5. Install Licensed Internal Code and Upgrade Load Source

Select the correct option and press the Enter key.

Attention!
Be sure you consult the correct recovery checklist before selecting an
option from the Install Licensed Internal Code (LIC) display. Some options
remove all data from your system.

2. If there is an alternate installation device defined and enabled, the Confirm

Alternate Installation Device display appears.
v To recover from the alternate installation device, press the Enter key.
v To recover from CD-ROM, press F12 to cancel. The Install Licensed Internal
Code display appears. Select option 3 (Define alternate installation device).
Perform steps 8 through 10 and disable the alternate installation device.
3. One of the following Install LIC and Initialize System–Confirmation screens is
shown if you chose option 2, 3, 4, or 5 from the LIC installation display. You
must press F10 to continue the installation; pressing the enter key just
re-displays the confirmation screen.

Install LIC and Initialize System - Confirmation

Warning:
All data on this system will be destroyed and the Licensed
Internal Code will be written to the selected disk if you
choose to continue the initialize and install.

Return to the install selection screen and choose one of the

other options if you want to perform some type of recovery
after the install of the Licensed Internal Code is complete.

Press F10 to continue the install.

Press F12 (Cancel) to return to the previous screen.
Press F3 (Exit) to return to the install selection screen.

Chapter 5. Recovering the Licensed Internal Code 131

 Install LIC and Recover Configuration - Confirmation

Warning:
All data on the selected disk will be destroyed and the Licensed
Internal Code will be written to this disk if you choose to
continue the install. When the install is complete, an IPL
will be done and you will be prompted to continue the recovery
of the DASD configuration.

Press F10 to continue the install.

Press F12 (Cancel) to return to the previous screen.
Press F3 (Exit) to return to the install selection screen.

Install LIC and Restore Disk Unit Data - Confirmation

Warning:
All data on the selected disk will be destroyed and the Licensed
Internal Code will be written to this disk if you choose to
continue the install. When the install is complete, an IPL
will be done and you will be prompted to restore the disk unit
data that you previously saved.

Press F10 to continue the install.

Press F12 (Cancel) to return to the previous screen.
Press F3 (Exit) to return to the install selection screen.

Install LIC and Upgrade Load Source - Confirmation

Warning:
All data on the selected disk will be destroyed and the Licensed
Internal Code will be written to this disk if you choose to
continue the install. When the install is complete, an IPL
will be done and you will be prompted to complete the upgrade.

Press F10 to continue the install.

Press F12 (Cancel) to return to the previous screen.
Press F3 (Exit) to return to the install selection screen.

The Initialize the Disk–Status screen is shown if you chose option 2, 3, 4, or 5

on the install selection menu, and then pressed F10 on the confirmation screen.
The actual time to initialize the disk can be considerably less than the estimated
time, depending on the current state of the disk.

132 OS/400 Backup and Recovery V5R1

 Initialize the Disk - Status

The load source disk is being initialized.

Estimated time to initialize in minutes : ___

Elapsed time in minutes . . . . . . . . : ___._

4. You are shown the Install Licensed Internal Code–Status display. You do not
need to respond to this display. The system shows this display for
approximately 30 minutes.

Licensed Internal Code Installation Status

Installation of the Licensed Internal Code in progress.

+-----------------------------------------+
Percent complete | XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX |
+-----------------------------------------+
0 20 40 60 100

Elapsed time in minutes . . . . . . . . : x.x

Please wait.

______________________________________________________________________

5. If an error occurs, you may see a display that requires a response.

Stop!
You have finished loading your Licensed Internal Code.

If you are using logical partitions, and you have installed the Licensed Internal
Code to the primary partition, you will receive the following message on the Disk
Configuration Error Report display:
This message indicates that the partitioning configuration should be recovered.
Disk Configuration Error Report

Type option, press Enter

5=Display Detailed Report

OPT Warning
___ Unit has incorrect logical partition configuration

Perform the steps that are listed below in How to Recover Your Logical Partition
Configuration.

Chapter 5. Recovering the Licensed Internal Code 133

 Note: There may be multiple logical partition configuration error messages for
different disk units. The steps that are listed will resolve all of these
messages.

How to Recover Your Logical Partition Configuration

If you are using logical partitions, use the following steps to recover the primary
partition.
1. Select the Use Dedicated Service Tools option.
2. Sign on to Dedicated Service Tools. The system displays the Use Dedicated
Service Tools menu.
3. From the Use Dedicated Service Tools (DST) menu, select option 11, Work with
system partitions.
Use Dedicated Service Tools

Select one of the following:

1. Perform an IPL
2. Install the operating system
3. Work with licensed internal code
4. Work with disk units
5. Work with DST environment
6. Select DST console model
7. Start a service tool
8. Perform automatic installation of the operating system
9. Work with save storage and restore storage
10. Work with remote DST support
11. Work with system partitions

You are shown the Work with System Partitions display.

4. From the Work with System Partitions display, select option 4, Recover
configuration data.
5. Choose option 1, Recover primary partition configuration data.
6. The system will look through all non-configured disk units for partition
configuration data. The disk unit with the newest configuration data for the
current system will be listed.
Select Disk Unit for Configuration Data Recovery
System: xxxxxxxx
Type option, press Enter:
1=Select

I/O Resource --Last Updated--- System

Opt Description Type-Model Date Time Serial Number
_ _____________________ ____-___ ________ _______ _________

7. Review the information for the displayed disk unit. Ensure that the Last
Updated and System Serial Number fields contain reasonable information. Type
a 1 to select the disk, and press the Enter key.
8. Press Enter to accept the recovery.
The system automatically copies the configuration data to the primary
partition’s load source, and performs an IPL to DST.

If you are restoring one partition with a previously mirrored load source, you may
continue to receive an error message after the IPL to DST. The message text is
″Unit has incorrect logical partition configuration″.

134 OS/400 Backup and Recovery V5R1

 If you do not receive this message, stop here. You have completed recovering your
logical partition configuration. Consult your recovery checklist to determine the
next step in your recovery process.

If you receive this message, you must clear this obsolete configuration by
performing the following steps:
1. After receiving the error message, use option 5 to determine which disk unit
has the obsolete partition configuration.
2. Exit the configuration error by pressing F3 to go to the DST menu.
3. From the Use Dedicated Service Tools menu, select option 11, Work with
system partitions.
4. Select option 4 (Recover configuration data).
5. Select option 3 (Clear non-configured disk unit configuration data).
6. Select the disk unit that was originally reported for the partition configuration
error.
7. Press F3 to return to the DST menu.
8. Select option 7, Start a service tool.
9. At the Start a Service Tool display, select option 7, Operator panel functions.
10. At the Operator Panel Functions display, press F8 to restart.

Stop!
You have completed recovering your logical partition configuration. Select
from the following options:
v If you are loading the Licensed Internal Code as part of the steps of
“Chapter 14. How to Restore the System from the Save Storage Media” on
page 311 you will see the Disk Configuration Attention Report displayed.
Select F3=Exit to Dedicated Service Tools (DST). Return to Chapter 14
and continue the Restore Storage procedures.
v If you selected option 2 from the Install Licensed Internal Code display,
continue with “How to Recover Your Disk Configuration After Installing
the Licensed Internal Code and Initializing the System”.
v If you selected option 3 from the Install Licensed Internal Code display,
continue with “How to Recover Your Disk Configuration” on page 142.
v If you selected option 4 from the Install Licensed Internal Code display,
continue with the recovery steps to restore disk unit data to the new load
source disk unit.
v If you selected option 5 from the Install Licensed Internal Code display,
continue with the recovery steps to upgrade the load source disk unit from
an IMPI-based system to a PowerPC-based system.
v If you do not need to restore the operating system, continue with “How to
Start Your System After Restoring Licensed Internal Code” on page 145.

How to Recover Your Disk Configuration After Installing the Licensed

Internal Code and Initializing the System
When you install the Licensed Internal Code by using option 2 from the Install
Licensed Internal Code (LIC) menu, the system does the following:

Chapter 5. Recovering the Licensed Internal Code 135

 v The system clears disk unit 1. Disk unit 1 contains information about how all the
other disk units on your system are configured. If you are using logical
partitioning, disk unit 1 also contains your partitioning configuration data.
v The system prepares to delete all data in the system ASP. The system ASP is not
actually cleared until you perform the IPL after installing the Licensed Internal
Code.
1. When you complete installing the Licensed Internal Code, you are shown the
Disk Configuration Attention Report display on the A or B mode IPL:

Disk Configuration Attention Report

Type option, press Enter

5=Display Detailed Report

OPT Warning
___ New disk configuration

2. If you type a 5 in the option column (OPT), you are shown the following
display:

The current configuration indicates a single unit system. You

can choose to accept it or do one of the following:

Following are the possible causes and recovery procedures:

v You can define a new configuration by adding units.

v Press F3 to exit to Dedicated Service Tools (DST) and if

necessary, take the right option to get to
'Use Dedicated Service Tools (DST)' display.
On the 'Use Dedicated Service Tools (DST)' display,
- Select option 4, Work with disk units.
- Select option 1, Work with disk configuration.
- Select option 3, Work with ASP configuration.
- Select option 3, Add units to ASPs.

v If you are performing 'Recover mirror load source'

utility, press F3 to exit to Dedicated Service Tools (DST).
If necessary, take the right option to get to 'Use
Dedicated Service Tools (DST)' display.
On the 'Use Dedicated Service Tools (DST)' display,
- Select option 4, Work with disk units.
- Select option 2, Work with disk unit recovery.
- Select option 16, Recover mirrored load source.

Press Enter to accept the current configuration and continue.

3. Press F10 or enter to accept the new disk configuration and continue.
4. Perform the following steps:
a. Create all logical partitions. Refer to “Chapter 30. Creating Logical
Partitions” on page 747 for information on creating logical partitions.
b. Initialize all of the non-load source disk units on your system.
c. Define which ASP each disk unit is configured in.
d. Determine which ASPs to start mirrored protection on.

136 OS/400 Backup and Recovery V5R1

 See “Configuring Disks on a New System–Checklist 1” on page 644 for a
checklist on adding disk units to the system ASP, adding disk units to the user
ASPs, starting mirrored protection on ASPs, and starting device parity
protection.

Stop!
You have now completed installing your Licensed Internal Code.
Continue with the next step on your recovery checklist, which is
restoring the operating system.

5. Press F3 (Exit to Use Dedicated Service Tools (DST) menu). The Use Dedicated
Service Tools (DST) menu display is shown:

Dedicated Service Tools (DST) Sign On

Type choice, press Enter.

DST user . . . . . . . . . .
DST password . . . . . . . .

6. From the Use Dedicated Service Tools (DST) menu, select option 4 (Work with
disk units).
7. From the Work with Disk Units menu, select option 2 (Work with disk unit
recovery).
8. From the Work with Disk Unit Recovery menu, select option 5 (Recover disk
configuration).
Press F10 to ignore problems and continue.
Problem Report

Note: Some action for the problems listed below may need to
be taken. Please select a problem to display more detailed
information about the problem and to see what possible
action may be taken to correct the problem.

Type option, press Enter.

5 = Display Detailed Report

OPT Problem
_ Load Source has been re-built
_ ASPs will be cleared

Chapter 5. Recovering the Licensed Internal Code 137

| Confirm Recover Configuration
|
| ATTENTION: There are problems in the system that could cause
| some of the system data to be destroyed.
|
| Press F11 to display the problems.
| Press F10 to confirm your choice to recover configuration.
| Press F12 to return to change your choice.
|
| Possible configuration found in the system records:
|
| Serial Resource
| ASP Unit Number Type Model Name Status
| __ ____ __________ ____ ___ _____________ ______________________
| __ ____ __________ ____ ___ _____________ ______________________
| __ ____ __________ ____ ___ _____________ ______________________
| __ ____ __________ ____ ___ _____________ ______________________
| __ ____ __________ ____ ___ _____________ ______________________
| __ ____ __________ ____ ___ _____________ ______________________
| More...
|
| F10=Confirm recover F12=Cancel
| ______________________________________________________________________________
||
|
9. Check the configuration of disk units on the display. The display shows the
disk units that are assigned to each user ASP and to the system ASP (ASP 1).
The warning on the display means that the system will clear all data on disk
units in the system ASP.
If this configuration is not correct, contact a service representative or software
support for assistance. Do not proceed further without getting help.
If the configuration that is shown is correct, press F10 to confirm the
configuration. The system builds the configuration information and returns to
the DST menu.
10. Press F12 to cancel the DST menu. You are shown the IPL or Install the
System menu.

Stop!
You have now completed installing your Licensed Internal Code. Continue
with the next step on your recovery checklist, which is restoring the operating
system.

| How to Recover Your Disk Configuration Using Operations Navigator

| at DST
| When you install the Licensed Internal Code by using option 2 from the Install
| Licensed Internal Code (LIC) menu, the system does the following:
| v The system clears disk unit 1. Disk unit 1 contains information about how all the
| other disk units on your system are configured. If you are using logical
| partitioning, disk unit 1 also contains your partitioning configuration data.
| v The system prepares to delete all data in the system ASP. The system ASP is not
| actually cleared until you perform the IPL after installing the Licensed Internal
| Code.

| These steps allow you to use the dedicated service tools (DST) debug mode to
| access disk management functions in Operations Navigator where you can
| configure disk units in system, basic, and independent auxiliary storage pools

138 OS/400 Backup and Recovery V5R1

| (ASPs) while your server is still in DST mode. When you are finished configuring
| disk units you can complete the step-mode IPL.

| Note: You must have configured the Service Tools Network Interface in order to
| do these steps.
| 1. You may have received a Disk Configuration Attention Report like the one
| below after you loaded the Licensed Internal Code. If so, press F10 to accept
| the problems and continue.
| DISK CONFIGURATION ATTENTION REPORT
|
| TYPE OPTION, PRESS ENTER.
| 5=DISPLAY DETAILED REPORT
|
| PRESS F10 TO ACCEPT ALL THE FOLLOWING PROBLEMS AND CONTINUE.
| THE SYSTEM WILL ATTEMPT TO CORRECT THEM.
|
| OPT PROBLEM
| NEW DISK CONFIGURATION
|
|
| 2. From the IPL or Install the System menu, select option 3 (Use Dedicated
| Service Tools (DST).
| IPL or Install the System
|
| Select one of the following:
| 1. Perform an IPL
| 2. Install the operating system
| 3. Use Dedicated Service Tools (DST)
| 4. Perform automatic installation of the operating system
| 5. Save Licensed Internal Code
||
| 3. On the Dedicated Service Tools (DST) Sign On display, sign on as QSECOFR /
| QSECOFR.
| Dedicated Service Tools (DST) Sign On
|
| Type choices, press Enter.
|
| Service tools user . . . . . . . . . . . QSECOFR
| Service tools password . . . . . . . . . _
|
|
|
| 4. Change the password for the QSECOFR user profile on the resulting screen, as
| the password is expired after the first use.
| Change Service Tools User Password
|
| Service tools user profile name . . . . . : QSECOFR
| Password last changed . . . . . . . . . . : 02/05/01
|
| Type choices, press Enter.
| Current password . . . . . . . . . . . . _
|
| New password . . . . . . . . . . . . . .
|
| New password (to verify) . . . . . . . .
|
||
| 5. On the Use Dedicated Service Tools (DST) menu, select option 6, Select DST
| console mode.

Chapter 5. Recovering the Licensed Internal Code 139

| Use Dedicated Service Tools (DST)
|
| Select one of the following:
| 1. Perform an IPL
| 2. Install the operating system
| 3. Work with Licensed Internal Code
| 4. Work with disk units
| 5. Work with DST environment
| 6. Select DST console mode
| 7. Start a service tool
| 8. Perform automatic installation of the operating system
| 9. Work with save storage and restore storage
| 10. Work with remote service support
|
|
|
| 6. On the Select DST Console Mode display, select option 2 (Start DST debug
| mode on IPL).
| Select DST Console Mode
|
| Attention: Incorrect use of DST debug mode can cause damage
| to data in this system. Contact your service representative
| for assistance.
|
| Select one of the following:
|
| 1. Exit Dedicated Service Tools (DST) on IPL
| 2. Start DST debug mode on IPL
|
|
| 7. On the IPL or Install the System menu, choose option 1 (Perform an IPL).
| IPL or Install the System
|
| Select one of the following:
| 1. Perform an IPL
| 2. Install the operating system
| 3. Use Dedicated Service Tools (DST)
| 4. Perform automatic installation of the operating system
| 5. Save Licensed Internal Code
|
|
| 8. On the Add All Disk Units to the System display, select option 1 (Keep the
| current disk configuration).
| Add All Disk Units to the System
|
| Select one of the following:
| 1. Keep the current disk configuration
| 2. Perform disk configuration using DST
| 3. Add all units to the system auxiliary storage pool (ASP)
| 4. Add all units to the system ASP and balance data
||
| 9. On the Install Required for Operating System display, press ENTER to
| continue.

140 OS/400 Backup and Recovery V5R1

| Install Required for Operating System
|
| The system ASP has been cleared, which requires an install
| of the operating system.
|
| To instal the operating system , do the following:
| - Load the install media in the device used to install the
| operating system and make the device ready.
| - Press Enter when the device is ready to install the
| operating system.
|
| -OR-
|
| Press F11 to display the Dedicated Service Tools sign on or
| main menu and not install the operating system.
|
|
| 10. On the Select Type of IPL display, select option 2 (Step-mode IPL).
| Select Type of IPL
|
| Select one of the following:
| 1. Normal IPL
| 2. Step-mode IPL
||
| 11. Step through the IPL by pressing ENTER. The last IPL step before you
| configure disk units is Storage Management Recovery. Press enter on the
| Storage Management Recovery screen shown below.
| Licensed Internal Code IPL in Progress
|
| IPL:
| Type . . . . . . . . . . . . . . . . . . : Attended
| Start date and time . . . . . . . . . . .: 00/00/00 00:00:00
| Previous system end . . . . . . . . . . .: Abnormal
|
| IPL step . . . . . . . . . . . . . . . . . : Storage Management Recovery
|
|
| 12. Stop at the resulting IPL step Start LIC Log screen.
| Licensed Internal Code IPL in Progress
|
| IPL:
| Type . . . . . . . . . . . . . . . . . . : Attended
| Start date and time . . . . . . . . . . .: 00/00/00 00:00:00
| Previous system end . . . . . . . . . . .: Abnormal
|
| IPL step . . . . . . . . . . . . . . . . . : Start LIC Log
||
| 13. Open Operations Navigator to begin disk unit configuration.
| 14. In the Environmental tasks panel, click Open Operations Navigator service
| tools window.
| 15. Enter the system name or IP address on the resulting window.
| 16. Expand the server you want to restore.
| 17. Expand Disk Units.
| 18. Right-click All Disk Units and select Add Disk Unit. The wizard will guide
| you through the process of adding disk units to disk pools and starting disk
| protection. Refer to a printed copy of your disk configuration to create all the
| necessary disk pools.
| 19. When your disk unit configuration is complete, continue through the step
| mode IPL. Press ENTER on all of the following IPL steps through Start the
| Operating System.
| 20. After the IPL steps complete, the Install the Operating System menu appears:

Chapter 5. Recovering the Licensed Internal Code 141

| Install the Operating System
|
| Type options, press Enter.
|
| Install
| option . . . . __ 1=Take defaults (No other
| options are displayed)
| 2=Change install options
|
| Date:
| Year . . . . . __ 00-99
| Month . . . . __ 01-12
| Day . . . . . __ 01-31
|
| Time:
| Hour . . . . . __ 00-23
| Minute . . . . __ 00-59
| Second . . . . __ 00-59
|
||
|

| Stop!
| You have now completed installing your Licensed Internal Code. Continue
| with the next step on your recovery checklist, which is restoring the operating
| system.
|
|
|
How to Recover Your Disk Configuration
When you install the Licensed Internal Code by using option 3 from the Install
Licensed Internal Code (LIC) menu, the system does the following:
v Clears disk unit 1. Disk unit 1 contains information about how all the other disk
units on your system are configured.
v Prepares to delete all data in the system ASP. The system ASP is not actually
cleared until you perform the IPL after installing the Licensed Internal Code.

Every disk unit on your system contains information about how it is configured.
Dedicated services tools (DST) provides an option to recover the disk configuration
on your system by using this information. The system reads every disk, assigns it
to the correct auxiliary storage pool (ASP), and rebuilds the disk configuration
information on unit 1.

In many cases, you can recover your disk configuration and avoid having to reload
all your user ASPs. To recover your disk configuration, do the following:
1. When you complete installing the Licensed Internal Code, you are shown the
Disk Configuration Error Report display on the A or B mode IPL:

Disk Configuration Error Report

Type option, press Enter

5=Display Detailed Report

OPT Error
___ Missing disk configuration

142 OS/400 Backup and Recovery V5R1

2. If you type a 5 in the option column (OPT), you are shown Missing Disk
Configuration display:
From either display, press F3 (Exit to Use Dedicated Service Tools (DST)). The
Missing Disk Configuration

The system disk configuration has been lost. The IPL

cannot be continued.

The following are the recommended recovery procedures:

o If the original system has more than one disk unit
configured and you want to keep the configuration
currently on the system, use Recover configuration
under Work with Disk Units.

Press F3 to exit to Dedicated Service Tools (DST) and

if necessary, take the right option to get to the 'Use
Dedicated Service Tools' (DST) display. On the 'Use
Dedicated Service Tools' (DST) display:
- Select option 4, Work with disk units.
- Select option 2, Work with disk unit recovery.
- Select option 5, Recover configuration.
o If the original system had only one disk unit configured
or you do not want to save the data currently on the
system, re-install the Licensed Internal Code, then
re-create the configuration, restore your data.

F3=Exit to use Dedicated Service Tools (DST) F12=Cancel

Use Dedicated Service Tools Sign On display is shown:

Dedicated Service Tools (DST) Sign On

Type choice, press Enter.

DST user . . . . . . . . . .
DST password . . . . . . . .

3. Sign on to Dedicated Service Tools. The system displays the Use Dedicated
Service Tools menu. If you are using logical partitions, and you wish to recover
the primary partition, continue with the following steps. If you are not using
logical partitions, continue with step 4.
4. From the Use Dedicated Service Tools (DST) menu, select option 4 (Work with
disk units).
5. From the Work with Disk Units menu, select option 2 (Work with disk unit
recovery).
6. From the Work with Disk Unit Recovery menu, select option 5 (Recover disk
configuration).

Chapter 5. Recovering the Licensed Internal Code 143

 Problem Report

Note: Some action for the problems listed below may need to
be taken. Please select a problem to display more detailed
information about the problem and to see what possible
action may be taken to correct the problem.

Type option, press Enter.

5 = Display Detailed Report

OPT Problem
_ Load Source has been re-built
_ ASPs will be cleared

Press F10 to ignore problems and continue.

Confirm Recover Configuration

ATTENTION: There are problems in the system that could cause

some of the system data to be destroyed.

Press F11 to display the problems.

Press F10 to confirm your choice to recover configuration.
Press F12 to return to change your choice.

Possible configuration found in the system records:

Serial Resource
ASP Unit Number Type Model Name Status
__ ____ __________ ____ ___ _____________ ______________________
__ ____ __________ ____ ___ _____________ ______________________
__ ____ __________ ____ ___ _____________ ______________________
__ ____ __________ ____ ___ _____________ ______________________
__ ____ __________ ____ ___ _____________ ______________________
__ ____ __________ ____ ___ _____________ ______________________
More...

F10=Confirm recover F12=Cancel

______________________________________________________________________________

7. Check the configuration of disk units on the display. The display shows the
disk units that are assigned to each user ASP and to the system ASP (ASP 1).
The warning on the display means that the system will clear all data on disk
units in the system ASP.
If this configuration is not correct, contact a service representative or software
support for assistance. Do not proceed further without getting help.
If the configuration that is shown is correct, press F10 to confirm the
configuration. The system builds the configuration information and returns to
the DST menu.
8. Press F12 to cancel the DST menu. You are shown the IPL or Install the System
menu.

Stop!
You have now completed installing your Licensed Internal Code. Continue
with the next step on your recovery checklist, which is restoring the operating
system.

144 OS/400 Backup and Recovery V5R1

How to Start Your System After Restoring Licensed Internal Code
Perform this procedure if you do not need to restore the operating system. After
you complete loading the Licensed Internal Code, you should see the IPL or Install
the System display:
IPL or Install the System

Select one of the following:

1. Perform an IPL
2. Install the operating system
3. Use Dedicated Service Tools (DST)
4. Perform automatic installation of the operating system
5. Save Licensed Internal Code

Do the following:
1. Select option 1 (Perform an IPL) on the IPL or Install the System menu. When
the IPL completes, the Sign On display is shown.
2. If your operator panel has a keylock switch, turn the key in the keylock switch
to the normal position.
3. Sign on the system as QSECOFR.
4. If the Select Product to Work with PTFs display appears, press F3 (Exit) to
continue the IPL.
5. Press the Enter key in response to any messages that are displayed.
6. When you are shown the IPL options display, type your choices and press the
Enter key.

IPL Options

Type choices, press Enter.

System date . . . . . . . . . . . . . . 07 / 26 / 88
System time . . . . . . . . . . . . . . 12 : 00 : 00
Clear job queues . . . . . . . . . . . . N
Clear output queues . . . . . . . . . . N
Clear incomplete job logs . . . . . . . N
Start print writers . . . . . . . . . . Y
Start system to restricted state . . . . N

Set major system options . . . . . . . . Y

Define or change system at IPL . . . . . N

Stop!
You have now completed starting your system after recovering the Licensed
Internal Code. Consult your recovery checklist to determine the next step in
your recovery process.

Disabling and Enabling the High-Speed Feature on the 2440 Tape Unit
If you have a 2440 Tape Unit with the high-speed feature enabled, it must be
disabled before you can install or restore the Licensed Internal Code. After the
restore operation, you can enable the high speed again. The high-speed feature is
disabled or enabled from the control panel on the 2440 Tape Unit.

Chapter 5. Recovering the Licensed Internal Code 145

 To find the control panel, open the front door of the 2440 Tape Unit. The control
panel is located in the upper right-hand corner. The following figure illustrates the
control panel.

A B C
1 2 3

D E F
4 5 6

CMD
7 8 9

CLR
0 EXEC

RV2W422-0

Disabling the High-Speed Feature: To disable the high-speed feature before the
restore operation, do the following from the control panel:
1. Press the arrow key and then the CMD 7 key.
2. Press the 9 key and then the 2 key.
3. Press the EXEC key.
4. Press the arrow key and then the CMD 7 key.
5. Press the 9 key and then the 3 key.
6. Press the EXEC key.
7. Press the arrow key and then the CMD 7 key.
8. Press the 6 key twice.
9. Press the EXEC key.
10. Press the 1 key.
11. Press the EXEC key.

Enabling the High-Speed Feature: To enable the high-speed feature after the restore
operation, do the following from the control panel:
1. Press the arrow key and then the CMD 7 key.
2. Press the 9 key and then the 2 key.
3. Press the EXEC key.
4. Press the arrow key and then the CMD 7 key.
5. Press the 9 key and then the 3 key.
6. Press the EXEC key.
7. Press the arrow key and then the CMD 7 key.
8. Press the 6 key twice.
9. Press the EXEC key.
10. Press the CLR 0 key.
11. Press the EXEC key.

146 OS/400 Backup and Recovery V5R1

Chapter 6. Restoring the Operating System
This chapter describes the procedures for recovering the operating system. The
procedure that is described here assumes that the Licensed Internal Code is
already installed on your system. Normally, the Licensed Internal Code is installed.
However, if the Licensed Internal Code is not on your system or is damaged, use
the charts in Chapter 4 to determine the correct recovery procedure for your
situation.

Why You Restore the Operating System: You might need to restore the operating
system for several reasons, such as:
v You are encountering problems with the operating system, such as damaged
objects.
v The software support center recommends it.
v You have replaced a disk unit in the system ASP.
v You are updating your system to a new release. See the following books for
procedures to install a new release of the iSeries server:
– Software Installation
– AS/400 Road Map for Changing to PowerPC Technology
– iSeries 940x RISC-to-RISC Road Map

Find These Things Before You Begin:

v Your most recent SAVSYS media. One of the following creates a SAVSYS media
volume:
– Running the Save System (SAVSYS) command.
– Using option 21 from the Save menu.
– Using option 22 from the Save menu.
– Use of option 11 from the Run Backup menu.

Attention!
DO NOT use a media volume that you created through DST by using
option 5=Save Licensed Internal Code from the IPL or Install the System
menu unless Software Services instructs you to do so. This process creates
a media volume that does not contain the Licensed Internal Code PTF
Inventory information or the OS/400 Operating System. If you perform the
recovery process using this media volume, you will have to re-install the
Licensed Internal Code from either a SAVSYS media volume or from your
distribution media before you can load any PTFs onto your system.

v If you do not have current SAVSYS media or they are damaged, you need the
following:
– The distribution media that is supplied by IBM
– All the media for program temporary fixes (PTFs) that you have applied.

© Copyright IBM Corp. 1997, 2000, 2001 147

 Attention!
Use the distribution media only if you do not have SAVSYS media. If
you use the distribution media to restore the operating system, the
version you restore will not have any of the PTFs that you have applied.
In addition, the restore sets the following default values that ship with
the OS/400 licensed program:
- System information
- Network attributes
- Configuration lists
- Edit descriptions
- Reply list entries
- IBM-supplied subsystem descriptions
- Passwords for IBM-supplied profiles

v The list of all the PTFs applied to your system at the time you saved the entire
system. You should attach this list to your backup log or keep it with your
SAVSYS media.
v The key or keystick for the system.
v The DST password, if you have set up your system with the installation of the
operating system secured.
v The QSECOFR password that is associated with the SAVSYS media that you are
using.

Do These Things Before You Begin:

v Clean the read and write heads of the tape unit if you are using a tape unit.
v If your source system (the system that was saved and needs recovery) is
operational, print a list of all the PTFs currently on the system, unless you
printed the list before restoring the Licensed Internal Code. Type the following
and press the Enter key:
DSPPTF LICPGM(*ALL) OUTPUT(*PRINT)

Choosing the Right Procedure for Restoring the Operating System

You can restore the operating system in several different ways. At several points
during the restore process, you need to make decisions based on which of these
operations you are using:
Complete Restore
Use a complete restore operation if the operating system is not on your
system or if it has damaged objects. This restores all the IBM-supplied
objects in the QSYS library and the language libraries from media.
Abbreviated Install
Use an abbreviated install operation to replace parts of the operating
system or system information, such as system values or the system reply
list.

Use the recovery checklist that you selected in Chapter 4 to determine the correct
procedure for your situation. You also need to know whether you are restoring
from SAVSYS media or the IBM-supplied distribution media. Use the distribution
media only if you do not have usable SAVSYS media.

148 OS/400 Backup and Recovery V5R1

 How to Begin Restoring the Operating System: The steps you have already
performed determine where you start:
v If you have just restored or installed the Licensed Internal Code, you are doing a
complete restore operation of the OS/400 program. You should see the IPL or
Install the System display. Start with the steps described in “How to Restore the
OS/400 Licensed Program”.
v If restoring the operating system is the first step or only step in your recovery
process, start by performing a manual IPL. The following section describes how
to do this.

How to Load the Operating System Using a Manual IPL

Follow these steps to begin loading the operating system. Use these steps only if
you have not just restored the Licensed Internal Code, as described in Chapter 5.

To perform a manual IPL, do the following:

1. Ensure that the tape unit, optical drive, or the CD-ROM unit is ready. For more
information on loading the tape, optical media, or the CD-ROM, see the setup
manual for the device.
2. Load the first media volume from your most recent SAVSYS media in the
appropriate device. If you do not have SAVSYS media or they are unusable,
load the first CD-ROM from the distribution media. Use the distribution media
only if no SAVSYS media exists.

Attention!
DO NOT use media that was created through DST by using option
5=Save Licensed Internal Code from the IPL or Install the System menu
unless you have been instructed to do so by Software Services. Media
created through this process does not contain the Licensed Internal Code
PTF Inventory information or the OS/400 Operating System. If you
perform the recovery process using this media, you will have to re-install
the Licensed Internal Code from either SAVSYS media or from your
distribution media before you can load any PTFs onto the system.

3. Ensure the key or keystick is in the system unit.

4. Place the system in manual mode.
5. Ensure that all jobs are ended and all users are signed off.
6. Power down the system.

Attention logical partitioning users!

If you are going to use this command on the primary partition, be sure to
power off all secondary partitions before running the command.

PWRDWNSYS OPTION(*IMMED) RESTART(*YES) IPLSRC(B)

7. Continue with “How to Restore the OS/400 Licensed Program”.

How to Restore the OS/400 Licensed Program

You are ready to begin these steps when you have completed an IPL. Either you
have just restored the Licensed Internal Code or you have just performed a manual
IPL from your alternate IPL device.

Chapter 6. Restoring the Operating System 149

 Note: If you use Operations console, perform the following steps to reset your
Operations console:
__ 1. On the IPL or Install the System screen, select 3, Use Dedicated Service
Tools (DST). Press Enter to continue.
| __ 2. Sign on to DST using a service tools user profile that has security
| officer authority and the assigned password.
__ 3. On the Use Dedicated Service Tools (DST) screen, select 5, Work with
DST environment. Press Enter to continue.
__ 4. On the Work with DST Environment screen, select 2, System Devices.
Press Enter to continue.
__ 5. On the Work with System Devices screen, select 6, Console Mode. Press
Enter to continue.
| __ 6. On the Select Console Type screen, select 2, Operations Console
| (Direct), or select 3, Operations Console (LAN). Press Enter to
| continue.

| Note: If you had to replace the load-source disk unit in the primary
| partition, you must use the directly attached console, option 2,
| to perform the restore operation.
| __ 7. Press F3 or F12 to get back to the IPL or Install the System screen.

. You should see the IPL or Install the System display:

IPL or Install the System

Select one of the following:

1. Perform an IPL
2. Install the operating system
3. Use Dedicated Service Tools (DST)
4. Perform automatic installation of the operating system
5. Save Licensed Internal Code

Task 1–Starting to Restore the Operating System

1. Load the first media volume from your most recent SAVSYS media in the
appropriate device. If you do not have SAVSYS media or they are unusable,
load the first CD-ROM from the distribution media. Use the distribution
media only if no SAVSYS media exists.

Attention!
DO NOT use media that you created through DST by using option
5=Save Licensed Internal Code from the IPL or Install the System menu
unless you have been instructed to do so by Software Services. Media
created through this process does not contain the Licensed Internal Code
PTF Inventory information or the OS/400 Operating System. If you
perform the recovery process using this media, you will have to re-install
the Licensed Internal Code from either SAVSYS media or from your
distribution media before you can load any PTFs onto the system.

2. From the IPL or Install the System display, select option 2 (Install the
operating system).

Note: Do not use option 4 (Perform automatic installation of the operating

system) to restore the operating system. This option can only be used
for installing the system and not for system recovery.

150 OS/400 Backup and Recovery V5R1

 3. Press the Enter key.
The Confirm Install of the Operating System display is shown.
Confirm Install of Operating System

Press Enter to confirm your choice to install the operating system. Press
F1 to return and cancel your choice to install the operating system.

4. Press the Enter key. If you see the Dedicated Service Tools (DST) Sign On
display, continue with step 5. If you see the Select a Language Group display,
skip to step 6.
5. If your system is set up to prevent unauthorized installation of the operating
system, you are shown the following display:
Dedicated Service Tools (DST) Sign On

Type choice, press Enter.

DST user . . . . . . . . . . ______
DST password . . . . . . . . ______

Type the DST user and the DST password and press the Enter key. You see the
Select a Language Group display.
Notes:
| a. The DST user and DST password are case sensitive. Type QSECOFR, the
| default user and password, in all capital letters.
b. The DST user is the same as the shipped value for the password at the
DST level. The DST user for security level DST is QSECOFR.
c. If your current DST password does not work, the password may have
been reset to the shipped value. Try QSECOFR as the DST password.
d. For more information about preventing unauthorized installation of the
operating system, see the iSeries Security Reference book.
6. You are shown the Select a Language Group display:
Select a Language Group

Note: The language feature shown is the language feature installed on the
system.

Attention: To keep the same primary language, ensure that the media you use
for installing the operating system matches the language feature shown.
If the operating system media does not match what is shown, the
installation process will attempt to install the operating system in a
different language feature than Licensed Internal Code.
This is undesirable. Type choice, press Enter.

Language feature . . . . . . . . . . . . . . . 2924

This display shows the primary language currently on the save media that
you are restoring.

This value should match the value that is already on your system. If it does
not, check to ensure that you have the correct save media. If you change the
value on the display, you will be prompted to insert different media to load a
different language feature. Press the Enter key. You are shown the Confirm
Language Feature Selection display.

Chapter 6. Restoring the Operating System 151

 Confirm Language Feature Selection

Language feature . . . . . . . . . . . . . . . . . : 2924

Press Enter to confirm your choice for language feature. Installing the
system will continue. Press F12 to return to change your choice for language
feature.

7. Press the Enter key to confirm the information.

Note: If you have to change your system’s primary language, see the Software
Installation book for more information.

If you see the Add All Disk Units to the System display, continue with step 8.
If you see an IPL status message display, skip to step 10 on page 153.
8. This display is shown only if disk units are in nonconfigured status:
9.
Add All Disk Units to the System

Select one of the following:

1. Keep the current disk configuration
2. Perform disk configuration using DST
3. Add all units to the system auxiliary storage pool (ASP)
4. Add all units to the system ASP and balance data

Disk units may be in nonconfigured status for these reasons:

v The Licensed Internal Code was installed using option 2 or option 3. The
recover disk configuration procedure was not run. All disk units except unit
1 appear in nonconfigured status.
v You have new or spare disk units that have not yet been configured on
your system.
Use the information in Table 38 to determine how to respond to this display:
Table 38. Configuring Disk While Installing the Operating System
Your Recovery Situation How To Respond to the Display
Restoring your entire system to a v If you plan to have user ASPs or mirrored protection you can select option
different system or to an upgraded 2 to configure your disks now. Or, you can select option 1 now and
system. configure the disks after you have restored the operating system. Use the
instructions in the section Part 8. Disk Configuration and Protection —
Procedures if you plan to configure disk protection or user ASPs.
v If you want all disks in the system ASP and do not plan to have mirrored
protection, select option 3.
The Licensed Internal Code was v If you want all disks in the system ASP and do not plan to have mirrored
installed on your own system using protection, select option 3.
option 2 or option 3 during a
v If you had user ASPs or mirrored protection on your system before the
recovery.
failure, you can select option 2 to reconfigure your disks. This removes any
data from disks that show as not configured.
v You can select option 1 and configure your disks later. However, the
system will not be able to recover the data on the disks that are not
configured.

After selecting option 3, you receive an Attention Report display. Take the
directed action for more detailed information, if necessary. Otherwise, press
F10 to accept the problems and continue.

152 OS/400 Backup and Recovery V5R1

 If you installed the Licensed Internal Code using option 2, you receive an
Attention report display. Take the directed action for more detailed
information, if necessary. Otherwise, press F10 to accept the problems and
continue.
10. Following is an example of a status display. You may see several similar
displays before the install the Operating System display is shown. These
status displays do not require any action by the user.
IPL Step in Progress

IPL step . . . . : Storage Management Recovery

The following list shows some of the IPL steps that are shown on the IPL Step
in Progress display:
v Authority Recovery
v Journal Recovery
v Database Recovery
v Journal Synchronization
v Start the Operating System

Some of the IPL steps could take a long time.

While the system is performing the IPL, system reference codes (SRCs) are
displayed on the control panel of the system unit to indicate what step is in
progress. The iSeries Service Functions book describes these SRCs. If the same
SRC is displayed for a long time in solid (not flickering) lights, your system
may have a problem completing the IPL. Look up the SRC in the iSeries
Licensed Internal Code Diagnostic Aids - Volume 1 book or contact software
support.

The system may prompt you for additional volumes of your SAVSYS media or
the distribution media. Follow the instructions on the display.

After the IPL steps complete, the Install the Operating System menu appears:
Install the Operating System

Type options, press Enter.

Install
option . . . . __ 1=Take defaults (No other
options are displayed)
2=Change install options

Date:
Year . . . . . __ 00-99
Month . . . . __ 01-12
Day . . . . . __ 01-31

Time:
Hour . . . . . __ 00-23
Minute . . . . __ 00-59
Second . . . . __ 00-59

11. Continue with Task 2.

Chapter 6. Restoring the Operating System 153

 Task 2–Selecting the Installation Options
1. Type your choice for the Install option prompt based on the following:
v If you are doing a complete restore operation, select option 1 (Take defaults).
This restores the entire operating system. Use this option if any of the
following is true:
– You are recovering from a failure of the load source unit.
– You are restoring your entire system.
– You are upgrading to a new system.
– You are recovering damaged objects in your operating system.

Note: If you are doing a complete restore operation and restoring a primary
language other than English or if you have changed the shipped
value of the QDATFMT system value, you must select option 2
(Change install options). This ensures the language-dependent system
values are restored correctly.

Restoring to a different system?

If you are restoring to a different system (with a different serial
number) and you want your network attributes to be restored, select
option 2 (Change install options). This will allow you to select to
restore your network attributes from your save media.

v If you are doing an abbreviated install operation, select option 2 (Change

install options). This allows you to specify which parts of the operating
system you want to restore. Use this option if you are recovering damaged
system information, such as system values.
2. If the date and time are not correct, type new values.
3. Press the Enter key.
If you selected install option 1 (Take defaults), skip to step 14 on page 158.
If you selected install option 2 (Change install options), you are shown the
Specify Install Options display:
Continue with step 4.
Specify Install Options

Type choices, press Enter.

Restore option . . . . __ 1=Restore programs and language objects

from current media set
2=Do not restore programs or
language objects
3=Restore only language objects
from current media set
4=Restore only language objects
from a different media set using the
current install device

Job and output

queues option . . . . __ 1=Clear, 2=Keep

Distribute OS/400 on
available disk units __ 1=Yes, 2=No

4. Type your choice for the Restore option prompt based on the following:

154 OS/400 Backup and Recovery V5R1

 Note: If you are performing a complete system recovery, you must select
option 1.
1 = Restore programs and language objects
This option restores system objects from the media that you are using.
Use this option if damage to a system user profile is found. If you
select the option, you will be prompted to specify whether to restore
system information, edit descriptions, or the system reply list.
Notes:
a. If your system had access path recovery times for user ASPs and
the user ASPs have not yet been reconfigured, the system cannot
restore access path recovery times for the ASPs. Use the Edit
Recovery Times for Access Paths (EDTRCYAP) command to set the
times after you have reconfigured your ASP configuration.
b. If you are restoring a primary language other than English, you
must select option 1.
2 = Do not restore programs or language objects
This option leaves the current programs and language objects on the
system. Select this option to do an abbreviated install of the operating
system.
When you select this option, the system does the following:
v Nothing is restored from media. Any damaged objects that the
system deletes and creates again are empty.
v All libraries on the system are checked for damage. Damaged
libraries are deleted and then created again.
v All system libraries (including QSYS) are created if they do not
exist.
v Information associated with the user profiles is created if it does not
exist or is damaged.
v The system entry-point table is created again.
3 = Restore only language objects from current media
This option loads only those objects that make up the national
language. The search for the language files begins on the current
media. Select this option only if you need to change your primary
language because you are recovering using the distribution media.
4 = Restore only language objects from a different media
This option loads only those objects that make up the national
language. The system prompts you to insert the language media.
Select this option only if you need to change your primary language
because you are recovering using the distribution media.

Note: Consult the Software Installation book if you want to change the
primary language. You should avoid changing the primary
language during a recovery.
5. Type your choice for the Clear Job and Output Queues prompt based on the
following:
1 = Clear
If you do not want to keep any spooled files or entries on job queues
after the installation or if you know they are damaged, select this
option. The system removes all jobs on job queues and spooled files. It
re-creates any internal objects associated with them. You should select

Chapter 6. Restoring the Operating System 155

 this option if you are performing an abbreviated installation of the
operating system. This option also resets the counter for assigning
unique job numbers.
2 = Keep
Any entries on job queues and output queues remain after the
installation. This is the normal option for this prompt.
| 6. Type your choice for the Distribute OS/400 on available disk units prompt based
| on the following:
| 1 = Yes
| This option will take more time. You should only use this option in
| specific recovery situations as directed by your service representative.
| This option specifies to distribute OS/400 objects on available disk
| units in the system ASP during the installation process.
| 2 = No
| Specifies to not distribute OS/400 objects on available disk units. This
| option will restore OS/400 objects from the media over the existing
| objects on the system.
| 7. Type your choices on the display and press the Enter key.
If you did not specify 1 for the Restore option prompt, skip to step 14 on
page 158.
If you specified 1 for the Restore option prompt, you are shown the Specify
Restore Options display:
Continue with step 8.
Specify Restore Options

Type choices, press Enter.

Restore from the installation media:

System information. . . _ 1=Restore, 2=Do not restore

Edit descriptions . . . _ 1=Restore, 2=Do not restore
Message reply list. . . _ 1=Restore, 2=Do not restore
Job descriptions. . . . _ 1=Restore, 3=Keep customization
Subsystem descriptions. _ 1=Restore, 3=Keep customization

8. Type your choice for the System information prompt based on the following:

Note: If you are performing a complete system recovery to a different system

(with a different serial number), and you want to restore the network
attributes from the save media, you must select option 1 (Restore).
1 = Restore
The system restores the system values and system management
objects, such as access path recovery times, from media. Select this
option if any of the following is true:
v You received a message during a previous IPL stating that the
system value object was created again.
v You want to restore them to their values from your last save
operation.
v You restored the operating system using option 2 or option 3 from
the Install Licensed Internal Code (LIC) display.
v You are restoring a primary language other than English, or if you
changed the shipped value of the QDATFMT system value.

156 OS/400 Backup and Recovery V5R1

 If you have changed your primary language since your last save
operation, the system may change some language-dependent system
values during the restore operation.
2 = Do not restore
The system values and system management objects that are currently
on the system are not changed.
A system value object must always be present on an operational
system. If the system value object does not exist, the system restores
it, even if you select option 2.

Note: For more information about system values, see the Work Management
book. For more information about access path recovery times, see
Chapter 18.
9. Type your choice for the Edit descriptions prompt based on the following:
1 = Restore
The system restores the edit descriptions from media. Select this
option if:
v The edit descriptions are damaged.
v You want to restore them to their values from your last save
operation.
v You installed the Licensed Internal Code by using option 2 or
option 3 from the Install Licensed Internal Code (LIC) display.
2 = Do not restore
The edit descriptions currently on the system are not changed.
10. Type your choice for the Message Reply List prompt based on the following:
1 = Restore
The system restores the reply list from media. Select this option if:
v The message reply list is damaged.
v You want to restore it to its values from your last save operation.
v You installed the Licensed Internal Code by using option 2 or
option 3 from the Install Licensed Internal Code (LIC) display.
2=Do not restore
The message reply list currently on the system is not changed.

The default for these options is 2 if the operating system is loaded on the
system. The default is 1 if the operating system is not already loaded.
11. Type your choice for the Job descriptions prompt based on the following:
1 = Restore
The system restores the job descriptions from media.
3 = Keep customization
The system restores the objects from media and customizes them with
the values from the same objects that are already on the system.
12. Type your choice for the Subsystem descriptions prompt based on the following:
1 = Restore
The system restores the subsystem descriptions from media.
3 = Keep customization
The system restores the objects from media and customizes them with
the values from the same objects that are already on the system.

Chapter 6. Restoring the Operating System 157

 13. Type your choices on the display and press the Enter key.
14. The Installation Status display indicates how many program or language
objects have been restored. They are for your information only and require no
response.
| Message ID. . . : CPI2070 OS/400 Installation Status
|
| Stage 2 +-----------------------------------------------------+
| |XXXXXXX |
| +-----------------------------------------------------+
| 0 20 40 60 80 100
|
| Installation Objects
| Stage Completed Restored
|
| 1 Creating needed profiles and libraries . . . . : X
|
| >> 2 Restoring programs to library QSYS . . . . . . : XXXX
|
| 3 Restoring language objects to library QSYS . . :
|
| 4 Updating program table. . . . . . . . . . . . :
|
| 5 Installing database files. . . . . . . . . . . :
|
| 6 Completing iSeries installation . . . . . . . . :

15. Continue loading media in sequence when messages are shown that ask you
to load additional media. The system searches through the media and loads
the necessary programs and language information. After processing all the
system save media or distribution media, the system may display the
following message at the bottom of a blank display:
Operating system has been installed. IPL in progress.
When the IPL is complete, the IPL Sign On display is shown and the system
in ready to complete the IPL. Continue with the next task.

Task 3–Selecting IPL Options

1. Sign on as QSECOFR. The password for QSECOFR depends on the recovery
steps you have performed:
v If you restored the operating system without first restoring the Licensed
Internal Code, the QSECOFR password was not changed during the restore
process.
v If you used option 1 to restore the Licensed Internal Code before you
restore the operating system, the system associates the QSECOFR password
with your SAVSYS media.
v If you used option 2 or option 3 to install the Licensed Internal Code before
you restore the operating system, the system requires no password at this
time. The system displays the Change Password display. The system sets
the QSECOFR user profile to *EXPIRED, and the system sets the password
to QSECOFR. The system restores the system security level after you install
the operating system and the IPL completes. When the IPL completes, the
QSECOFR password is the password associated with the SAVSYS media
you used.

Note: If you do not know the QSECOFR password, you can use DST to reset
the password to its shipped value of QSECOFR.

158 OS/400 Backup and Recovery V5R1

 Change Password

Password last changed . . . . . . xx/xx/xx

Type choices, press Enter.

Current password . . . . . . . QSECOFR

New password . . . . . . . . . _______

New password (to verify) . . . _______

2. Press the Enter key. Informational messages are displayed.

3. If the Select Product to Work with PTFs display appears, press F3 (Exit) to
continue.
Select Product to Work with PTFs

Position to . . . . . . ._____________ Product

Type options, press Enter. Press F21 to select all.

1=Select

Product
Opt Product Option Release
_ 5769999 *BASE V4R5M0
_ 5769SS1 *BASE V4R3M0

4. You are shown the IPL Options display:

IPL Options

Type choices, press Enter.

System date . . . . . . . . . . . . . . 07 / 26 / 88
System time . . . . . . . . . . . . . . 12 : 00 : 00
Clear job queues . . . . . . . . . . . . N
Clear output queues . . . . . . . . . . N
Clear incomplete job logs . . . . . . . N
Start print writers . . . . . . . . . . Y
Start system to restricted state . . . . N

Set major system options . . . . . . . . Y

Define or change system at IPL . . . . . Y

The values that appear as defaults depend on the recovery steps you have
performed.
5. If the system date and system time are not correct, type the correct values. If
you installed the Licensed Internal Code using option 2 or option 3, the date
and time may be blank. The system date must have a year value in the range
of 87 to 99, or 00 to 22.
6. Type your choice for the Start print writers prompt based on the following:
N = No
Select this value if you are going to restore user profiles, device
configuration objects, user libraries, and authorities.
Y = Yes
Select this value if you have completed your recovery.

Chapter 6. Restoring the Operating System 159

 7. Type your choice for the Start system to restricted state prompt based on the
following:
Y = Yes
Select this value if you are going to restore user profiles, device
configuration objects, user libraries, and authorities. Only the console
is started (varied on).
N = No
Select this value if you have completed your recovery. All devices are
started.
8. Type Y (Yes) for the Set major system options prompt.
9. Type Y (Yes) at the Define or change system at IPL prompt.
10. Type your choices on the display and press the Enter key. Continue with the
next task.

Task 4–Setting Major System Options

1. You are shown the Set Major System Options display:

Set Major System Options

Type choices, press Enter.

Enable automatic configuration . . . . . . Y Y=Yes, N=No

Device configuration naming . . . . . . . *NORMAL *NORMAL, *S36,
*DEVADR
Default special environment. . . . . . . . *NONE *NONE, *S36

2. If you choose to enable automatic configuration, the system will create a

device description for every device that is attached to your system. The device
description will be named according to the value that you specify for Device
configuration naming. You may need to change the names and descriptions of
these device descriptions later.
If you choose not to enable automatic configuration, you will need to configure
at least one device later in your recovery. You must configure the device after
you finish restoring the operating system and before you restore any other
information. You may also need to correct the device configuration for the
system console and respond to SRC A900 2000 (step 10 on page 164). The
instructions to recover from SRC A900 2000 are provided.
3. Type your choices and press the Enter key.
4. If you specified Y for the Define or change system at IPL prompt in step 4 of Task
3, continue with “Task 5–Defining or Changing the System at IPL”. If you
specified N for the Define or change system at IPL prompt in step 4 of Task 3, skip
to “Task 6–Completing the IPL” on page 162.

Task 5–Defining or Changing the System at IPL

1. If you specified Y for Enable automatic configuration on the Set Major System
Options display, skip to step 3 on page 161. If you specified N, continue with
step 2.
2. If you have chosen not to enable automatic configuration, you must change the
QIPLTYPE system value. Do the following:

160 OS/400 Backup and Recovery V5R1

 a. From the Define or Change the System at IPL menu, select option 3 (System
value commands). Press the Enter key.
b. Select option 3 (Work with system values) and press the Enter key.
c. Type a 2 in the Option column next to the system value QIPLTYPE and
press the Enter key.
d. Change the value to 2 and press the Enter key.
e. Press F12 until you return to the Define or Change the System at IPL menu.
3.
When you are recovering your system, there are some system values that must
be set correctly to prevent the recovery from failing. Also, if you restore the
system from distribution media, your system values will be reset to the
IBM-supplied defaults.
Use whatever documentation you have to set the system values to the correct
settings for your installation.

Attention!
If you are restoring to a system with a different processor or memory, you
must ensure that the QMCHPOOL, QBASPOOL, and QPFRADJ system
values are correct.

As a general rule, if the main storage size is 64M or larger, change the
QMCHPOOL system value to be 15 percent of the main storage size. If the
main storage size is less than 64M, change the QMCHPOOL system value
to be 20 percent of the main storage size. For a more precise setting of the
QMCHPOOL system value, refer to the Work Management book.

The QBASPOOL system value should be equal to the size of all the
unallocated storage.

The QPFRADJ system value should be set to 2.

a. From the Define or Change the System at IPL menu, select option 3 (System
value commands) and press the Enter key.
b. Select option 3 (Work with system values) and press the Enter key.
c. Type a 2 in the column next to the QALWOBJRST, QJOBMSGQFL, and any
other system values that you want to change and press the Enter key.
d. Change the QALWOBJRST system value to *ALL, and change the
QJOBMSGQFL system value to *PRTWRAP. Change any other system
values that you want to change and press the Enter key.
e. Press F12 to return to the Define or Change the System

Note: Some system values cannot be changed at this time. You will need to
change these values later in the recovery process. After the IPL
completes, you should check to ensure that the system values you
changed in 3 are correct.

If you are restoring to the same system from your SAVSYS media, skip to 5 on
page 162.
4. If you are restoring to a different system with a different serial number, and
you selected install option 1 (Take defaults) on the Install the Operating
System menu, the following network attributes are reset to the shipped values.

Chapter 6. Restoring the Operating System 161

 If you selected install option 2 (Change install options) on the Install the
Operating System menu, and selected option 1 (Restore) at the System
Information field on the Specify Restore Options display, the network attributes
will be restored.
v System name
v Local network ID
v Local control point name
v Default local location name
v Default node
v Default type
v Maximum number of intermediate sessions
v Route addition resistance
v Network node servers
v Alter primary focal point
v Alert default focal point

If you are restoring from distribution media and have previously changed the
network attributes from the IBM-supplied defaults, you need to reset them. Do
the following:
a. From the Define or Change the System at IPL menu, select option 4
(Network attributes commands) and press the Enter key.
b. Select option 2 (Change network attributes). Press the Enter key to display a
list of network attributes.
c. Change the values to the correct network attributes and press the Enter key.
d. Press F12 (Cancel) to return to the Define or Change the System at IPL
menu.
5. If you are partially restoring (only some libraries), continue with step 6.
Otherwise, skip to step 7.
6. If you are partially restoring, you need to make sure that all libraries specified
in the QSYSLIBL and QUSRLIBL system values are on the system. Do the
following:
a. From the Define or Change the System at IPL menu, select option 3 (System
value commands). Press the Enter key.
b. Select option 3 (Work with system values) and press the Enter key.
c. Type a 2 in the Option column next to the system values you want to
change and press the Enter key.
d. Change the values to the correct values and press the Enter key.
e. Press F12 to return to the Define or Change the System at IPL menu.
7. Continue with “Task 6–Completing the IPL”.

Task 6–Completing the IPL

1. Press F3 to continue the IPL.
2. The following display is shown during the IPL process (attended mode) when
system access paths are marked for rebuild:

162 OS/400 Backup and Recovery V5R1

 Edit Rebuild of Access Paths
05/12/90 13:49:34

IPL threshold . . . . . . . . . . . . . 50 0-99

Type sequence, press Enter.

Sequence: 1-99, *OPN, *HLD

------------Access Paths---------- Unique Rebuild

Seq Status File Library Member Keyed Time
25__ IPL QAPZSYM2 QSYS QAPZSYM2 NO 00:00:01

“Task 2–Using the Edit Rebuild of Access Paths Display” on page 171
describes how to interpret and update this display.

A status message is sent to notify the user that the system is performing
access path recovery.
3. Make any changes and press the Enter key. If you have made changes, the
Edit Rebuild of Access Paths display is shown again confirming your changes
or showing your error messages. Repeat this step until the Display Access
Path Status display is shown or the IPL continues.
4. The Display Access Path Status display is updated every 5 seconds while the
system is rebuilding access paths:

Display Access Path Status

IPL Threshold . . . . . . : 50

----------Access Paths---------- Rebuild Current

Status File Library Member Build Time Run Time
RUN QAPZSYM2 QSYS QAPZSYM2 00:00:01 00:00:01
JRN QAPZREQ2 QSYS QAPZREQ2 00:00:01
SYS QASULE03 QSYS QASULE03 00:00:01
IPL QASULE01 QSYS QASULE01 00:00:01

If you want to make changes, press F12 (Cancel) to return to the Edit Rebuild
of Access Paths display. If all access paths are rebuilt or you no longer want to
see the display, press F3 (Exit and continue IPL).

Note: The rebuild of access paths requires system memory. It is recommended

that you avoid a high level of system activity so that the rebuild of
access paths can complete.
5. The following display is shown if referential constraints need to be verified:
“Task 3–Using the Edit Check Pending Constraints Display” on page 173
Edit Check Pending Constraints
03/30/94 10:09:27
Type sequence, press Enter.
Sequence: 1-99, *HLD

----------Constraints----------- Verify Elapsed

Seq Status Cst File Library Time Time

75__ AFTIPL CSTF1 FILE567890 LIB4567890 00:00:56 00:00:00

describes how to interpret and update this display.

6. Make any changes and press the Enter key. The Edit Check Pending
Constraints display is shown again confirming your changes or showing your

Chapter 6. Restoring the Operating System 163

 error messages if you have made changes. Repeat this step until the Display
Constraint Status display is shown or the IPL continues.
7. The Display Constraint Status display is updated every 5 seconds while the
system is verifying constraints:

Display Constraint Status

IPL Threshold . . . . . . : 50

----------Constraints----------- Verify Elapsed

Status Constraint File Library Time Time
RUN CUST1 CUSTMAST CUSTLIB 00:00:04 00:00:01
RUN CUST2 CUSTMAST CUSTLIB 00:00:05 00:00:01
IPL ORDHST1 ORDHIST ORDLIB 00:00:23 00:00:00

If you want to make changes, press F12 (Cancel) to return to the Edit Check
Pending Constraints display. Press F3 (Exit and continue IPL) if all constraints
are verified or you no longer want to see the display.
8. Press the Enter key if QSYSOPR messages are displayed.
9. Press the Enter key to continue. If you restore the operating system from
distribution media, you may have a problem with sending messages or
creating documents if you have OfficeVision. To prevent errors, enter the
following command:
MRGMSGF QOFC/QZOFCMSG QSYS/QOFCMSG
10. You may receive A900 2000 on the control panel or message CPF0975, Console
did not vary on, on the console display. This occurs if your system
configuration was lost and you have disabled automatic configuration. The
system has created device description QCONSOLE to allow you to continue
the restore operation. You may also receive SRC A900 2000 if you perform an
IPL when the QIPLTYPE system value is set to 2. Do not create a user-defined
device description for the console display. This can cause unpredictable
results.
If you receive this message, perform the steps that are described in
“Recovering from SRC A900 2000” before continuing.
11. If you restored from the distribution media by using a 1/4-inch cartridge tape
drive, the light on the tape drive may still be on. After the system has finished
restoring the operating system, you may remove the tape while the light is on.

Stop!
When the Sign On display appears, you have completed restoring the
operating system. Consult your recovery checklist for the next step in your
recovery process.

Recovering from SRC A900 2000

When you restore the operating system, you may see SRC A900 2000. This happens
when you use option 2 or option 3 to install the Licensed Internal Code and
automatic configuration is not active while you are restoring the operating system.
Before you can continue your recovery operations, you must create a device
description and possibly a controller description to finish the restore operation. Do
not create a user-defined device description for the console display.

164 OS/400 Backup and Recovery V5R1

Creating a Configuration for 34xx Tape Units
If your tape unit is a 3422, 3430, 3480, or a 3490, and you want to use a tape
controller, do the following:

Note: The following steps do not apply to the 3490 Models E and F. For these
models refer to “Creating a Configuration for Other Tape Units” on
page 166.
1. Use the Work with Hardware Resource (WRKHDWRSC) command to
determine the location of the tape controller.
WRKHDWRSC TYPE(*STG)
2. Create the controller description for the tape controller by doing the following:
a. Locate the resource name for the tape controller on the Work with Storage
Resources display. The value 34xx is displayed in the Type column.
b. Write down the name of the resource.
c. Type a 9 (Work with resource) in the OPT column next to name of the tape
controller and press the Enter key. You see the Work with Storage Resources
display.
d. Type 5 (Work with controller descriptions) in the option column in front of
the tape controller. You see the Work with Controller Descriptions display.
e. Type 1 (Create) in the option column on the top line.
f. Type the controller name (such as TAPCTL01) in the description field and
press the Enter key. You see the Create Controller Description display.
g. If necessary, type additional information on the display. Then press the
Enter key. You return to the Work with Controller Descriptions display.
h. If the controller description that you created does not appear, press F5
(Refresh).
3. To create device descriptions for the tape units that are attached to the
controller, do the following:
a. On the Work with Controller Descriptions display, press F23 (More options).
The list of options changes.
b. Type 9 (Work with associated descriptions) in the option column in front of
the tape controller. You see the Work with Associated Descriptions display.
c. Locate the resource for the tape unit. Because no device description exists,
the description says *NONE.
d. Write down the name of the tape resource.
e. Type a 1 (Create) in the Opt column next to the description of *NONE and
press the Enter key. You see the Create Device Desc (Tape) (CRTDEVTAP)
screen.
f. In the Device description field, type a name such as TAP01.
g. In the Resource name prompt, type the name that you wrote down in step
3d. (If you did not write it down, press F12 to return to the display. Repeat
steps 3d through 3g.)
h. Press the Enter key.
i. Additional parameters appear on the display.
j. If necessary, type additional information on the display. Then press the Enter
key. You return to the Work with Associated Descriptions display.
k. Press F5 (Refresh). The name of the description that you created should now
be associated with the resource.

Chapter 6. Restoring the Operating System 165

 l. Type 8 (Work with configuration status) in front of the controller description
and the device description. You see the Work with Configuration Status
display.
m. Type 1 (Vary on or Make available) in front of the controller and the
devices.
4. Press F3 until your return to your original menu.

Creating a Configuration for Other Tape Units

If you are not using a 34xx tape unit, or want to create a 34xx (3490 Model E or F)
tape unit without a controller, do the following:
1. Use the Work with Hardware Resource (WRKHDWRSC) command to
determine tape controller name.
WRKHDWRSC TYPE(*STG)
2. Locate the tape controller on the Work with Hardware Resources display.
3. Type a 9 (Work with resource) next to tape controller name and press the
Enter key.
4. Locate the resource name for the tape unit (for example, TAP01).
5. Enter a 5 (Work with Configuration Descriptions) in the Opt column next to
the tape resource name and press the Enter key.
You are shown the Work with Configuration Descriptions display.
6. Type a 1 (Create) in the Option field and a device description name (for
example, TAP01) in the Description field. Press the Enter key. You are shown
the Create Device Description display.
7. Change any values that you want to change, then press the Enter key (twice)
to create the device description. You are shown the Work with Configuration
Descriptions display again. The device that you created should appear on the
display.
8. Type an 8 (Work with configuration status) in front of the new device
description. You are shown the Work with Configuration Status display.
9. Type a 1 (Vary on) in front of the new device. If the status does not change to
Varied On, wait a few minutes. Then press F5 (Refresh). If the status still does
not change to Varied On, follow normal problem determination procedures for
the device.
10. Press F3 until you return to the main menu.

SRC A900 2000 remains displayed on the control panel throughout the remaining
restore operations. When the final IPL of the system is complete, SRC A900 2000
disappears. The user-defined device description for the console display will be
restored when the Restore Configuration (RSTCFG) command is run later in the
recovery.

Stop!
When the Sign On display appears, you have completed restoring the
operating system. Consult your recovery checklist for the next step in your
recovery process.

166 OS/400 Backup and Recovery V5R1

Chapter 7. Starting the System After It Ends Abnormally
When your system stops normally, it does the following:
v Writes changed pages of information from memory to auxiliary storage.
v Closes access paths and files.
v Ends programs and jobs at the natural stopping points.

If your system stops without having time to do these things, it is called an

abnormal end. Your system might end abnormally for the following reasons:
v A power failure.
v A disk failure, if you do not have mirrored protection or device parity
protection.
v A failure in the processor.
v Failure of a critical operating system program.
v Operator action (forced power down).

What Happens When Your System Stops

The following list describes circumstances that cause your system to stop
unexpectedly and what happens:
v Power failure with uninterruptible power supply: When the system loses
normal power, the uninterruptible power supply system takes over and keeps
the system running. The system detects this change and sends a message to your
power-monitoring program. Your program can decide whether to keep the
system running until power returns or to begin an orderly shutdown.
v Power failure with no protection: If your system does not have an
uninterruptible power supply feature and the power fails, your system stops
immediately. The contents of main memory are lost. The system must
reconstruct information when power returns. This can be very time-consuming.
Whether the system starts automatically depends on how you have set the
QPWRRSTIPL system value.
v Disk failure with device parity protection or mirrored protection: In many
cases, the system can continue running without full disk protection until the
failed unit is replaced.
v Disk failure without disk protection: This is like a power failure without
protection. The system stops immediately. The system must reconstruct
information about jobs that were running and files that were open after the disk
is repaired or replaced.
v Failure of a critical operating system program: The system will stop
immediately, just as it does if an unprotected power failure or disk failure
occurs. The system attempts to copy the contents of main memory so that the
problem can be analyzed. This is called a main storage dump. When the system
stops, you will see the Main Storage Dump Manager Occurred display. See
“Using the Main Storage Dump Manager Occurred Display” on page 168 for
more information about this display.

© Copyright IBM Corp. 1997, 2000, 2001 167

 Using the Disk Configuration Error Report Display
When your system starts, it checks to ensure that it can access all of the disk units
that are configured. If it cannot access one or more disk units, you are shown the
Disk Configuration Error Report display:
Disk Configuration Error Report

Type option, press Enter.

5=Display Detailed Report

OPT Error
_ Missing disk units in the configuration

Following a temporary power outage, you may see this display because power has
been restored to the processor but not to the peripheral devices. Wait to respond to
the display until power is restored to all the disk units. The system’s ability to
access all the disk units when the system is starting is important for a successful
recovery. If disk units are not available, the system may not be able to recover
changed pages of memory. This can lengthen the time it takes to perform the IPL.

This screen may also be presented:

v After abnormal termination if the system is unable to activate all of the DASD
on the re-IPL.
v During any system IPL that has a similar problem, even if normal system
shutdown had taken the system down last.

Using the Main Storage Dump Manager Occurred Display

If your system encounters a serious software problem, you are shown the Main
Storage Dump Manager Occurred display:
Main Storage Dump Manager Occurred
S/N xxxxxxxx

The system has failed. Report the following information to

your IBM service representative.

Function 11 . . . . . : A1D03000
Function 12 . . . . . : 69B0015F
Function 13 . . . . . : 0000308F
Function 14 . . . . . : 3FFFDE00
Function 15 . . . . . : 0C211008
Function 16 . . . . . : 00000000
Function 17 . . . . . : 00000000
Function 18 . . . . . : 00D5A400
Function 19 . . . . . : 00CDA400
Type/Model/Feature . . : 9401 150 2270

Warning: The Main Storage Dump (MSD) must be copied for service.
Failure to copy the Main Storage Dump will limit
the ability to diagnose the failure.

Press Enter to copy the MSD for service or view the MSD.

F3=Exit F12=Cancel

Note: Some systems may display Word instead of Function in the Main Storage
Dump Manager Occurred screen.
Follow the instructions of your service provider in responding to this display. In
most cases, you should make a copy of the main storage dump. Save it either to
save media or to auxiliary storage (disk), to assist with diagnosing the problem.

168 OS/400 Backup and Recovery V5R1

 The iSeries Service Functions book has more information about the Main Storage
Dump Manager function.

How to Restart Your System

When you have solved whatever problem caused your system to stop, you must
start it again. In some cases, you start the initial program load (IPL) yourself. In
other cases, such as a power loss, the system starts automatically. When you start
your system again after it ends abnormally, the system tries to put things back in
order. It closes files that were in use, rebuilds access paths that were open, and
verifies file constraints. This process can take a long time.

If you want the system to determine when to rebuild and verify, perform a normal
(automatic) IPL to restart your system. If you want to view and change the
schedules for rebuilding access paths and verifying referential constraints, follow
the steps in this chapter:

Task 1–Performing an Attended IPL

Perform an attended IPL so that you have the opportunity to change rebuild
options.

Note: Your service representative may have started the IPL already. If so, skip to
the step in this task for the display that is currently shown on your system.

To perform an attended initial program load (IPL), you must use the control panel
on the system unit. The steps vary slightly based on the type of system unit that
| you have. Click on Getting started with iSeries under the System planning and
| installation topic in the Information Center if you are unsure of the procedures for
| your system. You can find the Information Center at the following Web site:
http://www.ibm.com/eserver/iseries/infocenter/

| Do the following:
1. If your system unit has a lock on the control panel, use the key or keystick to
unlock the control panel.
2. Place the system in Manual mode.
3. Ensure that any switches for all disk units are in the On position.
4. If your system is currently running, ensure that all users are signed off and all
jobs are ended.

Attention logical partitioning users!

If you are going to use this command on the primary partition, be sure
to power off all secondary partitions before running the command.

Then type:
PWRDWNSYS OPTION(*CNTRLD) DELAY(600) RESTART(*YES)

Note: For the delay parameter, specify a number of seconds that allows your
system time to bring most jobs to a normal end. On a large, busy
system, you may need a longer delay.
5. If your system is not running, power on your system.

Chapter 7. Starting the System After It Ends Abnormally 169

 6. When you see the IPL or Install the System display, select option 1 (Perform
an IPL). Following is an example of a status display. These status displays do
not require any action by the user.

IPL Step in Progress

IPL step . . . . : Storage Management Recovery

The following list shows some of the IPL steps that are shown on the IPL Step
in Progress display:
v Authority Recovery
v Journal Recovery
v Database Recovery
v Journal Synchronization
v Start the Operating System

Some of the IPL steps could take a long time.

While the system is performing the IPL, system reference codes (SRCs) are
displayed on the control panel of the system unit to indicate what step is in
progress. The iSeries Service Functions book describes these SRCs. If the same
SRC is displayed for a long time in solid (not flickering) lights, your system
may have a problem completing the IPL. Look up the SRC in the iSeries
Licensed Internal Code Diagnostic Aids - Volume 1 book or contact software
support.
7. Press the Enter key. Informational messages are displayed.
8. If the Select Product to Work with PTFs display appears, press F3 (Exit) to
continue.
Select Product to Work with PTFs

Position to . . . . . . ._____________ Product

Type options, press Enter. Press F21 to select all.

1=Select

Product
Opt Product Option Release
_ 5769999 *BASE V4R5M0
_ 5769SS1 *BASE V4R3M0

9. You are shown the IPL Options display:

170 OS/400 Backup and Recovery V5R1

 IPL Options

Type choices, press Enter.

System date . . . . . . . . . . . . . . 07 / 26 / 88
System time . . . . . . . . . . . . . . 12 : 00 : 00
Clear job queues . . . . . . . . . . . . N
Clear output queues . . . . . . . . . . N
Clear incomplete job logs . . . . . . . N
Start print writers . . . . . . . . . . Y
Start system to restricted state . . . . N

Set major system options . . . . . . . . Y

Define or change system at IPL . . . . . Y

The values that appear as defaults depend on the recovery steps you have
performed.
10. If the system date and system time are not correct, type the correct values. If
you installed the Licensed Internal Code using option 2 or option 3, the date
and time may be blank. The system date must have a year value in the range
of 87 to 99, or 00 to 22.
11. Specify these responses for the prompts on the display:

Clear job queues N

Clear output queues N
Clear incomplete job logs N
Start print writers Y
Start system to restricted state Y
Set major system options N
Define or change system at IPL N

12. Enter your choices and press the Enter key.

Task 2–Using the Edit Rebuild of Access Paths Display

If access paths are marked for rebuilding, you are shown the following display:
Edit Rebuild of Access Paths SYSTEMA
05/12/90 13:49:34

IPL threshold . . . . . . . . . . . . . 50 0-99

Type sequence, press Enter.

Sequence: 1-99, *OPN, *HLD

------------Access Paths---------- Unique Rebuild

Seq Status File Library Member Keyed Time
25__ IPL QAPZSYM2 QSYS QAPZSYM2 NO 00:00:01
25__ IPL QAPZREQ2 QSYS QAPZREQ2 NO 00:00:01
25__ IPL QAPZPTF3 QSYS QAPZPTF3 NO 00:00:01
25__ IPL QAPZPTF2 QSYS QAPZPTF2 NO 00:00:01
25__ IPL QAPZOBJ2 QSYS QAPZOBJ2 NO 00:00:01
*OPN OPEN QTWALL QSYS QTWALL NO 00:00:06
*OPN OPEN QASULTEL QSYS QASULTEL NO 00:00:01
*OPN OPEN QASULE05 QSYS QASULE05 NO 00:00:01
*OPN OPEN QASULE03 QSYS QASULE03 NO 00:00:01
*OPN OPEN QASULE01 QSYS QASULE01 NO 00:00:01
More...
F5=Refresh F11=Display member text F13=Change multiple F15=Sort by
F16=Repeat position to F17=Position to

Chapter 7. Starting the System After It Ends Abnormally 171

 Note: No access paths are listed if all access paths that are marked for rebuild
have a status of SYS, JRN, or SMAPP.

While you are working with this display, the system is rebuilding access paths. You
can use this display to:
v Change the sequence in which access paths are rebuilt.
v Delay rebuilding some access paths until after the IPL.
1. If you do not want to make changes to this display, press the Enter key. Skip to
step 5. If you want to make changes, continue with step 2.
2. You may change the value of the IPL threshold. All access paths with a
sequence (SEQ) less than or equal to the threshold are rebuilt during the IPL.
Access paths with a greater sequence number are rebuilt after the IPL
completes. The default threshold is 50.
3. You may change the sequence (SEQ) column on the display for specific access
paths. Initially, the sequence numbers are set this way:
25 Files with MAINT(*IMMED) and RECOV(*IPL)
75 Files with MAINT(*IMMED) and RECOV(*AFTIPL)
*OPN Files with MAINT(*DLY)

Within a group (same sequence numbers), the system rebuilds access paths
according to rebuild time, starting with the longest rebuild time.

Rebuild time is an estimate, based on the file size and key length. For journaled
access paths (status JRN) and access paths protected by system-managed
access-path protection (status SMAPP), the rebuild time shows as 0. The system
uses the journal entries to recover these access paths rather than rebuilding
them. The time that is required is minimal.

The estimate for rebuild time assumes that the rebuild job is not contending for
resources. If an access path is rebuilt after the IPL, the rebuild will probably
take longer.
4. Type your changes and press the Enter key. You are shown the Edit Rebuild of
Access Paths display again. You see error messages if the system could not
make some of the changes you requested. For example, you may have tried to
change the sequence number for an access path that the system rebuilt while
you were using the display.
If you have errors, return to step 2.
5. When you have finished the display, press the Enter key without making
changes. You are shown the Display Access Path Status display:
This display is updated every 5 seconds while the system is rebuilding access
Display Access Path Status

IPL Threshold . . . . . . : 50

----------Access Paths---------- Rebuild Current

Status File Library Member Build Time Run Time
RUN QAPZSYM2 QSYS QAPZSYM2 00:00:01 00:00:01
JRN QAPZREQ2 QSYS QAPZREQ2 00:00:01
SYS QASULE03 QSYS QASULE03 00:00:01
IPL QASULE01 QSYS QASULE01 00:00:01

paths.

172 OS/400 Backup and Recovery V5R1

 6. If you want to make changes to the IPL threshold or the sequence for
rebuilding access paths, press F12 (Cancel) to return to the Edit Rebuild of
Access Paths display. Repeat steps 2 through 5.
If you do not want to make changes, you can allow the Display Access Path
Status display to continue updating or you can press F3 (Exit and continue
IPL). In either case, the system completes rebuilding access paths before
continuing to the next step of the IPL.

Task 3–Using the Edit Check Pending Constraints Display

You can define required attributes for physical files on the system. The descriptions
for these values are referential constraints or simply constraints. When you
perform an IPL after the system ends abnormally or when you restore database
files, the system checks the validity of file constraints. Look in the Information
Center under the Database and File System topics for more information about
using referential constraints.

If database constraints are marked for verification, you are shown the following
display:
Edit Check Pending Constraints SYSTEMA
03/30/94 10:09:27
IPL threshold . . . . . . . . . . . . . 50_ 0-99

Type sequence, press Enter.

Sequence: 1-99, *HLD

----------Constraints----------- Verify Elapsed

Seq Status Cst File Library Time Time

75__ AFTIPL CSTF1 FILE567890 LIB4567890 00:00:56 00:00:00

75__ AFTIPL CSTF2 FILE567890 LIB4567890 00:00:56 00:00:00
*HLD INVAP CSTF5 FILE567890 LIB4567890 10:30:06 00:00:00
*HLD CHKPND CSTF6 FILE567890 LIB4567890 09:30:06 00:00:00
*HLD HELD CSTF7 FILE567890 LIB4567890 08:30:06 00:00:00
More...

You can use this display to do the following:

v Change the sequence in which constraints are verified.
v Have the system verify some constraints before the IPL completes.
v Delay verification for some constraints until you specifically request it.
1. If you do not want to make changes to this display, press the Enter key and
skip to step 5 on page 174. If you want to make changes, continue with step 2.
2. You may change the value of the IPL threshold. All constraints with a sequence
(SEQ) less than or equal to the threshold are verified during the IPL.
Constraints with a greater sequence number are verified after the IPL
completes. The default threshold is 50.
3. You may change the sequence (SEQ) column on the display for specific
constraints. Initially, the sequence for all constraints is set to 75.
Within a group (same sequence numbers), the system verifies constraints
according to verify time, starting with the longest estimated time.
Verify that time is an estimate. The estimate assumes that the verify job is not
contending for resources. If a constraint is verified after the IPL, the verify may
take longer.

Chapter 7. Starting the System After It Ends Abnormally 173

 If you want to delay the verification of a constraint indefinitely, specify *HLD for
the sequence. You can use the Edit Check Pending Constraint (EDTCPCST)
command later to set a sequence and have the system verify the constraint.
4. Type your changes and press the Enter key. You are shown the Edit Check
Pending Constraint display again. You see error messages if the system could
not make some of the changes you requested. For example, you may have tried
to change the sequence number for a constraint that the system verified while
you were using the display.
If you have errors, return to step 2 on page 173.
5. When you have finished the display, press the Enter key without making
changes. You are shown the Display Constraint Status display:
This display is updated every 5 seconds while the system is verifying
Display Constraint Status

IPL Threshold . . . . . . : 50

----------Constraints----------- Verify Elapsed

Status Constraint File Library Time Time
RUN CUST1 CUSTMAST CUSTLIB 00:00:04 00:00:01
RUN CUST2 CUSTMAST CUSTLIB 00:00:05 00:00:01
IPL ORDHST1 ORDHIST ORDLIB 00:00:23 00:00:00

F3=Exit and continue IPL F12=Cancel

constraints.
6. If you want to make changes to the IPL threshold or the sequence for verifying
constraints, press F12 (Cancel) to return to the Edit Check Pending Constraints
display. Repeat steps 2 through 5.
If you do not want to make changes, you can allow the Display Constraint
Status display to continue updating or you can press F3 (Exit and continue
IPL). In either case, the system completes verifying constraints before
continuing to the next step of the IPL.
7. When the IPL completes, continue with “Task 4–Recovering from Damaged
Objects and Unreadable Sectors”.

Task 4–Recovering from Damaged Objects and Unreadable

Sectors
If your system stops without warning or disk errors occur, some object description
information may not be updated correctly. If this happens, the object is considered
damaged. When you perform an IPL, the system attempts to locate damaged
objects and write them to the object recovery list. It writes a message (CPI18xx) to
the history (QHST) log for each damaged object that it finds. If any damaged
objects are added to the object recovery list during the IPL, message CPI8197 is
sent to the QSYSOPR message queue.

Note: Some damage, such as damage to the contents of a database file, may not be
detected until the object is used. If you suspect that a large number of
objects on your system have been damaged, contact your service
representative for advice on how to recover.

Do the following to check and recover damaged objects:

1. Display the QHST (history) log by typing DSPLOG and pressing F4 (Prompt).

174 OS/400 Backup and Recovery V5R1

 2. On the prompt display, fill in a starting date and time to limit the number of
entries you see.
3. On the display, fill in *PRINT for the Output prompt and press the Enter key.
4. Type: WRKSPLF. You are shown a list of spooled files for your job.
5. Locate the spooled file for the DSPLOG command. Use option 3 to hold the
spooled file.
6. Use option 5 to display the spooled file.
| 7. Look for entries for damaged objects that are not synchronized. You can use the
| Find function to search for lines that have these character strings: damage and
| sync. Following are some examples of messages you might see:
| CPF3113
| Member damaged
| CPF3175
| File is not synchronized
| CPF3176
| Data space is partially damaged
| CPF3171
| Journal is damaged
| CPF3172
| Objects are not synchronized with the journal
| CPF3173
| Journal receiver is damaged
| CPF3174
| Journal receiver is partially damaged
| CPF700C
| Object of type *object-type cannot be synchronized with journal.
| CPF81xx
| General messages about object damage
| 8. Write down the names and types of the objects you find. Consult Table 39 for
the correct recovery procedure, based on the type of object that is damaged.
Table 39. Recovery for Damaged Objects by Object Type
Type of Object Recovery Procedure

Operating system object in Contact software support for assistance. You may need to
QSYS library install the operating system again.
IBM-supplied user profile Perform an abbreviated installation of the operating system.
Job description that is If no other workstation entries exist for the controlling
specified on the workstation subsystem, the system is not usable. Contact software
entry for the console in the support for assistance.
controlling subsystem
Job queue Perform an IPL. Restore or re-create the damaged job queue.
All entries are lost.
Output queue Perform an IPL. If the output queue is the default output
queue for a printer, it is re-created and its entries are rebuilt.
Other output queues must be restored or re-created. Their
entries are not recovered.
Damaged file whose name Delete the file. Restore it from a backup copy. Run the
starts with QAOSS RCLDLO DLO(*DOCDTL) command.
Database file See “Recovering Damaged Database Files” on page 176

Chapter 7. Starting the System After It Ends Abnormally 175

 Table 39. Recovery for Damaged Objects by Object Type (continued)
Type of Object Recovery Procedure

Journal See “Recovering a Damaged Journal” on page 178.

Journal receiver See “Recovering a Damaged Journal Receiver” on page 179.
| Journaled object See “Recovering a Journaled Object That Is Damaged or Not
Synchronized” on page 179.
All others See “Recovering Other Types of Damaged Objects” on
page 180.

9. Watch for additional indications that objects have been damaged. Some
indications are:
v You cannot start the system because auxiliary storage is full.
v The system has ended abnormally several times since the last time you ran
the Reclaim Storage (RCLSTG) procedure.
v You see objects on the Work with Objects by Owner display that have no
library associated with them.
v The system status display shows an unexpectedly high percentage of
auxiliary storage that is used.
v You cannot access the data in a database file because a member is damaged.
Message CPF8113 indicates this.
v You cannot access objects because a damaged authorization list or authority
holder secures them.

If you see these indications on your system, run the RCLSTG procedure.
Running the procedure is described in “Reclaiming Storage” on page 46.

If you see these indications after a disk unit was replaced and the data was
restored from a partial pump, you should recover the entire ASP that contained
the failed disk unit. See the appropriate checklist.

Recovering Damaged Database Files

Performing a special IPL during which the system analyzes every disk segment for
parts of database objects can recover some types of object-level damage to database
files. Following are examples of object-level damage:
v Lost pointers between the index (access path) and the data.
v Unidentified objects on the object recovery list.

If you are experiencing problems with database files, you can display the Licensed
Internal Code log to determine whether a special IPL may resolve the problems.

Note: You must have *SERVICE special authority to perform the tasks that are
described in this topic.
Do the following:
1. Type STRSST and press the Enter key. You are shown the System Service Tools
(SST) menu menu.
2. Select option 1 (Start a service tool). You are shown the Start a Service Tool
display.
3. Select option 5 (Licensed Internal Code log). You are shown the Licensed
Internal Code Log display.
4. Select option 1 (Select entries from the Licensed Internal Code log). You are
shown the Specify Licensed Internal Code Log Selection Values display.

176 OS/400 Backup and Recovery V5R1

 Specify Licensed Internal Code Log Selection Values

Type choices, press Enter

Note ID:
Starting . . . . . . . . . . . FFFFFFFF 00000000-FFFFFFFF

Entry type:
Major code . . . . . . . . . . 0600 0000-FFFF
Minor code . . . . . . . . . . 145F 0000-FFFF

Starting:
Date. . . . . . . . . . . . . . 00/00/00 MM/DD/YY
Time. . . . . . . . . . . . . . 00:00:00 HH:MM:SS
Ending:
Date. . . . . . . . . . . . . . 00/00/00 MM/DD/YY
Time. . . . . . . . . . . . . . 00:00:00 HH:MM:SS

F3=Exit F12=Cancel

5. Type 0600 for the Major code prompt.

6. Type 145F for the Minor code prompt.
7. For the starting date and time, enter values that approximate when you first
started to have problems.
8. For the ending date and time, enter the current date and time.
9. Press the Enter key. If any errors have been recorded that may be resolved by a
special IPL, you are shown a list of the entries. Otherwise, you receive a
message that no log entries matched your criteria.

If you have log entries that suggest a special IPL, you need to schedule a time for
this IPL. It may take many hours for the system to analyze all the disk segments.
As a rough estimate, the analysis phase of the IPL will take approximately 1
second for each object on your system.

When you are ready to perform the IPL, do the following:

1. Place your system in a restricted state. See “Putting Your System in a
Restricted State” on page 45.
2. Type STRSST and press the Enter key. You are shown the System Service Tools
(SST) menu menu.
3. Select option 1 (Start a service tool). You are shown the Start a Service Tool
display.
4. Select option 4 (Display/Alter/Dump). You are shown the
Display/Alter/Dump Output Device display.
5. Select option 1 (Display/Alter storage). You are shown the Select Data display.
6. Select option 5 (Starting address). You are shown the Specify Starting Address
display:
Specify Starting Address

Output device . . . . . . : Display/Alter storage

Type choice, press Enter.

Address . . . . . . . . . 000000000E 000000

7. Type 000000000E 000000 for the address and press the Enter key. You are
shown the Display Storage display:

Chapter 7. Starting the System After It Ends Abnormally 177

 Display Storage

Control . . . . . . . nnnnn, Pnnnnn, Lcccccc, .cccccc, >

Address . . . . . . . 000000000E 000000

0000 20830048 00800000 00000000 0E00000000 * ................. *

0010 00010000 00000000 00000000 0000000000 * ................. *
0020 00000000 00000000 00000000 0000000000 * ................. *

8. On the third data line (offset 0020), type 8 in the first character. Press F11
(Alter storage) to make the change take affect.
9. Press F3 until you return to the Exit System Service Tools display.
10. Press the Enter key (continue ending SST).
11. On the command line, type
PWRDWNSYS OPTION(*IMMED) RESTART(*YES)

This causes the system to begin the special IPL.

Attention logical partitioning users!

If you are going to use this command on the primary partition, be sure
to power off all secondary partitions before running the command.

Recovering a Damaged Journal

Note: The following steps only apply to recovering a damaged local journal.

Do the following:
1. Type WRKJRN.
2. On the prompt display, type the name of the journal. You are shown the Work
with Journals display:

Work with Journals

Type options, press Enter.

2=Forward recovery 3=Backout recovery 5=Display journal status
6=Recover damaged journal 7=Recover damaged journal receivers
9=Associate receivers with journal

Opt Journal Library Text

JRNACC DSTA1 JOURNAL FOR ACCOUNTS

3. Select option 6 (Recover damaged journal).

| 4. Type: WRKJRNA JRN(library-name/journal-name) OUTPUT(*PRINT). You receive
| a listing that shows all the objects that are currently being journaled.
5. Start journaling for any physical files that should be journaled but are not on
the list by using the STRJRNPF command.
6. Start journaling for any access paths that should be journaled but are not on
the list by using the STRJRNAP command.
| 7. Start journaling for any integrated file system objects that should be journaled
| but are not on the list by using the STRJRN command.
| 8. Start journaling for any other object types that should be journaled but are not
| on the list by using the STRJRNOBJ command.

178 OS/400 Backup and Recovery V5R1

 9. If you previously had any remote journals that were associated with the
damaged journal, add those remote journals again. You can add the remote
journals by using the Add Remote Journal (QjoAddRemoteJournal) API or the
Add Remote Journal (ADDRMTJRN) command.
| 10. Save each journaled object. You should always save objects after you start
| journaling them.

Recovering a Damaged Journal Receiver

Note: The following steps only apply to recovering a damaged journal receiver
that was attached to a local journal.

Do the following:
1. Type WRKJRN.
2. On the prompt display, type the name of the journal that is associated with the
damaged journal receiver. You are shown the Work with Journals display:

Work with Journals

Type options, press Enter.

2=Forward recovery 3=Backout recovery 5=Display journal status
6=Recover damaged journal 7=Recover damaged journal receivers
9=Associate receivers with journal

Opt Journal Library Text

JRNACC DSTA1 JOURNAL FOR ACCOUNTS

3. Select option 7 (Recover damaged journal receivers).

Recovering a Journaled Object That Is Damaged or Not

Synchronized
| For a journaled object that is damaged, do the following:
| 1. Find your most recent saved copy of the object.
| 2. Delete the object.
| 3. Load the save media and restore the object.
| a. For journaled database physical files, data areas, or data queues, type:
| RSTOBJ OBJ(object-name) OBJTYPE(*object-type)
| SAVLIB(library-name)
| DEV(media-device-name)
| b. For journaled IFS objects, type:
| RST DEV('device-path-name')
| OBJ ('object-path-name')
| SUBTREE (*ALL)
| 4. Restore any journal receivers that are needed to recover the object, if they are
| not already on the system.
5. Use the APYJRNCHG command to apply journaled changes. “Applying and
Removing Journaled Changes” on page 444 provides more information about
applying journaled changes.

| Do the following for a journaled object that could not be synchronized:

| 1. Restore the object from your most recent saved copy.
2. Apply journaled changes to bring the object up to date.

Chapter 7. Starting the System After It Ends Abnormally 179

| Note: You may need to perform this procedure for all the objects if there are
| other objects related to the object that is not synchronized. Otherwise,
| the objects may not be synchronized with each other.

Recovering Damaged Objects in the Integrated File System (IFS)

To recover damaged objects in the IFS, run the Reclaim Storage (RCLSTG)
command. For more information on the RCLSTG command, refer to “Reclaiming
Storage” on page 46.

Recovering Other Types of Damaged Objects

Use the following procedure to recover most damaged objects on the system.
Table 39 on page 175 shows which types of objects require special procedures.
1. Find your most recent saved copy of the damaged object.

Note: If the damaged object is in the QSYS library, you may need to restore the
operating system. Contact software support for assistance.
2. Delete the object.
3. Load the save media and restore the object. Type
RSTOBJ OBJ(object-name)
OBJTYPE(object-type)
SAVLIB(library-name)
DEV(media-device-name)

180 OS/400 Backup and Recovery V5R1

Chapter 8. Recovering Information in a User Auxiliary Storage
Pool
When you have user ASPs on your system, you assign specific libraries or objects
to certain physical disk devices. One reason for having user ASPs is to limit the
amount of information you need to recover if a DASD device must be replaced.

The basic process for recovering a user ASP is:

1. Understand what was in the user ASP.
2. Choose the correct recovery procedure.
3. Plan your recovery.
4. Perform the recovery steps.

Describing the Contents of Your User Auxiliary Storage Pools

To choose the correct procedure for recovering the information on your user ASPs,
you must understand what they looked like before the failure. Figure 7 shows an
example of a user ASP configuration. This example is used throughout the
explanations that follow. You may want to begin by drawing a similar picture of
your configuration.

System ASP 1
┌───────────────────────────────────────────────────────────────────────┐
│ ┌───────────────────────────────────────────────────────────────────┐ │
│ │ QSYS Library │ │
│ │ │ │
│ └────┬─────────┬─────────┬─────────┬─────────────┬────────────┬─────┘ │
│ N │ N │ N N │
│ ┌─────────┐ │ ┌──────────┐ │ ┌─────────┐ ┌─────────┐ │
│ │ CUSTLIB │ │ │ ITEMLIB │ │ │ SAVFLIB │ │ $JRNLA │ │
│ │ │ │ │ │ │ │ │ │ JRNA │ │
│ └─────────┘ │ └──────────┘ │ └───────┬─┘ └────┬────┘ │
└────────────────┼───────────────────┼────────────────┼─────────┼───────┘
│ │ │ │
┌─────────────┐ │ ┌─────────────┐ │ ┌────────────┐ │ ┌──────N───────┐
│ ┌─────────┐ │ │ │ ┌─────────┐ │ │ │ ┌─────────┐│ │ │ ┌───────────┐│
│ │ ORDLIB │W┼──┤ │ │Documents│ │ │ │ │$RCVRB ││ │ │ │ RCVA0003 ││
│ └─────────┘ │ ├───┼S│Folders │ │ └─┼S│RCVB0004 ││ │ │ │ *JRNRCV ││
│ ┌─────────┐ │ │ │ │(QDOC0003│ │ │ │ *JRNRCV ││ │ │ └───────────┘│
│ │ TRANLIB │W┼──┤ │ │ library)│ │ │ └─────────┘│ │ │ ┌───────────┐│
│ └─────────┘ │ │ │ └─────────┘ │ │ │ └──┼S│ ORDSAV ││
│ ┌─────────┐ │ │ │ │ │ │ │ │ *SAVF ││
│ │ $JRNLB │W┼──┘ │ │ │ │ │ └───────────┘│
│ │ JRNB │ │ │ │ │ │ │ │
│ └─────────┘ │ │ │ │ │ │ │
└─────────────┘ └─────────────┘ └────────────┘ └──────────────┘
User ASP 2 User ASP 3 User ASP 4 User ASP 5

Figure 7. User ASP Configuration Before Failure

In the example:
v ASP 2 is a library user ASP. It contains these libraries: ORDLIB, TRANLIB, and
$JRNLB.

© Copyright IBM Corp. 1997, 2000, 2001 181

 v Files in the ORDLIB library and TRANLIB library are journaled to the JRNB
journal in the $JRNLB library.
v The journal receivers for the JRNB journal are in the $RCVRB library in ASP 4.
v ASP 3 is a library user ASP containing documents and folders.
v ASP 4 is a library user ASP. It contains the $RCVRB library.
v ASP 5 is a nonlibrary user ASP. It contains the ORDSAV save file. ORDSAV is in
the SAVFLIB library, which is in the system ASP. It also contains the RCVA0003
journal receiver, which is in the $JRNLA library.

Before the failure, the receiver directory for the JRNA journal looks like this:

Work with Receiver Directory

Journal . . . . . . : JRNA Library . . . . . . : $JRNLA

Total size of receivers . . . . . . . . . . . . . . . . . . . : 155648

Type options, press Enter.

4=Delete 8=Display attributes

Attach Save
Opt Receiver Library Number Date Status Date
_ RCVA0001 $JRNLA 00001 06/08/9x SAVED 06/08/9x
_ RCVA0002 $JRNLA 00002 06/09/9x SAVED 06/09/9x
_ RCVA0003 $JRNLA 00003 06/09/9x ATTACHED 00/00/00

Choosing the Procedure to Recover User ASPs

These basic situations may require you to recover information in a user ASP:
v You have replaced a disk unit in the system ASP. Although the data is still in
your user ASPs, you must recover the system’s ability to locate that data
(addressability). This process is described in “How to Recover a User ASP After
Recovering the System ASP”.
v You have replaced a disk unit in a user ASP. All the information that was in the
user ASP must be recovered. If this is your situation, follow the procedure
described in “How to Recover a Damaged User Auxiliary Storage Pool” on
page 196.
v You have replaced a disk unit in the system ASP. One of your user ASPs was in
overflowed status. You must recover the addressability to information in the
user ASPs that were not overflowed by using the procedure described in “How
to Recover a User ASP After Recovering the System ASP”. You must also recover
the information in the user ASP that was overflowed by using the procedure
described in “How to Recover a Damaged User Auxiliary Storage Pool” on
page 196.

How to Recover a User ASP After Recovering the System ASP

Perform this procedure after restoring your Licensed Internal Code and the
operating system.

When you replace a unit in your system ASP, the system loses addressability to the
objects in your user ASPs. The system in the example would look like this after the
operating system has been restored:

182 OS/400 Backup and Recovery V5R1

 System ASP 1
┌───────────────────────────────────────────────────────────────────────┐
│ ┌──────────────────────────────────────────────────────────────────┐ │
│ │ QSYS Library │ │
│ │ │ │
│ └──────────────────────────────────────────────────────────────────┘ │
└────────────────┬───────────────────┬────────────────┬─────────┬───────┘
│ │ │ │
│ │ │ │
┌─────────────┐ │ ┌─────────────┐ │ ┌────────────┐ │ ┌──────N───────┐
│ ┌─────────┐ │ │ │ ┌─────────┐ │ │ │ ┌─────────┐│ │ │ ┌───────────┐│
│ │ ORDLIB │W┼──┤ │ │Documents│ │ │ │ │$RCVRB ││ │ │ │ RCVA0003 ││
│ └─────────┘ │ ├───┼S│Folders │ │ └─┼S│RCVB0004 ││ │ │ │ *JRNRCV ││
│ ┌─────────┐ │ │ │ │(QDOC0003│ │ │ │ *JRNRCV ││ │ │ └───────────┘│
│ │ TRANLIB │W┼──┤ │ │ library)│ │ │ └─────────┘│ │ │ ┌───────────┐│
│ └─────────┘ │ │ │ └─────────┘ │ │ │ └──┼S│ ORDSAV ││
│ ┌─────────┐ │ │ │ │ │ │ │ │ *SAVF ││
│ │ $JRNLB │W┼──┘ │ │ │ │ │ └───────────┘│
│ │ JRNB │ │ │ │ │ │ │ │
│ └─────────┘ │ │ │ │ │ │ │
└─────────────┘ └─────────────┘ └────────────┘ └──────────────┘
User ASP 2 User ASP 3 User ASP 4 User ASP 5

Figure 8. User ASP Configuration After Restoring Operating System

The libraries and objects in the user ASPs are not known to the system.

You can use the procedures described in this topic to recover the objects in your
user ASPs. However, the system cannot recover ownership to the objects other than
document library objects (DLO) in the user ASPs because the addresses for all user
profiles are changed when you restore them. All object types except DLOs use the
address of the user profile to identify the owner.

Recovering object ownership for objects other than DLOs requires manually
assigning ownership for every object in every user ASP.

Task 1–Reclaiming Storage

1. Sign on the system with a user profile that is authorized to the RCLSTG
command. Either sign on at the console or use the Transfer Job (TFRJOB)
command to transfer your job to the controlling subsystem.
2. Type DSPSYSVAL QALWUSRDMN. If the current value does not include the QRCL
(Reclaim Storage) library or does not specify *ALL, use the CHGSYSVAL
command to add QRCL to the list of libraries for this system value. Write the
current value here: __________________
3. Type DSPSYSVAL QCTLSBSD to display the name of your controlling subsystem.
Write the current value here: _________________
4. Ensure your system is in a restricted state. If it is not, follow the procedure in
“Putting Your System in a Restricted State” on page 45.
5. Start the reclaim storage process by typing one of the following:

RCLSTG Reclaim storage of the entire system.

RCLSTG Reclaim storage of the database cross-reference table.
SELECT(*DBXREF)
RCLSTG OMIT(*DBXREF) Reclaim storage of the entire system except the database
cross-reference table.

Chapter 8. Recovering Information in a User Auxiliary Storage Pool 183

 6. Use the CHGSYSVAL command to set the QALWUSRDMN system value back
to its original setting. (You wrote the setting in step 2.)
7. When the reclaim storage procedure finishes, start the controlling subsystem by
typing the following:
STRSBS SBSD(controlling-subsystem)

(You wrote the name of the controlling subsystem in step 3.)

After the reclaim storage procedure, the example system looks like this:

System ASP 1
┌───────────────────────────────────────────────────────────────────────┐
│ ┌──────────────────────────────────────────────────────────────────┐ │
│ │ QSYS Library │ │
│ │ │ │
│ └─────────────┬───────────────────┬────────────────────┬───────────┘ │
│ │ │ N │
│ │ │ ┌─────────┐ │
│ │ │ │ QRCL │ │
│ │ │ └───┬─────┘ │
└────────────────┼───────────────────┼────────────────────┼─────────────┘
│ │ ┌───┴─────┐
│ │ │ │
│ │ │ N
┌─────────────┐ │ ┌─────────────┐ │ ┌────────────┐ │ ┌──────────────┐
│ ┌─────────┐ │ │ │ ┌─────────┐ │ │ │ ┌─────────┐│ │ │ ┌───────────┐│
│ │ ORDLIB │W┼──┤ │ │Documents│ │ │ │ │$RCVRB ││ │ │ │ RCVA0003 ││
│ └─────────┘ │ ├───┼S│Folders │ │ └─┼S│RCVB0004 ││ │ │ │ *JRNRCV ││
│ ┌─────────┐ │ │ │ │(QDOC0003│ │ │ │ *JRNRCV ││ │ │ └───────────┘│
│ │ TRANLIB │W┼──┤ │ │ library)│ │ │ └─────────┘│ │ │ ┌───────────┐│
│ └─────────┘ │ │ │ └─────────┘ │ │ │ └──┼S│ ORDSAV ││
│ ┌─────────┐ │ │ │ │ │ │ │ │ *SAVF ││
│ │ $JRNLB │W┼──┘ │ │ │ │ │ └───────────┘│
│ │ JRNB │ │ │ │ │ │ │ │
│ └─────────┘ │ │ │ │ │ │ │
└─────────────┘ └─────────────┘ └────────────┘ └──────────────┘
User ASP 2 User ASP 3 User ASP 4 User ASP 5

Figure 9. User ASP Configuration After Reclaiming Storage

The system recovers addressability to the objects in ASP 5, but it cannot recover
their original library assignments. They are placed in the QRCL (Recovery) library.
The objects in all user ASPs are owned by the QDFTOWN (Default Owner) user
profile.

“Reclaiming Storage” on page 46 provides more information about the RCLSTG

procedure.

Task 2–Restoring User Profiles

1. Sign on as QSECOFR.
2. Ensure the system is in a restricted state. See “Putting Your System in a
Restricted State” on page 45.
3. Find the most recent save media that has your user profiles. It may be a
SAVSYS media volume or a SAVSECDTA media volume. The file on the save
media volume is called QFILEUPR.
4. If you are using a SAVSYS media volume, type:
RSTUSRPRF DEV(media-device-name) USRPRF(*ALL)
ENDOPT(*LEAVE)

184 OS/400 Backup and Recovery V5R1

 If you are using a SAVSECDTA media volume, type:
RSTUSRPRF DEV(media-device-name) USRPRF(*ALL)
ENDOPT(*UNLOAD)

The time that this takes can vary significantly. “What Happens When You Restore
User Profiles” on page 215 describes what the system does when you restore user
profiles.

Task 3–Restoring the Configuration

1. Find the most recent save media that has your configuration. It may be a
SAVSYS media volume or a SAVCFG media volume. The file on the save media
volume is called QFILEIOC.
2. If you are using a SAVSYS media volume, type:
RSTCFG OBJ(*ALL) DEV(media-device-name)
OBJTYPE(*ALL)
ENDOPT(*LEAVE)

If you are using a SAVCFG media volume, type:

RSTCFG OBJ(*ALL) DEV(media-device-name)
OBJTYPE(*ALL)
ENDOPT(*UNLOAD)

Task 4–Recovering Journals and Journal Receivers in the

QRCL Library
1. Determine if any objects are in the QRCL library. Type: DSPLIB QRCL. You see
the Display Library display.
2. If no objects are listed on the display, skip to “Task 5–Restoring Libraries to the
System Auxiliary Storage Pool” on page 186.
3. If the QRCL library contains objects, save them before proceeding with your
recovery. Load a scratch media volume. Type the following:
SAVLIB LIB(QRCL) DEV(media-device-name)
ENDOPT(*UNLOAD)
4. If the QRCL library does not contain journals or journal receivers, skip to “Task
5–Restoring Libraries to the System Auxiliary Storage Pool” on page 186.
5. Create one or more libraries in the system ASP for the journals and journal
receivers from the QRCL library. The libraries you create must have the same
names as the original libraries that contained the journals and journal receivers.
In the example shown in Figure 9 on page 184, the QRCL library contains the
ORDSAV save file and the RCVA0003 journal receiver. At this point you need
to create the $JRNLA library. You would type: CRTLIB LIB($JRNLA).
6. Move the journals and journal receivers to the newly created libraries. This is
the only circumstance in which you can move journals and journal receivers
between libraries. You must use the MOVOBJ command. You cannot use save
and restore commands. The MOVOBJ command leaves the journal or journal
receiver in the user ASP but establishes its association with the correct library.
For the example shown in Figure 9 on page 184, you would type:
MOVOBJ OBJ(QRCL/RCVA0003) OBJTYPE(*JRNRCV) TOLIB($JRNLA)
7. Delete the QRCL library by typing DLTLIB QRCL.

Note: If the QRCL library contains save files, you will recover them in “Task
9–Recovering Save Files from the QRCL Library” on page 188. When you
recover them, you will use the media volume that you created in step 3.

Chapter 8. Recovering Information in a User Auxiliary Storage Pool 185

 At this point, the system in the example would look like this:

System ASP 1
┌───────────────────────────────────────────────────────────────────────┐
│ ┌──────────────────────────────────────────────────────────────────┐ │
│ │ QSYS Library │ │
│ │ │ │
│ └─────────────┬───────────────────┬────────────────────────────────┘ │
│ │ │ ┌─────────┐ │
│ │ │ │ $JRNLA │ │
│ │ │ │ │ │
│ │ │ └────┬────┘ │
└────────────────┼───────────────────┼──────────────────────────┼───────┘
│ │ │
│ │ │
│ │ N
┌─────────────┐ │ ┌─────────────┐ │ ┌────────────┐ ┌──────────────┐
│ ┌─────────┐ │ │ │ ┌─────────┐ │ │ │ ┌─────────┐│ │ ┌───────────┐│
│ │ ORDLIB │W┼──┤ │ │Documents│ │ │ │ │$RCVRB ││ │ │ RCVA0003 ││
│ └─────────┘ │ ├───┼S│Folders │ │ └─┼S│RCVB0004 ││ │ │ *JRNRCV ││
│ ┌─────────┐ │ │ │ │(QDOC0003│ │ │ │ *JRNRCV ││ │ └───────────┘│
│ │ TRANLIB │W┼──┤ │ │ library)│ │ │ └─────────┘│ │ │
│ └─────────┘ │ │ │ └─────────┘ │ │ │ │ │
│ ┌─────────┐ │ │ │ │ │ │ │ │
│ │ $JRNLB │W┼──┘ │ │ │ │ │ │
│ │ JRNB │ │ │ │ │ │ │ │
│ └─────────┘ │ │ │ │ │ │ │
└─────────────┘ └─────────────┘ └────────────┘ └──────────────┘
User ASP 2 User ASP 3 User ASP 4 User ASP 5

Figure 10. User ASP Configuration After Recovering Isolated Journal Receiver

Task 5–Restoring Libraries to the System Auxiliary Storage

Pool
1. Decide which libraries to restore. You should restore only the libraries in your
system ASP. Do not restore the libraries that are already on your system in user
ASPs.
If you are not sure which libraries are currently on your system, type DSPOBJD
OBJ(*ALL) OBJTYPE(*LIB).

Note: When you install the operating system, the system creates the QGPL
library and the QUSRSYS library. You should still restore these libraries
to restore the data from your saved copy.
2. Plan your restore sequence. If you restore in the wrong sequence, your
journaling environment may not be started again or some objects may not
restore successfully.
| For example, journals must be restored before the journaled objects. If journals
| and objects are in the same library, the system restores them in the correct
| order. If they are in different libraries, or the objects are integrated file system
| objects, you must restore them in the correct order. Similarly, physical files
must be restored before their associated logical files. Read “Sequence for
Restoring Related Objects” on page 45 for more information.
3. Choose the commands or menu options you will use. You can restore libraries
by name or in a group, such as *NONSYS. See “The Relationship Between Save
and Restore Commands” on page 41 for more information.
If you restore libraries in a group, omit the libraries in your user ASPs.

186 OS/400 Backup and Recovery V5R1

 4. Type the restore commands or menu options that you have chosen. In the
example shown in Figure 7 on page 181, libraries were saved using
SAVLIB(*ALLUSR). One way to restore them would be to type:
RSTLIB SAVLIB(*ALLUSR) DEV(media-device-name)
OMITLIB(ORDLIB TRANLIB $JRNLB $RCVRB)

If a media error occurs...

If you have an unrecoverable media error when you are restoring multiple
libraries, see “Recovering from an Error While Restoring Libraries” on
page 56.

Task 6–Restoring Document Library Objects to the System

Auxiliary Storage Pool
1. Find your most recent save media volume that you used to save all of the
documents in the system ASP. You may have specified ASP(1) or ASP(*ANY) for
the save operation. The media volume should have the library QDOC on it.
2. Use the following command to restore the DLOs:
RSTDLO DLO(*ALL) FLR(*ANY) ASP(1)

If a media error occurs...

If you have an unrecoverable media error when you are restoring DLOs,
see “Recovering from an Error While Restoring DLOs” on page 57.

Task 7–Restoring User-Defined File Systems to the User

Auxiliary Storage Pool
| If you are journaling you need to plan your restore sequence. If you restore in the
| wrong sequence, your journaling environment may not be started again or some
| objects may not restore successfully. For example, journals must be restored before
| the journaled objects. If the objects are integrated file system objects, you must
| restore them in the correct order. Read “Sequence for Restoring Related Objects” on
page 45 for more information.

Choose one of the three methods below based on the way in which your
User-Defined File Systems (UDFS) were saved.

Recovery Steps for Unmounted User-Defined File Systems

(UDFS)
1. Load the media volume with the most recent backup of your UDFS when they
were unmounted.
2. To restore a basic ASP, type RST OBJ(('/DEV/QASPxx')) where xx is the
Auxiliary Storage Pool number.
| 3. To restore an independent ASP, type RST OBJ(('/DEV/iasp-name')) where
| iasp-name is the name of the independent ASP.

Recovery Steps for Mounted User-Defined File Systems (UDFS) if

Data is not Restored
Use the following steps if all UDFS are not already restored. The UDFS
information is not saved or restored if the UDFS were saved mounted. You will
need to recreate this information in step 1.

Chapter 8. Recovering Information in a User Auxiliary Storage Pool 187

 1. Create the UDFS exactly as they were before the recovery by using the
CRTUDFS command. Be sure to include the authorities and object auditing.
2. Create the directory that each UDFS was mounted over at the time of the save
by using the CRTDIR command.
3. Mount the UDFS over the directory by using the MOUNT command.

Note: If you were instructed to refer to these steps from another checklist,
return to that checklist now.
4. Restore the UDFS by using the following command:
RST OBJ(('/directory_mounted_over'))

Recovery Steps for Mounted User-Defined File Systems (UDFS) if

Data is Restored

Attention!
This method is not recommended for recovery of UDFS. It is listed only as a
means of recovery if the data is already restored. The previous method,
“Recovery Steps for Mounted User-Defined File Systems (UDFS) if Data is not
Restored” on page 187, is the recommended method.

The UDFS information is not saved or restored if the UDFS were saved mounted.
You will need to recreate this information in step 1.
1. Create the UDFS exactly as they were before the recovery by using the
CRTUDFS command.
2. Create a temporary directory to use as a mount point, using the CRTDIR
command.
3. Mount the UDFS over the temporary directory, using the MOUNT command.
This now becomes your UDFS in the user ASP.
4. Create the directories that are currently in the restored mounted UDFS in the
UDFS that you created in the previous three steps. This tree structure must
exist in order to move or copy the objects.
5. Move or copy the objects in the new UDFS, using the MOV or CPY commands.
6. Unmount the UDFS, using the UNMOUNT command.

Task 8–Reclaiming Document Library Objects

1. If you do not have DLOs in any user ASPs, skip to “Task 9–Recovering Save
Files from the QRCL Library”.
2. Type:
RCLDLO DLO(*ALL) ASP(*ANY)

This procedure rebuilds the association between the DLOs in the user ASP and the
search index records. It also attempts to assign the DLOs to the correct owner.

Task 9–Recovering Save Files from the QRCL Library

If you did not have any save files in the QRCL library, skip to “Task
10–Associating Journal Receivers with Journals” on page 189.

Note: You displayed the QRCL library in “Task 4–Recovering Journals and Journal
Receivers in the QRCL Library” on page 185.

188 OS/400 Backup and Recovery V5R1

 1. Load the scratch media volume that you created in “Task 4–Recovering
Journals and Journal Receivers in the QRCL Library” on page 185.
2. Ensure that the original libraries for the save files were restored in “Task
5–Restoring Libraries to the System Auxiliary Storage Pool” on page 186. You
can check by typing DSPOBJD OBJ(library-name) OBJTYPE(*LIB).
3. Restore each save file from the scratch media volume to the correct library and
user ASP. In the example shown in Figure 7 on page 181, you would type:
RSTOBJ OBJ(ORDSAV) SAVLIB(QRCL)
RSTLIB(SAVFLIB)
OBJTYPE(*SAVF) RSTASP(5)

Task 10–Associating Journal Receivers with Journals

If you do not have any journals or journal receivers involved in the recovery, skip
to “Task 11–Restoring Object Ownership” on page 190.

Whenever you do a recovery involving journals and journal receivers, you should
ensure that your journal receivers are associated with the journal. This topic
provides basic information about how to associate your journals and journal
receivers.

Based on the steps performed so far, the receiver directory for journal JRNA in the
example would look like this:

Work with Receiver Directory

Journal . . . . . . : JRNA Library . . . . . . : $JRNLA

Total size of receivers . . . . . . . . . . . . . . . . . . . : 155648

Type options, press Enter.

4=Delete 8=Display attributes

Attach Save
Opt Receiver Library Number Date Status Date
_ RCVA0003 $JRNLA 00001 06/08/9x ONLINE 00/00/00
_ RCVA1002 $JRNLA 01001 06/09/9x ATTACHED 00/00/00

Notice that when JRNA was restored, the system created a new journal receiver
called RCVA1002 and attached it. The receiver name is based on the name of the
journal receiver that was attached when the journal was saved.

Do the following to associate journals and journal receivers:

1. Type WRKJRN on a command line and press the Enter key.
2. On the prompt display, type the name of the journal and the library name.
3. On the Work with Journals display, type a 9 (Associate receivers with journal)
in the Opt column next to the journal that you want to work with.
4. Press the Enter key.
The receivers are reassociated with the journal.

If any of the journal receivers in the user ASP were created before V3R1, using
option 9 from the Work with Journals display might not associate them in the
correct sequence. If you have journal receivers from a prior release or if any of the
journal receivers you need are not online, do this:
1. Save the journal receivers that are on the system to a scratch media volume:

Chapter 8. Recovering Information in a User Auxiliary Storage Pool 189

 SAVOBJ OBJ(*ALL) LIB(library-name)
DEV(media-device-name) OBJTYPE(*JRNRCV)
VOL(*MOUNTED) ENDOPT(*UNLOAD)
2. After you ensure that the receivers were saved successfully, delete the journal
receivers from the library:
a. Type WRKLIB library-name and press the Enter key. You are shown the work
with library display.
b. Type a 12 (Work with Objects) in the Opt column.
c. Type a 4 (Delete) in the Opt for each journal receiver you want to delete.
d. Press the Enter key.
3. Restore the journal receivers you need from the scratch media volume and
from your save media volumes. Restore them from newest to oldest by typing
the following command for each journal receiver:
RSTOBJ OBJ(receiver-name)
LIB(library-name) DEV(media-device-name)
OBJTYPE(*JRNRCV) VOL(*MOUNTED)
ENDOPT(*UNLOAD)

The receivers are reassociated with the journal.

At this point the receiver directory for JRNA looks like this:
Work with Receiver Directory

Journal . . . . . . : JRNA Library . . . . . . : $JRNLA

Total size of receivers . . . . . . . . . . . . . . . . . . . : 155648

Type options, press Enter.

4=Delete 8=Display attributes

Attach Save
Opt Receiver Library Number Date Status Date
_ RCVA0001 $JRNLA 00001 06/08/9x SAVED 06/08/9x
_ RCVA0002 $JRNLA 00002 06/09/9x SAVED 06/09/9x
_ RCVA0003 $JRNLA 00003 06/08/9x ONLINE 00/00/00
_ RCVA1002 $JRNLA 01002 06/09/9x ATTACHED 00/00/00

Task 11–Restoring Object Ownership

The RCLSTG procedure assigned ownership of all the objects in your user ASPs to
the QDFTOWN user profile. In “Task 8–Reclaiming Document Library Objects” on
page 188, you transferred ownership of DLOs to the correct user profiles. To
transfer ownership of the other objects to the correct user profiles, do the
following:
1. Type WRKOBJOWN USRPRF(QDFTOWN) and press the Enter key.
The Work with Objects by Owner display is shown:

190 OS/400 Backup and Recovery V5R1

 Work with Objects by Owner

User profile . . . . . . . : QDFTOWN

Type options, press Enter.

2=Edit authority 4=Delete 5=Display author
8=Display description 9=Change owner

Opt Object Library Type Attribute

9 ORDRCV001 JRNLIB *JRNRCV
9 ORDHDR ORDLIB *FILE
9 ORDDTL ORDLIB *FILE
9 ORDHST ORDLIB *FILE
9 ORDSAV SAVFLIB *SAVF
9 TRAN01 TRANLIB *FILE
..
.
Parameters or command
===> NEWOWN(OWNORD)
F3=Exit F4=Prompt F5=Refresh F9=Retrieve
F18=Bottom

Note: If you see document library objects on this list (type *DOC or *FLR), one
of the following has occurred:
v You forgot to run RCLDLO. See “Task 8–Reclaiming Document Library
Objects” on page 188.
v The user profile that owns the DLO has not been restored. Restore the
user profile. Then run the RCLDLO command.
v The DLO was owned by the QDFTOWN profile when it was saved.
Determine the correct owner for the DLO and transfer ownership.
2. To transfer ownership of objects individually:
a. Type a 9 in the Opt column for the object and press the Enter key. The
Change Object Owner display is shown.
b. Type the name of the correct owner in the New owner prompt and press the
Enter key.
c. Repeat steps 2a and 2b for each object on the display.
3. To transfer ownership of multiple objects that should have the same owner, use
the technique shown in the display:
a. Type 9 in the Opt column.
b. Type NEWOWN(owner-name) on the parameter line at the bottom of the
display.
c. Press the Enter key. The system transfers ownership of each object you
specified to the new owner.

Stop!
You have completed the recovery of information in your user auxiliary
storage pools. Consult your recovery checklist for the next step in your
recovery process.

How to Recover An Overflowed User Auxiliary Storage Pool

The topics in this section provide basic information about working with user ASPs
in recovery situations. Chapter 25, “Working with Auxiliary Storage Pools”,
provides more information about setting up and managing user ASPs.

Chapter 8. Recovering Information in a User Auxiliary Storage Pool 191

 When the disk units allocated to a user ASP become full, the user ASP is in
overflowed status. The system sends message CPI0953 to the QSYSOPR message
queue warning you that an ASP is approaching its storage threshold. The system
sends message CPI0954 when the storage threshold has been exceeded and the
ASP is in overflowed status.

You should reset a user ASP in overflowed status as soon as possible. An

overflowed ASP affects system performance. It also makes recovery more difficult
and may increase the amount of data lost if a failure occurs. Follow the procedure
in “Resetting An Overflowed User Auxiliary Storage Pool without an IPL”.

Note: To simplify future overflow recovery operations, you can enable automatic
overflow recovery for basic user ASPs with the Operations Navigator disk
management function. For more information, refer to the Information Center
at http://www.ibm.com/eserver/iseries/infocenter.

Resetting An Overflowed User Auxiliary Storage Pool without

an IPL
Do the following to reset a user ASP in overflowed status:
1. Determine which objects in the ASP have overflowed. Use one of these
methods:
v Use the DSPOBJD command to create an output file. Then run a query
against that output file:
a. For the first library in the user ASP, type:
DSPOBJD OBJ(library-name/*ALL) OBJTYPE(*ALL)
DETAIL(*FULL) OUTPUT(*OUTFILE)
OUTFILE(library-name/file-name)
b. For each additional library in the user ASP, type:
DSPOBJD OBJ(library-name/*ALL) OBJTYPE(*ALL)
DETAIL(*FULL) OUTPUT(*OUTFILE)
OUTFILE(library-name/file-name)
OUTMBR(*FIRST *ADD)
c. Create a query against the output file. Look for objects that have a 1 (Yes)
in the ODOASP (overflowed ASP) field.
v For a user ASP that contains only DLOs, use the QRYDOCLIB command. It
has a parameter to display overflowed DLOs.
2. Save each overflowed object to a scratch media volume.
3. Delete each overflowed object.
Some objects, such as journals and physical files, require that you take certain
actions before deleting them. Table 40. shows what to do before deleting these
objects.
Table 40. Object Types That Require Special Procedures for Deleting
Object Type Do This Before Deleting
Journal “Steps before Deleting a Journal” on page 249
Journal receiver “Steps before Deleting a Journal Receiver” on page 251
Physical file “Steps Before Deleting a Physical File” on page 248

4. Ensure that the ASP is no longer in overflowed status. You should have
received a message in the QSYSOPR message queue that the overflow
condition has been recovered. You can also use System Service Tools (SST) to
check:

192 OS/400 Backup and Recovery V5R1

 a. Type STRSST. You are shown the System Service Tools (SST) menu.
b. Select the option to work with disk units.
c. Select the option to display disk configuration.
d. Select the option to display disk configuration capacity. You are shown the
Display Disk Configuration Capacity display:
This display shows whether any ASPs are in overflowed status.
Display Disk Configuration Capacity

--Protected-- --Unprotected--
ASP Unit Type Model Threshold Overflow Size %Used Size %Used
1 90% No 0 0.00% 1400 8.22%
1 9332 400 0 0.00% 200 17.97%
.. 2 9332 400 0 0.00% 200 6.60%
.
2 Yes 0 0.00% 200 99.99%
8 9332 200 90% 0 0.00% 200 99.99%

If the user ASP is still overflowed, follow the procedure described in “Resetting
An Overflowed User Auxiliary Storage Pool during an IPL”.
5. Before you can restore the overflowed objects from a media volume, you must
make additional space available in the user ASP. Do one or more of the
following:
v Delete objects from the ASP if you no longer need them.
v Move one or more libraries to a different ASP.

Note: You cannot use the MOVOBJ command to do this. You must save the
library, delete it, and restore it to a different ASP.
v Move one or more folders to a different ASP by saving the folder, deleting it,
and restoring it to a different ASP.
v Add additional disk units to the ASP.
6. After you have made additional space available in the ASP, restore the objects
you saved in step 2 on page 192.
7. Check to make sure the user ASP has sufficient space and is not overflowed.
Repeat the procedure described in step 4 on page 192.

Resetting An Overflowed User Auxiliary Storage Pool during

an IPL
Sometimes you are not able to find all the overflowed objects in a user ASP. If you
have taken the steps described in “Resetting An Overflowed User Auxiliary
Storage Pool without an IPL” on page 192 and the user ASP is still overflowed, you
can perform a manual IPL to reset the user ASP. Do the following:
1. Ensure that you have enough space to reset the overflowed user ASP. Do the
following:
a. Type STRSST. You are shown the System Service Tools (SST) menu.
b. Select the option to work with disk units.
c. Select the option to display disk configuration.
d. Select the option to display disk configuration capacity. You are shown the
Display Disk Configuration Capacity display:

Chapter 8. Recovering Information in a User Auxiliary Storage Pool 193

 Display Disk Configuration Capacity

--Protected-- --Unprotected--
ASP Unit Type Model Threshold Overflow Size %Used Size %Used
1 90% No 0 0.00% 1400 8.22%
1 9332 400 0 0.00% 200 17.97%
.. 2 9332 400 0 0.00% 200 6.60%
.

2 Yes 0 0.00% 200 99.99%

8 9332 200 90% 0 0.00% 200 99.99%

This display shows whether any ASPs are in overflowed status.

e. Press F9 (Display ASP Overflow information) to display the overflow
amount and the additional amount of storage that is needed in the ASP to
recover the overflowed objects.

Display ASP Overflow Information

Overflow ----Amount Needed to Recover---- ASP Threshold Amount
To Capacity To Threshold
2 90% 14 0 0
3 90% 25 25 45

f. If the amount in the To Capacity field is greater than zero, the ASP will still
be overflowed when the recovery completes. There is not enough free space
in the user ASP to contain the overflowed data.
g. If you do not have enough space, repeat the instructions in step 5 on
page 193 to free more space.
2. Do the following to put your system in a restricted state:
a. Before putting your system in a restricted state, ensure that all users are
signed off and all jobs are ended.
b. To receive notification that the subsystems have ended, type the following
and press the Enter key:
CHGMSGQ MSGQ(QSYSOPR) DLVRY(*BREAK)
SEV(60)
c. To end all subsystems, type the following:
ENDSBS SBS(*ALL) OPTION(*CNTRLD)
DELAY(600)

Note: For the delay parameter, specify a number of seconds that allows
your system time to bring most jobs to a normal end. On a large,
busy system, you may need a longer delay.

A message is sent that indicates that the procedure for ending subsystems is in
progress. A final message is sent when the system is in a restricted state.
3. Perform a manual IPL and access DST:
Use this procedure to start DST. If the IPL or Install the System menu is already
displayed, start with step 5 on page 202.
a. Ensure the keystick is in the system unit control panel.
b. Place the system in manual mode.
c. Power down the system:
PWRDWNSYS OPTION(*CNTRLD) DELAY(600)
RESTART(*YES) IPLSRC(B)

194 OS/400 Backup and Recovery V5R1

 Attention logical partitioning users!
If you are going to use this command on the primary partition, be sure
to power off all secondary partitions before running the command.

Note: If you are sure that no jobs are running on your system, you can
specify OPTION(*IMMED) when you power down the system.
Otherwise, specify a delay time that is sufficient to allow jobs to end
normally.
d. When the IPL completes, the IPL or Install the System menu appears.

IPL or Install the System

Select one of the following:

1. Perform an IPL
2. Install the operating system
3. Use Dedicated Service Tools (DST)
4. Perform automatic installation of the operating system
5. Save Licensed Internal Code

4. Select option 1 (Perform an IPL). You are shown the Disk Configuration
Attention Report:
If you type 5 in the Option field, the following screen is displayed, listing the
Disk Configuration Attention Report

Type option, press Enter.

5=Display Detailed Report
Press F10 to accept all the following problems and continue.
The system will attempt to correct them.

Opt Problem
Overflowed ASPs

User ASPs that are overflowed.

Recover Overflowed User ASP

The following user ASPs are overflowed.

ASP
2
3

5. Press the F10 key to request recovery of the overflowed user ASPs. The
recovery takes place during the storage management recovery phase of the IPL.
The operation takes from several minutes to a few hours, depending on the
number of objects on the system and the amount of data to be recovered.
6. When the IPL of the system is complete, the Sign On display is shown.
7. Sign on and verify the results by checking the messages in the QSYSOPR
message queue.

Chapter 8. Recovering Information in a User Auxiliary Storage Pool 195

 How to Delete Overflowed Objects during Recovery
Use this procedure if you are recovering a user ASP that was in overflowed status.
1. After running the RCLSTG procedure, display the contents of the QRCL library
by typing: DSPLIB QRCL
2. Write down the names of the objects in the library. These objects were
overflowed into the system ASP at the time of the failure.
Although the initial disk extents for these objects may have been allocated in
the system ASP, portions of the objects may still have been lost. The integrity of
these objects cannot be predicted. They should be deleted and recovered.
3. Delete the overflowed objects. You must take special action before deleting
certain types of objects. See Table 40 on page 192 for more information.
4. When you run the RCLSTG command, any documents from the lost user ASP
that were in overflowed status are placed in the user ASP again. (The system
creates a new QDOCnnnn library, where nnnn is the number of the lost ASP,
and places the overflowed DLOs in it.)
Assuming you have not yet restored DLOs to the user ASP, type this to delete
the overflowed DLOs:
DLTDLO DLO(*ALL) FLR(*ANY) ASP(n)

where n is the number of the ASP whose data was lost.

How to Recover a Damaged User Auxiliary Storage Pool

Perform this procedure if one of the following is true:
v The service representative has replaced a failed disk unit in a user ASP. When
you lose a disk unit in an ASP, you must recover all of the information in that
ASP. The information in other ASPs on your system is not affected.
v The system has reassigned sectors on a disk unit, but object damage has
occurred.
v You have replaced a disk unit in the system ASP and one or more user ASPs
was overflowed.

Task 1–Restoring User Profiles

Even though user profiles are not lost when you replace a unit in a user ASP, they
must be restored to prepare for restoring authority to objects in the user ASP. Do
the following:
1. Sign on with the QSECOFR user profile.
| 2. End all subsystems with the ENDSBS command and go to a restricted state.
3. Load your most recent SAVSYS or SAVSECDTA media volume.
4. Restore all user profiles. Type:
RSTUSRPRF DEV(media-device-name) USRPRF(*ALL)
ENDOPT(*UNLOAD)
| 5. If you are restoring an independent user ASP, skip to “Task 7–Restoring
| User-Defined File Systems to the User Auxiliary Storage Pool” on page 200.
6. If you know what libraries or objects were in the user ASP that was lost, skip
to “Task 3–Determining Tasks to Restore Objects” on page 197.
If you do not know what was in the user ASP, continue with “Task
2–Determining the Contents of the Lost Auxiliary Storage Pool” on page 197.

196 OS/400 Backup and Recovery V5R1

Task 2–Determining the Contents of the Lost Auxiliary Storage
Pool
If your system has a small number of libraries and is well-documented, like the
one in Figure 7 on page 181, your task is relatively simple. In the example, if a disk
unit in ASP 2 is replaced, the user must recover the ORDLIB, TRANLIB, and
$JRNLB libraries. If a disk unit in ASP 5 is replaced, the user must recover all the
journal receivers in the $JRNLA library and the ORDSAV save file in the SAVFLIB
library.

If you are not sure what was on the user ASP, do this:
1. Sign on with a user profile that has *ALLOBJ special authority so that your
listings show all libraries.
2. Print a list of the libraries that are on the lost user ASP by doing the following:
a. Create a list of all the libraries in an output file:
DSPOBJD OBJ(QSYS/*ALL) OBJTYPE(*LIB) OUTPUT(*PRINT)
DETAIL(*FULL) OUTPUT(*OUTFILE)
OUTFILE(library-name/file-name)
b. Use a program or a query tool to display or print the output file. Select all
entries that have an ASP field that matches the ASP that is lost.
Notes:
1) When you lose a user ASP, you lose the contents of any libraries in the
ASP, not the libraries themselves. The library objects are in the QSYS
library, which is in the system ASP.
2) If you had documents in the user ASP, you should have a library on
your listing for the ASP. The library name is QDOCnnnn, where nnnn is
the number of the ASP.
3. If you have determined what must be recovered, continue with “Task
3–Determining Tasks to Restore Objects”. If you have not found any libraries to
recover, continue with step 4.
4. If you did not find any libraries to recover in step 2, the ASP was probably a
nonlibrary user ASP. A nonlibrary user ASP can contain only save files,
journals, and journal receivers.
Determining the objects that were in a nonlibrary user ASP can be very time
consuming. The following steps are one method. This method works only if
you have not yet run RCLSTG after losing the user ASP.
a. Type the following:
DSPOBJD OBJ(*ALL/*ALL)
OBJTYPE(*LIB *FILE *JRN *JRNRCV)
OUTPUT(*OUTFILE)
OUTFILE(library-name/file-name)
b. Use a program or query tool to list all the objects in the output file that are
in the ASP that is damaged.
5. When you have determined the objects that need to be recovered, continue
with “Task 3–Determining Tasks to Restore Objects”.

Task 3–Determining Tasks to Restore Objects

1. Use Table 41 on page 198 to determine how to recover the objects in your user
ASP. It shows the recovery tasks you must perform, based on the contents of
the user ASP you are recovering.
2. If you have different types of objects to recover, such as libraries and
documents, perform the tasks in the order shown in the table.

Chapter 8. Recovering Information in a User Auxiliary Storage Pool 197

 Table 41. Tasks for Restoring User ASP Objects
Type of ASP Contents Recovery Tasks

Library User ASP Libraries “Task 4–Restoring Libraries to the User Auxiliary
Storage Pool”
Nonlibrary User Journals “Task 5–Restoring Journals to the User Auxiliary
ASP Storage Pool”
Library User ASP Documents “Task 6–Restoring Documents to the User Auxiliary
Storage Pool” on page 200
| Library User ASP User-defined file “Task 7–Restoring User-Defined File Systems to the
| systems User Auxiliary Storage Pool” on page 200
Nonlibrary User Journal receivers “Task 8–Restoring Journal Receivers to the User
ASP Auxiliary Storage Pool” on page 201
Nonlibrary User Save files “Task 9–Restore Save Files to the User Auxiliary
ASP Storage Pool” on page 201

Task 4–Restoring Libraries to the User Auxiliary Storage Pool

1. Sign on with a user profile that has *SAVSYS and *JOBCTL special authority.
2. For each library you need to recover, load the correct volume from your latest
save media volumes.
3. Type:
RSTLIB SAVLIB(library-name) DEV(media-device-name)
ENDOPT(*LEAVE)

Note: You should restore changed objects and apply journaled changes for all
the ASPs included in your recovery at the same time. These steps appear
on the recovery checklist at the appropriate point.
4. Continue with the next task shown in Table 41. If you have completed all the
appropriate tasks in the table, continue with the next task in the recovery
checklist from Chapter 4.

Task 5–Restoring Journals to the User Auxiliary Storage Pool

1. Sign on with a user profile that has *SAVSYS and *JOBCTL special authority.
2. For each journal you need to recover, load the correct save media volume and
type:
RSTOBJ OBJ(journal-name) SAVLIB(library-name)
DEV(media-device-name) OBJTYPE(*JRN)

When you restore the journal, the system automatically creates and attaches a
new journal receiver. “Chapter 19. Planning and Setting Up Journaling” on
page 383 describes how the system names the journal receiver that is created
when you restore a journal.
3. Establish your journaling environment again. Do this:
a. For each database physical file that was journaled to the restored journal,
type:
STRJRNPF FILE(library-name/file-name)
JRN(library-name/journal-name)

| Note: To determine what options you specified for the file when you last
| journaled it, you can use the Display File Description (DSPFD) or
| Display Object Description (DSPOBJD) commands for the file to find
| out.
b. For each access path that was journaled to the restored journal, type:

198 OS/400 Backup and Recovery V5R1

 STRJRNAP FILE(library-name/file-name)
JRN(library-name/journal-name)
| c. For each IFS object that was journaled to the restored journal, type:
| STRJRN OBJ ('object-path-name')
| JRN('journal-path-name')

| Note: To determine what options you specified for the object when you last
| journaled it, you can use the Display Link (DSPLNK) command.
| d. For all other object types that were journaled, type one of the following:
| STRJRNOBJ OBJ(library-name/object-name)
| OBJTYPE(object-type)
| JRN(library-name/journal-name)

| Note: To determine what options you specified for the object when you last
| journaled it, you can use the Display Object Description (DSPOBJD)
| command.
| e. Save each object that you started journaling.
4. If you need to restore journal receivers for the journals, skip to “Task
8–Restoring Journal Receivers to the User Auxiliary Storage Pool” on page 201.
5. Associate journal receivers with the journals you restored. Do the following:
a. Type WRKJRN on a command line and press the Enter key.
b. On the prompt display, type the name of the journal and the library name.
c. On the Work with Journals display, type a 9 (Associate receivers with
journal) in the Opt column next to the journal that you want to work with.
d. Press the Enter key.
The receivers are reassociated with the journal.

If any of the journal receivers in the user ASP were created before V3R1, using
option 9 from the Work with Journals display might not associate them in the
correct sequence. If you have journal receivers from a prior release or if any of
the journal receivers you need are not online, do this:
a. Save the journal receivers that are on the system to a scratch media volume:
SAVOBJ OBJ(*ALL) LIB(library-name)
DEV(media-device-name) OBJTYPE(*JRNRCV)
VOL(*MOUNTED) ENDOPT(*UNLOAD)
b. After you ensure that the receivers were saved successfully, delete the
journal receivers from the library:
1) Type WRKLIB library-name and press the Enter key. You are shown the
work with library display.
2) Type a 12 (Work with Objects) in the Opt column.
3) Type a 4 (Delete) in the Opt for each journal receiver you want to delete.
4) Press the Enter key.
c. Restore the journal receivers that you need from both the scratch media
volume and from your save media volumes. Restore them from newest to
oldest by typing the following command for each journal receiver:
RSTOBJ OBJ(receiver-name)
LIB(library-name) DEV(media-device-name)
OBJTYPE(*JRNRCV) VOL(*MOUNTED)
ENDOPT(*UNLOAD)

The receivers are reassociated with the journal.

Chapter 8. Recovering Information in a User Auxiliary Storage Pool 199

 6. Continue with the next task shown in Table 41 on page 198. If you have
completed all the appropriate tasks in the table, continue with the next task in
the recovery checklist from Chapter 4.

Task 6–Restoring Documents to the User Auxiliary Storage

Pool
1. Sign on with a user profile that has *SAVSYS and *JOBCTL special authority.
2. Load the media volume with your last complete save of documents in the user
ASP.
3. Restore the documents to the user ASP by typing:
RSTDLO DLO(*ALL) SAVASP(ASP-number)
RSTASP(ASP-number)

This restores the documents and makes any changes necessary to the search
index database files.
4. Use the Query Document Library (QRYDOCLIB) to locate any documents that
were created on the user ASP since the last save operation. Query by ASP
number and creation date. Inform your users that these documents were lost
and develop a plan to re-create them.
5. Continue with the next task in the recovery checklist from Chapter 4.

Task 7–Restoring User-Defined File Systems to the User

Auxiliary Storage Pool
Choose one of the three methods below based on the way in which your
User-Defined File Systems (UDFS) were saved.

Recovery Steps for Unmounted User-Defined File Systems

(UDFS)
1. Load the media volume with the most recent backup of your UDFS when they
were unmounted.
2. To restore a basic ASP, type RST OBJ(('/DEV/QASPxx')) where xx is the auxiliary
storage pool number.
| 3. To restore an independent ASP, type RST OBJ(('/DEV/iasp-name')) where
| iasp-name is the name of the independent ASP.

Recovery Steps for Mounted User-Defined File Systems (UDFS) if

Data is not Restored
Use the following steps if all UDFS are not already restored. The UDFS
information is not saved or restored if the UDFS were saved mounted. You will
need to recreate this information in step 1.
1. Create the UDFS exactly as they were before the recovery by using the
CRTUDFS command. Be sure to include the authorities and object auditing.
2. Create the directory that each UDFS was mounted over at the time of the save
by using the CRTDIR command.
3. Mount the UDFS over the directory by using the MOUNT command.

Note: If you were instructed to refer to these steps from another checklist,
return to that checklist now.
4. Restore the UDFS by using the following command:
RST OBJ(('/directory_mounted_over'))

200 OS/400 Backup and Recovery V5R1

 Recovery Steps for Mounted User-Defined File Systems (UDFS) if
Data is Restored

Attention!
This method is not recommended for recovery of UDFS. It is listed only as a
means of recovery if the data is already restored. The previous method,
“Recovery Steps for Mounted User-Defined File Systems (UDFS) if Data is not
Restored” on page 187, is the recommended method.

The UDFS information is not saved or restored if the UDFS were saved mounted.
You will need to recreate this information in step 1.
1. Create the UDFS exactly as they were before the recovery by using the
CRTUDFS command.
2. Create a temporary directory to use as a mount point, using the CRTDIR
command.
3. Mount the UDFS over the temporary directory, using the MOUNT command.
This now becomes your UDFS in the user ASP.
4. Create the directories that are currently in the restored mounted UDFS in the
UDFS that you created in the previous three steps. This tree structure must
exist in order to move or copy the objects.
5. Move or copy the objects in the new UDFS, using the MOV or CPY commands.
6. Unmount the UDFS, using the UNMOUNT command.

Task 8–Restoring Journal Receivers to the User Auxiliary

Storage Pool
1. Sign on with a user profile that has *SAVSYS and *JOBCTL special authority.
2. For each journal receiver you need to recover, load the correct save media
volume and type:
RSTOBJ OBJ(receiver-name) SAVLIB(library-name)
DEV(media-device-name) OBJTYPE(*JRNRCV)

Note: For journal receivers created before V3R1, restore the journal receivers
from newest to oldest to ensure that they are associated with the journal
in the correct sequence. If all the journal receivers were created on V3R1
or later, you can restore them in any sequence.
3. Continue with the next task shown in Table 41 on page 198. If you have
completed all the appropriate tasks in the table, continue with the next task in
the recovery checklist from Chapter 4.

Task 9–Restore Save Files to the User Auxiliary Storage Pool

1. Sign on with a user profile that has *SAVSYS and *JOBCTL special authority.
2. For each save file you need to recover, load the correct save media volume and
type:
RSTOBJ OBJ(save-file-name) SAVLIB(library-name)
DEV(media-device-name) OBJTYPE(*SAVF)

Note: This command restores the description of the save file and its contents, if
you specified SAVFDTA(*YES) when you saved the save file. If you
specified SAVFDTA(*NO) when you saved the save file, this command
restores only the save file description.

Chapter 8. Recovering Information in a User Auxiliary Storage Pool 201

 3. Continue with the next task in the recovery checklist from Chapter 4.

How to Remove a Failed Disk Unit from the System ASP

Use this procedure to remove a disk unit from your configuration if the unit has
failed. This procedure allows you to return your system to operation if a
replacement disk unit is not immediately available. However, this procedure
removes all data from your system and requires a complete restore operation.

After performing this procedure, your system will have less disk capacity. You may
not be able to restore all user information until you have installed and configured
a replacement disk unit.

Before you perform this procedure, ensure that the remaining 2800-001 storage
units in your system ASP are large enough for a main storage dump. Consult
software support or refer to “Chapter 25. Working with Auxiliary Storage Pools” on
page 669.

Task 1–Access Dedicated Service Tools

Use this procedure to start DST. If the IPL or Install the System menu is already
displayed, start with step 5.
1. Ensure the keystick is in the system unit control panel.
2. Place the system in manual mode.
3. Power down the system:
PWRDWNSYS OPTION(*CNTRLD) DELAY(600) RESTART(*YES)
IPLSRC(B)

Attention logical partitioning users!

If you are going to use this command on the primary partition, be sure to
power off all secondary partitions before running the command.

Note: If you are sure that no jobs are running on your system, you can specify
OPTION(*IMMED) when you power down the system. Otherwise, specify a
delay time that is sufficient to allow jobs to end normally.
4. When the IPL completes, the IPL or Install the System menu appears.

IPL or Install the System

Select one of the following:

1. Perform an IPL
2. Install the operating system
3. Use Dedicated Service Tools (DST)
4. Perform automatic installation of the operating system
5. Save Licensed Internal Code

5. Select option 3 (Use Dedicated Service Tools (DST)) and press the Enter key.
The Dedicated Service Tools (DST) Sign On display is shown.

202 OS/400 Backup and Recovery V5R1

 Dedicated Service Tools (DST) Sign On
Type choice, press Enter.
DST user . . . . . . . . . . _______
DST password . . . . . . . . _______

6. In the DST user field, type QSECOFR. In the DST password field, type your DST
security-level or full-level password. On a new system, the password is
QSECOFR. The password is case sensitive; type in all capital letters. You should
change this password after you complete your installation procedures. The
iSeries Security Reference book has more information about DST passwords.
The Use Dedicated Service Tools (DST) menu is shown.

Use Dedicated Service Tools (DST)

Select one of the following:
1. Perform an IPL
2. Install the operating system
3. Work with licensed internal code
4. Work with disk units
5. Work with DST environment
6. Select DST console mode
7. Start a service tool
8. Perform automatic installation of the operating system
9. Work with save storage and restore storage
10. Work with remote DST support

Task 2–Delete the Auxiliary Storage Pool Data

1. From the Use Dedicated Service Tools (DST) menu, do the following:

┌────────────────────────┐ Select option 4

│ DST Menu │ (Work with disk units)
│ │
│ ┌───────────────────┴─────┐ Select option 1
│ 4 │ Work with Disk Units │ (Work with disk configuration)
└────┤ │
│ ┌────────────────────┴────┐ Select option 3
│ │ Work with Disk │ (Work with ASP configuration)
│ 1 │ Configuration │
└────┤ ┌────────────────────┴─────┐
│ 3 │ Work with ASP │
└────┤ Configuration │
│ │
│ │
└──────────────────────────┘

2. Select option 4 (Delete ASP data) on the Work with ASP Configuration display.

Note: Selecting this option deletes all data in the system ASP. Do not use this
procedure unless you have a failed disk unit and there is no immediate
replacement for the disk unit.

Chapter 8. Recovering Information in a User Auxiliary Storage Pool 203

 Select ASP to Delete Data From
Type options, press Enter

4=Delete ASP data

Option

ASP Threshold Overflow --Protected-- --Unprotected

Size %Used Size %Used
1 90% No 0.00 0.00% 1200 74.84%
2 90% Yes 0.00 0.00% 200 99.99%
3 90% Yes 0.00 0.00% 200 99.99%

3. Type a 4 in the Option column to select the ASP that you want to delete the
data from and press the Enter key. The following display is shown.

Confirm Delete ASP Data

Warning: All data will be deleted from the selected ASPs.

You have selected to delete data from ASP 1. This will prevent
you from changing the disk configuration in some ways until
the system is IPLed again to DST.

Press F10 to confirm your choice for 4=Delete ASP data.

Press F12=Cancel to return to change your choice.

--Protected-- --Unprotected--
Option ASP Threshold Overflow Size %Used Size %Used
4 1 90% No 0 0.00 1200 *

4. Press F10 (Confirm) to confirm your choice to delete the ASP data.
5. When the delete of the ASP data is complete, you return to the Use Dedicated
Service Tools (DST) menu.

Task 3–Remove the Disk Unit from the Auxiliary Storage Pool
Configuration
To remove the disk unit from the ASP, do the following:
1. If you are not already using DST, perform a manual IPL to start DST. See “How
to Start Dedicated Service Tools (DST)” on page 660.
2. From the Use Dedicated Service Tools (DST) menu, do the following:

┌────────────────────────┐ Select option 4

│ DST Menu │ (Work with disk units)
│ │
│ ┌───────────────────┴─────┐ Select option 1
│ 4 │ Work with Disk Units │ (Work with disk configuration)
└────┤ │
│ ┌────────────────────┴────┐ Select option 3
│ │ Work with Disk │ (Work with ASP configuration)
│ 1 │ Configuration │
└────┤ ┌────────────────────┴─────┐
│ 3 │ Work with ASP │
└────┤ Configuration │
│ │
│ │
└──────────────────────────┘

3. The Remove Units From Configuration) display is shown.

204 OS/400 Backup and Recovery V5R1

 Remove Units from Configuration

Type options, press Enter.

4=Remove unit from configuration

Serial Resource
OPT Unit ASP Number Type Model Name Status
2 1 10-00A7529 9332 400 DD010 Configured
3 1 10-00A4936 9332 400 DD012 Configured
4 1 10-00A4936 9332 400 DD014 Configured
4 5 1 10-00A7498 9332 400 DD015 Configured
4 6 1 10-00A7498 9332 400 DD017 Configured
7 1 10-00A7530 9332 400 DD018 Configured
8 1 10-00A7530 9332 400 DD021 Configured

4. Type a 4 (Remove unit from configuration) in the OPT column for each unit that
you want to remove and press the Enter key. If the remove operation would
leave the ASP with insufficient storage, you receive an error message.
If you see the Confirm Remove Disk Units display, skip to 6.
The Confirm Continuation display may be shown before the Confirm Remove
Disk Units display if the storage management directories are not usable.

Confirm Continuation

To proceed, the system must perform internal processing

that may take several minutes during which the system may
appear inactive.

Press Enter to continue.

Press F12=Cancel to return and change your choice.

5. Determine whether you want to cancel the procedure or continue. If you want
to continue, press the Enter key.
6. The Confirm Remove Disk Units display is shown:
Press F9 (Capacity information to display the resulting capacity.
Confirm Remove Disk Units

Removing disk units will take several minutes.

Press Enter to confirm remove of disk units.

Press F9=Capacity information to display the capacity information.
Press F12=Cancel to return to change your choice.

Serial Resource
OPT Unit ASP Number Type Model Name Status
4 5 1 10-00A7498 9332 400 DD010 Configured
4 6 1 10-00A7498 9332 400 DD012 Configured

Chapter 8. Recovering Information in a User Auxiliary Storage Pool 205

 Resulting Capacity

The configuration change that you requested would result in the

following ASP capacities.

Press Enter to continue.

-----------Current---------- ----------Modified--------
--Protected-- -Unprotected- --Protected-- -Unprotected-
ASP Threshold Size %Used Size %Used Size %Used Size %Used
1 90% 0 0.00% 1600 52.70% 0 0.00% 1200 70.26%

7. Press the Enter key to return to the Confirm Remove Disk Units display.
8. Press the Enter key on the Confirm Remove Disk Units display to remove the
selected units. The system moves the data off the units selected to be removed
to the remaining units in the source ASP. The remove can take several minutes
or several hours during which the system appears inactive.
Notes:
a. The time it takes to remove a unit depends on the disk unit type and
model.
b. If the data on the unit being removed is severely fragmented and the
amount of storage used is high, the remove operation could take several
hours.
9. When the remove operation is complete, you return to the Work with ASP
Configuration display.
Press F3 until you return to the Use Dedicated Service Tools (DST) display.

206 OS/400 Backup and Recovery V5R1

Chapter 9. The Restore Menu
The Restore menu provides many options for recovering information. Figure 11
shows the menu. Options marked with a plus sign (+) require the system to be in a
restricted state. When your system is in a restricted state, that does not prevent
client workstations from attempting to access information. If you have directories
managed by the Windows server on iSeries, you should vary off the network
server descriptions.

RESTORE Restore

Select one of the following:

Restore Data
1. Files
2. Libraries
3. Documents and folders
4. Programs
5. Other objects
6. Licensed programs
7. Configuration
+ 8. User profiles
9. Objects in directories

Figure 11. Restore Menu–First Display

You can page down on the Restore menu to see additional options:

Restore System and User Data™

+ 21. System and user data
+ 22. System data only
+ 23. All user data
Restore Office Data
30. All documents, folder, and mail
31. Documents and folders
32. Mail only
33. Calendars
Restore Libraries
+ 40. All libraries other than system library
41. All IBM libraries other than system library
42. All user libraries
Restore from Different Systems
50. Restore from System/36 format

© Copyright IBM Corp. 1997, 2000, 2001 207

What the Restore Menu Options Do
Following are the commands the system runs for the menu options that restore
either the system, the system data only, or all user data. The name of the CL
program that the system runs is in parentheses () following the description of the
menu option. You can change this CL program if you need different values than
the system-supplied defaults.

Option Number Description and Commands

21 System and user data (QMNRSTE):
ENDSBS SBS(*ALL) OPTION(*IMMED)
RSTUSRPRF USRPRF(*ALL)
RSTCFG OBJ(*ALL)
RSTLIB SAVLIB(*NONSYS)
RSTDLO DLO(*ALL) SAVFLR(*ANY)
RST DEV('/QSYS.LIB/media-device-name.DEVD')
OBJ(('/*') ('/QSYS.LIB' *OMIT)
('/QDLS' *OMIT))
RSTAUT
STRSBS SBSD(controlling subsystem)
22 System data only (QSRRSTI):
ENDSBS SBS(*ALL) OPTION(*IMMED)
RSTUSRPRF USRPRF(*ALL)
RSTCFG OBJ(*ALL)
RSTLIB SAVLIB(*IBM)
RST DEV('/QSYS.LIB/media-device-name.DEVD')
OBJ(('/QIBM/ProdData')
('/QOpenSys/QIBM/ProdData'))
STRSBS SBSD(controlling subsystem)
23 All user data (QSRRSTU):
ENDSBS SBS(*ALL) OPTION(*IMMED)
RSTUSRPRF USRPRF(*ALL)
RSTCFG OBJ(*ALL)
RSTLIB SAVLIB(*ALLUSR)
RSTDLO DLO(*ALL) SAVFLR(*ANY)
RST DEV('/QSYS.LIB/media-device-name.DEVD')
OBJ(('/*') ('/QSYS.LIB' *OMIT)
('/QDLS' *OMIT)
('/QIBM/ProdData' *OMIT)
('/QOpenSys/QIBM/ProdData' *OMIT))
RSTAUT USRPRF(*ALL)
STRSBS SBSD(controlling subsystem)

How to Use Restore Menu Options 21, 22, and 23

This topic describes the procedure for restoring information using option 21, 22, or
23 from the Restore menu. The basic steps are the same for each menu option.
Which option or options you use depends on which save menu option was used
and what other procedures, if any, you use to save information. This is discussed
in “Choosing the Procedure to Recover User Information” on page 107.

Before You Begin

v Clean the read and write heads of the tape unit if you are restoring from a
tape drive.

1. Sign on the system using a user profile with sufficient authority to do the
restore operation (such as QSECOFR).

208 OS/400 Backup and Recovery V5R1

2. Ensure that you load the correct volume of your last set of save media and
make the device ready. The save media should contain the file labeled
QFILEUPR.
a. If you use tape media, run the DSPTAP command and specify
DATA(*LABELS) to find the file labeled QFILEUPR.
b. If you use DVD-RAM optical media, perform the following steps:
1) From a command line run this command: DSPOPT VOL(*MOUNTED)
DEV(OPT01) DATA(*FILATR) PATH('QFILEUPR').
2) If the file is on the media, page down on your display to verify that
the file is on the first volume. If the display says Continued from
previous volume...NO, then the file QFILEUPR is on the first volume
of your save media set.
3. Ensure that any device configuration objects not used in the restore operation
are varied off. You can use the Work with Configuration Status (WRKCFGSTS)
command to display the status of devices.
4. Ensure that the devices that you are using for the restore operation
(workstations, devices, and device controllers) are varied on. These
configuration objects are excluded from the restore operation (message
CPF379C in the job log).
5. Display the Restore menu: GO RESTORE.
6. If you want to do an attended restore, skip to step 7. In most cases, you should
do an attended restore operation to monitor for messages and correct any
problems that occur. This helps your system return to operation as quickly as
possible. If you want to do an unattended restore, do the following steps. This
prevents your restore operation from stopping because of unanswered
messages:
a. Display the reply list sequence numbers to find what numbers are
available for use:
WRKRPYLE
b. If MSGID(CPA3709) is not already in your reply list, add it. For xxxx,
substitute an unused sequence number from 1 through 9999:
ADDRPYLE SEQNBR(xxxx) MSGID(CPA3709) RPY('G')
c. Change your job to use the reply list:
CHGJOB INQMSGRPY(*SYSRPYL)
7. Select the option (21, 22, or 23) from the Restore menu. After pressing the
Enter key, you are shown the Specify Command Defaults display:

Specify Command Defaults

Type choices, press Enter.

Devices . . . . . . . . . . . TAP01 Names

__________
__________
__________

Prompt for commands . . . . . Y Y=Yes, N=No

Message queue delivery . . . . *BREAK *BREAK, *NOTIFY

Restore to different system. . N Y=Yes, N=No

Chapter 9. The Restore Menu 209

 8. Type your choices for the Device prompt. You can specify up to four device
names. If you specify more than one device, the system automatically switches
to the next device after it finishes reading the current save media.
9. Type your choice for the Prompt for commands prompt. Specify N (No) if you
want to run an unattended restore. Specify Y (Yes) if you want to change the
defaults on the RSTxxx commands.
10. Type your choice for the Message queue delivery prompt. Specify *NOTIFY if
you want to do an unattended restore. This prevents communications
messages from stopping the restore procedure. If you specify *NOTIFY,
severity 99 messages that are not associated with the restore operation do not
interrupt the restore process. For example, messages that request a new
volume to be loaded interrupt the restore operation because they are
associated with the job. You cannot continue until you reply to these
messages.
Specify *BREAK, if you want to be interrupted for severity 99 messages that
require a reply.
11. Type your choice for the Restore to different system prompt. If you specify Y
(Yes), the following values will be specified. The system requires these values
in order to perform a system recovery to a different system or to a different
logical partition.
v SRM(*NONE) will be specified on the RSTCFG command
v ALWOBJDIF(*ALL) will be specified on all the restore commands
v MBROPT(*ALL) will be specified on the RSTLIB command
12. After you type your choices, press the Enter key.
13. If you responded Y to the Prompt for commands prompt, you are shown the
End Subsystem display. Type any changes and press the Enter key. While the
system is ending subsystems, you see and respond to these messages:
a. CPF0994 ENDSBS(*ALL) command being processed. Press the Enter key.
b. CPF0968 System ended to restricted condition. Press the Enter key.

If you responded N to the Prompt for commands prompt, skip to step 15 on

page 211.
14. When the system is ready to perform each major step in the restore process,
you are shown the prompt display for that step. The time between displays
may be quite long.
For option 21, you are shown these displays:
v ENDSBS SBS(*ALL) OPTION(*IMMED)
v RSTUSRPRF USRPRF(*ALL)
v RSTCFG OBJ(*ALL)
v RSTLIB SAVLIB(*NONSYS)
v RSTDLO DLO(*ALL) SAVFLR(*ANY)
v RST DEV('/QSYS.LIB/media-device-name.DEVD') OBJ(('/*')
('/QSYS.LIB' *OMIT) ('/QDLS' *OMIT))
v RSTAUT
v STRSBS SBSD(controlling subsystem)

For option 22 (System data only) you are shown these displays:
v ENDSBS SBS(*ALL) OPTION(*IMMED)
v RSTUSRPRF USRPRF(*ALL)
v RSTCFG

210 OS/400 Backup and Recovery V5R1

 v RSTLIB SAVLIB(*IBM)
v RST DEV('/QSYS.LIB/media-device-name.DEVD')
OBJ(('/QIBM/ProdData') ('/QOpenSys/QIBM/ProdData'))
v STRSBS SBSD(controlling-subsystem)

For option 23 (All user data) you are shown these displays:
v ENDSBS SBS(*ALL) OPTION(*IMMED)
v RSTUSRPRF USRPRF(*ALL)
v RSTCFG
v RSTLIB SAVLIB(*ALLUSR)
v RSTDLO DLO(*ALL) SAVFLR(*ANY)
v RST DEV('/QSYS.LIB/media-device-name.DEVD') OBJ(('/*')
('/QSYS.LIB' *OMIT) ('QDLS' *OMIT) ('/QIBM/ProdData'
*OMIT) ('/QOpenSys/QIBM/ProdData' *OMIT))
v RSTAUT
v STRSBS SBSD(controlling-subsystem)

Type your changes, if any, when the display is shown and press the Enter key.

Note: The RSTAUT command will run immediately after the RST commands
when you use option 21 or option 23. If you use option 22 only, you
must run the RSTAUT command. You are not shown the display for
the RSTAUT command because it has no parameters. You cannot
prevent it from running when your restore using the menu options. If
you have additional restore operations to run, you may need to restore
security data and restore authority again after those restore operations.
15. When the system sends a message asking you to load the next volume, load
the next media volume and respond to the message.

If a media error occurs...

If an error occurs during the restore operation, see “Recovery from an
Unsuccessful Restore Operation” on page 56.

If an unrecoverable error occurs when running the RSTDLO DLO(*ALL)

SAVFLR(*ANY) command, see “Recovering from an Error While
Restoring DLOs” on page 57.

16. If you used the distribution media to restore the operating system, some
information was not restored. If you are restoring to a different system, the
network attributes may have been reset to the IBM-supplied defaults. You
must create or change this information again. You should have lists of this
information that were created at the time you performed your save operation.
The following may need to be created or changed:
v Configuration lists
v Network attributes
v Edit descriptions
v Reply list entries
v IBM-supplied subsystem descriptions
a. For the configuration lists, do the following:

Chapter 9. The Restore Menu 211

 Use the Work With Configuration Lists (WRKCFGL CFGL(*ALL))
command to create the configuration lists to match the information in your
list.
b. For network attributes, do the following:
Use the Change Network Attributes (CHGNETA) command to change the
network attributes to match the information in your list.
c. For edit descriptions, do the following:
Use the Work with Edit Descriptions (WRKEDTD EDTD(*ALL)) command
to create edit descriptions to match the information in your list.
d. For reply list entries, do the following:
Use the Add Reply List Entry (ADDRPYLE) command to add reply list
entries to match the information in your list.
e. For IBM-supplied subsystem descriptions, do the following:
Use the Work with Subsystem Descriptions (WRKSBSD SBSD(*ALL))
command to change the IBM-supplied subsystem descriptions to match the
information in your list.
17. This completes the restore operation.
18. If you are unsure what the QSECOFR password is, change it now. To see if
the password has expired, type the following:
DSPUSRPRF QSECOFR

The passwords from your save media are now the current passwords. If the
password expiration is active for the QSECOFR user profile, you will see the
expiration date on the Date password expired field. If the date is the current
system date or prior, change the password for the QSECOFR user profile.
19. Check the job log to ensure all objects were restored.
The job log contains information about the restore operation. To verify that all
objects were restored, you should spool the job log for printing, along with
the job’s remaining spooled output, if any.
DSPJOBLOG * *PRINT

Or
SIGNOFF *LIST

Message CPC3703 is sent to the job log for each library that was successfully
restored. Message CPF3773 is sent to tell you how many objects were restored.
It also tells you how many objects were not restored. Objects are not restored
for various reasons. Check for any error messages, correct the errors, and then
restore those objects from the media.

212 OS/400 Backup and Recovery V5R1

Chapter 10. How to Restore Specific Types of Information
This chapter describes procedures for restoring particular types of information on
the system. It also describes considerations when you restore particular types of
information, whether you restore the information by using menu options or
commands. The topics are presented in the same order as the recovery operations
should occur.

Recovering System Information

You can customize some system information, such as edit descriptions and
network attributes. When you run the SAVSYS command this system information
is saved. It cannot be saved separately.

If you have SAVSYS media and need to restore system information, follow the
procedure that is described in Chapter 6. Restoring the Operating System. Do an
abbreviated installation of the operating system.

If you have restored your operating system from distribution media, you need to
reconstruct system information. “Printing system information” on page 26 describes
how to print the system information. Find the most recent listings you have.
Table 42 shows the commands for changing the system information to the correct
values:
Table 42. Commands for Changing System Information
Information Type Command
1
Access path recovery times EDTRCYAP
Configuration lists WRKCFGL
Edit descriptions WRKEDTD
IBM-supplied subsystem descriptions WRKSBSD
Network attributes CHGNETA
Reply list entries ADDRPYLE
Service attributes CHGSRVA
System values WRKSYSVAL
1
When you reset your access path recovery times, ensure that the ASP configuration
matches the configuration at the time that you printed the recovery times. If it does
not, make a note to reset your access path recovery times after recovering your ASP
configuration.

Sequence for Restoring Security Information

Security information on your system consists of:
v User profiles and group profiles
v Authorization lists
v Authority holders
v Authority information that is stored with objects:
– Owner
– Owner authority
– Primary group

© Copyright IBM Corp. 1997, 2000, 2001 213

 – Primary group authority
– Public authority
v Private authorities

It is essential that you restore security information in the correct sequence.

Otherwise, object ownership and authority information is not restored correctly
and your applications may not run correctly. The recovery checklists include the
correct sequence of steps for restoring security information. If you are developing
your own restore procedure, restore security information in the following sequence:
1. Restore user profiles. The user profile that owns an object must exist before the
object can be restored.
If you restore all user profiles (RSTUSRPRF USRPRF(*ALL)), you also restore
authorization lists and authority holders. Authorization lists and authority
holders must also be on the system before you restore objects.
2. Restore objects. This restores ownership and the authority information is stored
with the object.
3. Restore authority. This restores users’ private authorities to objects.

Restoring User Profiles

You can restore a single user profile, a list of user profiles, or all user profiles. You
restore a user profile to move a user from one iSeries server to another iSeries
server and to recover a damaged user profile.

| You can also use the (*NEW) value with the USRPRF parameter to restore only
| user profiles that are new to your system. Also, when you are restoring individual
| user profiles, you can specify SECDTA(*PWDGRP) to restore passwords and group
| connections. These values are useful if you are merging user profiles from multiple
| systems onto a single system.

| The OMITUSRPRF parameter allows you to limit the number of user profiles you
| restore. You can specify a list of up to 300 specific or generic user profile values
| that will not be restored. This value is helpful if you are restoring a subset of user
| profiles.

Note: You cannot delete an IBM-supplied user profile if it is damaged. You must
restore the operating system again to recover a damaged IBM-supplied user
profile.
Table 43. How User Profiles Are Restored
Method Restricted State?
1,3
| RSTUSRPRF command No
| Restore menu option 8 1,3 No
Restore menu option 21 1,2 Yes
Restore menu option 22 1,2 Yes
Restore menu option 23 1,2 Yes
1
You must have *SAVSYS special authority. You must have *ALLOBJ special
authority to specify ALWOBJDIF(*ALL).
2
These menu options restore all user profiles.
3
| You need to put the system in a restricted state if you specify USRPRF(*ALL).

214 OS/400 Backup and Recovery V5R1

 Do This to Restore All User Profiles
1. Sign on as QSECOFR.
2. Ensure the system is in a restricted state. See “Putting Your System in a
Restricted State” on page 45.
3. Find the most recent media volume that has your user profiles. It may be a
SAVSYS volume or a SAVSECDTA volume. The name of the file on the media
volume is QFILEUPR.
4. If you are using a SAVSYS media volume, type:
RSTUSRPRF DEV(media-device-name) USRPRF(*ALL)
ENDOPT(*LEAVE)

If you are using a SAVSECDTA media volume, type:

RSTUSRPRF DEV(media-device-name) USRPRF(*ALL)
ENDOPT(*UNLOAD)

What Happens When You Restore User Profiles

| When you restore a user profile, you restore all the attributes of the profile that
| you see on the Display User Profile display. The system builds a working table that
| holds that user’s private authorities to objects. You must use the Restore Authority
| (RSTAUT) command to restore the user’s private authorities. (See “Restoring Object
| Authorities” on page 219.) If you specify SECDTA (*PVTAUT), then only the
| working tables that hold the user’s private authorities are restored. The user
| profiles themselves are not restored.

Some values in a user profile may be changed when it is restored. Table 44 on

page 216 shows the actions the system takes when you restore user profiles:

Chapter 10. How to Restore Specific Types of Information 215

 Table 44. Results of Restoring User Profiles
Restore Procedure Used

Restore Individual User

User Profile Restore Individual User Profile Profile That Does Not Exist on
Attributes Restore *ALL User Profiles That Already Exists on System System
1 1
| Group profile Value is restored from media Value on system is not changed Value is set to *NONE
(GRPPRF)
Owner Value is restored from media Value on system is not changed Value is set to *USRPRF
(OWNER) of
new objects
1 1
| Group Value is restored from media Value on system is not changed Value is set to *NONE
authority
(GRPAUT) to
new objects
1 1
| Password Value is restored from media Value on system is not changed Value is set to *NONE
1 1
| Document Value is restored from media Value on system is not changed Value is set to *NONE
Password
1
| Date password Value is restored from media Value on system is not changed The current date is used.
was last
changed
Owner of user See “How the System Value on system is not changed Value is restored from the
profile Establishes Ownership for media. If the owning profile
Restored Objects” on page 218. does not exist, ownership is
assigned to the QDFTOWN
user profile.
Primary group See “How the System Value on system is not changed. Value is restored from the
of user profile Establishes the Primary Group media. If the primary group
for Restored Objects” on does not exist, the value in the
page 218 user profile is set to *NONE.
*ALLOBJ See “What You Should Know See “What You Should Know See “What You Should Know
special About Restoring User About Restoring User Profiles”. About Restoring User Profiles”.
authority Profiles”.
User Value on system is not Value on system is not changed. Value is restored from the
identification changed. media unless it is a duplicate of
number (UID) a UID on the system. In that
case, a new UID is generated.
Group Value on system is not Value on system is not changed. Value is restored from the
identification changed. media unless it is a duplicate of
number (GID) a GID on the system. In that
case, a new GID is generated.
Note: This situation does not apply when migrating from a System/38™. See the System/38 Migration Planning book
for more information.
1
| If you specify SECDTA (*PWDGRP) the value is restored from media.

What You Should Know About Restoring User Profiles

| When you are restoring user profiles from a source system to a target system, you
| must make sure that the password level values (QPWDLVL) are compatible. For
| example, restoring a user profile from the source system with a password value of
| 2 may result in an invalid password on the target system with a password value of
| 0 or 1. Password level 2 allows more characters than password level 0 or 1.

Keep these things in mind when you restore user profiles:

Restoring All Profiles: When you restore all profiles, the system does not first
delete all profiles, authorization lists, and authority holders on the system.
Therefore, the result is both of the following:

216 OS/400 Backup and Recovery V5R1

 v All the profiles, authorization lists, and authority holders on the media.
v Any profiles, authorization lists, and authority holders on the system that were
not on the save media.

Restoring all profiles is the only way to restore authorization lists and authority
holders. However, if an authorization list secures an object in library QSYS, the
association between the authorization list and the object is not restored
automatically. This is because the objects in QSYS library are restored before the
authorization lists. In other words, the object stores the name of the authorization
list it is associated with, and the authorization lists are stored with the user
profiles. Since QSYS is restored before the RSTUSRPRF command executes, the
authorization list is not on the system at the time the object in QSYS is restored.
The IBM publication, An Implementation Guide for iSeries Security and Auditing,
contains sample programs (ALLAUTL and FIXAUTL) that can be used to attach
authorization lists to the objects in library QSYS when the authorization lists are
restored. ALLAUTL must be run before the operating system is restored or
reinstalled in order to create a database of the objects secured by authorization
lists. FIXAUTL must run afterwards to re-establish the links. These programs may
need to be modified to meet your own requirements.

Security Note
If the IBM-supplied user profiles have the default passwords on your save
media, they will again have default passwords after you restore. This is a
security exposure. After a restore operation, verify that the IBM-supplied user
profiles do not have the default passwords.

Restoring *ALLOBJ Special Authority: *ALLOBJ special authority is removed from

user profiles being restored to a system at security level 30 or higher in either of
these situations:
v The profile was saved from a different system, and the person doing the restore
does not have *ALLOBJ and *SECADM authority.
v The profile was saved from the same system or a different system at security
level 10 or 20.

The systems keeps *ALLOBJ special authority for the following system user
profiles:
v QSYS
v QSECOFR
v QLPAUTO
v QLPINSTALL

Moving Users to Another System: To transfer user profiles and their authorities to
another system, do the following:
1. Save the user profiles and authorities by using the SAVSECDTA command.
2. Save the owned objects.
3. Restore the user profiles by using RSTUSRPRF USRPRF(*ALL)
ALWOBJDIF(*ALL).

| Note: You may want to consider using the USRPRF(*NEW) parameter to

| restore only user profiles which do not currently exist on the target
| system. Also, you can omit profiles that you do not want to restore with
| the OMITUSRPRF command.

Chapter 10. How to Restore Specific Types of Information 217

 4. Restore the needed objects by using the RSTLIB, RSTOBJ, RST, or RSTDLO
commands by specifying ALWOBJDIF(*ALL).
5. Restore the private authorities of the user profiles by using the RSTAUT
command.

“Chapter 15. Release-to-Release Support” on page 323 provides more information

about moving objects between systems running different releases of the operating
system. The iSeries Security Reference book provides more information about these
security features.

How the System Establishes Ownership for Restored Objects

Most objects on the system have an owner. The exception to this is objects in the
QNTC and QNetWare file systems because most PC applications do not have a
concept of object ownership.

When you restore an object, the system determines what profile owns the restored
object by using the following rules:
v Ownership is restored to that profile if the profile that owns the object is on the
system.
v If the owner profile does not exist on the system, ownership of the object is
given to the QDFTOWN (default owner) user profile.
v If the object exists on the system and the owner on the system is different from
the owner on the save media, the object is not restored unless
ALWOBJDIF(*ALL) is specified. In that case, the object is restored and the owner
on the system is used.
v See “How the System Restores Programs” on page 252 for additional
considerations when restoring programs.

How the System Establishes the Authorization List for a

Restored Object
Table 45 shows what happens when you restore an object that already exists if the
object is linked to an authorization list. These rules do not apply when you are
restoring a document or a folder.
Table 45. Restoring an Object Linked to an Authorization List
Authorization List on System Value of ALWOBJDIF
and Media Parameter Result

Same Any Data restored; link to authorization list not

changed.
Different *NONE Object is not restored
Different *ALL Data restored; linked to authorization list
associated with existing object.

How the System Establishes the Primary Group for Restored

Objects
Many objects on the system have a primary group. When you restore an object, the
system determines the primary group for the object by using the following rules:
v If the profile that is the primary group for the object is on the system, that
profile is the primary group for the object.
v If the profile that is the primary group for the object is not on the system, the
primary group is set to *NONE. Message CPI380E is sent to the job log.

218 OS/400 Backup and Recovery V5R1

 v If the object exists on the system and the primary group on the system is
different from the primary group on the save media, the system does not restore
the object unless you specify ALWOBJDIF(*ALL). In that case, the system
restores the object with the primary group on the system.

Restoring Object Authorities

Restoring a user’s private authorities to objects is a separate task from restoring
user profiles. When you restore user profiles, the system builds an authority
reference table for each user profile that you restore. The authority reference table
temporarily holds the user’s private authorities to objects.

Possible Method Restricted State?

RSTAUT command 1 No
1
Restore menu option 21 Yes
1
Restore menu option 22 Yes
1
Restore menu option 23 Yes
1
You must have *SAVSYS special authority.

Overview of Restoring Authorities

When you run the Restore Authority (RSTAUT) command, the system restores
authority for each user you specify. You can restore authority for a specific user
profile, a list of specific user profiles, or all user profiles. If you restore authority
for all users, the RSTAUT command restores authority by using every authority
reference table it finds on the system. If you restore a single user profile to the
system because it was damaged, deleted, or is being moved from another system,
you can also use RSTAUT and specify that profile name to restore authorities for
that user profile.

When you run RSTAUT USRPRF(*ALL), you will receive status message CPI3821
informing you of the current number of user profiles for which restore authority is
complete after each authority reference table is processed.

You may run the RSTAUT command regardless of whether or not a system is in a
restricted state. However, there are differences between running RSTAUT on
systems in a restricted state and running RSTAUT on systems in a non-restricted
state. These differences include system performance, job log appearance, and object
availability. More information is provided below.

Note: The system saves and restores authorities differently for objects in the QNTC
and QNetWare file systems. The system saves and restores all authorities,
including private authorities, with the object. “Completing Recovery for the
iSeries Integration for Windows Server Product” on page 263 provides more
information.

Restoring authorities should be the last thing you do before performing an IPL, in
a recovery.

You can also restore authority for a specific profile or a list of profiles. For
example, if you have restored a single user profile to the system because it was
damaged, you can also use the RSTAUT command and specify that profile name.

Chapter 10. How to Restore Specific Types of Information 219

 Restoring Authority On a System in a Non-Restricted State
The RSTAUT command uses prestarted jobs in order to process more than one
authority reference table at a time. The prestarted jobs that are used by RSTAUT
use subsystem description QSYSWRK in library QSYS, program QSRRATBL in
library QSYS, and class QINTER in library QGPL.

There are several advantages to running the RSTAUT command on a system in a

non-restricted state. These advantages include the following:
v Because more than one authority reference table is being processed at a time, the
RSTAUT command on a system in a non-restricted state is up to 30% faster in
most cases than the same command on a system in a restricted state. Generally,
the more user profiles for which RSTAUT is being run, the greater the
performance gain for the RSTAUT command overall.
v Subsystems do not have to be terminated when one or more user profiles are
restored without a full system recovery.
v Authority reference tables are not always deleted after RSTAUT is run for a user
profile. If all private authorities are successfully granted or an abnormal error
occurs then the authority reference table is deleted. If some of the private
authorities are not granted for any reason such as ’object not found’ or ’object in
use’, then the entries for those private authorities that were not granted are kept
in the authority reference table, and the RSTAUT command may be run again
for the user profile to try granting the failed private authorities before the next
IPL or the next restore of the user profile.

There are also some limitations to running the RSTAUT command on a system in a
non-restricted state. These limitations include the following:
v Because the system is not in a restricted state, all objects must be locked by
RSTAUT. This means that several objects could be in use during the processing
of any authority reference table. If the RSTAUT command is unable to lock an
object, a diagnostic message of CPF3736 or CPD3776 will be sent to the job log
of the prestarted job for each object that could not have authority granted. This
is most likely to occur when the object is a user profile or a message queue.
Since private authorities that are not granted are kept in the authority reference
table, the RSTAUT command my be run again before the next IPL to grant
authorities to objects that were in use.
v If you are running RSTAUT for a large group of user profiles that have private
authorities to the same few objects, it is recommended that you put the system
in a restricted state before running the RSTAUT command. This will minimize
the number of objects in use and consequently minimize the number of objects
that are found locked by the RSTAUT command.
v Only one RSTAUT command may be run on a system at a time.

What You Should Know Before Running RSTAUT

There are some general facts you should be aware of when running RSTAUT on a
system in a non-restricted state.
v This command may take a long time depending on how many private
authorities that you have on your system.
v During system recovery, you should not start all subsystems and allow all users
to sign on and use the system while RSTAUT is being run. The only subsystem
that is needed to run RSTAUT in a non-restricted state is QSYSWRK. Allowing
all users access to the system before RSTAUT completes may cause many more
objects to be locked, preventing a larger number of the private authorities from
being restored.

220 OS/400 Backup and Recovery V5R1

v All private authorities for all authority reference tables that are processed by one
prestarted job may or may not be successfully regranted. If they are regranted,
then the message logging level that is used for that prestarted job will be the
same as the logging level that is used by the user’s main job.
If one or more private authorities for an authority reference table are not
successfully regranted, then LOG(*SECVL) will be used for the message logging
for that prestarted job.
For example, you could run the RSTAUT command with the system default
logging level of LOG(4 0 *NOLIST). All prestarted jobs that are run by RSTAUT
that have all private authorities regranted successfully will use the LOG(4 0
*NOLIST) message logging level. The job log will not remain on the system after
the prestarted job completes. All prestarted jobs that are run by RSTAUT, that
have one or more private authorities that are not regranted, will instead use the
LOG(4 0 *SECLVL) logging level. The job log will remain on the system after the
prestarted job completes.
v Do not cancel any of the prestarted jobs RSTAUT has started. Doing so will
cause the entire RSTAUT command to cancel, similar to cancelling a RSTAUT
command on a restricted state system.
v One authority reference table is always processed by a single prestarted job.
v If the authority reference tables are small, one prestarted job may process more
than one authority reference table.

Subsystem QSYSWRK must be started in order for the prestarted jobs to start. The
RSTAUT command will start several prestarted jobs at once, and assign the restore
of authorities for one or more user profiles to each of the prestarted jobs. During
the RSTAUT command, when the prestarted jobs are running, an entry will appear
for each prestarted job on the Work with Active Jobs screen.
Work with Active Jobs MYSYSTEM
05/01/97 16:02:05
CPU %: 26.5 Elapsed time: 00:00:31 Active jobs: 94

Type options, press Enter.

2=Change 3=Hold 4=End 5=Work with 6=Release 7=Display message
8=Work with spooled files 13=Disconnect ...

Opt Subsystem/Job User Type CPU % Function Status

__ QSYSWRK QSYS SBS .1 DEQW
__ QSRRATBL QUSER PJ 1.2 RUN
__ QSRRATBL QUSER PJ 1.0 RUN

If subsystem QSYSWRK is active but the prestarted jobs cannot be started for any
reason, you should receive messages in your job log, including escape message
CPF386D, stating why the prestarted jobs could not be started.

Job Log Considerations

The job logs that are generated by a RSTAUT command running on a system in a
non-restricted state are significantly different than those for a system in a restricted
state. When the RSTAUT command runs on a system in a restricted state, one job
log is generated. When the RSTAUT command runs on a system in a non-restricted
state, each prestarted job (run by RSTAUT) generates a job log that contains some
of the information that is found in the single job log that is produced on a system
running in a non-restricted state.

You may encounter a situation where job logs that contain diagnostic messages
from prestarted jobs that ran during the RSTAUT get deleted. If this occurs, you

Chapter 10. How to Restore Specific Types of Information 221

 can rerun the RSTAUT command at any time before running another RSTUSRPRF
command or RCLSTG command. The system will attempt to regrant the failed
private authorities, and will generate new job logs.

Figures 12 through 14 show a sample job log and message information for a
RSTAUT USRPRF(QPGMR) command run on a system in a restricted state.

>RSTAUT USRPRF(QPGMR)
Authority not restored for user QPGMR.
Some authorities not restored for user profile QPGMR.
Not all user profiles had all authorities restored.

Figure 12. Sample Job Log for RSTAUT on a System in a Restricted State

The expanded text for message CPF3736 appears as follows:

Additional Message Information

Message ID......: CPF3736 Severity......: 20

Message type....: Diagnostic
Date sent ......: 04/24/97 Time sent......: 19:35:17

Message....: Authority not restored for user QPGMR.

Cause......: Private authority for DTAARA DTAARA1 in library QGPL was
not restored. Either the object does not exist, is damaged, or was not
available at the time authority was being granted.
Recovery...:
Do one of the following:
--If the system was dedicated while the RSTAUT command was running,
display the description of the object (DSPOBJD command). If the object was
damaged or not found, restore the user profile (RSTUSRPRF command), restore
the object (RSTOBJ command), and restore the authorizations (RSTAUT command).
If the object exists and is not damaged, report the problem (ANZPRB command).

Figure 13. Expanded Text for Message CPF3736

222 OS/400 Backup and Recovery V5R1

The expanded text for message CPF3845 appears as follows:
In the case of a system in a restricted state, all messages appear in the user’s main
Additional Message Information

Message ID......: CPF3845 Severity......: 20

Message type....: Diagnostic
Date sent ......: 04/24/97 Time sent......: 19:35:17

Message....: Some authorities not restored for user profile QPGMR.

Cause......: 1434 authorities were restored and 1 authorities were not
restored for user profile QPGMR at 04/24/97 19:21:36.
The prestart job name used to restore private authorities for this user
profile is *N. The prestart job name that contains messages about
authorities not restored is *N.
--If the job name above is *N then a prestart job was not used to restore
authorities for this user profile.
--If a job name other than *N is listed above, then a prestart job was
used to restore private authorities for this user profile and messages may
be found in the joblog for the job name listed. Use one of the following
commands to display the joblog for the prestarted job:

Figure 14. Expanded Text for Message CPF3845

job log. When the name of the prestarted job that is used in message CPF3845 is
*N, then no prestarted job was used.

Figures 15 and 16 show a sample job log message information for a RSTAUT
USRPRF(QPGMR) command run on a system in a non-restricted state.

>RSTAUT USRPRF(QPGMR)
Start of prestart jobs in progress.
Some authorities not restored for user profile QPGMR.
Start of prestart jobs in progress.
Not all user profiles had all authorities restored.

Figure 15. Sample Job Log for RSTAUT on a System in a Non-restricted State

The expanded text for message CPF3845 appears as follows:

Chapter 10. How to Restore Specific Types of Information 223

 Additional Message Information

Message ID......: CPF3845 Severity......: 20

Message type....: Diagnostic
Date sent ......: 04/24/97 Time sent......: 19:35:17

Message....: Some authorities not restored for user profile QPGMR.

Cause......: 1433 authorities were restored and 2 authorities were not
restored for user profile QPGMR at 04/24/97 19:21:36.
The prestart job name used to restore private authorities for this user
profile is 010648/QUSER/QSRRATBL. The prestart job name that contains
messages about authorities not restored is 010648/QUSER/QSRRATBL.
--If the job name above is *N then a prestart job was not used to restore
authorities for this user profile.
--If a job name other than *N is listed above, then a prestart job was
used to restore private authorities for this user profile and messages may
be found in the joblog for the job name listed. Use one of the following
commands to display the joblog for the prestarted job:

Figure 16. Expanded Text for Message CPF3845

In figure 16, the name of the prestarted job used is 010648/QUSER/QSRRATBL, and it
appears in the CPF3845 message. The CPF3736 message for the data area
DTAARA1 in library QGPL whose authority was not restored, does not appear in
the user’s main job log. Instead, all messages that are related to restoring of
individual private authorities are in the job log for the prestarted job. To view
these messages, you would run the command DSPJOB JOB(010648/QUSER/QSRRATBL),
and then select option 4 to view the job log for the prestarted job. The expanded
message text for CPF3736 appears in that job log.

You should pay special attention to any CPF3845 message that states that *N
authorities were not restored. This may indicate a problem such as damaged
objects or a function check. Any CPF3845 message with *N authorities that are not
restored should be investigated further by checking the job log of the prestarted
job that is named.

If all authorities in an authority reference table were restored successfully, then

message CPC3706 is sent for the user profile instead of CPF3845. CPC3706 will also
contain the name of the prestarted job that is used to restore authorities for the
user profile. If all authorities restored from a prestarted job were restored
successfully, then the job log for the prestarted job will contain only job start and
end messages.

The order of CPC3706 and CPF3845 messages depends on whether you run the
RSTAUT command on a system that is in a restricted or non-restricted state. These
messages are for user profiles with restored private authorities. The order of these
messages is as follows:
Restricted state system
The order will generally be alphanumeric because only one authority table is
being restored at a time, in alphanumeric order
Non-restricted state system
The order will generally be that these messages will appear first for the user
profiles with fewer private authorities, then later for the user profiles with
many private authorities. This is because multiple authority reference tables are
being restored at once and the smaller authority reference tables usually
complete first.

224 OS/400 Backup and Recovery V5R1

Restoring Authority On a System in a Restricted State
The RSTAUT command on a system running in a restricted state restores
authorities for each authority reference table, one table at a time. No prestarted
jobs are used.

When processing is complete for an authority reference table, the table is deleted
regardless of whether all private authorities were successfully restored or not.

What the System Does When You Restore Authority

When you run the RSTAUT command, the system grants all the private authorities
that it finds in each authority reference table. The user’s private authorities after
the command are both of the following:
v The authorities from the temporary authority reference table.
v Any authorities that are granted to the user since the save operation.

How the System Restores Authority–Example 1: Assume that the authority to

PRICES looks like this at the time of the save operation:

Display Object Authority

Object . . . . . . . : PRICES Owner . . . . .

Library . . . . . : CONTRACTS Primary group .
Object type . . . . : *FILE
Object secured by authorization list . . . . . . . . . .

Object
User Group Authority
OWNCP *ALL
DPTSM *CHANGE
DPTMG *CHANGE
WILSONJ *USE
*PUBLIC *EXCLUDE

Note: Your display looks different when your user profile has a user option setting
of *EXPERT.

After you save security information, you grant and revoke several authorities to
the PRICES file. Just before the restore operation, the authority looks like this:

Display Object Authority

Object . . . . . . . : PRICES Owner . . . . .

Library . . . . . : CONTRACTS Primary group .
Object type . . . . : *FILE
Object secured by authorization list . . . . . . . . . .

Object
User Group Authority
OWNCP *ALL
DPTSM *USE
DPTMG *CHANGE
WILSONJ *EXCLUDE
ANDERSP *USE
*PUBLIC *EXCLUDE

If authority is restored for all users, the authority to the PRICES file looks like this:

Chapter 10. How to Restore Specific Types of Information 225

 Display Object Authority

Object . . . . . . . : PRICES Owner . . . . .

Library . . . . . : CONTRACTS Primary group .
Object type . . . . : *FILE

Object secured by authorization list . . . . . . . . . .

Object
User Group Authority
OWNCP *ALL
DPTSM *CHANGE
DPTMG *CHANGE
WILSONJ *USE
ANDERSP *USE
*PUBLIC *EXCLUDE

Authorities for DPTSM and WILSONJ are restored to the values they have on the
save media. The authority for ANDERSP remains, even though it did not exist on
the save media.

How the System Restores Authority–Example 2: Assume that the authority for the
PRICES file looks like this just before the restore operation:

Display Object Authority

Object . . . . . . . : PRICES Owner . . . . .

Library . . . . . : CONTRACTS Primary group .
Object type . . . . : *FILE

Object secured by authorization list . . . . . . . . . .

Object
User Group Authority
OWNCP *ALL
DPTMG *CHANGE
WILSONJ *CHANGE
*PUBLIC *USE

If authority is restored for all users, the authority to the PRICES file looks like this:

Display Object Authority

Object . . . . . . . : PRICES Owner . . . . .

Library . . . . . : CONTRACTS Primary group .
Object type . . . . : *FILE

Object secured by authorization list . . . . . . . . . .

Object
User Group Authority
OWNCP *ALL
DPTSM *CHANGE
DPTMG *CHANGE
WILSONJ *CHANGE
*PUBLIC *USE

Notice that WILSONJ still has *CHANGE authority. The authority from the save
media (*USE) is granted to WILSONJ, but the authority WILSONJ already has is
not revoked. *USE authority is added to *CHANGE authority, so WILSONJ has
*CHANGE authority.

226 OS/400 Backup and Recovery V5R1

 Notice also that *PUBLIC authority is not affected by this process. Public authority
is stored with the object and is handled when the object is restored. If public
authority on the system is different from public authority on the save media, the
public authority on the system is used.

Authority is restored to the object with the same name in the same library. In some
cases, this could result in restoring authority to a different object.

Assume that you delete program PGMA in library CUSTLIB. You create a new
program with the same name but different function. If you restore authority, users
who were authorized to the original PGMA are now authorized to the new PGMA.
See “How the System Restores Programs” on page 252 for more information.

How to Restore Configuration Objects

You can restore:
v All configuration objects
v A group of configuration objects by generic name
v Only specific types of configuration objects, such as line descriptions or
connection lists.
v System resource management information

A configuration object must be varied off before you can restore it.

| Note: If you execute the restore configuration (RSTCFG) command against a

| printer device description and the output queue associated with that device
| description contains zero spooled files, the system will recreate the output
| queue. Any changes made to the output queue prior to the RSTCFG will be
| lost.
Table 46. How Configuration Objects Are Restored
Possible Method Restricted State?

RSTCFG command 1 No
Restore menu option 7 No
Restore menu option 21 Yes
Restore menu option 22 Yes
Restore menu option 23 Yes
1
You must have *ALLOBJ special authority to specify ALWOBJDIF(*ALL)

Do This to Restore All Configuration Objects:

1. Find the most recent media volume that has your configuration. It may be a
SAVSYS volume or a SAVCFG volume. The name of the file on the volume is
QFILEIOC.
2. If you are using a SAVSYS media volume, type:
RSTCFG OBJ(*ALL) DEV(media-device-name)
OBJTYPE(*ALL)
ENDOPT(*LEAVE)

If you are using a SAVCFG media volume, type:

RSTCFG OBJ(*ALL) DEV(media-device-name)
OBJTYPE(*ALL)
ENDOPT(*UNLOAD)

Chapter 10. How to Restore Specific Types of Information 227

 Restoring to a Different System?
You must specify ALWOBJDIF(*ALL) when you restore the configuration to a
different system. (An option is available on the restore menu that indicates
that you are restoring to a different system. If you selected this option, the
system automatically specifies ALWOBJDIF(*ALL) for you.)

The restoring of configuration objects to a different system whose

configuration objects exist overlays the existing configuration. In some cases,
the configuration description may not match the hardware on the system.

Do not restore system resource management objects to another system. This

may cause problems that can be fixed only by a service representative. When
you use the RSTCFG command to another system, specify SRM(*NONE).

Correcting Problems with the System Resource Management

Information
The system resource management (SRM) information provides a link between the
hardware on your system and the software descriptions of that hardware (the
configuration). When you restore your configuration to a different system, you
should not restore the SRM information because it will not match the hardware on
the target system. Sometimes during a system upgrade, you are instructed to
restore the SRM information to your system even though some of your hardware
has changed.

If you have restored the SRM information and the hardware configuration does not
match, use the following procedure to correct the SRM information:
1. Type STRSST and press the Enter key to access System Service Tools.
2. Select Option 1 (Start a service tool) from the System Service Tools menu and
press Enter.
3. Select Option 7 (Hardware service manager) from the Start a Service Tool menu
and press Enter.
4. Select Option 2 (Logical hardware resources) from the Hardware Service
Manager menu and press Enter.
5. Select Option 1 (System bus resources) from the Logical Hardware Resources
menu and press Enter.
6. Select F10 (Non-reporting resources) to display any non-reporting resources.
Any hardware resources that did not report during the last IPL or that were
created during the last Restore configuration (RSTCFG) will be displayed.
7. Type a 4 (Remove) in the Option column to delete any entry you are certain is
not valid for this system’s configuration.

Recovering Devices That Will Not Vary On

If you have a problem with your devices, such as not being able to vary on a
device, it may be because the system resource management (SRM) database that
was restored does not match the device descriptions on the system.

To correct the problem for a tape unit or a tape controller, do the following:
1. Type WRKHDWRSC TYPE(*STG). You are shown the Work with Storage Resources
display.

228 OS/400 Backup and Recovery V5R1

 2. Type a 9 (Work with resource) in the Opt column next to the resource name
that would not vary on. The Work with Storage Controller Resources display
is shown.
3. Write down the valid resource name for the device type and model that you
tried to vary on.
4. Press F12 (Cancel) until you return to a display with a command line.
5. If the problem is with a tape unit, other than a 3422, 3430, 3480, or 3490, skip
to step 8.
6. Type WRKCTLD CTLD(controller-name). You are shown the Work with
Controller Descriptions display.
7. Type 2 (Change) in the Opt column next to the controller that would not vary
on and press the Enter key. The Change Controller Description display is
shown. Skip to step 10.
8. Type WRKDEVD DEVD(media-device-name). The Work with Device Descriptions
display is shown.
9. Type 2 (Change) in the Opt column next to the device description that you
want to change and press the Enter key. The Change Device Description
display is shown.
10. Change the name in the Resource name prompt to the correct name for the
resource and press the Enter key. You return to the Work with Device
Descriptions display or the Work with Controller Descriptions display.
11. Type 8 (Work with status) in the Opt column next to the device or controller
that you changed and press the Enter key. The Work with Configuration
Status display is shown.
12. Type 1 (Vary on) in the Opt column next to the device description name or the
controller description name. Press the Enter key to vary it on.

Local Workstation Controller: To correct the problem for a workstation, do the

following:
1. Type the following and press the Enter key to display the Work with Local
Workstation Resources display.
WRKHDWRSC TYPE(*LWS)
2. Find the correct controller description for the device that would not vary on.
3. Type a 5 (Work with configuration description) in the Opt column next to the
controller description name and press the Enter key. The Work with
Configuration Description display is shown.
4. Type a 5 (Display) in the Opt column to display the valid resource name for
the workstation controller.
5. Press F12 (Cancel) until you return to a display with a command line.
6. Type the following and press the Enter key to display the device description
for the device that would not vary on.
WRKCTLD CTLD(controller-name)

The Work with Controller Descriptions display is shown.

7. Type a 2 (Change) in the Opt column next to the controller description that
you want to change and press the Enter key. The Change Controller
Description display is shown.
8. Change the name in the Resource name prompt to the correct name for the
resource and press the Enter key. You will return to the Work with Controller
Descriptions display.

Chapter 10. How to Restore Specific Types of Information 229

 9. Type an 8 (Work with status) in the Opt column next to the controller
description that you changed and press the Enter key. The Work with
Configuration Status display is shown.
10. Type a 1 (Vary on) in the Opt column next to the controller description name
and press the Enter key to vary on the device.

Note: It is possible that another device description is varied on for this

resource. Vary off the device first and then vary on the changed device
description. This situation can happen to the console device.

Recovering When You Change the Console Type

When you restore your information to a different system or an upgraded system,
you may have a different console type on the target system. After you have
restored user information, you need to create a new controller and device
description. Do the following:
1. Type WRKHDWRSC *LWS and press the Enter key. You are shown the Work with
Local Work Station Resources display.
2. Type a 5 (Work with controller descriptions) in the Opt column next to the
name of the first workstation controller. Press the Enter key. You are shown the
Work with Controller Descriptions display.

Note: The first workstation controller may not be CTL01.

3. Type a 1 in the Opt column and press the Enter key. You are shown the Create
Controller Description display.
4. For the New controller description prompt, type the name that you want for the
console. Press the Enter key.

Note: If you want to use the name that you had on your old system, you must
first delete the device configuration name and then re-create it.
5. Use the CRTDEVDSP command to create a device description for the console.

Recovering the System/36 Environment Configuration

If you are experiencing a problem with the System/36 environment after restoring
the system, it may be caused by the locking rules used during the installation
process. The QS36ENV configuration object in library #LIBRARY may have been
locked by the System/36 environment.

This object contains the System/36 environment names for the workstation, printer,
tape and diskette units on the system, and default System/36 environment values
used for all users. This object may have been modified by the Change S/36
Environment Configuration (CHGS36) command to customize the System/36
environment.

When the first subsystem is started on the system after the installation process is
complete, a new #LIBRARY and a new QS36ENV object in #LIBRARY are created
with the system defaults. In addition to creating the new objects, each subsystem
holds a lock on the QS36ENV configuration object to ensure that it is not deleted.
This lock will not allow the saved QS36ENV configuration object to be restored.

If the QS36ENV configuration object did not restore, start with step 1 on page 231.
If the configuration object did restore but you are experiencing problems with the
System/36 environment configuration, go to step 5 on page 231.

230 OS/400 Backup and Recovery V5R1

 1. Rename the newly created #LIBRARY to something else (for example,
#LIBNEW).
The locks held on the QS36ENV object remain with the renamed library. This
allows the saved System/36 environment configuration object to be restored.
2. Restore the saved copy of library #LIBRARY: RSTLIB SAVLIB(#LIBRARY)
3. Perform an IPL of the system.
The QS36ENV object in the restored copy of #LIBRARY is the System/36
environment configuration again.
4. Delete the earlier renamed version of #LIBRARY (for example, #LIBNEW).
5. Use the Change S/36 Environment Configuration (CHGS36) command to
refresh the configuration object.
a. Select each of the device types that you want to change.
v Workstation devices
v Printer devices
v Tape devices
v Diskette devices
b. For each device type that you want to change:
1) Press the F5 key to ensure the configuration object matches the device
descriptions on the system.
2) If any System/36 names are not specified, do one of the following:
v Press the F10 key to use the defaults for the System/36 names for
those devices.
v Update the System/36 names manually.
c. Save the changes to the configuration object.

See the topic on configuring the System/36 environment in the Concepts and
Programmer’s Guide for the System/36 Environment for more information about
configuring the System/36 environment.

Restoring Logical Partitions

Look for information on how to restore your Logical Partitions that is integrated
throughout this book. The integrated steps include information on how to recover
your logical partitions configuration data and your system and user data for each
partition.

Keep the following things in mind when you recover your system and user data:
1. Recover your primary partition first.
2. Recover each partition as if it is a stand-alone system.

| For more information about logical partitions, refer to the Information Center Web
| site at the following url:
| http://www.ibm.com/eserver/iseries/infocenter
|
Restoring Libraries
Restoring entire libraries is a common way to recover user information. Use the
Restore Library (RSTLIB) command to restore a single saved library or a group of
libraries. The RSTLIB command restores the entire library, including the library
description, object descriptions (only descriptions are restored for logical files, job
queues, message queues, output queues, user queues, and data queues), and the

Chapter 10. How to Restore Specific Types of Information 231

 contents of other objects. This command also restores status information for
programming temporary fixes (PTFs) that were in the library at the time the library
was saved.

When you use the RSTLIB command, you can use the OPTION parameter to
specify which objects in a library are restored:

Possible Values for the OPTION Parameter of the RSTLIB Command:

*ALL Old objects are replaced and new objects are added to a library.
*ALL is the default.
*OLD Only old objects that already exist on the system are replaced in a
library.
*NEW Only objects not found on the system are added to a library. The
old objects are not replaced.
*FREE Only those objects that have their storage freed on the system are
restored.

Restoring a Library From a Previous Release

When you are restoring a library that was saved on a system at an earlier release,
you can use the Force object conversion (FRCOBJCVN) parameter to specify whether
programs are translated when they are restored. This can significantly impact the
time it takes to restore the library. See “Restoring Programs to a Different Release”
on page 253 for more information.

Restoring Multiple Libraries

You can use the RSTLIB command to restore libraries in these groups:
*NONSYS
All libraries that were saved with SAVLIB LIB(*NONSYS) command,
including the IBM-supplied libraries QGPL, QUSRSYS, and licensed
program libraries.
*ALLUSR
All user libraries that were saved with SAVLIB LIB(*ALLUSR) or SAVLIB
LIB(*NONSYS).
*IBM All IBM-supplied libraries that were saved with SAVLIB LIB(*IBM) or
SAVLIB(*NONSYS). Only IBM-supplied libraries that contain IBM objects
are restored.

Figure 5 on page 40 shows which libraries are saved and restored in these groups.
If you are restoring the QGPL library or the QUSRSYS library, you must restore
them before restoring any other user libraries. If you use the special values
(*ALLUSR or *NONSYS), the system restores these libraries in the correct
sequence.

When you restore a group of libraries (*ALLUSR, *NONSYS, or *IBM), you can
omit up to 300 libraries by using the OMITLIB parameter. The libraries you omit
are not restored from the save media.

When you use a media definition to restore libraries that were saved in parallel
with one of the following groups specified, *ALLUSR, *IBM, *NONSYS, or a
generic value such as X*, you may have to perform some involved recovery
operations. You must first load each drive with the volume that contains the QFILE
so that the system can verify that each library resulted from the same save

232 OS/400 Backup and Recovery V5R1

 operation. You will then have to answer an inquiry message for each drive to
position it to the correct volume, if you are starting the recovery on a library other
than *FIRST.

Attention!
| If you have related objects, such as physical and logical files or journals and
| journaled objects, in different libraries, you must ensure that you restore them
| in the correct sequence. Read “Sequence for Restoring Related Objects” on
| page 45.

If you are restoring to a different system, specify ALWOBJDIF(*ALL) when

you are restoring libraries.

Considerations and Restrictions

The following restrictions and considerations apply to the RSTLIB command:
v You cannot restore a QDOCnnnn (Document) library by using the RSTLIB
command. Use the Restore Document Library Object (RSTDLO) command to
restore documents.
v You cannot restore the QSYS (System) library by using the RSTLIB command.
Use the procedures for restoring the operating system in Chapter 6 to restore
QSYS.
v A RSTLIB command may run concurrently with a RSTOBJ or SAVOBJ command
that uses the same library.
v You may not run multiple concurrent RSTLIB commands that use the same
library.
v A RSTLIB and SAVLIB command may not run concurrently using the same
library.

How to Restore All Libraries from a Single Save Operation

Follow this procedure for restoring all libraries that were saved with a single
command or menu option.
1. Sign on with a user profile that has *SAVSYS special authority. Using *SAVSYS
special authority ensures that you will not have authority problems during the
restore procedure and improves restore performance.
2. Ensure that the system is in a restricted state. For more information, see
“Putting Your System in a Restricted State” on page 45.
3. Find your most recent save media.
4. Use “Task 4–Restoring Libraries to the User Auxiliary Storage Pool” on
page 198. Type your choice and press F4 (prompt).
Table 47. Methods for Restoring All Libraries–Single Save Operation
How Your Libraries Were Saved Type This to Restore Them
Save Menu option 21 RSTLIB SAVLIB(*NONSYS)
SAVLIB LIB(*NONSYS) RSTLIB SAVLIB(*NONSYS)

5. Fill in your choices for other parameters, such as device and whether to rewind
the tape in a tape device. Press the Enter key.
6. If you receive messages to load a media volume, load the correct media volume
and respond to the messages.
7. When the restore operation completes, check your job log to see which libraries
were restored and whether any objects were not restored.

Chapter 10. How to Restore Specific Types of Information 233

 How to Restore All Libraries from Multiple Save Operations
Following is the procedure for restoring all libraries if they were saved with
multiple menu options or commands. Adapt the examples to fit your own save
procedures and recovery situation. Before restoring multiple libraries, make sure
that you read about “Sequence for Restoring Related Objects” on page 45.
1. Sign on with a user profile that has *SAVSYS special authority.
2. Ensure that the system is in a restricted state.
3. Find your most recent save media.
4. Use Table 48 repeat this step and step 5 for each command. Type your choice
and press F4 (prompt).
Table 48. Methods for Restoring All Libraries–Multiple Save Operations
How Your Libraries Were Saved Type This to Restore Them

Save Menu options 22 and 23 RSTLIB SAVLIB(*IBM)

RSTLIB SAVLIB(*ALLUSR)
Save Menu options 21 and 23 RSTLIB SAVLIB(*IBM)
RSTLIB SAVLIB(*ALLUSR)
SAVLIB *NONSYS followed by RSTLIB SAVLIB(*NONSYS)
SAVLIB LIB(LIBA LIBB LIBC) OMITLIB(LIBA LIBB LIBC)
RSTLIB LIB(LIBA)
RSTLIB LIB(LIBB)
RSTLIB LIB(LIBC)

5. Fill in your choices for other parameters, such as device and whether you want
to rewind the tape in a tape device or not. Press the Enter key.
6. If you receive messages to load a media volume, load the correct media volume
and respond to the messages.
7. When the restore operation completes, check your job log to see which libraries
were restored and whether any objects were not restored.

How to Restore Objects

You can use the Restore Object (RSTOBJ) command to restore individual objects or
an entire library. When you restore a library by using the RSTOBJ command, the
library description is not restored.

The following conditions apply when using the RSTOBJ command:

v The RSTOBJ command restores objects to only one library.
v Multiple concurrent RSTOBJ commands may be run against a single library.
v Multiple concurrent RSTOBJ commands may run against a single library
concurrently with the following commands:
– The SAVLIB command
– The RSTLIB command
– One or more SAVOBJ commands
– The RSTLIB command and the SAVOBJ command
– The SAVLIB command and the SAVOBJ command

Attention!
Do not use RSTOBJ to restore licensed programs to library QSYS.
Unpredictable results can occur.

234 OS/400 Backup and Recovery V5R1

| Restoring Objects That Are Journaled
| If the journal exists on the system before the journaled objects are restored, all
| objects that were saved while being journaled are journaled again provided one of
| the following is true:
| v The objects are not on the system at restore time.
| v The objects are on the system and journaling was not ended for the objects.
| v The journal is on the system, it is not damaged, and its journal state is *ACTIVE.
| To find out which object types can be journaled and therefore have these restore
| considerations, refer to “Chapter 19. Planning and Setting Up Journaling” on
| page 383.

| When you restore an object that was being journaled at the time of the save
| operation, an entry is written to the journal to indicate that it was restored.

| When you restore access paths that were being journaled at the time of the save
| operation, no journal entry is written to the journal to indicate that it was restored.

| If the journal is not on the system at the time a journaled object is being restored,
| the restore operation for the object causes a warning message to be sent and
| journaling is not resumed. This warning message causes a diagnostic message to
| be sent at the end of the restore operation. (See the topic “How to Verify That
| Objects Are Restored Successfully” on page 54.)

| What Happens When You Restore Journaled Objects to a

| Different Library
| The system assigns a unique internal journal identifier (JID) to every object that is
| journaled. If you restore a journaled object to a library or directory other than the
| original library or directory, and the object still exists on the system and continues
| to be journaled to the same journal, the JID of the restored object is changed. No
| message is sent telling the user that the JID of the restored object is changed.

| All the journal entries associated with the media copy of the object have the
| original JID. You cannot apply these journal entries to the object that was restored
| to a different library or directory because it has a different JID. For this reason, you
| should avoid restoring a journaled object to a different library or directory.

| For example, in Figure 17 on page 236, the original object FILEA in LIBX library
| has an internal journal identifier of Z that is recorded with every journal entry
| associated with FILEA in LIBX. When FILEA is restored from the media to LIBC
| library, it is assigned the journal identifier of Y because FILEA still exists in LIBX
| and continues to be journaled.
|

Chapter 10. How to Restore Specific Types of Information 235

 ┌────────────────┐
│ Journal Entries│
│ with JID of Z │
│ │
└───┬───────┬────┘
│ │
┌────────────┐ │ │ Journal ┌────────────┐
│ LIBX/FILEA │ │ │ entries │ LIBC/FILEA │
├────────────┤ │ │ cannot be ├────────────┤
│ JID Z │W──────┘ │ applied │ JID Y │
│ │ │ │
└────────────┘ └────────────┘

Figure 17. Example: Restoring a Journaled Object to a Different Library

| Any journal operation that references an object by name and involves using journal
| entries requires that the journal identifier of the object and the journal identifier
| recorded in the journal entries be the same. Because FILEA in LIBC has journal
| identifier Y, journal entries with journal identifier Z are not associated with the
| restored FILEA in LIBC. As a result, journal changes that are recorded for FILEA in
| LIBX cannot be applied to FILEA in LIBC. For the same reason, if you are
| referencing FILEA in LIBC on the Display Journal (DSPJRN), Receive Journal Entry
| (RCVJRNE), or Retrieve Journal Entry (RTVJRNE) commands, or on the Retrieve
| Journal Entries (QjoRetrieveJournalEntries) API, the entries for FILEA in LIBX are
| not returned.

| To display or retrieve the journal entries of the original object:

| 1. Save and then delete the existing object on the system.
| 2. Restore the original object to the system.
| 3. Display or retrieve the journal entries.
| 4. Delete the original object.
| 5. Restore the existing object back to the system.
|
Restoring Database Files
You can restore one or more database files or one or more members of database
files by using the RSTOBJ command. Figure 18 on page 237 shows, conceptually,
how a database file with two members looks to the system. It has multiple parts:

236 OS/400 Backup and Recovery V5R1

 FILE A (Physical File)
┌────────────────────────┐
│ Attributes │
│ Access Path Definition │
┌──┤ Members │
│ └────────────────────────┘
│
├──SMEMBER1
│ ┌───────────────────────┐
│ │ Attributes │
│ ├───────────────────────┤
│ │ Data │
│ ├───────────────────────┤
│ │ Keyed Access Path │
│ └───────────────────────┘
│
└──SMEMBER2
┌───────────────────────┐
│ Attributes │
├───────────────────────┤
│ Data │
├───────────────────────┤
│ Keyed Access Path │
└───────────────────────┘

Figure 18. Example of a Database File with Two Members

If FILEA exists on the system and you restore it, the system restores the data and
access paths for FILEA’s two members. The attributes for the file and its members
are not changed on the system.

If you want to restore the file attributes as they existed at the time of the save
operation, delete the file, then restore it. If you want to restore the member
attributes, remove the member (RMVM) and then restore it, specifying
MBROPT(*NEW).

When you restore a database file, the system uses information that is stored with
the file and the parameters you specify to make decisions. The topics that follow
describe special considerations when restoring database files and members.

Chapter 10. How to Restore Specific Types of Information 237

 Unique File Identification: You can restore a file only to itself. A saved version and
a copied version of the same file are not the same and cannot be used
interchangeably in a restore operation. Figure 19 illustrates this:
File Locking during Restore Operation: When you restore a file, no member of the
┌─────────┐ Copied to ┌─────────┐
│ FILE A │ ──────────────S │ FILE B │
└─────────┘ └─────────┘
│ SAVOBJ │ SAVOBJ
N N
┌─────────┐ RSTOBJ ┌─────────┐
│ FILE A │ ┌─────────│ FILE B │
│ on tape │ │ Not │ on tape │
└─────────┘ │ allowed └─────────┘
│ RSTOBJ │ │ RSTOBJ
N │ N
┌─────────┐ │ ┌─────────┐
│ FILE A │W──────┴ │ FILE B │
└─────────┘ └─────────┘

Figure 19. Restoring a Copy of a File

file can be used during the restore operation, even through logical files. The file is
exclusively locked during the restore operation.

Comparing File Attributes during a Restore Operation

When you restore a database file or member that exists on the system, the system
expects the creation dates for the system copy and the media copy to be the same.
If they are not the same, the system cannot ensure that the contents of the saved
copy match the format of the copy on the system.

If you specify ALWOBJDIF(*NONE) on the restore command, the system does not
restore the file or member if the creation dates do not match. A message is sent to
the user to indicate that the file or member could not be restored from the media.
ALWOBJDIF(*NONE) is the default.

The creation dates on the system and media might be different because:
v A file or member was deleted and created again after the save operation.
v The file or member on the media was created on another system, but it has the
same name as an existing file or member.

If you really want to restore a file or member whose creation date differs from the
system version, you have three choices:
v Delete the file or member from the system. Then restore.
| v Specify ALWOBJDIF (*FILELVL) on the restore command. This value lets you
| attempt to restore physical file data even though its creation date is different
| than the system copy creation date.
v Specify ALWOBJDIF(*ALL) on the restore command. However, this can cause
problems. You should be aware of what the system does when you specify
ALWOBJDIF(*ALL).

How the System Restores Database Files with ALWOBJDIF(*ALL): Figure 20 on

page 239 shows what the system does when creation dates for a database file are
different on the system and media copies:

238 OS/400 Backup and Recovery V5R1

 Files on System after
FILEA on the System Restore Operation
┌──────────────────┐ ┌─────────────────────┐
│ ┌────────────┐ │ │ ┌───────────────┐ │
│ │ FILEA │ │ │ │ FILEA0001 │ │
│ │ 940715 │ │ │ │ 940715 │ │
│ └────────────┘ │ │ └───────────────┘ │
│ │ │ │
│ ┌────────────┐ │ │ ┌───────────────┐ │
│ │ FILEA/MBR1 │ ├────────S│ │ FILEA0001/MBR1│ │
│ │ 940715 │ │ │ │ 940715 │ │
│ └────────────┘ │ │ └───────────────┘ │
│ │ │ │
│ ┌────────────┐ │ │ ┌───────────────┐ │
│ │ FILEA/MBR2 │ │ │ │ FILEA0001/MBR2│ │
│ │ 940730 │ │ │ │ 940730 │ │
│ └────────────┘ │ │ └───────────────┘ │
└──────────────────┘ └─────────────────────┘

FILEA on the Media

┌──────────────────┐ ┌─────────────────────┐
│ ┌────────────┐ │ │ ┌───────────────┐ │
│ │ FILEA │ │ │ │ FILEA │ │
│ │ 940708 │ │ │ │ 940708 │ │
│ └────────────┘ │ │ └───────────────┘ │
│ │ │ │
│ ┌────────────┐ │ │ ┌───────────────┐ │
│ │ FILEA/MBR1 │ ├────────S│ │ FILEA/MBR1 │ │
│ │ 940708 │ │ │ │ 940708 │ │
│ └────────────┘ │ │ └───────────────┘ │
│ │ │ │
│ ┌────────────┐ │ │ ┌───────────────┐ │
│ │ FILEA/MBR2 │ │ │ │ FILEA/MBR2 │ │
│ │ 940730 │ │ │ │ 940730 │ │
│ └────────────┘ │ │ └───────────────┘ │
└──────────────────┘ └─────────────────────┘

Figure 20. Restoring Database Files with Different Creation Dates

The file on the system is renamed. The media version is restored. A message is sent
to the user.

Figure 21 on page 240 shows what the system does when the creation date for one
of the members in the file is different:

Chapter 10. How to Restore Specific Types of Information 239

 File on System after
FILEA on the System Restore Operation
┌──────────────────┐ ┌─────────────────────┐
│ ┌────────────┐ │ │ ┌───────────────┐ │
│ │ FILEA ├──┼─────────┼─S│ FILEA │ │
│ │ 940715 │ │ │ │ 940715 │ │
│ └────────────┘ │ │ └───────────────┘ │
│ │ │ │
│ ┌────────────┐ │ │ ┌───────────────┐ │
│ │ FILEA/MBR1 ├──┼─────────┼─S│ FILEA/MBR1 │ │
│ │ 940715 │ │ │ │ 940715 │ │
│ └────────────┘ │ │ └───────────────┘ │
│ │ │ │
│ ┌────────────┐ │ │ ┌───────────────┐ │
│ │ FILEA/MBR2 ├──┼─────────┼─S│ FILEA/MBR20001│ │
│ │ 940730 │ │ │ │ 940730 │ │
│ └────────────┘ │ │ └───────────────┘ │
└──────────────────┘ │ │
│ ┌───────────────┐ │
FILEA on the Media ┌────┼─S│ FILEA/MBR2 │ │
┌──────────────────┐ │ │ │ 940725 │ │
│ ┌────────────┐ │ │ │ └───────────────┘ │
│ │ FILEA │ │ │ └─────────────────────┘
│ │ 940715 │ │ │
│ └────────────┘ │ │
│ │ │
│ ┌────────────┐ │ │
│ │ FILEA/MBR1 │ │ │
│ │ 940715 │ │ │
│ └────────────┘ │ │
│ │ │
│ ┌────────────┐ │ │
│ │ FILEA/MBR2 ├──┼────┘
│ │ 940725 │ │
│ └────────────┘ │
└──────────────────┘

Figure 21. Restoring Database Files with Different Creation Dates

The member on the system is renamed. All members from the media are restored.
A message is sent to the user.

When you specify ALWOBJDIF(*ALL) and additional members are created during
a restore operation, the system ignores the MAXMBRS (maximum members)
parameter for the file. After the restore operation, you may have more than the
allowed members in the file.

If a logical file is associated with a file or member that is renamed, the logical file
is still associated with the renamed file or member, not the restored member.

In both examples, specifying ALWOBJDIF(*ALL) can result in duplicate

information, additional files, and additional members. Your system becomes
cluttered and your applications may produce unexpected results. If you specify
ALWOBJDIF(*ALL), carefully check the messages you receive and analyze your
files and members after the restore operation.
Notes:
1. The ALWOBJDIF parameter also affects object ownership. This is described in
“How the System Establishes Ownership for Restored Objects” on page 218.

240 OS/400 Backup and Recovery V5R1

 2. When you specify MBROPT(*MATCH) on a restore command, you cannot
specify ALWOBJDIF(*ALL). See “How the System Matches File Members
during a Restore Operation”.

How the System Matches File Members during a Restore

Operation
When you are restoring to an existing database file, you use the member option
(MBROPT) parameter on either the RSTOBJ command or the RSTLIB command to
tell the system what to do if the members do not match. The choices are:
*MATCH
If the set of members on the save media and on the database are not
identical, the restore operation fails. *MATCH is the default.
*ALL All members on the save media are restored, whether or not they exist on
the system copy.
*NEW Only those members on the save media that do not exist in the database
file are restored.
*OLD Only those members on the save media that already exist in the database
file are restored.

Note: The ALWOBJDIF parameter determines what the system does if creation
dates on the members do not match. See “Comparing File Attributes during
a Restore Operation” on page 238.

Restoring Members to a File

You can restore a list of members for a database file by using the FILEMBR
parameter of the RSTOBJ command. This list may consist of specifically named
members, generically named members, or a combination of both specifically and
generically named members.

The FILEMBR parameter is used to specify:

v A list of file members (specific or generic) for a specific database file
v The same group of members from more than one file

The default value *ALL causes all file members of files specified with the OBJ
parameter to be restored.

Restrictions on the File Member Parameter

The following restrictions apply to the FILEMBR parameter:
v Each database file that is specified in the FILEMBR parameter must also be
specified in the OBJ parameter by its complete name, a generic name, or *ALL.
v Generic names are not valid for the database file name.
v Generic names are valid for the member name.
If a generic file member name is used, and the file does not have members that
fit the generic name, the file is not restored. If all files specified by the FILEMBR
parameter are not restored, a diagnostic message is sent and the restore
operation ends with an escape message giving the number of files not restored.
If a name that is not generic is used, the specific members must exist in the file
for any part of the file to be restored.
v The OBJTYPE must be *ALL or include *FILE.
v The MBROPT parameter must not have the *MATCH value.

Chapter 10. How to Restore Specific Types of Information 241

 Restoring Logical Files
When you restore a logical file, the system uses the description for the logical file
to establish its relationship with the based-on physical files and logical files. All
based-on files must exist before you can restore the logical file.

You can restore a logical file to a library different than the library for the associated
physical file. However, the associated physical file must remain in or be restored to
its original library location.

If you try to restore a logical file to a library in which it does not exist, the restore
operation fails if any of the associated physical files have had their storage freed.

When a logical file is restored, it must be dependent on the same physical files as
it was when it was saved.
v The logical file is created over the physical files in the library where they are
being restored if any of the following occur:
– The logical file and the associated physical files existed in the same library at
the time of the save operation.
– The logical file and the associated physical files are present in the library
where the files are being restored.
– The logical file and the associated physical files are being restored to the same
library.
v If the files are not present in the restore library, then the logical files are created
over the physical files in the original saved library.
v If the correct physical files are not found in either library, then the restore
operation of the logical file fails. To correct the problem, run the RSTOBJ
command again and specify OBJ(*NEW). If the restore operation is successful, an
informational message (CPF3291) is sent to indicate which library was used for
associated physical files.

The creation dates of the physical files must not have changed since the logical file
was saved. If the date has changed, an informational message (CPF3293) is sent
indicating that the physical file has been changed since the save operation, but the
restore operation continues.

Restore physical or logical files with dependent logical files before the dependent
logical files, unless the physical and logical files already exist on the system. The
following considerations apply to restoring logical files:
v If the dependent physical or logical files are in the same library, the system
provides the proper sequencing.
v If the files are in different libraries, you must restore the libraries in order, so
that the physical or logical files that have logical files built on them are restored
first.
v If the depended-on physical or logical files are not restored before you attempt
to restore the logical files, restoring the logical files fails.
v This sequencing also applies to other requirements between files, such as shared
formats. You can restore those logical files that failed by using the RSTOBJ
command.

How the System Restores Access Paths

The description for a database file contains a description of its access path, if it has
one. When you save a database file, you may save the access path with the file.

242 OS/400 Backup and Recovery V5R1

This depends on the type of file, the type of access path, and how you performed
the save operation. For more information, see the Backing up your system topic
on the Information Center at the following Web site:
http://www.ibm.com/eserver/iseries/infocenter

When you restore a file, the system either restores the access path with the file or
rebuilds the access path based on the information in the file description. The
process of rebuilding the access path for a large database file can take a long time.
This topic describes when the system restores access paths and when it cannot. If
possible, you should plan your save operations to avoid having to rebuild access
paths during a restore operation.

The system always restores the access path for a keyed physical file of type *DATA
unless the access path was not saved. The access path for a keyed physical file is
always saved, unless the access path is not valid at the time of the save.

Normally, source physical files are not keyed. The default for the CRTSRCPF is to
create a non-keyed file. When you restore a keyed source physical file, the access
path is rebuilt after the restore operation.

Access paths that are owned by logical files are restored if all of the following
conditions are true:
v The system saved the access path. Although this seems obvious, the system only
saves access paths if the certain conditions are met. For more information, see
the Backing up your system topic on the Information Center at the following
Web site:
http://www.ibm.com/eserver/iseries/infocenter
v All based-on physical files are in the same library and are being restored at the
same time on the same restore command.
v If the logical file exists on the system, it does not specify MAINT(*REBLD).
v The logical file owned the access path at the time it was saved.
v If the logical file is created again by the restore operation and it shares an access
path that already exists, the key length for the access path must be equal to the
maximum key length of the logical file or you receive an error.

If you meet these conditions, you minimize the rebuilding of access paths.
However, during the restore operation, the system checks the integrity of each
access path. If it detects any discrepancy, the access path is rebuilt.

In a few cases, the system may decide to rebuild access paths even though they
were saved. For example, you may have defined a new logical file that specified
the same key as the physical file but also specified UNIQUE. The based-on
physical file was in use at the time that the logical file was created. Therefore, the
system had to create a new access path for the logical file. Assume that you save
these two files with a single command. If you restore them with a single
command, the system will determine that they can share a single access path.
Instead of restoring the two access paths, it builds a new, shared access path for
the two files.

Restoring a File Network–Examples

Figure 22 on page 244 shows a physical file and two logical files:

Chapter 10. How to Restore Specific Types of Information 243

 LIB1/FILEA (Physical) LIB1/FILEB (Logical)
┌─────────────────┐ ┌─────────────────┐
│ Attributes │ │ Attributes │
│ Access Path │ │ Access Path │
│ Definition │ │ Definition │
┌──┤ Members │ ┌──┤ Members │
│ └─────────────────┘ │ └─────────────────┘
│ │
├─SMEMBER1 └─SMEMBER1
│ ┌─────────────────┐ ┌─────────────────┐
│ │Attributes │ │Attributes │
│ ├─────────────────┤ ├─────────────────┤
│ │Data │W──────────┤Access Path │
│ ├─────────────────┤ └─────────────────┘
│ │Keyed Access Path│
│ └─────────────────┘
│
└─SMEMBER2 LIB2/FILEC (Logical)
┌─────────────────┐ ┌─────────────────┐
│Attributes │ │ Attributes │
├─────────────────┤ │ Access Path │
│Data │W──┐ │ Definition │
├─────────────────┤ │ ┌──┤ Members │
│Keyed Access Path│ │ │ └─────────────────┘
└─────────────────┘ │ │
│ └─SMEMBER1
│ ┌─────────────────┐
│ │Attributes │
│ ├─────────────────┤
└───────┤Access Path │
└─────────────────┘

Figure 22. Restoring Access Paths

Assume that these files were saved with this command:

SAVLIB LIB(LIB1 LIB2) ACCPTH(*YES)

The save media volume contains all three files (FILEA, FILEB, and FILEC) and
three access paths, each owned by a different file. Table 49 shows what the system
does when you restore these libraries by using different methods. These examples
assume that none of the files are on the system when the system restores them:
Table 49. Restoring a File Network
Sequence of Restore Commands What the System Does
Example 1: Results of Example 1:
1. RSTLIB SAVLIB(LIB1) 1. FILEA and FILEB are restored. The access paths owned by
2. RSTLIB SAVLIB(LIB2) FILEA and FILEB are restored.
2. FILEC is restored. The access path owned by FILEC is
rebuilt.
Example 2: Results of Example 2.
1. RSTLIB SAVLIB(LIB2) 1. FILEC is not restored because FILEA is not on the system.
2. RSTLIB SAVLIB(LIB1) 2. FILEA and FILEB are restored. The access paths owned by
FILEA and FILEB are restored.

These examples highlight the problems that can occur when logical files and
based-on physical files are in different libraries. Access paths are restored when
physical files are restored because they are built over data that is contained in the
physical file. In the first example, FILEC owned the access path but FILEC was not

244 OS/400 Backup and Recovery V5R1

 on the system when the physical file was restored. Therefore, the access path was
not restored. In the second example, FILEC could not be restored because its
based-on physical file (FILEA) was not on the system.

How to Prevent the System from Rebuilding a Large Access Path

If the situation that is shown in Table 49 on page 244 occurs on your system and
you want to prevent the system from rebuilding a large access path, do the
following:
1. Restore the physical file or the library that contains the physical file. In the case
of example 2, restore FILEA or LIB1.
2. Restore the logical file (FILEC) using the RSTOBJ command.
3. Immediately after restoring the logical file, type EDTRBDAP. You are shown the
Edit Rebuild of Access Paths display.
4. Change the value in the Seq column for the logical file to *HLD.
5. Restore the physical file (FILEA) again using the RSTOBJ command. Because
the logical file (FILEC) is now on the system, the system will restore the access
path that is owned by FILEC.
6. Type EDTRBDAP. You are shown the Edit Rebuild of Access Paths display.
7. Change the sequence number for FILEC to a value from 1 through 99 to
remove the access path from the display.

How the System Restores Files with Shared Formats

When a database file is restored and that file, before it was saved, had shared the
record format of another file, an attempt is made to find the file whose format was
shared, and reestablish the original format sharing.

The search for restoring the shared format starts in the library to which the
restored file is directed and continues in the library from which the restored file
was saved. Following are the results of the search:
v If the sharing file is found and has not been changed (level check) since the
save, then no new format is created for the restored file.
v If the sharing file is not found, or it is found but fails the level check, then a
new format for the restored file is created with the same definition as the one it
initially shared.
v If a format sharing file has been renamed, deleted, or moved to a library other
than the save or restore library, a new format is created for the dependent file
when the dependent file is restored.

How the System Restores Files with Referential Constraints

Information about DB2/400* database files is kept in system cross-reference files.
This includes information about constraints that are defined. When you define a
referential constraint, you specify that a record with a certain primary key must
exist in the parent file before a record with the same values in a foreign key can
exist in the dependent file. For example, you cannot add an order to the order file
(dependent file) unless a record exists for the customer in the customer file (parent
file).

A referential constraint is defined, stored, and saved with the dependent file. Each
referential constraint has a name, which must be unique for the library that
contains the dependent file. When you restore a file that has a referential constraint
name that already exists in the library, the system generates a new name for the
referential constraint that is being restored.

Chapter 10. How to Restore Specific Types of Information 245

 When you restore a database file that already exists on the system, the referential
constraints that are defined for the system copy of the file are used. If the saved
version of the file has additional referential constraints that are not on the system
copy, these additional constraints are not restored.

When you restore a database file that does not exist, you should ensure that any
referential constraints that were not on the saved copy are reestablished.
Otherwise, you lose the data integrity checking that was on your system before a
failure occurred.

Files that are related by referential constraints form a database network similar to
the network formed by logical files and the based-on physical files. You should try
to save an entire referential constraint network in one operation. If this is not
possible, you should at least save the files with consecutive operations where no
activity occurs in between. This ensures that the files are synchronized.

If you journal database files, you should journal all physical files that are part of a
referential constraint network. This ensures that your referential constraints remain
valid after you have applied journaled changes. “Chapter 19. Planning and Setting
Up Journaling” on page 383 provides more information about journaling and
referential constraints.

Referential Constraint Network–Example

Figure 23 shows an example of a referential constraint network.

ORDER File COMPANY File

┌──────────────────┐ ┌──────────────────┐
│ Order Number │ ┌─S│ Company Number │
├──────────────────┤ │ │ (Primary Key) │
│ Customer Number ├────────┐ │ └──────────────────┘
│ (Foreign Key) │ │ │
├──────────────────┤ │ │ CUSTOMER File
│ Item Number ├──────┐ │ │ ┌──────────────────┐
│ (Foreign Key) ├──┐ │ │ └──┤ Company Number │
├──────────────────┤ │ │ │ │ (Foreign Key) │
│ Warehouse ├──┤ │ │ ├──────────────────┤
│ (Foreign Key) │ │ │ └────S│ Customer Number │
└──────────────────┘ │ │ │ (Primary Key) │
│ │ └──────────────────┘
│ │
INVENTORY File │ │ ITEM File
┌──────────────────┐ │ │ ┌──────────────────┐
│ Item Number + │ │ └──────S│ Item Number │
│ Warehouse Number │W─┘ ┌──────S│ (Primary Key) │
│ (Primary Key) │ │ └──────────────────┘
├──────────────────┤ │
│ Item Number ├──────┘
│ (Foreign Key) │
└──────────────────┘

Figure 23. Restoring a Referential Constraint Network

You can restore the files in this network in any sequence. When you restore the
files, the system reestablishes the relationships and attempts to determine whether
the constraints are still valid.

For example, if you restore both the ITEM file and the INVENTORY file, the
system checks the internal information stored with the files to determine whether
the indexes for the two files are synchronized.

246 OS/400 Backup and Recovery V5R1

 If the internal information does not match, the system validates the constraint for
the INVENTORY file. It does this by reading every record in the INVENTORY file
and ensuring that a record with that item number exists in the ITEM file. If this
process is successful, the constraint is valid. If this process is not successful, the
status of the constraint is set to Check pending. When the status of a constraint is
check pending, you must take action to correct the situation, either by restoring
one of the files or using a program to update the files.

If you restore one of the files, the system again attempts to validate the constraint.
If you use a program to update the information, you must use the Edit Check
Pending Constraints (EDTCPCST) command to force the system to revalidate the
constraint. The topic “Task 3–Using the Edit Check Pending Constraints Display”
on page 173 describes how to determine the status of files that have referential
constraints.

Look in the Information Center under the Database and File System topics for
more information about using referential constraints.

How the System Restores Files with Triggers

You can define one or more trigger programs for a file. When a certain event
occurs in the file, the system calls the trigger program. When you save a file that
has trigger programs, you are saving only the definitions of the trigger programs,
not the programs themselves. You must ensure that the programs are also saved,
perhaps by placing them in the library with the file.

When you restore a database file that already exists, the system does not restore
any trigger program definitions from the save media. When you restore a database
file that does not exist, you should ensure that any definitions for trigger programs
that were not on the saved copy are reestablished. Otherwise, you lose the data
integrity checking that was on your system before a failure occurred.

The system does not stop restoring a database file if its trigger programs cannot be
found. Therefore, you must ensure that files and trigger programs are saved and
restored correctly. Otherwise, an error may occur.

Table 50 shows examples of actions the system takes when you restore the physical
file FILEA and the trigger program PGMA:
Table 50. Restoring Files That Have Trigger Programs
How the Trigger Program Is
Save Procedure That Is Restore Procedure That Is Defined after the Restore
Used Used Operation
FILEA is saved from LIBX. PGMA is restored to LIBY. The trigger is defined as
PGMA is saved from FILEA is restored to LIBX. LIBX/PGMA. When an event
LIBX. The trigger is occurs that causes this trigger, the
defined as LIBX/PGMA. program will not be found.
FILEA is saved from LIBX. PGMA is restored to LIBY. The trigger is defined as
PGMA is saved from FILEA is restored to LIBY. LIBY/PGMA.
LIBX. The trigger is
defined as LIBX/PGMA.
FILEA is saved from LIBX. PGMA is restored to LIBZ. The trigger is defined as
PGMA is saved from LIBY. FILEA is restored to LIBZ. LIBY/PGMA. When an event
The trigger is defined as occurs that causes this trigger, the
LIBY/PGMA. program will not be found.

Chapter 10. How to Restore Specific Types of Information 247

 Look in the Information Center under the Database and File System topics for
more information about using trigger programs. “Chapter 19. Planning and Setting
Up Journaling” on page 383 describes special considerations when you journal
database files that have triggers defined. You must make special provisions to
ensure the integrity of your data because trigger programs are not called when you
apply journaled changes.

Steps Before Deleting a Physical File

In some situations, you must delete a physical file as part of your recovery. For
example, the physical file may be damaged. Or a physical file in a user ASP may
have overflowed into the system ASP. You cannot delete a physical file if other files
are dependent on it, such as logical files or files that share the record format.

Before deleting a physical file, do the following:

1. Use the Display Database Relationships (DSPDBR) command to list all the files
that are dependent on the physical file.
2. Save and delete each file that is dependent on the physical file.

After you have recovered the physical file, restore all the dependent files.

Restoring Journals and Journal Receivers

Note: If you are restoring remote journals or journal receivers that are associated
with remote journals, refer to “Save and Restore Considerations” on
page 500.

You can restore journals or journal receivers only to the same library from which
they were saved. Use the RSTOBJ and RSTLIB commands to restore journals and
journal receivers. When you are restoring multiple objects with one of these
commands, journals and journaled objects are restored before the journal receivers.

When you use several commands to restore several objects, restore the objects in
this order:
1. Journals
| 2. Based-on physical files
| 3. Other journaled objects associated with those journals
4. Dependent logical files
5. Journal receivers

| Note: Journal receivers can be restored at any time after the journals. They do
| not have to be restored after the journaled objects.

If you have journal receivers that were created before V3R1, you must restore them
in the order of newest to the oldest to establish the receiver chain correctly.

Restoring Journals
When you restore a journal, the system creates a new journal receiver and attaches
it. The characteristics of the new journal receiver are based on the journal receiver
that was attached when the journal was saved:
v The system creates a name that is not likely to conflict with other journal
receivers that may be on the system. The topic “Naming Journal Receivers” on
page 399 describes how the system generates a name.

248 OS/400 Backup and Recovery V5R1

 v The system attempts to assign the same owner and to create the journal receiver
in the same library. If the owner of the receiver is not found, the receiver is
| assigned to the default owner (QDFTOWN) user profile. If the library is not
| found, the journal receiver is placed in the journal’s library.
v The system starts a new receiver chain. “Chapter 20. Working with Journal
Entries, Journals, and Journal Receivers” on page 427 discusses receiver chains.

Note: At the time a new journal receiver is created and attached, private
authorities have not been restored on the system. Therefore, private
authorities will not be assumed by the new journal receiver. After the
Restore Authority (RSTAUT) command is run, users will receive private
authority to the receiver that was attached before the restore operation.
Users will not receive private authority to the new receiver. Users must be
manually granted private authority to the new receiver.

| You cannot restore a journal to a library that contains the same journal. If a journal
must be restored (because of damage) to a library, the existing journal must be
deleted first.

Steps before Deleting a Journal

| In some situations, you must delete a journal as part of your recovery. For
| example, the journal may be damaged. Or a journal in a user ASP may have
| overflowed into the system ASP. You cannot delete a journal while objects are
| being journaled to it.

You use the Delete Journal (DLTJRN) command to delete a journal. Before deleting
a journal, try to do the following steps. You may not be able to perform these steps
successfully if the journal is damaged.
1. Type
WRKJRNA JRN(library-name/journal-name)
OUTPUT(*PRINT)

| and press the Enter key. You receive a listing that shows all the objects that are
| currently being journaled.
2. End journaling for all the access paths assigned to the journal by typing:
ENDJRNAP FILE(*ALL)
JRN(library-name/journal-name)
3. End journaling for all the physical files assigned to the journal by typing:
ENDJRNPF FILE(*ALL)
JRN(library-name/journal-name)
| 4. End journaling for all the integrated file system objects assigned to the journal
| by typing:
| ENDJRN OBJ(*ALL)
| JRN(/QSYS.LIB/library-name.LIB/journal-name.JRN)
| 5. End journaling for all other object types assigned to the journal by typing:
| ENDJRNOBJ OBJ(*ALL) OBJTYPE(*ALL)
| JRN(library-name/journal-name)
| 6. Deactivate any remote journals that are associated with the journal by using the
Change Journal State (QjoChangeJournalState) API or the CHGRMTJRN
command.

When you try to delete the journal, you may receive message CPF7021 indicating
that the journal is being used for commitment control. If this occurs, end the jobs
that are using commitment control and then try to delete the journal again. You

Chapter 10. How to Restore Specific Types of Information 249

 can use the End Job (ENDJOB) command or you can use the End option from the
Work with Active Jobs (WRKACTJOB) display.

| After you restore the journal or create it again, you must start journaling again for
| each object. Use the following commands to begin journaling for each object type
| listed below:
| v Database physical files — STRJRNPF
| v Access paths — STRJRNAP
| v IFS objects — STRJRN
| v All other object types — STRJRNOBJ

You should save the objects after you start journaling, in case the system assigned
a new journal identifier (JID) to an object. If you previously had any remote
journals associated with the journal, add them again by using the Add Remote
Journal (ADDRMTJRN) command or the Add Remote Journal
(QjoAddRemoteJournal) API. If you added any remote journals, you should save
the journal in order to preserve that information.

Restoring Journal Receivers

If you have journal receivers that were created before V3R1, restore them from
newest to oldest to establish the receiver chain correctly. If the journal receivers
were created on V3R1 or later, you can restore them in any sequence. If you restore
journal receivers in a single command, the system restores them in the correct
sequence.

The system will not restore a journal receiver over the journal receiver that is
currently attached. The system will not restore a journal receiver over an existing
journal receiver that contains more entries. If you use the SAVCHGOBJ command
to save journal receivers, this is likely to occur. The journal receiver that is attached
at the time of the save operation is a changed object and is saved by the command.
When you restore, you receive message CPF3706 and the system continues with
the next journal receiver.

If your save procedure saves the currently attached journal receiver, you may try
to restore a journal receiver with fewer entries than the journal receiver on the
system. For example, assume that you save your journal receivers when receiver
RCVR0006 is attached. RCVR0006 has 1500 entries. Later, you use the CHGJRN
command to create and attach a new receiver. Now receiver RCVR0007 is attached.
Receiver RCVR0006 is still on the system and has 4300 entries. If you attempt to
restore receiver RCVR0006 from your save media volume, the operation fails
because the saved copy has only 1500 entries.

If the library you specify on the restore command for a journal receiver does not
exist, the system restores the journal receiver to the library that contains the
journal. If you specify RSTASP(*SAVASP) and the ASP does not exist, the system
usually restores the journal receiver to the same ASP as the library that contains
the journal.

Placing Journal Receivers in the Correct Auxiliary Storage Pool: If the attached
journal receiver is not in the desired ASP after the restore operation, do the
following:
1. Create a journal receiver in the desired ASP. Follow your existing naming
convention and use the same journal receiver attributes.
2. Use the CHGJRN command to attach the new journal receiver to the journal.

250 OS/400 Backup and Recovery V5R1

 How to Resolve Name Conflicts When Restoring Journal
Receivers
When you restore a journal, the system creates and attaches a new journal receiver.
The system attempts to name this journal receiver so that a name conflict does not
occur. However, in rare cases, this new journal receiver may have a name that
matches the name of a journal receiver that you need to restore. If this occurs, do
the following:
1. Create a new journal receiver with a name separate from your normal naming
convention. For example, type: CRTJRNRCV JRNRCV(library-name/TMP0001).
2. Use the CHGJRN command to attach the temporary journal receiver: CHGJRN
JRN(library-name/journal-name) JRNRCV(library-name/TMP0001).
3. Delete the journal receiver that has the name conflict. This journal receiver
should not have any entries you need for your recovery because it was created
when the journal was restored.
4. Restore the journal receivers.
5. Create a journal receiver that continues your naming convention and has the
same journal receiver attributes.
6. Use the CHGJRN command again to attach the journal receiver that you
created in step 5.

How to Correct the Journal Receiver Directory

Every journal has a directory of journal receivers. The sequence of journal receivers
is called the receiver chain. Before you begin a recovery by using journal receivers,
you should ensure that this receiver directory is current and correct.

Do the following:
1. Type WRKJRNA JRN(library-name/journal-name) and press the Enter key.
2. From the Work with Journal Attributes display, press F15 (Work with receiver
directory). You are shown the Work with Receiver Directory display.
3. If the receiver directory is not correct, do the following:
a. Type WRKJRN and press the Enter key.
b. On the prompt display, enter the name of the journal.
c. On the Work with Journals display, type a 9 (Associate receivers with
journal) in the option column in front of the journal. The system establishes
the receiver chain for the journal.

Note: If some of the journal receivers were created prior to Version 3 Release 1,
you may need to restore all the journal receivers, from newest to oldest, to
establish the receiver chain correctly.

Steps before Deleting a Journal Receiver

In some situations, you must delete a journal receiver as part of your recovery. For
example, the journal receiver may be damaged. Or a journal receiver in a user ASP
may have overflowed into the system ASP.

You cannot delete a journal receiver that is currently attached to a local journal.
You also cannot delete a journal receiver if later journal receivers in the receiver
chain are still on the system, unless one of the following conditions exists:
v The receiver being deleted is damaged
v The journal is a remote journal
v The journal is system-managed

Chapter 10. How to Restore Specific Types of Information 251

 If you need the journal receiver for recovery, you should not delete it without first
saving it. If you do, the system warns you but does not prevent you from deleting
the journal receiver.

Before deleting a journal receiver, do the following:

1. If the journal receiver is attached, detach it by typing:
CHGJRN JRN(library-name/journal-name)
JRNRCV(*GEN)
Notes:
a. If the current journal receiver is damaged, you cannot specify JRNRCV(*GEN).
Use the Create Journal Receiver (CRTJRNRCV) command to create a new
journal receiver that follows your naming convention and has the same
attributes. Specify that receiver name on the CHGJRN command.
2. If earlier journal receivers are on the system, save them and delete them. You
can print the receiver chain by typing WRKJRNA JRN(library-name/journal-
name) OUTPUT(*PRINT).

How the System Restores Programs

Restoring programs to your system represents a security exposure. A restored
program may have been altered to perform functions that you do not intend, or
the program may adopt the authority of a powerful user profile.

| When the QSECURITY (security level) system value on your system is 40 or

| higher, the system checks for restricted instructions in all programs that are
| restored. You can use the QALWOBJRST system value to allow or prevent the
| restoration of certain types of objects on your system. You can also set the
| QVFYOBJRST (verify object on restore) system value to specify how the system
| verifies program-object signatures during a restore operation. See “Controlling
| Restoration of Security-Sensitive Objects” on page 50.

| The system stores a validation value for all programs. When a program is restored,
| the system calculates the validation value and compares it to the value on the
| media. If they are different, but the QALWOBJRST system value allows objects
| with validation errors to be restored, the system creates the program again from
| the program object. The system contains all the information necessary to re-create
| an AS/400 or iSeries program. Starting with V5R1 the object does not have to have
| program observability in order to be restored. If a program validation error exists
| at the time the program is restored, the program will be recreated in order to
| correct the program validation error.

Table 51 shows what the system does when restoring programs. The iSeries Security
Reference book has more information about protecting your system from programs
that might circumvent security.
Table 51. System Actions When Restoring Programs
Audit Version of Private and
Journal Program Public
Situation Description Entry Message Owner Restored Authorities

Program created before V1R3; security level less None None Unchanged Original Unchanged
than 40
Program created before V1R3; security level 40 or None None Unchanged Original Unchanged
higher; ALWOBJDIF(*ALL)

252 OS/400 Backup and Recovery V5R1

Table 51. System Actions When Restoring Programs (continued)
Audit Version of Private and
Journal Program Public
Situation Description Entry Message Owner Restored Authorities

Program created before V1R3; security level 40 or None None Unchanged Re-created Unchanged
higher; ALWOBJDIF(*NONE); retranslation
successful
Program created before V1R3; security level 40 or Yes CPF375B QDFTOWN Original Revoked
higher; ALWOBJDIF(*NONE); retranslation not
successful
Program created on V1R3 or later; validation None None Unchanged Original Unchanged
value is valid
Program created on V1R3 or later; validation Yes CPF375C Unchanged Re-created Unchanged
value is not valid, retranslation successful
Program created on V1R3 or later; validation Yes CPF375A Unchanged Original Unchanged
value is not valid; retranslation is not successful;
security level less than 40
Program created on V1R3 or later; validation Yes CPF375D Unchanged Original Unchanged
value is not valid; retranslation is not successful;
security level 40 or greater; ALWOBJDIF(*ALL)
Program created on V1R3 or later; validation Yes CPF375B QDFTOWN Original Revoked
value is not valid; retranslation is not successful;
security level 40 or greater; ALWOBJDIF(*NONE)

Restoring Programs to a Different Release

Servers that run Version 3 Release 2 Modification 0 or earlier of the OS/400
licensed programs are IMPI (internal microprogramming interface) processors.
IMPI refers to the low-level instruction set and the Licensed Internal Code. iSeries
or AS/400e processors that run V3R6 or higher of the OS/400 licensed programs
are PowerPC AS processors. When you move a program object (*MODULE, *PGM,
*SRVPGM, *SQLPKG) between a system with an IMPI processor and a system with
a PowerPC AS processor, the system must create the program object again from the
information that is stored with the program.

Object conversion occurs at one of the following times:

v When the object is used for the first time. This is the default.
v When you use the Start Object Conversion (STROBJCVN) command to convert
objects. This is normally done for an entire library.
v When you restore the object. The Force Object Conversion On Restore
(QFRCCVNRST) system value and the Force Object Conversion (FRCOBJCVN)
parameter on the restore command determine whether an object is converted
when it is restored.

Following are the possible values for the QFRCCVNRST system value. The default
value is 0.

Possible Values for the QFRCCVNRST System Value:

0 Objects are not converted when they are restored. This is the recommended value
when restoring all your user libraries.
1 A restored object is converted if the object is not already in the format required
by the target system. If a program does not have the necessary creation data, it is
restored but not converted. A message is sent.

Chapter 10. How to Restore Specific Types of Information 253

 Following are the possible values for the FRCOBJCVN parameter on the RSTLIB
command, the RSTOBJ command, and the RST command. The default value is
*SYSVAL. If you specify *YES, you must specify a second part of the parameter.

Possible Values for the FRCOBJCVN Parameter:

*SYSVAL The value specified for the QFRCCVNRST system value is used.
*NO Objects are not converted when they are restored.
*YES *RQD A restored object is converted if the object is not already in the format required by the target
system. If a program does not have the necessary creation data, it is restored but not
converted. A message is sent.
*YES *ALL A restored object is converted even if it is already in the format required by the target
system. If a program does not have the necessary creation data, it is not restored. You might
choose this option after your new system is operational if you have high security
requirements.

The AS/400 Road Map for Changing to PowerPC Technology has more information
about moving from an IMPI system to a system with a PowerPC AS processor.

Restoring OPM Program Objects

If you create and save OPM (Original Program Model) programs on a system with
a PowerPC AS processor and specify an IMPI target release, the objects are saved
in a form that requires conversion when restored to either an IMPI or PowerPC AS
system.

Note: As of V5R1M0, you can no longer specify an IMPI target release.

This conversion can take significant time, depending on the model and size of the
system on which the objects are restored. OPM programs that are saved with a
PowerPC AS target release, do not require conversion when restored to a PowerPC
AS system. For these reasons, we recommend that you do not specify an IMPI
target release for routine backups from a system with a PowerPC AS processor.

If you are creating a single distribution medium for both IMPI and PowerPC AS
systems, you might want to place two separate save files on the same distribution
medium. The first save file would contain OPM programs that were saved with
the IMPI target release. The second save file would contain the OPM files that
were saved with the PowerPC AS target release. Files with the IMPI target release
will still require conversion, but the files with the PowerPC AS target release can
be restored to PowerPC AS systems without conversion. To avoid conversions for
both IMPI and PowerPC AS restore operations, create the IMPI save file on an
IMPI system and the PowerPC AS save file on a PowerPC AS system.
Notes:
1. These conversions are required because of significant differences between OPM
objects on PowerPC AS systems and IMPI systems.
2. ILE objects that are saved on a PowerPC AS system require no conversion
when restored on a PowerPC AS system, even if they were saved with an IMPI
target release.

254 OS/400 Backup and Recovery V5R1

Restoring Save File Data
You can save a save file to tape, optical media, or diskette with the SAVSAVFDTA
command. When you restore the save file, it appears as though the data originally
came from the same type of save media. You can use the RSTOBJ, RSTLIB, RST,
RSTDLO, RSTCFG, or RSTUSRPRF commands to restore the data.

You can save file data to tape, optical media, or diskette by using the SAVLIB,
SAVOBJ, or SAVCHGOBJ commands. If you specified SAVFDTA(*YES) on the save
command, you must restore the save file before you can restore the objects in the
save file.

Restoring Spooled Output Files

You cannot directly save and restore spooled files on an output queue. If you use
the technique described in the Backing up your system topic in the Information
Center at the following Web site: http://www.ibm.com/eserver/iseries/infocenter,
you can restore the spooled files by first restoring the database files with a restore
command, such as Restore Object (RSTOBJ) or Restore Library (RSTLIB), and then
copy the database file members to the spooled output files by using the Copy File
(CPYF) command and specifying TOFILE(QSYSPRT).

Restoring Licensed Programs

Use the RSTLICPGM command to add or replace licensed programs on the system.
Refer to the Software Installation book for more information about installing licensed
programs.

Restoring Documents and Folders

Use the Restore Document Library Object (RSTDLO) command to restore
documents, folders, and mail. To use this command most efficiently, you should
know how documents were saved. To determine this, use the output that was
printed for the SAVDLO procedures, the DSPTAP command, or the DSPOPT
command. RSTDLO performance is also better if you have *SAVSYS special
authority.

RSTDLO Command Options

The RSTDLO command provides many options. You can restore any of the
following:
v A specific document or system object whose name you specify.
v All the documents and folders you saved by typing: RSTDLO DLO(*ALL)
SAVFLR(*ANY). If you saved DLOs from more than one ASP, you must specify
SAVASP(*ANY). You must also specify the sequence numbers (SEQNBR parameter)
for the files on the save media.

Note: When you use RSTDLO DLO(*ALL), this includes the folders that are
used by IBM-supplied programs, such as Client Access. Ensure that you
saved these folders from the current release, or you may need to install
the licensed programs again.
v 1 to 300 documents from the same media file by specifying the names of the
documents or the system object names.
v 1 to 300 folders from the same media file.

Chapter 10. How to Restore Specific Types of Information 255

 v All filed documents that are not in any folder on the save media. See “Restoring
Folders” on page 258 for more information.

Using Multiple Concurrent DLO commands

Multiple concurrent SAVDLO or RSTDLO commands may be used in specific
situations.

No two of the following commands may be run on one system at the same time:
v RCLDLO DLO(*ALL)
v RCLDLO DLO(*DOCDTL)
v RCLDLO DLO(*INT)
v DLTDLO DLO(*ALL)
v RNMDIRE

An attempt to run these commands at the same time results in the message
CPF8A47: Internal system objects are in use. An attempt to run a SAVDLO or
RSTDLO operation while one of these commands is running will also result in
MSGCPF8A47 and no objects will be saved or restored.

Output from the RSTDLO Command

You can use the OUTPUT parameter on the RSTDLO command to show
information about the restored documents, folders, and mail. You can either print
the output (OUTPUT(*PRINT)) or save it to a database file (OUTPUT(*OUTFILE)).

If you print the output, you should be aware of device dependencies:

v The heading information in the output is device-dependent. All information does
not appear for all devices.
v The print file for the RSTDLO command uses a character identifier (CHRID) of
697 500. If the printer you are using does not support this character identifier,
you will receive message CPA3388. To print the RSTDLO output and not receive
message CPA3388, specify the following before specifying *PRINT on the
RSTDLO command:
CHGPRTF FILE(QSYSOPR/QPRSTDLO) CHRID(*DEV)

For more information about character identifiers (CHRID), see the Printer Device
Programming book.

If you use an output file, the system uses the file format
QSYS/QAOJRSTO.OJRDLO. The file layout is described in the Office Services
Concepts and Programmer’s Guide book.

Considerations and Restrictions

You should be aware of the following additional factors when using the RSTDLO
command.

Moving Documents
When you restore documents, you can rename them, restore them to a different
folder, or have the system assign new system object names. The folder for a
document determines its ASP location. You can move a document to a different
ASP by doing the following:
1. Save the document.
2. Delete it with the DLTDLO command.

256 OS/400 Backup and Recovery V5R1

3. Restore it into a folder in a different ASP.

Searching Tape Files

When you restore documents or folders from a list and specify
SEQNBR(*SEARCH), the system restores from the first tape file that contains any
of the documents or folders that you specified. If the tape file does not contain all
the documents and folders in your list, the system does not search other tape files
for the additional documents and folders. You can specify SEQNBR(starting-
sequence ending-sequence) to search more than one tape file.

Selecting files from DVD-RAM optical media

The OPTFILE and SAVASP parameters control which file(or files) that the system
uses. If you specify a file pathname, the system uses that file. If you specify the
default of OPTFILE(’*’) or OPTFILE(’directory-path-name/*’), the system uses files
named QDOC or QDOCnnnn in the directory that you specify, depending on the
SAVASP value.

Search Index Database Errors

When you restore DLOs, the system updates the search index database information
for the DLOs. If you receive error messages during the restore procedure because
the information in the database does not match the DLOs, run the Reclaim
Document Library Object (RCLDLO) command. Then try the restore procedure
again.

Note: The message tells you if the RCLDLO procedure is necessary. Use RCLDLO
only if you are instructed by a message or by the recovery checklist you are
using.

Authority Required to Restore DLOs

If you are restoring DLOs into a folder, you must have authority to the folder. If
you are restoring existing DLOs, you must have authority to those DLOs. Certain
combinations of the RSTDLO command require additional authority. The iSeries
Security Reference book provides information about the specific authorities that are
required for the RSTDLO command.

How the System Restores New DLOs

When you restore new DLOs, the system files them. The DLO is treated as new to
the system if any of the following is true:
v It has been previously deleted.
v It is being restored to a different system.
v It is being restored with the NEWOBJ(*NEW) parameter.

How the System Restores Existing DLOs

When you are restoring an existing DLO, the system skips the DLO and continues
with the next one if either of the following is true:
v The DLO is in use.
v You do not have the necessary authority.

If the existing document is damaged, some of the security information may be lost.
The restore operation continues and a message is sent informing you that the
document is damaged and some of the security information is lost.

Size Limitations When Restoring Document Library Objects

On V2R3 or later, you cannot restore more than 349 000 objects to a single library.
Before V2R3, the limit is 250 000 objects from a single library. Because DLOs are

Chapter 10. How to Restore Specific Types of Information 257

 nominally stored in libraries, this limit applies to the QDOC library in the system
ASP and to the QDOCnnnn libraries in user ASPs.

Restoring Folders
To restore a folder object, the entire folder (the folder object plus all document and
folder objects within it) must also be restored. However, if the specific folder being
restored was stored in other folders at the time it was saved, those higher level
folders do not have to be restored to restore the specific folder.

When you restore a folder, the fully qualified folder path name you are restoring
must exist unless you are restoring a first-level folder. For example, if you save
folder A and then delete it, you can enter RSTDLO DLO(*ALL) SAVFLR(A) and
restore folder A in addition to all the documents and folders in it. However, if you
want to restore folder A/B/C/D, you must create folder A, then folder B in folder
A, then folder C in folder A/B, before you can restore folder D in folder C. You
only have to create the folders that comprise the A/B/C path, and you do not
have to create folder D in folder A/B/C before you can restore it.

If you try to restore a folder that is in use, the system bypasses restoring the folder
and all the DLOs in it.

If you try to restore into an existing folder but the folder is damaged and cannot
be reclaimed, you receive a message informing you that the folder is damaged and
not restored. The folder and all documents and folders in it are not restored.

Renaming Documents When Restoring

You can use the RENAME parameter to give documents a different name when
they are restored. You can also place them in a different folder by using the
RSTFLR parameter.

If renaming a document when it is restored would result in a duplicate name in a

folder, the system does the following:
v If ALWOBJDIF(*NONE) is specified, the document is not restored.
v If ALWOBJDIF(*ALL) is specified, the document is restored and replaces the
existing document in the folder.

You can specify more than one value for the RENAME parameter. The system
matches the RENAME values with the DLO values until it runs out of values for
one or the other. Assume that you specify:
RSTDLO DLO(A B C D) SAVFLR(X) RENAME(J K L) RSTFLR(Y)

After the restore operation, you would have these documents:

v Document J in folder Y
v Document K in folder Y
v Document L in folder Y
v Document D in folder Y

Restoring OfficeVision/400 Mail and Distribution Objects

You can restore OfficeVision/400 mail by specifying RSTDLO DLO(*MAIL). If you
specified SAVDLO DLO(*MAIL) when you saved, you can specify RSTDLO DLO(*ALL)
SAVFLR(*ANY) to restore OV/400 mail.

258 OS/400 Backup and Recovery V5R1

 Specifying RSTDLO DLO(*MAIL) restores only those filed documents that have a
OV/400 mail log reference at the time they are saved. This also saves the
distribution objects and distribution documents from the save media or online save
file. Specifying RSTDLO DLO(*ALL) SAVFLR(*ANY) restores all distribution
objects, all documents, and all folders from the save media or online save file.

Distribution documents and objects cannot be restored individually. If you specify

any other form of the RSTDLO command, such as RSTDLO DLO(*ALL)
SAVFLR(A) and RSTDLO DLO(X) SAVFLR(A/B), then no distribution documents
and objects are restored. If you restore the filed documents using these other forms
of the RSTDLO command that contain OV/400 mail log references, then the
OV/400 mail log references are restored if the distribution objects exist on the
system.

Mail log references are updated for all existing local recipients of a restored
document. Mail log references on remote systems for remote recipients are not
restored. If a document being restored still exists in a mail log at the time it is
restored, then the contents of the document are restored and the status of the
document in the mail log is not changed. If the document being restored has been
deleted from a mail log, then the status of the restored document is either filed for
a filed document or opened for a distribution document.

OV/400 mail log references are restored for a local sender of a document if there
was an entry in the sender’s mail log at the time the distributions were saved.
Entries in the OV/400 mail logs of remote senders are not saved or restored.

How the System Restores Descriptive Information for DLOs

The creation date, file date, and revision date for restored documents and folders
are set as follows:
v The creation date of the document or folder on the save media is restored with
the document or folder.
v When the RSTDLO command replaces a document or folder, the file date of the
document or folder being replaced on the system is used.
v The object revision date is set to the current date when the document or folder is
restored.
v The content revision date of the document on the save media is restored with
the document.
v The content revision date is set to the current date when replacing a folder.
v The content revision date of the folder on the save media is restored with the
folder if the folder is new.

How the System Restores Authority and Ownership for DLOs

“How the System Establishes Ownership for Restored Objects” on page 218 and
“Restoring Object Authorities” on page 219 describe how the system handles
ownership and authority when restoring objects. These same rules apply when
restoring DLOs, with these additions:
v If the user profile that owns a DLO is not in the system distribution directory,
ownership is assigned to the QDFTOWN user profile.
v When you restore a DLO that does not exist on the system, any access codes and
explicit users are removed. If you have restored user profiles and you later run
the RSTAUT command, the private authorities to the DLO are restored. The
access codes are not restored.

Chapter 10. How to Restore Specific Types of Information 259

 When to Run the Rename Directory (RNMDIRE) Command
When you need to run the Rename Directory Entry (RNMDIRE) command for a
local user, schedule it just before you perform the following operations:
v Saving mail
v Saving the system distribution directory

If the rename operation is performed just before saving the mail and the directory,
the changed information is saved and the information will be the same as what is
on the system. If the information on the media does not match the information on
the system, the mail will not be restored during the restore operation.

When to Run the Rename Document Library Object (RNMDLO)

Command
When you need to run the Rename Document Library Object (RNMDLO)
command, schedule it just before you back up document library objects. If the
rename operation is performed just before saving the document library object, the
changed name is saved and the information on the media will be the same as what
is on the system.

If you rename a document library object after a save operation, the document
library object name on the system is different than the name on the media.
However, the system object names remain the same. The restore operation fails
because the system thinks that the document library object already exists. Message
CPF90A3 or CPF909C is sent indicating that the document or folder already exists.

Do one of the following:

v To create a new document or folder, specify NEWOBJ(*NEW).
v To replace an existing document, specify RENAME(document-name), where
document-name is the name that is given to the document by the RNMDLO
command.
v To replace an existing folder, specify RSTFLR(folder-name), where folder-name is
the name that is given to the folder by the RNMDLO command.

Recovery of Text Index Files for Text Search Services

The text index database files are a part of the Text Search Services. The text index
recovery process must ensure that a consistent, usable set of index files. The text
search index contents must be consistent with the document library contents. The
last version indexed date for all documents is recorded in both the text index and
the individual documents. The dates are used in the recovery process to ensure
that the text index and the document library content match.

The text index database files are saved when library QUSRSYS is saved. A list of
the files that are saved when library QUSRSYS is saved is shown in a table in the
Office Services Concepts and Programmer’s Guide book.

If you are restoring the text index files, then all of the files must be restored
together from the same backup media. If they are not restored from the same
media, their association to each other is lost. The loss of the association to each
other can cause unpredictable results. If you do not have saved copies of the files,
you must delete the files and then restore them from your distribution media.

260 OS/400 Backup and Recovery V5R1

 The text index details are kept in the administration table file. Pointers to the
current index are stored in the table. The administration table file must be restored
with the other files. If you have changed the defaults for the text index details,
then before restoring the files, write down the current text index details (if
available). To display the text index details, type WRKTXTIDX on a command line and
press the Enter key. Then select option 5 (Display details) on the Work with Text
Index display. You can enter the values again after the table is restored.

If the scheduling queue (file QABBIQTB) is damaged and there are documents on
the system, you can restore the scheduling queue and can get back some of the
requests that were lost if the saved scheduling queue is a very recent copy.
Retaining the requests on the restored scheduling queue may not be a benefit if it
is not a recent copy.

If you recover all text index search files, and documents from the same set of save
media, you should not have problems. If you recover your system in pieces,
consult Appendix F. Procedures for Recovering the Text Index for possible
problems and their solutions.

For more information about Text Search Services, see the Office Services Concepts
and Programmer’s Guide.

Restoring Objects in Directories

Use the RST (Restore) command to restore objects that you have saved with the
SAV command. These commands are most commonly used to save and restore
objects the QNTC file system, the QOpenSys file system, and the Root file system.

Attention!
| If you have related objects, such as journals and journaled objects, you must
| ensure that you restore them in the correct sequence. Read “Sequence for
| Restoring Related Objects” on page 45.

| If you are restoring to a different system, specify ALWOBJDIF(*ALL) when

| you are restoring directories.

You can use the RST command to restore:

v A specific object
v A directory or subdirectory
v An entire file system
v Objects that meet search criteria
v A list of object path names

You can also restore the items in the preceding list by using the QsrRestore API.
For more information, look in the Information Center under the Programming
topic at the following Web site: http://www.ibm.com/eserver/iseries/infocenter.

For example, to restore all objects (or changed objects) in directories, use the
following:
RST DEV('/QSYS.LIB/media-device-name.DEVD')
OBJ(('/*') ('/QSYS.LIB' *OMIT) ('/QDLS' *OMIT))

Chapter 10. How to Restore Specific Types of Information 261

 Note: This example is the same restore (RST) command that is issued under option
21 on the Restore menu.

You can rename an object or restore it to a different directory by using the

new-name element of the object (OBJ) parameter.

The OBJ parameter on the RST command supports the use of wildcard characters
and the directory hierarchy. For more information about how to specify object
names when you use integrated file system commands, look at the Information
Center at the following Web site:
http://www.ibm.com/eserver/iseries/infocenter.

Some file systems allow you to name the same physical object different ways by
using aliases and links. For examples of objects with links and how those objects
are saved, go to the Backing up your system topic on the Information Center.

In the example of Figure 24, FILEA in the JCHDIR directory and FILEB in the
DRHDIR directory are both hard links to the same file. They point to the same
object. They can have the same name or different names for the objects.

┌──────────────┐
│ Root │
└──────┬───────┘
N
┌──────────────┐
│ UserDir │
└──────┬───────┘
│
┌──────────────┴──────────────┐
N N
┌──────────────┐ ┌──────────────┐
│ JCHDIR │ │ DRHDIR │
└─────┬────────┘ └──────┬───────┘
│ │
N N
┌──────────────┐ ┌──────────────┐
│ FILEA │ │ FILEB │
└──────────────┘ └──────────────┘
│ │
┌──────────────┐
└ ─ ─ ─S│ Object A │W─ ─ ─┘
└──────────────┘

Figure 24. An Object with Hard Links–Example

Table 52 shows several examples of how the system restores these objects. These
examples assume that you use this SAV command: SAV OBJ('/UserDir/*'). The
media volume contains OBJECT A and both hard links that point to the object.
Table 52. Restoring Objects That Have Hard Links
Objects That Are on the
System before the RST
Object Parameter on RST Command Command Objects after the RST Command

RST OBJ('/UserDir/*') JCHDIR/FILEA The saved data is restored. The object

DRHDIR/FILEB is created on the
system. It points to the same object as
JCHDIR/FILEA.

262 OS/400 Backup and Recovery V5R1

Table 52. Restoring Objects That Have Hard Links (continued)
Objects That Are on the
System before the RST
Object Parameter on RST Command Command Objects after the RST Command

RST OBJ('/UserDir/DRHDIR/*') JCHDIR/FILEA A new object, DRHDIR/FILEB, is

created. The JCHDIR/FILEA that exists
on the system is not affected by the
restore operation.
OBJ('/UserDir/*'), or JCHDIR/FILEA, Data from the media copy of FILEA
OBJ('/UserDir/JCHDIR/*'), or DRHDIR/FILEB and FILEB is restored over the system
OBJ('/UserDir/DRHDIR/*') copy because the same name is
specified as a name that already exists
on the system.

Figure 25 shows the symbolic link called customer that points in the CUSTLIB
library.

Figure 25. An Object with a Symbolic Link–Example

If you restore the customer object (RST OBJ('/customer')), you are restoring only
the fact that it points to the CUSTMAS file, not the file itself. If the CUSTMAS file
does not exist, the restore operation succeeds. However, if you try to use the
customer object, you receive an error message. If you restore the CUSTMAS file or
create it again, the symbolic link between customer and the CUSTMAS file is
re-established.

Completing Recovery for the iSeries Integration for Windows Server

Product

Recovery for save performed with Integrated xSeries Server

varied off
If you completely saved your directories with the Integrated xSeries Server varied
off, your system restores Windows server on iSeries data. You need to perform the
following steps to complete the recovery of these products:
1. To add the links for the server descriptions, type the following for each server
description:
ADDNWSSTGL NWSSTG(Storage_Name)
NWSD(Server_Description)
2. Vary on your Integrated xSeries Servers by typing WRKCFGSTS *NWS and selecting
option 1 to vary on each Integrated xSeries Server.

Chapter 10. How to Restore Specific Types of Information 263

 Note: If you saved the Server Storage space beneath QFPNWSSTG (by using the
command SAV OBJ('/QFPNWSSTG/Server_Storage'), you must create the
QFPNWSSTG first. Create /QFPNWSSTG by performing the following
steps:
1. Create the server storage by using the CRTNWSSTG command.
2. RST OBJ(’/QFPNWSSTG/Server_Storage’)
3. Add the storage link by using the ADDNWSSTGL command.
4. Vary on the Integrated xSeries Server by typing WRKCFGSTS *NWS and
selecting option 1 to vary on.

Recovery for save performed with Integrated xSeries Server

varied on
Perform the following steps for Windows server on iSeries:
1. If you have any Integrated xSeries Servers that are varied on, vary them off by
using the WRKCFGSTS *NWS command and selecting option 2.
2. Create any needed Network Storages by using the CRTNWSSTG command.
3. Add the storage links by using the ADDNWSSTGL command.
4. Vary on your Integrated xSeries Servers by using the WRKCFGSTS *NWS
command and selecting option 1.
5. Restore the Windows server data by typing RST OBJ('/QNTC') and press the
Enter key.

| Recovering Linux in a Partition

| If you completely saved your directories with the network server description
| (NWSD) for Linux varied off, your system restores Linux data. You need to
| perform the following steps to complete the recovery of Linux:
| 1. To add the links for the server descriptions, type the following for each server
| description:
| ADDNWSSTGL NWSSTG(Storage_Name)
| NWSD(Server_Description)
| 2. Vary on your NWSD for Linux by typing WRKCFGSTS *NWS and selecting option
| 1 to vary on each NWSD for Linux.

| Note: If you saved the Server Storage space beneath QFPNWSSTG (by using the
| command SAV OBJ('/QFPNWSSTG/Server_Storage'), you must create the
| QFPNWSSTG first. Create /QFPNWSSTG by performing the following
| steps:
| 1. Create the server storage by using the CRTNWSSTG command.
| 2. RST OBJ(’/QFPNWSSTG/Server_Storage’)
| 3. Add the storage link by using the ADDNWSSTGL command.
| 4. Vary on the NWSD for Linux by typing WRKCFGSTS *NWS and selecting
| option 1 to vary on.

| Recovery Steps for OS/400 Enhanced Integration for Novell

NetWare
OS/400 Enhanced Integration for Novell NetWare runs on a remote server. Your
iSeries server communicates with the remote server through /QNetWare, but it
saves all of the Netware data on remote server storage.

264 OS/400 Backup and Recovery V5R1

 The previous OS/400 Integration for Novell NetWare ran on an Integrated xSeries
Server which meant that you restored both the /QNetWare subdirectory and
Netware server storage when you restored your server.

Since the new OS/400 Enhanced Integration for Novell NetWare product does not
store any data on your server, you have two backup options. First, you can easily
backup the /QNetWare subdirectory and restore the /QNetWare subdirectory with
your server in a restricted state or a non-restricted state.

Your second option is to restore your server to the point that you can start your
network descriptions and save the data from your remote Netware server through
/QNetWare. This is very slow, however.

A better choice is to consider your remote server like a personal computer

workstation and save the Netware data with workstation backup software. You can
save the remote directories on your Netware Server with ARCserve or SBACKup
utilities after you vary the Integrated xSeries Server on.

Refer to the ARCserve or SBACKup documentation for the recovery steps.

For additional information about recovering your NetWare environment, refer to

the Information Center at the following Web site:
http://www.ibm.com/eserver/iseries/infocenter.

Recovering a Domino™ Server

The Domino product resides in libraries in the QSYS.LIB file system on your
server. All of your Domino databases reside in the integrated file system in a
directory path that you specify when you configure your server. Your backup
strategy for your Domino server should include saving both the libraries
(infrequently) and the database directories (frequently).

You might need to recover Domino for a variety of reasons, for example:
v Damage to your server, such as fire or flooding
v Hardware problems, such as a disk failure
v User or operator error, such as deleting a database or running a month-end
procedure twice.
Sometimes, you must recover your entire server. Other times, you must restore a
specific directory. The following topics provide general information about recovery
procedures for Domino.
v “Recovering an entire Domino server”
v “Recovering Domino mail” on page 266
v “Recovering specific Domino databases” on page 267
v “Restoring changed objects to a Domino server” on page 267

Recovering an entire Domino server

If you are faced with a system disaster, such as a site loss or the failure of an
unprotected disk unit, you must recover (restore) your entire server from a backup.
Because the iSeries and AS/400e servers are highly integrated systems, you must
restore objects in the correct sequence to rebuild the proper links between objects.
Consult other parts of this book for complete instructions for performing a full
system recovery.

Chapter 10. How to Restore Specific Types of Information 265

 If you are faced with a problem that requires restoring only your Domino server,
you can use the Restore (RST) command to restore your Domino directories from
save media. Following is an example of the steps.

Example
1. Start an iSeries session with a user profile that has *JOBCTL and *SAVSYS
special authorities.
2. To ensure that no one is using the server that you plan to restore, stop the
server. Use the End Domino Server (ENDDOMSVR) command.
3. Mount the media volume that has the most recent backup copy of the
directories for the server.
4. Type the appropriate restore (RST) command for your Domino directory. For
example, if your Domino directory is /NOTES/DATA, type the following
command:
RST DEV('/QSYS.LIB/media-device-name.DEVD')
OBJ('/NOTES/DATA/*')

Note: Consult your Domino documentation for any special recovery activities that
you might need to perform after you have restored the directories.

Recovering Domino mail

If you need to recover one or more mail databases from your backup save media,
use the Restore (RST) command. Following is an example of the steps:
1. Start an iSeries session with a user profile that has *JOBCTL and *SAVSYS
special authorities.
2. Stop the server that contains the mail databases that you want to restore. Use
the End Domino Server (ENDDOMSVR) command.
3. Mount the media volume that has the most recent backup of the mail
databases.
4. Type the appropriate Restore (RST) command for the mail databases that you
want to restore. For example, to restore all the databases to the MAIL
subdirectory, type the following command:
RST DEV('/QSYS.LIB/media-device-name.DEVD')
OBJ('/NOTES/DATA/MAIL/*')

Examples
v The name of a user’s mail database is usually the user’s ID (short name) with
the .NSF extension. (The Domino administrator has the option to use different
names for mail database files.) To restore a specific user’s mail database, such as
the mail database for user GNELSON, use the following command:
RST DEV('/QSYS.LIB/media-device-name.DEVD')
OBJ('/NOTES/DATA/MAIL/GNELSON.NSF')

v You can specify more than one file on the restore command. To restore mail
databases for GNELSON, LSMITH, and JPETERS, use the following command:
RST DEV('/QSYS.LIB/media-device-name.DEVD')
OBJ(('/NOTES/DATA/MAIL/GNELSON.NSF')
('/NOTES/DATA/MAIL/LSMITH.NSF')
('/NOTES/DATA/MAIL/JPETERS.NSF'))
Notes about the examples:
1. All of the examples assume that the directory for your Domino server is
/NOTES/DATA.

266 OS/400 Backup and Recovery V5R1

 2. You cannot restore over a database that is in use. All users must close the
database before you can restore a backup copy.
3. Consult your Domino documentation for any special recovery activities that
you might need to perform after you have restored Domino mail.

Recovering specific Domino databases

You might need to restore a specific Domino database or a group of databases. Use
the Restore (RST) command. Following is an example of the steps for restoring all
the files to the HRDPT subdirectory:
1. Start an iSeries session with a user profile that has *JOBCTL and *SAVSYS
special authorities.
2. Stop the server that contains the databases that you want to restore. Use the
End Domino Server (ENDDOMSVR) command. (You can restore a database
when the server is running. However, you need to make sure that no one is
using the database. Stopping the server is the best way to ensure that no one is
using the database.)
3. Mount the media volume that has the most recent backup of the databases.
4. Type the appropriate Restore (RST) command for the mail files that you want
to restore. For example, to restore all the files to the MAIL subdirectory, type
the following command:
RST DEV('/QSYS.LIB/media-device-name.DEVD')
OBJ('/NOTES/DATA/HRDPT/*.NSF')

Examples
v To restore a specific database that is named HRINFO to the HRDPT subdirectory
(folder), type the following:
RST DEV('/QSYS.LIB/media-device-name.DEVD')
OBJ('/NOTES/DATA/HRDPT/HRINFO.NSF')

v To restore all the Domino databases to the CUSTSVC subdirectory, type the
following:
RST DEV('/QSYS.LIB/media-device-name.DEVD')
OBJ('/NOTES/DATA/CUSTSVC/*.NSF')

v To restore all the Domino databases whose name begins with INV to the main
directory for your server, type the following:
RST DEV('/QSYS.LIB/media-device-name.DEVD')
OBJ('/NOTES/DATA/INV*.NSF')
Notes about the examples:
1. All of the examples assume that the directory for your Domino server is
/NOTES/DATA.
2. You cannot restore over a database that is in use. All users must close the
database before you can restore a backup copy.
3. Consult your Domino documentation for any special recovery activities that
you might need to perform after you have restored a Domino database.

Restoring changed objects to a Domino server

To reduce the length of your backup window, your save strategy might include
saving only changed objects from your Domino during the business week. When
you need to use this save media to recover, you must decide your recovery

Chapter 10. How to Restore Specific Types of Information 267

 sequence and determine the location of the most recent copy of each database.
Following are examples of different recovery scenarios and an overview of the
recovery steps for each.

Example: Restoring changed Domino objects from a cumulative

backup
Assume that your strategy for saving changed objects is cumulative (each night
you save everything that changed since the last complete backup). To recover your
entire Domino directory, do the following:
1. Start an iSeries session with a user profile that has *JOBCTL and *SAVSYS
special authorities.
2. To ensure that no one is using the databases, stop the Domino server. Use the
End Domino Server (ENDDOMSVR) command.
3. Locate the save media from your most recent complete backup. Mount the
correct media volume in the save device.
4. To restore the entire Domino database directory, use the Restore (RST)
command. For example,
RST DEV('/QSYS.LIB/media-device-name.DEVD')
OBJ('/NOTES/DATA/*')

5. Locate your most recent save media (from saving changed objects).
6. To restore all the objects on the save media (everything that has changed since
your full backup), type the following:
RST DEV('/QSYS.LIB/media-device-name.DEVD')
OBJ('/NOTES/DATA/*')
Notes about the example:
1. All of the examples assume that the directory for your Domino server is
/NOTES/DATA.
2. You cannot restore over a database that is in use. All users must close the
database before you can restore a backup copy.
3. Consult your Domino documentation for any special recovery activities that
you might need to perform after you have restored a Domino database.

Example: Restoring changed Domino objects from a nightly

backup
Assume that your strategy for saving changed objects is nightly (each night you
save only objects that changed since the previous night). To recover your entire
Domino directory, do the following:
1. Start an iSeries session with a user profile that has *JOBCTL and *SAVSYS
special authorities.
2. To ensure that no one is using the databases, stop the Domino server. Use the
End Domino Server (ENDDOMSVR) command.
3. Locate the save media from your most recent complete backup. Mount the
correct media volume in the save device.
4. To restore the entire Domino database directory, use the Restore (RST)
command. For example,
RST DEV('/QSYS.LIB/media-device-name.DEVD')
OBJ('/NOTES/DATA/*')

5. Locate your first save media volume (from saving changed objects). For
example, if you save everything on Saturday night, locate your save media
from Sunday night.

268 OS/400 Backup and Recovery V5R1

6. To restore all the objects on the save media (everything that has changed since
the previous night), type the following:
RST DEV('/QSYS.LIB/media-device-name.DEVD')
OBJ('/NOTES/DATA/*')

7. Repeat steps 5 on page 268 and 6 for each nightly save media until your
directory is current. For example, if you are restoring on Thursday, you will
need to use the media volumes for Monday, Tuesday, and Wednesday nights.
Notes about the example:
1. All of the examples assume that the directory for your Domino server is
/NOTES/DATA.
2. You cannot restore over a database that is in use. All users must close the
database before you can restore a backup copy.
3. Consult your Domino documentation for any special recovery activities that
you might need to perform after you have restored a Domino database.

Example: Restoring Domino databases from an incremental

backup
To restore a specific database that is named HRINFO to the HRDPT subdirectory
(folder), do the following:
1. Start an iSeries session with a user profile that has *JOBCTL and *SAVSYS
special authorities.
2. To ensure that no one is using the databases, stop the Domino server. Use the
End Domino Server (ENDDOMSVR) command.
3. Locate the most recent save media that has the database. Do one of the
following:
Consult the log that the system creates during the save operation.
Use the Display Tape (DSPTAP) or the Display Optical (DSPOPT) command
to display the contents of the save media volume.
4. Mount the save media volume in the save device.
5. To restore the database, type the following:
RST DEV('/QSYS.LIB/media-device-name.DEVD')
OBJ('/NOTES/DATA/HRDPT/HRINFO.NSF')

Notes about the example:

1. All of the examples assume that the directory for your Domino server is
/NOTES/DATA.
2. You cannot restore over a database that is in use. All users must close the
database before you can restore a backup copy.
3. Consult your Domino documentation for any special recovery activities that
you might need to perform after you have restored a Domino database.

Example: Restoring changed objects from a specific Domino

subdirectory
To restore all the Domino databases to the CUSTSVC subdirectory, use the same
approach that you use to recover the entire server. Do the following:
1. Start an iSeries session with a user profile that has *JOBCTL and *SAVSYS
special authorities.
2. To ensure that no one is using the databases, stop the Domino server. Use the
End Domino Server (ENDDOMSVR) command.
3. Locate the save media from your most recent complete backup. Mount the
correct media volume in the save device.

Chapter 10. How to Restore Specific Types of Information 269

 4. To restore the entire directory from the media volumes from your last full save,
use the RST (Restore) command:
RST DEV('/QSYS.LIB/media-device-name.DEVD')
OBJ('/NOTES/DATA/CUSTSVC/*')

5. If your incremental backup media volumes are cumulative, mount your most
recent incremental backup media volume. Use the same restore command (step
4) to restore the changes.
Otherwise, if your backup media volumes are nightly, repeat step 4 for each
incremental backup media volume. Start with the oldest volume and work
forward.
Notes about the example:
1. All of the examples assume that the directory for your Domino server is
/NOTES/DATA.
2. You cannot restore over a database that is in use. All users must close the
database before you can restore a backup copy.
3. Consult your Domino documentation for any special recovery activities that
you might need to perform after you have restored a Domino database.

Restoring a Windows server

Restoring the NWSD and disk drives for Windows server on

iSeries
One method of restoring your Windows server data is to restore the network
server description and that OS/400 associates with that server. On pre-V4R5
systems, this is your primary OS/400 recovery option. It remains the faster method
for restoring large amounts of data. For Windows servers running on V4R5 and
later systems, if you used file-level backup, you can also restore specific Windows
server files. You can find information about that procedure in the iSeries
Information Center.

When you restore saved objects from OS/400, you need to be aware of these
considerations:
Notes:
1. Treat a network server description (NWSD) of type *WINDOWSNT, its
predefined disk drives, and any user-defined disk drives that are linked to it as
a unit. Restore them at the same time. Otherwise, Windows server may not be
able to re-establish items such as Windows server File System permissions.
2. To have OS/400 automatically relink restored disk drives in the integrated file
system to the appropriate NWSD, restore the NWSD after you restore the disk
drives.
3. If you restore an NWSD of type *WINDOWSNT before restoring the predefined
and user-defined disk drives in the integrated file system, you need to relink
those disk drives. You can do this by using the Add Network Server Storage
Link (ADDNWSSTGL) command for each disk drive that is associated with the
NWSD:
ADDNWSSTGL NWSSTG(Storage_Name) NWSD(NWSD_Name)
4. When you restore a domain controller, ensure that the domain database held on
the server is synchronized with the other domain controllers. Follow normal
Windows server procedures to do this and refer to documentation from
Microsoft as necessary.

270 OS/400 Backup and Recovery V5R1

To restore the the NWSD and disk drives for Windows server on iSeries, refer to
these pages:
v “Restoring predefined disk drives for Windows servers created on V4R5 and
later systems”
v “Restoring predefined disk drives for Windows servers created on pre-V4R5
systems”
v “Restoring user-defined disk drives for Windows servers on iSeries” on page 272
v “Restoring NWSDs for Windows server on iSeries” on page 273

Restoring predefined disk drives for Windows servers created on

V4R5 and later systems
For Windows servers that you create on V4R5 or later systems, disk drives that
contain the Windows server operating system and registry are in the integrated file
system. You restore these predefined disk drives just as you do user-defined disk
drives. To restore disk drives in the integrated file system on OS/400, use the
Restore (RST) command:
1. If you are restoring from save media, ensure that you have mounted your
media.
2. If there are no network server storage spaces that currently exist on the system
(none appear when you use the WRKNWSSTG command), you must create the
/QFPNWSSTG directory before you can restore network server storage spaces
that you saved beneath that directory. To create the /QFPNWSSTG directory,
complete these steps:
a. On the OS/400 command line, type CRTNWSSTG to create a server storage
space and press F4.
b. Provide a name for the storage space.
c. Use the minimal size allowed and specify the appropriate disk pool (ASP).
d. Press Enter to create the storage space. OS/400 creates the storage space in
the /QFPNWSSTG directory.
3. To restore the storage spaces, type RST and press F4.
4. If you saved the storage space to a save file instead of to tape, use *SAVF for
the device. Otherwise, specify the device name.
5. In the Name field under Objects:, specify '/QFPNWSSTG/stgspc', in which stgspc
is the name of the network server storage space.
To restore the system (C) drive, use /QFPNWSSTG/nwsdname1. To restore the D
drive, use /QFPNWSSTG/nwsdname2.
6. Specify values for any other parameters that you want and press Enter to
restore the storage space.
7. You also need to restore any user-defined disk drives that are associated with
the server and restore the NWSD. When you are done restoring the NWSD and
all its associated disk drives, vary on the Windows server.

Restoring predefined disk drives for Windows servers created on

pre-V4R5 systems
Earlier versions of iSeries Integration for Windows Server created disk drives for
the C, D, and E drives in the QUSRSYS library. Those disk drives contain the
Windows server operating system and registry, boot and system drives. Even after
you upgrade your system to V4R5, these storage spaces remain where OS/400
created them unless you reinstall Windows server. You restore those storage spaces
with the Restore Object (RSTOBJ) command. System drives that are larger than
1007 megabytes also have data in a network storage space that you need to restore.

To restore server storage spaces, you use the Restore Object (RSTOBJ) command:
1. On the OS/400 command line, type RSTOBJ and press F4.
2. If you are restoring from save media, ensure that you have mounted your
media.

Chapter 10. How to Restore Specific Types of Information 271

 3. In the Objects field, specify the name the storage space. (If you want to
restore all the predefined storage spaces, first type + and press Enter.)
v To restore the C drive, specify the name of the NWSD followed by 1.
v To restore the D drive, specify the name of the NWSD followed by 2.
v To restore the E drive, specify the name of the NWSD followed by 3.
4. In the Save Library field, specify QUSRSYS.
5. In the Device field, specify either the name of the device that contains the save
media or specify *SAVF if you are restoring from a save file.
6. In the Object types field, specify *SVRSTG.
7. If you are restoring from a save file, specify the name and library for the save
file.
8. Press Enter to restore the storage spaces.
9. If your system drive (E) is not larger than 1007 megabytes, go directly to step
10. If your system drive is larger than 1007 megabytes, you need to restore
data that you saved from an additional disk drive in the integrated file
system:
a. If no disk drives currently exist on the system (none appear when you use
the WRKNWSSTG command), you must create the /QFPNWSSTG
directory before you can restore disk drives that you saved beneath that
directory. To create the /QFPNWSSTG directory, complete these steps:
1) On the AS/400 command line, type CRTNWSSTG to create a server
storage space and press F4.
2) Provide a name for the storage space.
3) Use the minimal size allowed and specify the appropriate ASP.
4) Press Enter to create the storage space. OS/400 creates the storage
space in the /QFPNWSSTG directory.
b. To restore the storage space, type RST and press F4.
c. If you saved the storage space to a save file instead of to tape, use *SAVF
for the device. Otherwise, specify the device name.
d. In the Name field under Objects:, specify '/QFPNWSSTG/nwsdname3', in
which nwsdname3 is the name of the network server storage space for the E
drive.
e. Specify values for any other parameters that you want and press Enter to
restore the storage space.
10. You also need to restore any user-defined disk drives that are associated with
the server and restore the NWSD. When you are done restoring the NWSD
and all its associated disk drives, vary on the Windows server.

Restoring user-defined disk drives for Windows servers on

iSeries
Although, you can now backup and restore individual Windows server files, the
fastest way to restore large amounts of data is to restore the entire storage space. If
you saved your user storage space from the \QFPNWSSTG directory, you can
restore only the entire storage space. You cannot restore individual files from this
backup.

To restore disk drives in the integrated file system, do this:

1. If you are restoring from save media, ensure that you have mounted your
media.
2. If there are no network server storage spaces currently exist on the system
(none appear when you use the WRKNWSSTG command), you must create the
/QFPNWSSTG directory before you can restore network server storage spaces
that you saved beneath that directory. To create the /QFPNWSSTG directory,
complete these steps:
a. On the OS/400 command line, type CRTNWSSTG to create a server storage
space and press F4.

272 OS/400 Backup and Recovery V5R1

 b. Provide a name for the storage space.
c. Use the minimal size allowed and specify the appropriate ASP.
d. Press Enter to create the storage space. OS/400 creates the storage space in
the /QFPNWSSTG directory.
3. To restore the storage spaces, type RST and press F4.
4. In the Objects: name field, specify '/QFPNWSSTG/stgspc' and
'dev/QASPnn/stgspc.UDFS', where stgspc is the name of the network server
storage space and nn is the number of the ASP.

Note: To restore the .UDFS object to an independent ASP, the ASP device must
be varied on.
5. Specify values for any other parameters that you want and press Enter to
restore the storage space.
6. You also need to restore any predefined disk drives that are associated with the
server and restore the NWSD. When you are done restoring the NWSD and all
its associated disk drives, vary on the Windows server.

Restoring NWSDs for Windows server on iSeries

In a disaster recovery situation, you would restore all the configuration objects,
which include the network server description (NWSD) for your Windows server. In
some situations, for example when you migrate to new Integrated xSeries Server
hardware, you need to specifically restore the NWSD. To have OS/400
automatically relink disk drives within the integrated file system to the restored
NWSD, restore those disk drives first. To restore the NWSD, you use the Restore
Configuration (RSTCFG) command:
1. On the OS/400 command line, type RSTCFG and press F4.
2. In the Objects field, specify the name of the NWSD.
3. In the Device field, specify the device name if you are restoring from media. If
you are restoring from a save file, specify *SAVF and identify the name and
library for the save file in the appropriate fields.
4. Press Enter to have OS/400 restore the NWSD.
5. When you are done restoring the NWSD and all its associated storage spaces,
vary on the Windows server.

Note: When you restore a NWSD, you must also restore any line, controller, and
device description objects that are associated with the NWSD.

Recovering Windows server files

For Windows servers at V4R5 and later iSeries Integration for Windows Server:

Release V4R5 of iSeries Integration for Windows Server supports file-level backup
and recovery of your Windows server files. Now you can recover a particular file
from your OS/400 backup without restoring the entire disk drive. Before using this
method, however, consider the amount of data you need to restore. For large
amounts of data, restoring an entire disk drive object is much faster than restoring
all the individual files in the disk drive. To restore a smaller amount of data, this
method works great.

You should restore the directory first, then files, then the registry, then reboot
Windows server for new registry entries to take effect. To restore files that you
saved by this method, use the RST command:
1. Ensure that the Windows server and TCP/IP are running.
2. On the OS/400 command line, type RST and press F4.
3. In the Device field, specify the device on which the data is available. (For
example, 'QSYS.LIB/TAP01.DEVD' restores the data from tape.)

Chapter 10. How to Restore Specific Types of Information 273

 4. In the Object field, specify what you want OS/400 to restore in the form
’/QNTC/servername/sharename’
You can use wildcard characters. Refer to “Examples: How to address parts of
Windows server” for how to specify particular parts of Windows server. Avoid
restoring Windows system files by this method because the restored files may
behave unpredictably.
5. In the Name field, specify the pathname of the object to restore.
6. You can use the Include or omit field to include or omit objects with the
pattern that you specify in the Name portion of the Object parameter.
7. In the New object name field, leave the object name the same or specify a new
pathname. The new pathname must be referenced by a sharename that exists
on Windows server.

Note: When you save a directory that has shares defined over it, OS/400
saves the share information with the directory. If you specify a new
object name when you restore the directory, OS/400 does not recreate
these shares.
8. Use the Directory subtree field to specify whether you want to restore
subtrees under a directory. The default is to restore all directories.
9. To specify that you want to restore files that were saved during a particular
period, specify starting and ending dates and times in the Change period field.
10. Provide any other information on the screen that you want OS/400 to use to
restore the files and press Enter.
11. When the files are restored, reboot Windows server for new registry entries to
take effect.

Examples: How to address parts of Windows server

These examples show how to refer with the SAV or RST commands to specific
parts of Windows server for a server that is named server1:

To save or restore this: Specify this:

All Windows server objects. OBJ('/QNTC/*') SUBTREE(*ALL)
All objects for server1. OBJ('/QNTC/server1/*') SUBTREE(*ALL)
All objects for server1 that OBJ('/QNTC/server1/*') SUBTREE(*ALL)
changed since you last saved CHGPERIOD(*LASTSAVE)
the files.
All objects for server1 that OBJ('/QNTC/server1/*') SUBTREE(*ALL)
changed during a certain CHGPERIOD('10/19/99' '00:00:00' '10/25/99'
period (in this case between '23:59:59')
10/19/99 and 10/25/99).
All directories, files, and OBJ('/QNTC/server1/fshare/*') SUBTREE(*ALL)
shares to which a particular
share (for example, ’fshare’)
refers. OS/400 does not save
and restore the directory over
which the share is built.
Only files to which the OBJ('/QNTC/server1/fshare/pay*')
specified share (for example,
’fshare’) refers that match the
specified pattern (pay*).
OS/400 does not save
directories nor shares.

274 OS/400 Backup and Recovery V5R1

 To save or restore this: Specify this:
Only directories and shares OBJ('/QNTC/server1/fshare') SUBTREE(*DIR)
(no objects) for ’fshare’ and its
immediate children.
Directories, shares, and files OBJ('/QNTC/server1/fdrive/terry/*') SUBTREE(*ALL)
for ’terry’ and its subtrees (not
directory ’terry’).
Only the specific file OBJ('/QNTC/server1/gdrive/myfile.exe')
’myfile.exe’.
The Windows server registry. OBJ('/QNTC/server1/$REGISTRY')

Restrictions When Using the Restore Command

The RST command can be used to restore objects to any file system. The topics that
follow describe restrictions that apply when using the RST command.

Restrictions When Restoring Objects to Multiple File Systems: When you use the
RST command to restore objects to more than one file system at the same time and
the file systems include the QSYS.LIB file system or the QDLS file system, the
following restrictions apply:
v Different file systems support different types of objects and different methods of
naming objects. Therefore, when you restore objects from more than one file
system with the same command, you cannot specify object names or object
types. You can restore all objects from all file systems, or you can omit some file
systems. These combinations are valid:
– Restoring all objects on the system: OBJ('/*')

Note: Using this command is not the same as using option 21 from the
Restore menu. Following are the differences between SAV OBJ(’/*’) and
option 21:
- RST OBJ(’/*’) does not put the system in a restricted state.
- RST OBJ(’/*’) does not start the controlling subsystem when it finishes.
- RST OBJ(’/*’) does not provide prompting to change default options.
– Restoring all objects in all file systems except the QSYS.LIB file system and
the QDLS file system: OBJ(('/*') ('/QSYS.LIB' *OMIT) ('/QDLS' *OMIT))
– Restoring all objects in all file systems except the QSYS.LIB file system, the
QDLS file system, and one or more other file systems: OBJ(('/*')
('/QSYS.LIB' *OMIT) ('/QDLS' *OMIT) ('/other values' *OMIT))
v Values for other parameters of the RST command are supported only for some
file systems. You must choose values that are supported by all file systems.
Specify the following parameters and values:
OPTION
*ALL
ALWOBJDIF
*NONE or *ALL
LABEL
*SEARCH
OUTPUT
*NONE

Chapter 10. How to Restore Specific Types of Information 275

 SUBTREE
*ALL
SYSTEM
*LCL
DEV (Must be a tape device or an optical device)
VOL *MOUNTED
v When you specify RST OBJ(’/*’), the following applies:
– The system restores only objects that are saved by SAV OBJ(’/*’).
– The system must be in a restricted state.
– You must have *SAVSYS or *ALLOBJ special authority.
– You cannot specify diskette, or save file for the DEV parameter.
– You must specify SEQNBR(*SEARCH).

Note: RST OBJ(’/*’) is not the recommended method for restoring the entire
system. Chapter 4. Selecting the Right Recovery Strategy describes how to
determine the recovery procedure for your situation.

Restrictions When Restoring Objects to the QSYS.LIB File System: When you
use the RST command to restore objects to the QSYS.LIB (library) file system, the
following restrictions apply:
v The OBJ parameter must have only one name.
v You specify objects in the same way that you specify them on the RSTOBJ
command and the RSTLIB command. Table 53 shows the valid options for the
Object (OBJ) parameter when restoring objects to the QSYS.LIB file system and
the equivalent RSTOBJ or RSTLIB command:
Table 53. Using the RST Command for QSYS.LIB Objects
Object Parameter on RST Command Equivalent RSTxxx Command

OBJ('/QSYS.LIB/library-name.LIB') RSTLIB SAVLIB(library-name)

OBJ('/QSYS.LIB/library-name.LIB/*') RSTOBJ SAVLIB(library-name) OBJ(*ALL)
OBJTYPE(*ALL)
OBJ('/QSYS.LIB/library-name.LIB/*.object-type') RSTOBJ SAVLIB(library-name) OBJ(*ALL)
OBJTYPE(object-type)
OBJ('/QSYS.LIB/library-name.LIB/object-name.object-type') RSTOBJ SAVLIB(library-name) OBJ(object-name)
OBJTYPE(object-type)

OBJ('/QSYS.LIB/library-name.LIB/file-name.FILE/*') RSTOBJ SAVLIB(library-name) OBJ(file-name)

OBJTYPE(*FILE)
OBJ('/QSYS.LIB/library-name.LIB/file-name.FILE/*.MBR') RSTOBJ SAVLIB(library-name) OBJ(file-name)
OBJTYPE(*FILE)
OBJ('/QSYS.LIB/library-name.LIB/file-name.FILE/ RSTOBJ SAVLIB(library-name) OBJ(file-name)
member-name.MBR') OBJTYPE(*FILE) FILEMBR((*ALL) (member-name))

v You can specify only object types that are allowed on the RSTOBJ command. For
example, you cannot use the RST command to restore user profiles because
OBJTYPE(*USRPRF) is not allowed on the RSTOBJ command.
v Some libraries in the QSYS.LIB file system cannot be restored with the RSTLIB
command because of the type of information they contain. Following are
examples:
– The QDOC library, because it contains documents.

276 OS/400 Backup and Recovery V5R1

 – The QSYS library, because it contains system objects.
You cannot use the RST command to restore these entire libraries:

QDOC QSRV QSPLxxxx1

QDOCxxxx1 QSPL
QRECOVERY QSYS
QRPLOBJ QTEMP
1
Where xxxx is a value from 0002 to 0032, corresponding to an ASP.

v You can use the new-name element of the object parameter to rename an object
in a directory, restore an object to a different directory, or restore an object to a
different library. Table 54 shows some examples:
Table 54. New Name Options on the RST Command–Examples
Object Parameter on RST Command Results

OBJ(('/DBSDIR/FILEB' *INCLUDE '/DBSDIR/FILEX')) FILEX is created in the DBSDIR directory. The data
that was saved with FILEB is restored to FILEX. If
FILEB still exists on the system, it is not changed.
OBJ(('/DBSDIR/FILE*' *INCLUDE LMSDIR)) Restores all objects from the DBSDIR whose names
begin with FILE to the LMSDIR directory.
OBJ(('/QSYS.LIB/LIB1.LIB' *INCLUDE '/QSYS.LIB/LIB2.LIB')) Library LIB1 (and all objects) are restored as library
LIB2.
OBJ(('/QSYS.LIB/LIB1.LIB/*' *INCLUDE All the objects of library LIB1 are restored into
'/QSYS.LIB/LIB2.LIB')) library LIB2.
OBJ(('/QSYS.LIB/LIB1.LIB/*.type' All objects of type ’type’ from library LIB1 are
*INCLUDE '/QSYS.LIB/LIB2.LIB')) restored into library LIB2.

v For database file members, OPTION(*NEW) restores members for new files only.
v Other parameters must have these values:
SUBTREE
*ALL
SYSTEM
*LCL
OUTPUT
*NONE
ALWOBJDIF
*ALL or *NONE
v You can only rename the library, you cannot rename the object. The new name
must be *SAME or
/QSYS.LIB/libname.LIB

where the library that is specified by libname must exist.

Restrictions When Restoring Objects to the QDLS File System: When you use
the RST command to restore objects to the QDLS (document library services) file
system, the following restrictions apply:
v The OBJ parameter must have only one name.
v The OBJ and SUBTREE parameters must be one of the following:
– OBJ('/QDLS/path/folder-name') SUBTREE(*ALL)
– OBJ('/QDLS/path/document-name') SUBTREE(*OBJ)
v Other parameters must have these values:

Chapter 10. How to Restore Specific Types of Information 277

 SYSTEM
*LCL
OUTPUT
*NONE
ALWOBJDIF
*ALL or *NONE
OPTION
*ALL

How to Restore Program Temporary Fixes

If you have restored the Licensed Internal Code or the operating system, you need
to ensure that the PTFs on your system are current. Do the following:
1. Print a list of all the program temporary fixes (PTFs) currently on the system.
Type the following and press the Enter key:
DSPPTF LICPGM(*ALL) OUTPUT(*PRINT)
2. Compare this list of PTFs with the list you printed when you saved the system.
If the lists are the same, return to your recovery checklist. If PTFs are missing
from the list you printed in step 1, you must apply them. Continue with the
next step.
3. Find the most recent cumulative program temporary fix media. This package
could be on distribution media or on a stand-alone media volume.

Note: If you do not have the PTFs you need, order them and apply them later.
Continue with your recovery checklist.
4. You can use option 8 (Install program temporary fix package) on the Program
Temporary Fix menu. All of the PTFs in the cumulative PTF package will be
installed for the licensed programs you have installed on your system. Refer to
the iSeries System PTF Shipping Information Letter for special instructions that are
required.
If you want to restore individual PTFs, see the System Operation book for more
information about applying individual PTFs.

278 OS/400 Backup and Recovery V5R1

Chapter 11. How to Restore Changed Objects and Apply
Journaled Changes
Figure 26 shows a typical time line for recovering your system.

Point 1
─────────────────────────────S
Known point (last save) ┌───────────────────┐
│ Activity occurs │
│ on system │
Point 2 └───────────────────┘
─────────────────────────────S
Failure occurs ┌───────────────────┐
│ Hardware repair │
│ or IPL │
Point 3 └───────────────────┘
─────────────────────────────S
Hardware is available ┌───────────────────┐
│ Information is │
│ restored from │
│ backup │
Point 4 └───────────────────┘
─────────────────────────────S
System is recovered to ┌───────────────────┐
known point 1 │ Transactions from │
│ point 1 to │
│ point 2 are │
│ recovered │
Point 5 └───────────────────┘
─────────────────────────────S
System is recovered to ┌───────────────────┐
failure point 2 │ Business activity │
│ from failure │
│ point 2 to │
│ recovery point 5 │
│ is recovered │
│ │
Point 6 └───────────────────┘
─────────────────────────────S
System is current

Figure 26. Sample Recovery Time Line

Chapter 5 through Chapter 10 describe what is required to reach point 4 in the

time line. This returns your system to the point of the last complete save operation.

This chapter describes two procedures that are available to reach point 5 in the
time line:
v Restoring changed objects
v Applying journal changes

These procedures are designed to recover activity that has occurred since the last
complete save operation.

© Copyright IBM Corp. 1997, 2000, 2001 279

 Task 1–Restoring Changed Objects
The Backing up your system topic on the Information Center at the following Web
site: http://www.ibm.com/eserver/iseries/infocenter describes two methods of
saving changed objects. Table 55 shows the two methods and the correct restore
procedures for each:
Table 55. Restore Procedures for Changed Objects
Save Method Description Restore Procedure

| Note: The SAVCHGOBJ command does not apply to objects in directories. If you are
| restoring changed objects in directories, go to “Task 2–Restoring Changed Objects in
| Directories” on page 281 for restore instructions from both the cumulative and not
| cumulative save methods.

Cumulative You save all changes since the last “Restoring Changed
complete save operation. Objects by Library”
Not cumulative You save changes since the last “Restoring Changed
SAVCHGOBJ operation. Objects Individually”

If you save journal receivers using the SAVCHGOBJ command, read “Restoring
Journal Receivers” on page 250 for special considerations that may apply when
restoring them.

Restoring Changed Objects by Library

Do the following to restore changed objects by library:
1. Load the SAVCHGOBJ media volume.
2. Type DSPTAP DEV(media-device-name) OUTPUT(*PRINT) for tape media. Type
DSPOPT VOL(*MOUNTED) DEV(OPT01) DATA(*SAVRST) PATH(*ALL) OUTPUT(*PRINT)
for DVD-RAM optical media. Press the Enter key.
3. For each library on the list, type:
RSTOBJ OBJ(*ALL) DEV(media-device-name) SAVLIB(library-name)
OBJTYPE(*ALL) ENDOPT(*LEAVE) MBROPT(*ALL)

Repeat this step for each library on the volume.

Attention!
If you have changed objects that do not restore because of creation date
mismatches for files or members, please refer to “Comparing File
Attributes during a Restore Operation” on page 238.

4. If you have journaled changes to apply, continue with “Task 4–Determining

What Journal Receivers to Use” on page 282. If you do not need to apply
journaled changes, skip to “Task 7–Restoring Changed Documents and Folders”
on page 287. If you are not sure whether you need to apply journaled changes,
continue with “Task 3–Determining Whether You Need to Apply Journaled
Changes” on page 282.

Restoring Changed Objects Individually

If your method for saving changed objects is not cumulative, you may have the
same object on more than one set of SAVCHGOBJ save media. You may choose to

280 OS/400 Backup and Recovery V5R1

 restore each set of SAVCHGOBJ save media completely, starting from the oldest
media volume. This is the simplest method. However, it may be time-consuming if
you have the same large objects on more than one SAVCHGOBJ media volume.

If you want to restore each set of SAVCHGOBJ save media completely, follow the
procedure described in “Restoring Changed Objects by Library” on page 280 for
each set of save media. If you want to restore each object only once, follow this
procedure:
1. Load each SAVCHGOBJ media volume.
2. Type DSPTAP DEV(media-device-name) OUTPUT(*PRINT) and press the Enter key.
3. Compare the listings and find the most recent saved copy of each object.
4. For each object, load the correct media volume and type:
RSTOBJ OBJ(object-name)DEV(media-device-name)
SAVLIB(library-name) OBJTYPE(*ALL)
ENDOPT(*LEAVE) MBROPT(*ALL)

Repeat this step for each object you need to restore.

5. If you have journaled changes to apply, continue with “Task 4–Determining
What Journal Receivers to Use” on page 282. If you do not need to apply
journaled changes, skip to “Task 7–Restoring Changed Documents and Folders”
on page 287. If you are not sure whether you need to apply journaled changes,
continue with “Task 3–Determining Whether You Need to Apply Journaled
Changes” on page 282.

Task 2–Restoring Changed Objects in Directories

Perform this task if you saved changed objects in directories. If you do not need to
do this task, continue with the next step in your recovery checklist.

If you use a cumulative method when you saved changed objects from directories
(your save media contain all objects that have changed since the last complete save
operation), do the following:
1. Mount your most recent save media from saving changed objects in directories.
2. Type:
RST DEV('/QSYS.LIB/media-device-name.DEVD')
OBJ(('/*') ('/QSYS.LIB' *OMIT) ('/QDLS' *OMIT))
3. If you have journaled changes to apply, continue with “Task 4–Determining
What Journal Receivers to Use” on page 282. If you do not need to apply
journaled changes, skip to “Task 7–Restoring Changed Documents and Folders”
on page 287. If you are not sure whether you need to apply journaled changes,
continue with “Task 3–Determining Whether You Need to Apply Journaled
Changes” on page 282.

If your save media from saving changed objects in directories is not cumulative,
repeat the following steps for each set of save media since your last complete save
operation. Start with the oldest save media volumes and end with the most recent
volumes.
1. Mount the media volume.
2. Type:
RST DEV('/QSYS.LIB/media-device-name.DEVD')
OBJ(('/*') ('/QSYS.LIB' *OMIT) ('/QDLS' *OMIT))
3. If you have journaled changes to apply, continue with “Task 4–Determining
What Journal Receivers to Use” on page 282. If you do not need to apply

Chapter 11. How to Restore Changed Objects and Apply Journaled Changes 281
 journaled changes, skip to “Task 7–Restoring Changed Documents and Folders”
on page 287. If you are not sure whether you need to apply journaled changes,
continue with “Task 3–Determining Whether You Need to Apply Journaled
Changes”.

Task 3–Determining Whether You Need to Apply Journaled Changes

You may have set up journaling yourself, or you may be using applications that
use journaling. For example, the OfficeVision program and the Client Access
program use the QUSRSYS/QAOSDIAJRN journal. Some applications provided by
software vendors also use journaling.

If you know you have journaled changes to apply, continue with “Task
4–Determining What Journal Receivers to Use”. If you are not sure, do the
following:
1. Type DSPOBJD OBJ(*ALL/*ALL) OBJTYPE(*JRN) OUTPUT(*PRINT) and press the
Enter key. This command prints a list of all the journals on your system.
2. For each journal on the list, do the following:
a. Type: WRKJRNA JRN(library-name/journal-name). You are shown the Work
with Journal Attributes display.
| b. Press F19 to display the objects being journaled.
c. Press F12 to return to the Work with Journal Attributes display.
| d. Press F15 to display the receiver directory. Note the attach and detach times
| for the journal receivers in relation to your journaled objects change dates.
| Additionally, you can use option 8 to display specifics about each journal
| receiver.
e. Press F12 to return to the Work with Journal Attributes display.
| f. From the information you have seen, you should be able to determine
| whether any objects are being journaled and whether any journal entries
| exist that are more recent than your most recent saved copies of the objects.
| You can also determine which receivers are on the system for the journal.
| Repeat these steps for each additional journal.
3. If you need to apply journaled changes, continue with “Task 4–Determining
What Journal Receivers to Use”. If you do not need to apply journaled changes,
skip to “Task 7–Restoring Changed Documents and Folders” on page 287.

Task 4–Determining What Journal Receivers to Use

The next several topics describe the general procedure for applying journaled
changes.

Follow this procedure:

1. Ensure that all the journal receivers required for the apply journaled changes
operation are available on the system. In general, you will need all journal
receivers that were attached to the journal for the length of time for which
journaled changes are now to be applied to the restored files.
2. Restore all necessary journal receivers that are not already on the system. Use
the Display Journal Receiver Attributes (DSPJRNRCVA) command to
determine when a journal receiver was attached to and detached from a
journal.
3. Determine the name of the last journal receiver (the last receiver restored) and
whether there are chain breaks by printing the receiver chain:

282 OS/400 Backup and Recovery V5R1

 a. Type WRKJRNA JRN(library-name/journal-name) OUTPUT(*PRINT) and press
| the Enter key. You receive a listing that shows the receiver directory and
| all the objects being journaled.
b. Look at the receiver directory part of the listing. If you saved the currently
attached journal receiver, your journal receiver directory should look
similar to Figure 27. The journal receiver that was attached during the save
procedure shows a status of Partial. The following example shows the
displayed version of the receiver directory:

Receiver Directory
Total size of receivers (in kilobytes). . . . . . : 1507
Attach Save Size
Number Receiver Library Date Date Status (K)
00001 RCVA0001 DSTJRN 06/08/9x 06/08/9x SAVED 42
00002 RCVA0002 DSTJRN 06/09/9x 06/09/9x SAVED 900
00003 RCVA0003 DSTJRN 06/09/9x 06/09/9x PARTIAL 92
01001 RCVA1003 DSTJRN 06/10/9x 00/00/00 ATTACHED 473

Figure 27. Receiver Directory–Saving Attached Receivers

If you save only detached journal receivers, your receiver directory should
look similar to Figure 28:

Receiver Directory
Total size of receivers (in kilobytes). . . . . . : 1507
Attach Save Size
Number Receiver Library Date Date Status (K)
00001 RCVA0001 DSTJRN 06/08/9x 06/08/9x SAVED 42
00002 RCVA0002 DSTJRN 06/09/9x 06/09/9x SAVED 900
00003 RCVA0003 DSTJRN 06/09/9x 06/09/9x SAVED 92
01001 RCVA1003 DSTJRN 06/10/9x 00/00/00 ATTACHED 473

Figure 28. Receiver Directory–Saving Detached Receivers

4. On the listing, mark the name of the last receiver with a status of SAVED or
PARTIAL.
5. Determine the chain of receivers to be used in the APYJRNCHG command
| from the Work with Receiver Directory listing. Mark the first and last receiver
| that you need, based on the date that you saved the objects being recovered.
Notice that the first and last receiver are the same if only one journal receiver
was restored.

Note: While looking at the receiver directory, you should also look for any
receiver chain breaks. You can determine a chain break by looking at
the first two digits in the Number column on the Work with Receiver
Directory display. You cannot apply journaled changes across receiver
chain breaks. Therefore, you must write down the beginning and
ending receiver names for each receiver chain. Then you need to run a
series of apply journaled changes operations, one for each chain using
these receivers. A chain break may mean that you are missing all or
part of a journal receiver. (It was on the system and was not saved
before the failure occurred.) You should evaluate how applying
journaled changes across a change break may affect the integrity of
your data. “Journal Receiver Chains” on page 440 has more information
about receiver chain breaks.
6. If the ending receiver has a status of PARTIAL (saved-while-attached),
determine the last entry to be applied for the last receiver:

Chapter 11. How to Restore Changed Objects and Apply Journaled Changes 283
 a. Type 8 (Display attributes) in the Opt field next to the receiver name.
b. Write down the value for the Last Sequence Number field.
| 7. Look at the part of the listing that shows what objects are currently being
| journaled. (You printed the listing in step 3a on page 283.) Compare it to your
| records of what objects should be journaled. Follow the procedures in
“Printing system information” on page 26 before you save your system.
8. For each physical file that should be journaled and does not appear on the
current listing, type:
STRJRNPF FILE(library-name/file-name)
JRN(library-name/journal-name)
9. For each access path that should be journaled and does not appear on the
current listing, type:
STRJRNAP FILE(library-name/file-name)
JRN(library-name/journal-name)
| 10. For each integrated file system object that should be journaled and does not
| appear on the current listing, type:
| STRJRN OBJ('object-path-name')
| JRN('journal-path-name')
| 11. For all other object types that should be journaled and do not appear on the
| current listing, type the following:
| STRJRNOBJ OBJ(library-name/object-name)
| OBJTYPE(object-type)
| JRN(library-name/journal-name)
| 12. The journal receiver that is currently attached may not match your naming
conventions. Usually this is because the journal receiver was created when
you restored the journal. If this is the case, create a new receiver that follows
the same naming convention and receiver attributes as the last receiver but
assign it a number of one greater. In the example shown on the Work with
Receiver Directory display, you would type:
CRTJRNRCV JRNRCV(DSTJRN/RCVA0004)
13. Use the CHGJRN command to detach the current receiver and attach the
journal receiver you just created. In the example, you would type:
CHGJRN JRN($JRNLA/JRNA)
JRNRCV(DSTJRN/RCVA0004)

Task 5–Applying Journaled Changes for User Journals

Do the following steps if you need to apply journaled changes to user journals. If
you do not need to apply journaled changes, skip to “Task 6–Applying Journaled
Changes for the QAOSDIAJRN Journal” on page 286.
1. If you have a single receiver chain for the journal entries that you need to
| apply and the status of the last receiver you are using is SAVED, do one of the
following:
| a. For objects in libraries type the following:
| APYJRNCHG JRN(library-name/journal-name)
| OBJ((library-name/*ALL object-type))
| RCVRNG(lib-name/first-receiver
| lib-name/last-receiver)
| FROMENT(*LASTSAVE) TOENT(*LASTRST)
| b. For objects in directories type the following:

284 OS/400 Backup and Recovery V5R1

| APYJRNCHG JRN(jrnlib/jrnname)
| OBJPATH('object-path-name')
| RCVRNG(lib-name/first-receiver
| lib-name/last-receiver)
| FROMENT(*LASTSAVE) TOENT(*LASTRST)

| Note: If you want to apply journaled changes to library and directory objects
in the same command invocation, you can use both the OBJ and
OBJPATH parameters in one APYJRNCHG command invocation.

If you have a single receiver chain for the journal entries that you need to
| apply and the status of the last receiver you are using is PARTIAL, do one of the
following:
a. For objects in libraries type the following:
| APYJRNCHG JRN(library-name/journal-name)
| OBJ((library-name/*ALL object-type))
| RCVRNG(lib-name/first-receiver
| lib-name/last-receiver)
| FROMENT(*LASTSAVE) TOENT(last-sequence-number)
| b. For objects in directories type the following:
| APYJRNCHG JRN(jrnlib/jrnname)
| OBJPATH('object-path-name')
| RCVRNG(lib-name/first-receiver
| lib-name/last-receiver)
| FROMENT(*LASTSAVE) TOENT(last-sequence-number)

| where the last-sequence-number is the number you wrote down in step 6 on

page 283 of “Task 4–Determining What Journal Receivers to Use” on page 282.

Note: If you want to apply journaled changes to library and directory objects
in the same command invocation, you can use both the OBJ and
OBJPATH parameters in one APYJRNCHG command invocation.
2. If you have determined that this journal had receiver chain breaks, you must
decide whether you are actually missing journal receivers and necessary journal
entries or whether the chain breaks were caused by something else. You should
evaluate how applying journaled changes across a chain break may affect the
integrity of your data. “Journal Receiver Chains” on page 440 provides more
information about receiver chain breaks.
If you decide to apply journal entries across chain breaks, you must use a
APYJRNCHG command for each chain. Type the APYJRNCHG command and
use these values in place of the values shown in step 1 on page 284.
For the first (earliest) receiver chain:
RCVRNG
First and last receivers in this chain
FROMENT
*LASTSAVE
TOENT
*LAST

For each middle receiver chain:

RCVRNG
First and last receivers in this chain
FROMENT
*FIRST

Chapter 11. How to Restore Changed Objects and Apply Journaled Changes 285
 TOENT
*LAST

For the last receiver chain:

RCVRNG
First and last receivers in this chain
FROMENT
*FIRST
TOENT
last-sequence-number

where the last-sequence-number is the number you wrote down in step 6 on

page 283.

Task 6–Applying Journaled Changes for the QAOSDIAJRN Journal

If you have document library objects, you may need to apply journaled changes
from the receivers associated with the QAOSDIAJRN journal. If you are not sure,
determine when you last saved the QUSRSYS library. Then perform the steps
through step 1e to determine whether you have any journal entries for the
QAOSDIAJRN journal that are more recent than your save media for the QUSRSYS
library.

You cannot apply all journaled changes in the QAOSDIAJRN journal in library
QUSRSYS. You must specify individual files on the FILE parameter instead of
*ALL. Do not apply journal changes to the document and folder search index
database files (QAOSSS10 through QAOSSS15, QAOSSS17, and QAOSSS18) for
journal QAOSDIAJRN in library QUSRSYS.
1. Display the receiver chain for the QAOSDIAJRN journal by doing the
following:
a. Type: WRKJRNA JRN(QUSRSYS/QAOSDIAJRN) and press the Enter key.
b. From the Work with Journal Attributes display, press F15 (Work with
receiver directory). Examine the receiver directory to determine whether any
chain breaks exist. (See the note on page 5 on page 283.)
c. Type 8 (Display attributes) in the Opt field next to the receiver name of the
last receiver that is not attached.
d. On the Display Journal Receiver Attributes display, write down the value
for the Last Sequence Number field.
e. Press F12 three times to return to a command line.
2. If no chain breaks exist, type the following to apply journaled changes for the
QAOSDIAJRN journal to individual files:
APYJRNCHG JRN(QUSRSYS/QAOSDIAJRN)
FILE((QUSRSYS/QAOKPLCA) (QUSRSYS/QAOSAY05)
(QUSRSYS/QAOKPX4A) (QUSRSYS/QAOSAY07)
(QUSRSYS/QAOKP01A) (QUSRSYS/QAOKP02A)
(QUSRSYS/QAOKP03A) (QUSRSYS/QAOKP04A)
(QUSRSYS/QAOKP05A) (QUSRSYS/QAOKP06A)
(QUSRSYS/QAOKP08A) (QUSRSYS/QAOKP09A))
RCVRNG(lib-name/first-receiver
lib-name/last-receiver)
FROMENT(*LASTSAVE)
TOENT(last-sequence-number)

where last-sequence-entry is the number you wrote down in step 1d.

286 OS/400 Backup and Recovery V5R1

 3. If chain breaks exist, you must determine whether any journal receivers are
missing and how that may affect the integrity of your recovery. If you decide to
apply journaled changes, use the command shown in step 2 of this topic.
Repeat the command for each receiver chain, substituting the correct receiver
range, from-entry, and to-entry parameters. Step 2 in the topic “Task
5–Applying Journaled Changes for User Journals” on page 284 describes how to
use these parameters.

Task 7–Restoring Changed Documents and Folders

Perform this task if you save changed documents and folders. If you do not need
to do this task, skip to “Task 2–Restoring Changed Objects in Directories” on
page 281.

Do the following:
1. If your procedure for saving changed DLOs is cumulative, load the last daily
SAVDLO media volume. If your procedure is not cumulative, start with your
earliest daily save volume and repeat these steps for each set of SAVDLO save
media.
2. If you have documents in user ASPs, display the save media volumes to find
the sequence numbers for each ASP. Type DSPTAP DEV(media-device-name)
OUTPUT(*PRINT) for tapes. Mark the names and sequence numbers of the files
on the listing. They will be named QDOC for the system ASP and QDOCnnnn
for each user ASP that contains DLOs, where nnnn is the number of the ASP.
3. To restore the DLOs to a single ASP, type:
RSTDLO DLO(*ALL) DEV(media-device-name) SAVFLR(*ANY)
SAVASP(ASP-number) RSTASP(*SAVASP)
4. To restore the DLOs to all ASPs, type:
RSTDLO DLO(*ALL) DEV(media-device-name) SAVFLR(*ANY)
SAVASP(*ANY) RSTASP(*SAVASP)

5. If an unrecoverable error occurs when running the RSTDLO DLO(*ALL)

SAVFLR(*ANY) command, see “Recovering from an Error While Restoring
DLOs” on page 57.

Chapter 11. How to Restore Changed Objects and Apply Journaled Changes 287
288 OS/400 Backup and Recovery V5R1
Chapter 12. Mirrored Protection Recovery Actions
In considering aspects of recovery, you need to distinguish between errors and
failures in the disk subsystem.

A disk error refers to an unexpected event during an I/O operation which can
cause the loss or corruption of the data that is being transferred. Most disk errors
are caused by a failure in some part of the component chain from the I/O
processor to the disk surface. Environmental effects such as power abnormalities or
severe electrostatic discharges can also cause disk errors. Included in the definition
of disk errors is a failure of the Licensed Internal Code that controls the disk
subsystem.

When the system detects an error, generally the occurrence is logged and the
operation is attempted again. Temporary errors are those from which the system
can recover and complete the I/O operation successfully. When the error is so
severe that the I/O operation cannot succeed, it is a permanent error.

When the system detects a permanent error, it classifies it as a failure in that

hardware subsystem. In an ASP that does not have mirrored protection, a failure
causes the system to become unusable. The system displays an error message
which contains a System Reference Code (SRC) of A6xx 0244, A6xx 0255, or A6xx
0266 where xx is incremented every minute. During this time, the system will retry
the operation in which the failure occurred. If the condition that caused the failure
can be corrected (for example, by powering on a disk unit or replacing an
electronic component), then normal system operations are resumed.

On a system with mirrored protection, errors and failures have different effects.
When a failure occurs on a system with mirrored protection, the recovery
procedure is affected by the level of protection that is configured.

System Actions for Permanent Errors

When a permanent error occurs and mirroring is active, the system attempts to
recover. The topics that follow describe the actions the system takes for different
types of permanent errors.

Device Error: If the system detects a device, I/O processor or bus failure on a
mirrored unit, it does the following:
1. The system disables the failing unit and suspends mirroring for the pair. If the
other unit in the pair is also failing or is already suspended, the first unit is
considered unprotected.
2. The system sends a message that identifies the failing unit and indicates that
mirroring has been suspended. You can use problem analysis on this message
for more information.
3. When a disk unit is suspended following an error, the system records all
updates that are performed on the active unit of the mirrored pair. If the
suspended disk unit becomes usable within a short period of time, the system
automatically synchronizes the data between the mirrored units.
4. After the failed unit is replaced, the system synchronizes the pair and resumes
mirrored protection. The system sends a message that indicates that mirrored
protection has been resumed.

© Copyright IBM Corp. 1997, 2000, 2001 289

 Read Error:
1. The system reads from the other storage unit of the mirrored pair. If the
permanent read error occurs on the other storage unit as well, the original read
request completes with a permanent read error.
2. If the read operation from the other storage unit is successful, the data is
written back to the first unit of the mirrored pair, assigning an alternate sector.
Only then does the system signal that the original read request is complete.

Connection Failure: If the system cannot communicate with a device, it does the
following:
1. The system attempts to recover from the communications error. Any job
requesting the disk unit waits during the period that the system is attempting
recovery.
2. If the recovery is successful, normal system operations continues.
3. If the system cannot recover within the time limit for the reset command, the
unit is considered to have a device error. The system performs the steps that
are described in 289.

Load Source Unit Failure: If an error occurs on the load source unit before the
Storage Management Recovery portion of the IPL, the system does the following:
1. The system determines if the other mirrored unit in the load source mirrored
pair is usable. If not, the system fails.
2. If the system is able to continue, it starts an IPL from the remaining usable unit
in the load source mirrored pair.

Suspending Mirrored Units

If you have to suspend a mirrored unit, you can do so using the Suspend Mirrored
Protection option on the Work With Disk Unit Recovery display using SST or DST.

To suspend mirrored protection, do the following:

1. Type:
STRSST
2. From the System Service Tools (SST) menu, do the following:

┌────────────────────────┐ Select option 3

│ SST Menu │ (Work with disk units)
│ │
│ ┌───────────────────┴─────┐ Select option 3
│3 │ Work with Disk Units │ (Work with disk
└────┤ │ unit recovery)
│ ┌─────────────────────┴──┐
│3 │ Work with Disk Unit │
└───┤ Recovery │
│ │
│ │
└────────────────────────┘

3. Select option 3 (Suspend mirrored protection) on the Work with Disk Unit
Recovery display and press the Enter key.

290 OS/400 Backup and Recovery V5R1

 Suspend Mirrored Protection

Type option, press Enter.

1=Suspend Mirrored Protection

Serial Resource
OPT Unit ASP Number Type Model Name Status
_ 1 1 00-31297 6109 030 DD002 Resuming
_ 3 1 00-0184097 6602 050 DD011 Active
_ 3 1 00-0125986 6602 050 DD005 Active

4. Type a 1 (Suspend Mirrored Protection) in the Option column for each unit that
you want to suspend mirrored protection. You can suspend protection only on
units that have both units in an Active or Resuming status. If one of the units is
in Resuming status, then it is the only unit that can be selected to suspend. It
takes several minutes to suspend a resuming unit that is using SST.

If you suspend a mirrored unit that is using SST, the system starts to keep a list of
disk pages that are changed. If you resume mirrored protection on the suspended
mirrored unit before this list becomes full, the system uses this list to copy data
from only those disk pages that were changed instead of copying an entire disk.

Resuming Mirrored Units

If you have to resume a mirrored unit, you can do so using the Resume Mirrored
Protection option on the Work With Disk Unit Recovery display using SST or DST.

To resume mirrored protection, do the following:

1. Type:
STRSST
2. From the System Service Tools (SST) menu, do the following:

┌────────────────────────┐ Select option 3

│ SST Menu │ (Work with disk units)
│ │
│ ┌───────────────────┴─────┐ Select option 3
│3 │ Work with Disk Units │ (Work with disk
└────┤ │ unit recovery)
│ ┌─────────────────────┴──┐
│3 │ Work with Disk Unit │
└───┤ Recovery │
│ │
│ │
└────────────────────────┘

3. Select option 4 (Resume mirrored protection) on the Work with Disk Unit
Recovery display and press the Enter key.

Resume Mirrored Protection

Type option, press Enter.
1=Resume Mirrored Protection
Serial
Resource OPT Unit ASP Number Type Model Name Status
_ 2 3 00-59681F7 6602 050 DD004 Suspended

Chapter 12. Mirrored Protection Recovery Actions 291

 4. Type a 1 (Resume Protection) in the Option column for each unit that you want
to resume mirrored protection. You can select only a unit in Suspended status to
resume.

Replacing a Mirrored Unit

A unit selected to replace the failed mirrored unit must satisfy all of the mirrored
protection configuration rules and restrictions when it is paired with the remaining
unit in the mirrored pair. (See “Mirrored Protection–Configuration Rules” on
page 719.)

You can replace mirrored units by using the Replace Disk Unit option in either
DST or SST. To do this, you need to have a spare storage unit that can be paired
with the mirrored unit of the storage unit that is being replaced. The storage unit
being replaced can have a status of active or suspended. However, one of the
storage units in the pair must be suspended. The results of the replace operation is
different for each status. Replacing a suspended storage unit causes that storage
unit to go to a status of resuming after the replace operation. Replacing an active
unit causes data in the ASP to be lost, so you must first delete the data in ASP
(using the DST ’Delete ASP Data’ option). The storage unit being replaced can also
be missing or not missing. To replace a unit with a status of resuming, you must
suspend it. If the status of unit 1 is unknown, replace operations are not allowed
until the status of the mirrored units for unit 1 is known. The unit selected to
replace another mirrored unit must satisfy all of the mirrored protection
configuration rules and restrictions when it is paired with the remaining unit in the
mirrored pair. (See “Mirrored Protection–Configuration Rules” on page 719.)

If a storage unit fails, and if the same storage unit that failed has been repaired, it
is not necessary to replace it. The failed disk should have a status of suspended
and can be resumed after the repair is complete.

If the storage unit that is being replaced is active, it can only be replaced at DST
before the IPL to the OS/400 licensed program. It should never be necessary to
replace an active unit unless both units of the mirrored pair have failed. If this
situation does occur, the service representative should first try to recover the data
from the failed units using the Save Disk Unit Data option on the Work with Disk
Unit Recovery display. When an active unit is replaced, the last good copy of the
data has been lost. The data in the ASP that contains the unit being replaced must
be deleted using the DST ’Delete ASP Data’ option before an active unit is allowed
to be replaced.

Replacing unit 1 requires special handling. If the system ASP has mirrored
protection, one of the units in the mirrored pair for unit 1 is selected as the IPL
device. It is the only unit that is used until the system performs an IPL to the
OS/400 licensed program. Until then, it cannot be replaced or even suspended.
However, its mirrored unit can be both suspended and replaced. After the IPL to
the OS/400 licensed program, the IPL device can be suspended and then replaced.

Replacing a unit may cause the level of protection for a mirrored pair to change. If
a lower level of protection results from a replacement operation, it displays a
warning screen. At certain times, especially when missing units are involved in the
replace operation, the system may not be able to accurately calculate the level of
protection and the same warning display is shown.

To replace a disk unit by using SST, do the following:

1. Type:

292 OS/400 Backup and Recovery V5R1

 STRSST
2. From the System Service Tools (SST) menu, do the following:

┌────────────────────────┐ Select option 3

│ SST Menu │ (Work with disk units)
│ │
│ ┌───────────────────┴─────┐ Select option 3
│3 │ Work with Disk Units │ (Work with disk
└────┤ │ unit recovery)
│ ┌─────────────────────┴──┐
│3 │ Work with Disk Unit │
└───┤ Recovery │
│ │
│ │
└────────────────────────┘

3. Select option 1 (Replace configured unit) on the Work with Disk Unit Recovery
display and press the Enter key.
The Select Configured Unit to Replace display is shown.

Select Configured Unit to Replace

Type option, press Enter.

1=Select
Serial
Resource OPT Unit ASP Number Type Model Name Status
_ 1 1 00-0163477 6602 030 DD019 Suspended
1 2 1 00-17900 6602 030 DD002 Suspended

4. Type a 1 in the Option column on the Select Configured Unit to Replace display
and press the Enter key.

Select Replacement Unit

Serial
Resource Unit ASP Number Type Model Name Status
2 1 00-17900 6602 030 DD002 Suspended

Type option, press Enter. 1=Select

Serial
Resource Option Number Type Model Name Status
_ 00-0330477 6602 030 DD005 Non-configured
1 00-0323200 6602 030 DD033 Non-configured

5. Type a 1 in the Option column on the Select Replacement Unit display and
press the Enter key.

Chapter 12. Mirrored Protection Recovery Actions 293

 Confirm Replace of Configured Unit

This screen allows the confirmation of the configured unit to be replaced

with the selected replacement unit. Press Enter to confirm your choices for
Replace Press F12 to return to change your choices. The configured unit
being replaced is:
Serial
Resource Unit ASP Number Type Model Name Status
2 1 00-17900 6602 030 DD002 Suspended

The replacement unit will be:

Serial
Resource Unit ASP Number Type Model Name Status
2 1 00-0323200 6602 030 DD033 Resuming

6. Press Enter to Confirm.

7. The replacement function runs for several minutes. Wait until the replacement
function completes.

Using Spare Nonconfigured Units for Replacement

If mirrored units become suspended as a result of a hardware failure, the system
continues to run. However, one or more storage units will be suspended and
therefore unprotected until your service representative can repair or replace the
failed hardware. If you have spare nonconfigured units, you may be able to
resume mirrored protection before the repair actions are done.

Call your service representative. You may be directed to examine the Service
Action Log for information that is concerning the failure. Use the Display Disk
Configuration Status option by using SST or the Work with Disk Status
(WRKDSKSTS) command to determine which units are suspended. If all disk units
under an I/O processor are suspended, the I/O processor probably has failed. If
you have enough spare units of the right type and model, and if the spare units
are not on the I/O processor that has failed, you may be able to use the spare
nonconfigured units to resume mirrored protection.

After your service representative repairs a failed storage unit, you may want to use
it instead of the spare to restore the previous level of protection. To use the
repaired unit, do the following:
1. Suspend the active storage unit that was previously used as the spare by
typing the following on a command line and pressing the Enter key.
STRSST
2. From the System Service Tools (SST) menu, do the following:

┌────────────────────────┐ Select option 3

│ SST Menu │ (Work with disk units)
│ │
│ ┌───────────────────┴─────┐ Select option 3
│3 │ Work with Disk Units │ (Work with disk
└────┤ │ unit recovery)
│ ┌─────────────────────┴──┐
│3 │ Work with Disk Unit │
└───┤ Recovery │
│ │
│ │
└────────────────────────┘

3. Select the option 3 (Suspend mirrored protection).

294 OS/400 Backup and Recovery V5R1

 Suspend Mirrored Protection

Type option, press Enter.

1=Suspend Mirrored Protection
Serial
Resource OPT Unit ASP Number Type Model Name Status
_ 1 1 00-0193825 6602 030 DD001 Active
_ 1 1 00-0184097 6602 030 DD019 Active
_ 2 1 00-0125986 6602 030 DD036 Active
_ 2 1 00-0125986 6602 030 DD002 Active

4. Type a 1 (Suspend Mirrored Protection) in the Option column. The original

spare unit is the same disk type and model as the repaired disk unit.
5. Return to the Work with Disk Unit Recovery display by pressing F12 (Cancel)

Work with Disk Unit Recovery

Select one of the following:

1. Replace configured unit
2. Disk unit problem recovery procedures
3. Suspend mirrored protection
4. Resume mirrored protection
5. Delete disk unit data
6. Rebuild disk unit data

6. Select option 1 (Replace configured unit).

Select Configured Unit to Replace

Type option, press Enter.
1=Select
Serial
Resource OPT Unit ASP Number Type Model Name Status
_ 1 1 00-0163477 6602 030 DD019 Suspended
1 2 1 00-17900 6602 030 DD002 Suspended

7. Type a 1 in the Option column on the Select Configured Unit to Replace

display and press the Enter key.

Select Replacement Unit

Serial
Resource Unit ASP Number Type Model Name Status
2 1 00-17900 6602 030 DD002 Suspended

Type option, press Enter. 1=Select

Serial
Resource Option Number Type Model Name Status
_ 00-0330477 6602 030 DD005 Non-configured
1 00-0323200 6602 030 DD033 Non-configured

8. Type a 1 in the Option column on the Select Replacement Unit display and
press the Enter key.

Chapter 12. Mirrored Protection Recovery Actions 295

 Confirm Replace of Configured Unit
This screen allows the confirmation of the configured unit to be replaced
with the selected replacement unit. Press Enter to confirm your choices for
Replace Press F12 to return to change your choices. The configured unit
being replaced is:
Serial
Resource Unit ASP Number Type Model Name Status
2 1 00-17900 6602 030 DD002 Suspended

The replacement unit will be:

Serial
Resource Unit ASP Number Type Model Name Status
2 1 00-0323200 6602 030 DD033 Resuming

9. Press Enter to confirm.

10. The replacement function runs for several minutes. Wait until the replacement
function completes.

Mirrored Protection Recovery Actions Performed by the

Service Representative
The procedures that are described here are overviews of the steps and
considerations involved in disk unit repair in the mirrored environment. Although
your service representative performs these steps, they are included here for your
information.

Actions When Concurrent Maintenance is Possible

1. Perform problem analysis on the failing storage unit.
Performing problem analysis may cause mirrored protection to be suspended
on the failing storage unit, and in some cases, on additional storage units.
2. Power down the failing storage unit.
3. Repair or replace the failing storage unit.
4. If the Replace Configured Unit option is necessary, the new storage unit is
formatted and initialized, and mirrored protection is automatically resumed.
5. Resume mirrored protection on the repaired unit, if necessary, and on any other
units that were suspended as part of the repair action.
Synchronization begins on the resuming storage units immediately, and a
message is sent to the QSYSOPR message queue when synchronization
completes.

Actions When Concurrent Maintenance is Not Possible

1. Power down the system.
2. If unit 1 has failed, see “Mirrored Protection–Configuration Rules” on page 719
for restrictions that apply.
3. Perform an attended IPL to DST.
4. Perform problem analysis on the failing storage unit.
Performing problem analysis may cause mirrored protection to be suspended
on the failing storage unit, and in some cases, on additional storage units.
5. Power down the failing storage unit.
6. Repair or replace the failing storage unit.
7. If the Replace Configured Unit option is necessary, the new storage unit is
formatted and initialized, and mirrored protection automatically resumes.
8. Resume mirrored protection on the repaired unit, if necessary, and on any other
storage units that were suspended as part of the repair action.

296 OS/400 Backup and Recovery V5R1

 9. Continue the IPL to command entry. Synchronize the resuming storage units
during the IPL.

Other Recovery Considerations for Mirrored Protection

Message Handling: When a system with mirrored protection experiences a disk
failure, the only external indication of the failure is a message that is sent to the
system operator message queue (QSYSOPR). If there is a message queue that is
named QSYSMSG in the QSYS library, the messages are sent there also.

When suspended units exist, the system sends a message every hour to the
QSYSOPR message queue as a reminder.

You should have a method of bringing these messages to the attention of the
system administrator. If the interactive job at the console allocates the QSYSMSG
message queue and places it in break mode, it notifies you of any problems. For
more information on QSYSMSG, see the CL Programmer’s Guide.

Synchronization: When the system is synchronizing (resuming) a disk unit, the

system response time becomes longer.

When mirrored protection is resumed on a suspended disk unit at DST, the

synchronization to the OS/400 licensed program is done during the IPL.

Mirrored Protection Disk-Error Handling

Mirrored protection handles disk errors as follows:

Unrecoverable device error:

1. The system suspends the failing storage unit and mirrored protection is
suspended for the mirrored pair.
2. The system continues operations by using the operating storage unit of the
mirrored pair.
3. A message sent to the QSYSOPR message queue identifies the failing storage
unit. It informs you that mirrored protection is suspended for the mirrored pair.

Permanent read error:

1. The system reads from the other storage unit of the mirrored pair. If the
permanent read error occurs on the other storage unit as well, the original read
request completes with a permanent read error.
2. If the read operation from the other storage unit is successful, the data is
written back to the first unit of the mirrored pair, assigning an alternate sector.
Only then does the system signal that the original read request is complete.

Not operational storage unit:

1. The system attempts recovery. If it is successful, normal system operations
continue with mirrored protection and without suspending or synchronizing
the unit.
2. If recovery is unsuccessful, the unit is considered to have an unrecoverable
device error, which is processed as described previously.

Time-out:

Chapter 12. Mirrored Protection Recovery Actions 297

 1. The system attempts to recover from the timeout. If it is successful, normal
system operations continue with mirrored protection and without suspending
or synchronizing this unit.
2. If recovery is unsuccessful, the storage unit is considered to have an
unrecoverable device error, which is processed as described previously.

I/O processor or bus failure:

1. The system suspends each disk unit that is attached to the failing I/O processor
or bus in the same way it is done for an unrecoverable error.
2. The system saves a copy of the failing I/O processor’s storage so the problem
can be diagnosed. The system continues without the failing I/O processor.

Disk-related failure of unit 1 before the IPL to the Operating System/400: See
“Mirrored Protection–Configuration Rules” on page 719 for restrictions that apply.

Missing Disk Units

If a disk unit, a controller, or an I/O processor fails during an IPL, the system
detects the failure and does one of the following:
v Displays an SRC on the control panel if the keylock switch is not in the Manual
position.
v Displays the Missing Disk Unit display on the console if the keylock switch is in
the Manual position.

If the failing unit has mirrored protection and its mirrored unit is active, the
following display is shown.
Disk Configuration Warning Report

Type option, press Enter.

5=Display Detailed Report

Press F10 to accept all the warnings and continue the IPL.
The system will attempt to correct the warnings.

OPT Warning
5 Missing mirror protected units in the configuration

Type a 5 in the option column and press the Enter key.

Suspend Missing Disk Units

The following disk units are missing from the disk configuration:
Serial
Resource Reference ASP Unit Type Model Number Name Code
1 2 6602 030 00-0190494 DD036 1713

You can suspend mirrored protection on the affected units and continue the IPL.
An entry is written in the problem log. You can run problem analysis on the failing
unit at a later time. The type and reference code fields can be used with the unit
reference code guide to determine the cause of the problem. If the keylock switch
is not in the Manual position, a system reference code is displayed on the control
panel. If the affected units do not report to the system within six minutes, the
system automatically suspends mirrored protection on the affected units and
continues the IPL.

298 OS/400 Backup and Recovery V5R1

 If the suspended disk units become ready before the system is powered down, the
system will automatically resume mirrored protection on these units.

Saving a Unit
The system allows you to save data from storage units that use the DST Save Disk
Unit Data option.

The following rules apply to saving units on a system with mirrored protection:
v Only configured units can be saved.
v The save operation is not allowed when both mirrored units of a mirrored pair
are active. Only one of the mirrored units can be saved. Therefore, one mirrored
unit must be suspended.
v Only the active unit of a mirrored pair can be saved because the active unit
contains the current data.
v If multiple failures cause the state of unit 1 to be unknown, saving of any
storage unit is not allowed.

Restoring a Unit
In the mirrored environment, the system allows you to restore data to storage
units.

The following rules apply to restoring units on a system with mirrored protection:
v The restore is possible only for an active device.
v This option can restore to either a configured or nonconfigured disk unit.
v The restore operation requires that the unit restored to is as large as or larger
than the unit saved.
v The restore operation is not permitted if the state of a unit is unknown. You can
restore unit 1 only to the IPL device.
v After a unit is restored, the system performs an IPL to DST.
v The unit being restored must satisfy all mirrored protection configuration rules
and restrictions.

Active Mirrored Load Source Failure

If unit 1 is mirrored, the system attempts to IPL from a load source mirrored unit
that contains Licensed Internal Code and system data. The mirrored unit state of
that storage unit will be active.

System Cannot Find Active Mirrored Load Source for IPL

If the system cannot find a load source unit that contains current data and can
only find a load source unit that is suspended or resuming, the system will IPL on
the suspended/resuming unit. That unit contains back-level data. The system
cannot be used until it finds or repairs the active mirrored load source.

If the system has not been able to IPL on an active mirrored load source, it is
presumed to be broken in some way, and the following screens are displayed.

Chapter 12. Mirrored Protection Recovery Actions 299

 Disk Configuration Error Report

Type option, press Enter.

5=Display Detailed Report

OPT Error
5 Load source failure

Type a 5 in the option column and press the Enter key.

Display Load Source Failure

The system could not use the load source disk unit that
contains correct data.

The following disk unit contains the correct data:

Disk unit:
Type . . . . . . . . . . . . . . . . . . : 6603
Model . . . . . . . . . . . . . . . . . : 030
Serial number . . . . . . . . . . . . . : 00-0193825
Resource name . . . . . . . . . . . . . : DD001

Press Enter to use Dedicated Service Tools (DST).

Active Mirrored Load Source Being Used for IPL Fails

If the system is IPLing on an active mirrored load source and that storage unit fails
during the IPL to DST or at DST, the system attempts to perform a directed IPL to
the other storage unit (tries to re-IPL on the remaining load source).
v If the directed IPL fails, the system ends abnormally and displays a system
reference code.
v If the remaining storage unit of the load source mirrored pair is active and the
original load source is still broken on the re-IPL, the broken load source is
treated like any other missing mirrored unit, and the following are displayed:
Disk Configuration Warning Report
Type option, press Enter.
5=Display Detailed Report

Press F10 to accept all the warnings and continue the IPL.
The system will attempt to correct the warnings.

OPT Warning
5 Missing mirror protected units in the configuration

Type a 5 in the option column and press the Enter key.

Suspend Missing Disk Units
The following disk units are missing from the disk configuration:
Serial
Resource Reference ASP Unit Type Model Number Name Code
1 2 6602 030 00-0190494 DD036 1713

v If the remaining storage unit of the load source mirrored pair does not contain
current data (it is suspended or resuming), it is treated as if the system cannot
find an active mirrored load source for IPL, as described earlier. The IPL is not
allowed to continue past DST until the active load source is found or repaired.

300 OS/400 Backup and Recovery V5R1

 Active Mirrored Load Source Fails Late in the IPL or at Runtime
When an active mirrored load source fails after Storage Management Recovery is
completed, it is treated like a failure in any other mirrored pair:
v If the other storage unit in the mirrored pair is present and active, the failing
unit is suspended and the system continues to run using data the data on the
remaining active unit of the pair.
v If the failing storage unit is the last active unit of the mirrored pair (the other
unit of the pair is either suspended or resuming), the system displays a DASD
Attention system reference code and becomes unusable.

Cannot Read System Configuration Data from Active Mirrored

Load Source
If the system cannot read the system configuration data from the active mirrored
load source that is being used for the IPL, one of the following screens is
displayed.
Accept Load Source Warning Report

Some of the configuration information on the load source is

missing. The system can rebuild this information using
the default values.

Press Enter to let the system rebuild the configuration

information on the load source.

If you were performing any disk unit recovery actions, go to

Work with Disk Units and complete those actions.

Disk Configuration Warning Report

Type option, press Enter.
5=Display Detailed Report

Press F10 to accept all the warnings and continue the IPL.
The system will attempt to correct the warnings.

OPT Warning
5 Bad load source configuration

Unknown Unit 1 Status

If both the service processor and one unit of the mirrored pair for unit 1 have
failed, the following display is shown.
Disk Configuration Error Report

Type option, press Enter.

5=Display Detailed Report

OPT Error
5 Unknown load source status

Type a 5 in the option column and press the Enter key.

Chapter 12. Mirrored Protection Recovery Actions 301

 Display Unknown Mirrored Load Source Status

The system can not determine which disk unit of the load
source mirrored pair contains the correct level of data.

The following disk unit is not available:

Disk unit:
Type . . . . . . . . . . . . . . . . . . : 6603
Model . . . . . . . . . . . . . . . . . : 030
Serial number . . . . . . . . . . . . . : 00-0193825
Resource name . . . . . . . . . . . . . : DD001

Press Enter to continue.

If the keylock switch is not in the Manual position, the control panel displays a
system reference code .

The missing unit must be repaired or the state of the unknown load source
recovered. If the missing unit can be repaired without losing the data on the unit,
then the state of the load source will become known when the system is IPLed. If
the missing unit cannot be repaired or if the data on it is lost, then you can
possibly recover the state of the unknown load source and avoid restoring the
entire system.

You should only attempt to recover the state of the unknown load source when
you know its mirrored unit state was active before the failures that caused the state
to become unknown. Since the state is unknown, the system cannot verify that
your choice is correct. If you recover the state of the unknown load source when
the actual state of the disk unit used to IPL was not active, you will cause data loss
or damaged objects on your system.

To Recover the State of the Unknown Load Source

1. From the DST main menu, select option 4, Work with disk units.
2. From the Work with disk units menu, select option 2, Work with disk unit
recovery.
3. From the Work with disk unit recovery menu, selection option 15, Recover
unknown load source.
This will display a confirmation screen, showing the disk configuration and
mirrored unit states the system will have after the recovery.
4. If the configuration and states are as you expect, press Enter to confirm.
The state of the load source mirrored pair is changed so that the load source
just used to IPL is active and the other (missing) load source is suspended.

If you cannot recover the state of the unknown load source and if the missing unit
cannot be repaired, you must install the Licensed Internal Code and restore the
entire system.

Display Incorrect Licensed Internal Code Install

When the Licensed Internal Code is restored on a mirrored unit for unit 1, one of
the mirrored units may have the incorrect level of data stored on it. If this
condition occurs, and the disk unit that contains the correct data is not available,
the Licensed Internal Code is restored to the disk unit with the incorrect data.
When a disk preforms an IPL and the correct disk unit is available, the following
display is shown. If the keylock switch is not in the Manual position, system
reference code (SRC) is shown on the control panel.

302 OS/400 Backup and Recovery V5R1

 Display Incorrect Licensed Internal Code Install

Licensed Internal Code has been installed on the incorrect

disk unit of the load source mirrored pair.

If you continue the IPL, the previously installed Licensed Internal

Code installed on the incorrect disk unit of the mirrored
load source pair will be deleted. The Licensed Internal Code
will be replaced by the Licensed Internal Code from the correct
disk unit.

The following disk unit is the correct disk unit.

Disk unit:
Type . . . . . . . . . . . . . . . . . : 6602
Model . . . . . . . . . . . . . . . . : 030
Serial number. . . . . . . . . . . . . . : 00-0163477_
Resource name . . . . . . . . . . . . . : DD019

Press Enter to continue.

Chapter 12. Mirrored Protection Recovery Actions 303

304 OS/400 Backup and Recovery V5R1
Chapter 13. How to Restore Your System Using Operational
Assistant Tapes
The topic “Recovering User Information Using Tapes from Operational Assistant
Backup–Checklist 27” on page 117 provides a list of the steps necessary to recover
user information on your system. This chapter describes specific tasks associated
with restoring information from Operational Assistant backup tapes. The
descriptions assume that you are recovering all the data on your system. If you are
recovering a single library or a single ASP, adapt the procedures to your situation.

Figure 29 on page 306 shows the parts of your system and how they are saved with
Operational Assistant. Refer to it in the topics that follow.

© Copyright IBM Corp. 1997, 2000, 2001 305

 Options from
RUNBCKUP menu
┌─────S┌───────────────────────────────┐
│ │ Licensed Internal Code │ Saved using
│ ├───────────────────────────────┤ RUNBCKUP command
│ │ OS/400 Objects in QSYS │ or RUNBCKUP menu,
│ └───────────────────────────────┘ option 1, 2, or 3
│
│ ┌───────────────────────────────┐W─────────────┐
│ │ User Profiles │ │
│ ├───────────────────────────────┤ │
│ │ Private Authorities │ │
│ └───────────────────────────────┘ Can be │
│ included │
11 ┌───────────────────────────────┐ on │
│ │ Configuration Objects │ *DAILY, │
│ └───────────────────────────────┘ *WEEKLY, │
│ or │
│ ┌──S┌───────────────────────────────┐ *MONTHLY │
│ │ │ OS/400 Optional Libraries │ backup │
│ 10 │ QHLPSYS QUSRTOOL │ options │
│ │ ├───────────────────────────────┤ │
│ │ │ Licensed Program Libraries │ │
│ │ │ QRPG QCBL Qxxxxx │ │
│ └──S└───────────────────────────────┘ │
│ │
│ ┌───────────────────────────────┐ │
│ │ IBM Libraries with User Data │ │
│ │ QGPL QUSRSYS QS36F #LIBRARY │ │
│ ├───────────────────────────────┤ │
│ │ User Libraries │ │
│ │ LIBA LIBB LIBC LIBxxx │ │
│ └───────────────────────────────┘ │
│ │
│ ┌───────────────────────────────┐ │
│ │ Documents and Folders │ │
│ ├───────────────────────────────┤ │
│ │ Distribution Objects │ │
│ └───────────────────────────────┘ │
│ │
│ ┌───────────────────────────────┐ │
│ │ User Directories │ │
│ │ DIRA DIRB DIRC DIRxxx │ │
│ ├───────────────────────────────┤W─────────────┘
│ │ IBM Directories │
└─────S└───────────────────────────────┘

Figure 29. How the System Is Saved with Operational Assistant Backup

How to Restore Your Libraries

To recover your entire system, you must restore IBM-supplied libraries and user
libraries. To restore IBM-supplied libraries, do the following:
1. Find the tapes that you used most recently to save IBM-supplied libraries. You
saved them by using one of these methods:
v Option 10 from the Run Backup menu.
v Option 11 from the Run Backup menu.
v The SAVLIB LIB(*IBM) command.
v The SAVLIB LIB(*NONSYS) command.
v Option 21 from the Save menu.

306 OS/400 Backup and Recovery V5R1

 v Option 22 from the Save menu.
v Option 41 from the Save menu.
2. Mount the first tape and type: RSTLIB SAVLIB(*IBM) DEV(media-device-name).
Press the Enter key.

To restore user libraries, do the following:

1. Find the tapes that you used most recently to save all user libraries. You saved
them by using one of these methods:
v Option 1, 2, or 3 from the Run Backup menu and specifying 2 (All) for the
User libraries option.
v Option 11 from the Run Backup menu.
v The SAVLIB LIB(*ALLUSR) command.
v The SAVLIB LIB(*NONSYS) command.
v Option 21 from the Save menu.
v Option 23 from the Save menu.
v Option 40 from the Save menu.
v Option 42 from the Save menu.

If you are not sure which tapes have your user libraries on them, do the
following for each possible tape:
a. Mount the tape.
b. Type DSPTAP DEV(media-device-name)
c. Page through the displays, looking for the file called QFILE.
d. When you find the tape with the QFILE file on it, write down the sequence
number for that file on the tape.
e. Leave the tape in the tape unit and type: DSPTAP DEV(media-device-name)
LABEL(QFILE) SEQNBR(sequence-number) DATA(*SAVRST) OUTPUT(*PRINT).
f. If the listing contains user libraries, it was created by either the
SAVLIB(*NONSYS) command or the SAVLIB(*ALLUSR) command. The
libraries from the tape can be restored using the RSTLIB SAVLIB(*ALLUSR)
command.
2. Mount the first tape that has user libraries and type: RSTLIB SAVLIB(*ALLUSR)
DEV(media-device-name). Press the Enter key.

You have now restored all the libraries on your system to the point where they
were all saved completely. Return to “Recovering User Information Using Tapes
from Operational Assistant Backup–Checklist 27” on page 117.

How to Restore Libraries That You Saved by Using a Backup List

This topic describes how to restore libraries that you saved using a backup list,
either for daily or weekly backup. It assumes that you save all libraries for your
monthly backup. This topic describes how to restore libraries, not changed objects.
Use this procedure if all of the following are true:
v You have an Operational Assistant backup that is more recent than the last time
you saved the entire system or all libraries.
v You specified 1 (Selected from list) for the User libraries option for your
Operational Assistant backup.
v You specified N (No) for the Save changed objects only option for your Operational
Assistant backup.

Chapter 13. How to Restore Your System Using Operational Assistant Tapes 307
 If you have both a weekly and a daily backup that meet these conditions, do the
following:
v If your daily backup and weekly backup both save exactly the same libraries
from the backup list, perform steps 2 through 4 once, using your most recent set
of tapes (daily or weekly).
v If your daily backup saves fewer libraries than your weekly backup, do the
following:
– If your most recent backup is a weekly backup, perform steps 2 through 4
once, using your most recent set of weekly tapes.
– If your most recent backup is a daily backup, perform steps 2 through 4 once,
using your most recent set of weekly tapes. Repeat step 2 through 4, using
your most recent set of daily tapes.
1. Mount the first tape.
2. Find the printed copy of the backup list associated with the save tapes. If you
have the list, skip to step 4
3. If you do not have the list, display the contents of the save tapes by typing:
DSPTAP DEV(media-device-name) OUTPUT(*PRINT) DATA(*SAVRST).
4. Use the listing from step 2 or step 3. For each library that was saved, do the
following:
a. Type: RSTLIB SAVLIB(library-name) DEV(media-device-name).
b. Check off the library name on the list.

Note: Restore the user libraries to each user ASP that you are recovering. If
you are restoring the QGPL library and the QUSRSYS library and doing
partial recovery, restore these libraries before any other libraries. When
recovering the entire system, there is no need to restore QGPL and
QUSRSYS libraries first.

How to Restore Changed Objects That You Saved by Using

Operational Assistant
If you save only changed objects for either your weekly or daily backup, use this
procedure. If you save changed objects both weekly and daily, use your most
recent set of tapes. If you save complete libraries on your weekly backup and
changes on your daily backup, perform this procedure only if your daily backup is
more recent than your weekly backup.

Do the following:
1. Mount the first tape from your most recent backup of changed objects.
2. Determine if any objects are on the tape for libraries that do not exist on the
system:
a. Print a list of libraries on the system by typing: DSPBCKUPL OUTPUT(*PRINT).
b. Print the contents of the tape by typing: DSPTAP DEV(media-device-name)
OUTPUT(*PRINT) DATA(*SAVRST).
c. Compare the two lists. Mark any libraries on the DSPTAP listing (from step
2b) that do not appear on the DSPBCKUPL listing (from step 2a).
d. For any libraries you marked in step 2c, type the following: CRTLIB
LIB(library-name).
3. Restore the changed objects from the tapes. For each library that appears on the
DSPTAP listing (from step 2b), type:

308 OS/400 Backup and Recovery V5R1

RSTOBJ OBJ(*ALL) SAVLIB(library-name)
OBJTYPE(*ALL) DEV(media-device-name)

Chapter 13. How to Restore Your System Using Operational Assistant Tapes 309
310 OS/400 Backup and Recovery V5R1
Chapter 14. How to Restore the System from the Save
Storage Media
When you recover your system from the Save Storage (SAVSTG) media, you reset
your system to the point when the SAVSTG procedure was run. Your system will
not be available for use until the restore process completes successfully.

The disk configuration of the restoring system must be the same as the disk
configuration of the saving system. There must be at least as many disk units on
the restoring system as there were on the saving system. Each disk unit capacity
on the restoring system must be equal to or greater than the capacity of the disk
unit on the saving system. Serial numbers and physical addresses do not have to
be the same. All disk units that were saved are required for the restore operation.

If your system has mirrored protection now, when the restore storage procedure
runs, your system will not have mirrored protection on any Auxiliary Storage Pool
(ASP).

Find These Things Before You Begin:

v The list of all the Licensed Internal Code fixes applied to your system at the
time you saved storage. This list should be attached to your backup log or
found with the SAVSTG tapes.
v If you applied any PTFs since the last save storage operation, you will need
your most recent cumulative PTF tape.
v A recent SAVSYS or SAVCFG media volume. The SAVSYS or SAVCFG media
contains configuration information that you will have to restore after the restore
storage is completed.

Do These Things Before You Begin:

v Clean the read and write head of the tape unit.
v Print a list of all the Licensed Internal Code fixes currently on the system. Type
the following and press the Enter key:
DSPPTF LICPGM(*ALL) OUTPUT(*PRINT)

Task 1–Powering Down the System and Loading the Licensed Internal
Code
1. Ensure that all users are off the system.
2. Type the following to power down the system:
PWRDWNSYS OPTION(*IMMED)

Attention logical partitioning users!

If you are going to use this command on the primary partition, be sure to
power off all secondary partitions before running the command.

3. Load the first SAVSTG tape in the tape unit that is your alternative IPL device.
4. Install the Licensed Internal Code by using the procedure that is described in
“Task 2–Powering Down the System” on page 123 through “How to Load the
Licensed Internal Code” on page 130. Select option 2 (Install Licensed Internal

© Copyright IBM Corp. 1997, 2000, 2001 311

 Code and Initialize System) from the Install Licensed Internal Code (LIC)
display. When the procedure asks for SAVSYS media, use your SAVSTG tapes
instead.

Task 2–Restoring the Save Storage Tapes

1. When the IPL following the installation of the Licensed Internal Code
completes, the Disk Configuration Attention Report display is shown. Press
F10 to accept the new configuration. The IPL or Install the System menu is
displayed.

IPL or Install the System

Select one of the following:
1. Perform an IPL
2. Install the operating system
3. Use Dedicated Service Tools (DST)
4. Perform automatic installation of the operating system
5. Save Licensed Internal Code

2. Select option 3 (Use Dedicated Service Tools (DST)) and press the Enter key.
The Dedicated Service Tools (DST) Sign On display is shown.

Dedicated Service Tools (DST) Sign On

Type choice, press Enter
DST user . . . . . . . . . . ______
DST password . . . . . . . . ______

3. Sign on DST with the DST security-level or full-level password. The iSeries
Security Reference book has more information about DST passwords.
The Use Dedicated Service Tools (DST) menu is shown.

Use Dedicated Service Tools

Select one of the following:

1. Perform an IPL
2. Install the operating system
3. Work with licensed internal code
4. Work with disk units
5. Work with DST environment
6. Select DST console mode
7. Start a service tool
8. Perform automatic installation of the operating system
9. Work with save storage and restore storage
10. Work with remote DST support

Note: If you are able to use logical partitions on your system, the Use
Dedicated Service Tools screen will include option 11, Work with
system partitions.
4. If you are using logical partitioning, and you are restoring to the primary
partition, you must restore your partition configuration before you restore
storage. For secondary partitions, you will not restore the partition
configuration — this step is only for primary partitions. Refer to “How to
Recover Your Logical Partition Configuration” on page 134 for instructions on
restoring your partition configuration. Then return here and continue with the
next step.

312 OS/400 Backup and Recovery V5R1

 5. Select option 9 (Work with save storage and restore storage) and press the
Enter key.
6. Select option 1 (Restore storage) and press the Enter key. The Specify Volume
Identifier display is shown:

Specify Volume Identifier

Type choice, press Enter.

Volume identifier . . . . . . . . . . ______

7. Type the volume name in the Volume Identifier prompt. The volume name is
SAVEDS. This is the volume that is currently loaded. You will be shown one
of the following displays. Continue with the step that is indicated:

Display Name Continue With This Step

Select Tape Unit Step 8

Device Intervention Required Step 9
Confirm Restore Storage

8. If the Select Tape Unit display appears, select the proper unit and press the
Enter key.
Continue with step 12 on page 314.
Select Tape Unit

Type option, press Enter.

1=Select

Serial
Option Type Model Number Resource Name
_ ____ ____ __________ _____________
_. ____ ____ __________ _____________
..

9. If the wrong volume is loaded, the following display appears:

Device Intervention Required

Device type. . . . . . . . . . . . . . . . . . . : _____

Device
.. model . . . . . . . . . . . . . . . . . . : _____
.
If the wrong volume was loaded, type change, press Enter.

Type choice, press enter

New
.. volume or file . . . . . . . . . . . . . . . _____________________
.

Wrong volume loaded

10. Type the name of the correct volume or file, and press the Enter key. The
following display is shown:

Chapter 14. How to Restore the System from the Save Storage Media 313
 Device Intervention Required

Device type. . . . . . . . . . . . . . . . . . . : _____

Device
.. model . . . . . . . . . . . . . . . . . . : _____
.
Type choice, press enter
Action . . . . . . . . . . . . . . . . . . . . . 1=Cancel
__________________________
3=Retry
__________________________

11. Select option 3 (Retry), and press the Enter key.

12. There is a delay while the tape is read to determine what has been saved on
the tape. The Confirm Restore Storage display is shown.

Confirm Restore Storage

Warning: A restore of storage will destroy the current data

on the system. The restore will take several minutes for each
unit saved. An automatic IPL is part of the restore.

Press F10 to confirm your choice to restore all storage.

Press F12 to return to change your choice.

------ Restore To ------------------ ------ Saved From ---------

Serial Resource Serial Resource
Unit ASP Type Model Number Name Number Address
1 1 6602 030 00-0261624 DD003 00-0261624 DD003
3. 5 6602 030 00-0211957 DD002 00-0211957 DD002
..

13. Press F10 (Confirm restore) to confirm. The restore status display on the
console continually displays the progress of the restore operation.

Function Status

You selected to restore storage.

51% Complete
12 pages not readable

The display indicates what percent of the total system sectors have been
restored. However, this is an estimate and cannot be used to predict how long
the entire restore procedure will take.
14. If no errors occur, the system performs a programmed IPL when the restore
storage process completes, go to “Task 4–Completing the Restore Storage
Operation” on page 315 otherwise, continue to “Task 3–Responding to
Messages”.

Task 3–Responding to Messages

While you are performing the restore storage operation, you may see the Device
Intervention Required display or the Handle Tape Intervention display.

314 OS/400 Backup and Recovery V5R1

 When one of these displays appears, look for messages at the bottom of the
display or for an I/O manager code on the display. Respond to the display by
using the information in Table 56:
Table 56. Handling Messages When Restoring Storage
Message or Code Your Action

End of tape encountered. Load next volume. Load the next tape volume. Select option 3 (Continue), and
press the Enter key.
Tape unit not ready Make the tape unit ready, select option 3 (Continue), and press
the Enter key.
Wrong volume loaded Remove the tape. Load the correct tape. Selection option 3
(Retry) and press the Enter key.

If the tape can not be read because of a media error, the following display is
shown:
Restore Storage

Status of restore . . . . . . . . . . . . . . . : Ended

A media error was found on tape.

If this is the first time the restore storage has ended because
a media error occurred on this tape, do the following:
1. Remove the tape from the tape device.
2. Clean the tape path using the cleaning procedure
described in the tape device operator's guide.
3. Press Enter, F3, or F12 to continue. The system will
perform an IPL, an then display either the IPL or Install
the System menu or the Missing Disk Units display.
4. Select the option to use Dedicated Service Tools (DST)
5. Select the option to Work with Save Storage and Restore
Storage.
6. Select the option Resume restore storage.
7. Insert the tape which had the media error into the tape
device.
8. Make the tape device ready, if necessary.

Media Error While Restoring?

For information on how to recover, see “How to Resume the Restore Storage
Operation” on page 318.

Task 4–Completing the Restore Storage Operation

1. When the IPL completes after the restore storage operation, the IPL or Install
the System menu appears.

IPL or Install the System

Select one of the following:
1. Perform an IPL
2. Install the operating system
3. Use Dedicated Service Tools (DST)
4. Perform automatic installation of the operating system
5. Save Licensed Internal Code

2. Perform an attended IPL by selecting the Perform an IPL option.

Chapter 14. How to Restore the System from the Save Storage Media 315
 3. If the following display is shown, disk units have been attached to the system
and are in nonconfigured status.
Select option 3 (Add all disk units to the system auxiliary storage pool) and
Add All Disk Units to the System
Select one of the following:
1. Keep the current disk configuration
2. Perform disk configuration using DST
3. Add all units to the system auxiliary storage pool (ASP)
4. Add all units to the system ASP and balance data

press the Enter key.

As the disk units are being configured, the following display is shown:

..
.
Function status

You selected to add units

10% complete
..
.

Adding disk units takes several minutes. The time it takes depends on the
size of each unit and the number of units to be added.
4. The Sign On display appears. Sign on as QSECOFR.

Note: It is important that the following steps are performed so that the device
resource names are updated correctly.
5. At the IPL Options display, set the Start system to restricted state option
to Y (yes).

Note: As the IPL continues, SRC A900-2000 may appear. See “Recovering
from SRC A900 2000” on page 164. This section describes how to create
a tape device descriptor so that the system hardware configuration can
be restored in a later step of this procedure.
6. When the IPL is complete, ensure that the system is in a restricted state. See
“Putting Your System in a Restricted State” on page 45.
7. You have to restore the configuration of your system. Use the most recent
media volume that has your saved configuration. If you performed the
Restore Storage on the same system that you did the Save Storage (SAVSTG)
on, you were instructed to create a SAVCFG media volume before the
SAVSTG ran. If your system configuration has changed since the Save Storage
was performed, use the most recent SAVCFG or SAVSYS media volume. If
you performed the Restore Storage on a system different than the system the
Save Storage (SAVSTG) ran on, use the most recent SAVCFG or SAVSYS
media volume from the system you restored to. The file on the tape is called
QFILEIOC.
| Before you perform the RSTCFG command, you need to vary off all
| unnecessary configuration objects. Do not vary off the workstation and media
| drive that you are using to perform the resotre operation.
With the SAVSYS or SAVCFG media volume loaded, type:
RSTCFG OBJ(*ALL) DEV(media-device-name) OBJTYPE(*ALL)

316 OS/400 Backup and Recovery V5R1

 8. If you want the system to automatically configure new devices on subsequent
IPLs, change the system value for QIPLTYPE to allow an unattended IPL.
Type:
CHGSYSVAL QIPLTYPE '0'
9. It may be necessary to update the Network Attributes on the system. Get the
latest list of the system’s Network Attributes. The instructions for the Save
Storage (SAVSTG) command directed you to print the list of the Network
Attributes and keep that list with the Save Storage tapes. To update the
Network Attributes on the system, type the following and press the Enter key:
CHGNETA

Use the list of Network Attributes to enter the values on the input fields.
10. Change the system value for QAUTOCFG to allow automatic configuration to
run. Type:
CHGSYSVAL QAUTOCFG '1'
11. Do a PWRDWNSYS *IMMED RESTART(*YES).

Attention logical partitioning users!

If you are going to use this command on the primary partition, be sure
to power off all secondary partitions before running the command.

If you have a problem with your devices, such as not being able to vary on a
device, see “Recovering Devices That Will Not Vary On” on page 228.

When you restore your information to a different system or an upgraded

system, you may have a different console type on the target system. See
“Recovering When You Change the Console Type” on page 230.
12. While the system is IPLing, you may see an error message about the
System/36 environment, such as CPF3761. Refer to “Recovering the
System/36 Environment Configuration” on page 230 for the procedure to
follow after you have completed restoring storage.
13. When the IPL completes, the restore history information for data area
QSAVSTG in library QSYS is updated to show the date and time of the last
restore storage operation. Use the Display Object Description (DSPOBJD) to
display the last date and time of the restore storage operation.
14. Use the Display Log (DSPLOG) command to display the QHST log or use the
Display Messages (DSPMSG) command to display the QSYSOPR messages.
Look at the restore storage message CPC3735 to determine if:
v The system found any sectors that had data that could not be restored. The
data may have been unreadable during the Save Storage operation.
v The restore storage process is complete.
15. If you have additional information to restore, such as SAVCHGOBJ tapes or
journaled changes to apply, continue with “Task 5–Restoring Additional
Information” on page 318. Otherwise, skip to “Task 6–Restoring Program
Temporary Fixes (PTFs)” on page 318.

Chapter 14. How to Restore the System from the Save Storage Media 317
Task 5–Restoring Additional Information
If you are restoring changed objects, changed DLOs, or changed objects in
directories, you must first restore user profiles. This builds the authority
information for any new objects that you restore. If you are applying only
journaled changes, start with step 4.
1. Sign on as QSECOFR.
2. Put your system in a restricted state. See “Putting Your System in a Restricted
State” on page 45.
3. Restore user profiles. See “Restoring User Profiles” on page 214.
4. Restore changed objects and apply journaled changes. Follow the instructions
in “Chapter 11. How to Restore Changed Objects and Apply Journaled
Changes” on page 279.
5. Restore authority by typing: RSTAUT.

Task 6–Restoring Program Temporary Fixes (PTFs)

If you have applied PTFs since your SAVSTG procedure, follow the instructions in
“How to Restore Program Temporary Fixes” on page 278.

Stop!
You have now completed restoring your system from SAVSTG media.

How to Resume the Restore Storage Operation

You can use this procedure to resume the restore storage operation that ended
before the entire restore operation of the disk unit data was complete.

To start the restore storage operation again, do the following:

1. From the Dedicated Service Tools (DST) menu, select option 9 (Work with save
storage and restore storage) and press the Enter key.
2. Select option 2 (Resume restore storage) and press the Enter key.
3. If the following display is shown, load the tape that is indicated and press the
Enter key.

318 OS/400 Backup and Recovery V5R1

 Resume Restore Storage

Do the following:

1. Locate the tape to resume the restore on. The tape

that was being read when the restore storage was
interrupted has the following identification:
Volume identifier . . . . . . . . : ________
Sequence number . . . . . . . . . : ____
2. Insert the tape in the tape device.
3. Make the tape device ready, if necessary.

Note:
If the restore storage was interrupted because of a media
error on a tape, you may want to resume the restore
storage on the tape following the failing tape. If you
resume the restore storage on that tape, the system will
have damaged objects, and the system might not be able to
perform and IPL to OS/400 when the restore storage is complete.

Press Enter to continue.

4. If the wrong volume is loaded, the Device Intervention Required display with a
message at the bottom is shown. Type the name of the correct volume or file,
and press the Enter key.
5. The restore storage operation starts again.

If the restore storage operation continues to fail on the same tape with a tape
media failure, you have three options:
v Use a previous copy of your save storage tapes to completely restore storage.
v Resume the restore storage operation by using the tape following the tape with
the media error. If the tape that has the media error is the last tape to restore in
the set, option 3 (Force end of an interrupted restore storage) on the Restore
Storage menu should be selected.

Attention!
Some disk unit data is not restored. There may also be many objects that
are damaged on the system when the restore operation completes. An
initial program load of the operating system may not be successful. You
should restore the operating system again.

v Initialize your system and then begin a restore of your system from the tapes
that were created using SAVSYS and SAVLIB commands or options from the
Save menu.

Chapter 14. How to Restore the System from the Save Storage Media 319
320 OS/400 Backup and Recovery V5R1
Part 4. Release-to-Release Support
Chapter 15. Release-to-Release Support . . . 323
Current Release-to-Previous Release Support . . . 323
Creating the Object for the Previous Release . . 324
Saving the Object for the Previous Release . . 325
Testing the Object on the Current Release . . . 330
Restoring and Using the Object on the Previous
Release . . . . . . . . . . . . . . 331
Restrictions for Current Release-to-Previous
Release Support . . . . . . . . . . . 331
Previous Release-to-Current Release Support . . . 331
Considerations when Moving System
Customization Information . . . . . . . . 332
Restoring previous release user data to a new
system . . . . . . . . . . . . . . 332
Prerequisites for the recovery... . . . . . 333
Restoring previous release user data to a new
system: Step-by-step instructions . . . . . 334
Saving spooled files . . . . . . . . . 347
Restrictions when going from previous release
to current release . . . . . . . . . . . 348

Chapter 16. System Synchronization-Planning

and Procedures . . . . . . . . . . . . 351
Synchronization Methods: Overview . . . . . 352
Moving Changed Objects . . . . . . . . . 353
Steps for Saving Changed Objects . . . . . 354
Steps for Restoring Changed Objects . . . . 355
Problems When Restoring Changed Objects . . 357
Problems Restoring Journal Receivers . . . 357
Problems Restoring Database File Members 357
Problems with Object Authority or
Ownership . . . . . . . . . . . . 358
Moving Entire Libraries . . . . . . . . . . 358
Considerations for Moving Entire Libraries . . 360
Moving Individual Objects . . . . . . . . . 360
Applying Journaled Changes . . . . . . . . 361
Refreshing Your new system . . . . . . . . 363
Additional Synchronization Tips . . . . . . . 363

© Copyright IBM Corp. 1997, 2000, 2001 321

322 OS/400 Backup and Recovery V5R1
 Chapter 15. Release-to-Release Support
The release-to-release support on the iSeries and AS/400e servers allows you to
move data from the current release to a previous-release system. This support also
allows you to move data from a previous-release system to a current-release
system.

Releases of licensed programs on the iSeries or AS/400e servers have a three-part

| name that consists of a version, a release, and a modification. For example, the
| current version is Version 5 Release 1 Modification 0. The short form of the current
| release name is V5R1M0. This chapter uses the short form for release names.

Please read “Restrictions for Current Release-to-Previous Release Support” on

page 331 and “Restrictions when going from previous release to current release” on
page 348 for important information.

Current Release-to-Previous Release Support

This support enables objects (such as programs and files) that are created and
saved on the current release to be restored and used on a previous release. Object
compatibility is provided for many languages, and most object types are supported
on both release levels as long as the objects use only functions from a previous
release.

You can enable current release-to-previous release support by using the target
release (TGTRLS) parameter on create or save commands.

Table 57 illustrates the TGTRLS parameter and values available for the current and
previous releases. The values in the table are used throughout this chapter. Refer to
this table to determine the valid values for the release currently on your system.
| Table 57. Values for TGTRLS Parameter
| Current OS/400
| Release *CURRENT *PRV Other Valid Values
| V5R1M0 V5R1M0 V4R5M0 V4R4M0
| V4R3M0 V4R2M0
| V4R5M0 V4R5M0 V4R4M0 V3R2M0
| V4R4M0 V4R4M0 V4R3M0 V4R2M0 V3R2M0
|

| Note: As of V5R1M0, you cannot target a save for an internal microprogram

| instruction (IMPI) based system.

This support is extremely useful to:

v A network enterprise with a central site development system on the current
release and with remote sites still on the previous release.
v An application development business with a single system on the current release
that supports customers who may still be on the previous release.

Current release-to-previous release support provides great savings and

productivity improvements to application developers. By using this support, most

© Copyright IBM Corp. 1997, 2000, 2001 323

 network enterprises and application development businesses no longer need to
maintain two development systems. (For example, two development systems could
contain a previous-release system that contains previous-release objects, or a
current-release system that contains current-release objects.) In most cases, this
support enables previous- and current-release objects to exist on one development
system.

The following sections describe how to create and save objects on the current
release, and how to restore and use them on the previous release.

Creating the Object for the Previous Release

If you wish to run CL programs for a previous release, you must install option 9
(*PRV CL Compiler Support) from OS/400.

The following object types must be created specifically for a target release:
v Program (*PGM)
v Service program (*SRVPGM)
v Module (*MODULE)
v C locale description (*CLD)
v SQL package (*SQLPKG)
Create the object on the current release by using the appropriate create command
with the TGTRLS parameter. All other object types can skip this step. If the object
was created on, or is restored from, the previous release, and is not created again
on the current release, you can skip this step. To determine what release the object
was created on, use the DSPOBJD command and specify DETAIL(*SERVICE) to
display the System-level value.

It is recommended that previous-release programs and current-release programs be

stored in separate libraries to simplify maintenance. Using the DSPPGM command,
the Earliest release that program can run field shows if a program can be saved to the
previous release. This recommendation also applies to previous-release and
current-release modules and service programs. To determine if *MODULE objects
can be saved to a previous release, use the DSPMOD command. To determine if
*SRVPGM objects can be saved to a previous release, use the DSPSRVPGM
command.

Table 58 shows the languages and commands that support the TGTRLS parameter:
Table 58. Language Support for the Target Release Parameter
Language Command
ILE C CRTBNDC
CRTCMOD
CRTCLD

| ILE C++ CRTBNDCPP

CRTCPPMOD

CICS® CRTCICSC
CRTCICSCBL
CRTCICSGRP
CRTCICSMAP

324 OS/400 Backup and Recovery V5R1

 Table 58. Language Support for the Target Release Parameter (continued)
Language Command
CL CRTBNDCL
CRTCLMOD
CRTCLPGM

ILE COBOL CRTBNDCBL

CRTCBLMOD
CRTCBLPGM
CRTS36CBL

ILE RPG CRTBNDRPG

CRTRPGMOD
CRTRPGPGM
CRTRPTPGM
CRTS36RPG
CRTS36RPT

SQL CRTSQLCI
CRTSQLCBL
CRTSQLCBLI
CRTSQLCPPI
CRTSQLPLI
CRTSQLRPG
CRTSQLRPGI

Other CRTPGM
CRTSRVPGM

| The following compilers are not available for V5R1M0 and later or do not support
| the TGTRLS parameter:
v BASIC
v CL (S/38 Environment)
v COBOL/74 (S/38 Environment)
v FORTRAN/400
v AS/400 Pascal
v AS/400 PL/I
v RM/COBOL-85
v RPG/III (S/38 Environment)

Saving the Object for the Previous Release

You must save the object on the current release by using the TGTRLS parameter
before restoring it on the previous release or previous modification. This saves the
object in a format that is known to the previous release or previous modification.

Use communication lines or removable storage media (tape, optical media volume,
or diskette to move the objects from the current-release system).

It is recommended that you store previous-release and current-release objects in

separate libraries to simplify maintenance.

The following save commands support the TGTRLS parameter:

Chapter 15. Release-to-Release Support 325

 v Save (SAV)
v Save Calendar (SAVCAL)
v Save Changed Objects (SAVCHGOBJ)
v Save CICS Group (SAVCICSGRP)
v Save Document Library Objects (SAVDLO)
v Save DLO using BRM (SAVDLOBRM)
v Save Folder List using BRM (SAVFLRLBRM)
v Save Library (SAVLIB)
v Save Library using BRM (SAVLIBBRM)
v Save Licensed Program (SAVLICPGM)
v Save Media Information using BRM (SAVMEDIBRM)
v Save Object (SAVOBJ)
v Save Object by using BRM (SAVOBJBRM)
v Save Object List by using BRM (SAVOBJLBRM)
v Save/Restore Objects (SAVRST)
v Save/Restore Changed Object (SAVRSTCHG)
v Save/Restore Document Library Object (SAVRSTDLO)
v Save/Restore Library (SAVRSTLIB)
v Save/Restore Object (SAVRSTOBJ)
v Save Ultimedia System Facilities Container (SAVUSFCNR)

The System Manager licensed program uses the previous-release support provided
by the SAVLICPGM command. It provides the capability to package software for
multiple releases from the same system.

Object compatibility is provided for most object types that are supported on both
levels as long as the object only uses the previous-release function.

| Table 59 shows what object types can and cannot be specifically created or saved
| for a previous release. IBM does not support saving IBM-supplied objects (such as
| system commands and programs) from the current release and restoring them on a
| previous-release system. Refer to Table 57 on page 323 for a list of supported
| TGTRLS values.

| Table 59 uses these values:

| v All means the object can be saved to all TGTRLS values that are supported on
| the current version of the operating system.
| v VvRrMm indicates the earliest release to which an object could be saved.
| However, you may need to refer to Table 57 on page 323 to find the earliest
| TGTRLS value that is supported on the current version of the operating system.
| v *CURRENT means the object can only be saved to the current release,
| TGTRLS(*CURRENT).
| v None means the object is saved by a command, such as SAVSECDTA or
| SAVCFG, that does not support the TGTRLS parameter.
| Table 59. Previous-Release Support by Object Type
| Object Type Earliest Target Release
| *ALRTBL All
| *AUTHLR None

326 OS/400 Backup and Recovery V5R1

| Table 59. Previous-Release Support by Object Type (continued)
| Object Type Earliest Target Release
| *AUTL None
| *BLKSF All
| *BNDDIR All
| *CFGL None
| *CHTFMT All
| *CLD All
| *CLS All
| *CMD All
| *CNNL None
| *COSD None
| *CRG V4R4M0
| *CRQD All
| *CSI All
| *CTLD None
| *DDIR All
| *DEVD None
| *DIR All
| *DOC All
| *DSTMF All
| *DTAARA All
| *DTADCT All
| *DTAQ All6
| *EDTD All
| *EXITRG All
| *FCT All
| *FIFO V5R1M0
| *FILE (database, device, save) All
| *FLR All
| *FNTRSC All
| *FNTTBL V3R7M0
| *FORMDF All
| *FTR All
| *GSS All
| *IGCDCT All
| *IGCSRT All
| *IGCTBL All
| *IPXD None
| *JOBD All
| *JOBQ All
| *JOBSCD All

Chapter 15. Release-to-Release Support 327

| Table 59. Previous-Release Support by Object Type (continued)
| Object Type Earliest Target Release
| *JRN All
5
| *JRNRCV All
| *LIB All
| *LIND None
| *LOCALE V3R7M0
| *MEDDFN V4R4M0
| *MENU All
| *MGTCOL V4R4M0
| *MODD None
1
| *MODULE

|| v ILE C All

|| v ILE C++ All

|| v ILE CL All

|| v ILE COBOL All

|| v ILE RPG/400® All

7
| *MSGF All
| *MSGQ All
| *M36 V3R7M0
| *M36CFG V3R7M0
| *NODGRP V3R2M0
| *NODL All
| *NTBD None
| *NWID None
| *NWSD None
| *OUTQ All
| *OVL All
| *PAGDFN All
| *PAGSEG All
| *PDG All
1, 2
| *PGM:
|| v BASIC *CURRENT

|| v CL (S/38 environment) *CURRENT

|| v CL (iSeries environment) All

|| v COBOL (iSeries environment) All

|| v COBOL/74 (S/38 environment) *CURRENT

|| v COBOL/74 (S/36 environment) All

|| v ILE C All

328 OS/400 Backup and Recovery V5R1

| Table 59. Previous-Release Support by Object Type (continued)
| Object Type Earliest Target Release

|| v ILE C++ All

|| v ILE CL All

|| v ILE COBOL All

|| v ILE RPG All

|| v PASCAL *CURRENT

|| v PL/I *CURRENT

|| v RPG/II (S/36 environment) All

|| v RPG/III (S/38 environment) *CURRENT

|| v RPG (iSeries environment) All

| *PNLGRP All
| *PRDAVL *CURRENT
| *PRDDFN All
| *PRDLOD All
| *PSFCFG V3R2M0
| *QMFORM All
| *QMQRY All
| *QRYDFN All
| *RCT *CURRENT
| *SBSD All
| *SCHIDX All
| *SOCKET None
| *SPADCT All
| *SQLPKG All
| *SQLUDT V4R4M0
1, 3
| *SRVPGM
|| v ILE C All

|| v ILE C++ All

|| v ILE CL All

|| v ILE COBOL All

|| v ILE RPG/400 All

| *SSND All
4
| *STMF All
| *SVRSTG V3R2M0
| *SYMLNK All
| *S36 *CURRENT
| *TBL All
| *USRIDX All

Chapter 15. Release-to-Release Support 329

| Table 59. Previous-Release Support by Object Type (continued)
| Object Type Earliest Target Release
| *USRPRF None
| *USRQ All
| *USRSPC All
| *VLDL V4R1M0
| *WSCST All

1
| To move a program, module, or service program from a system that is running a PowerPC based release
| (Version 3 Release 6 Modification 0 or later) to a system that is running a non-PowerPC based release
| (V3R2 or earlier), the object must have observability. You can use the following commands to determine
| whether the observable information is stored with the object or has been removed:
| v Display Program (DSPPGM)
| v Display Module (DSPMOD)
| v Display Service Program (DSPSRVPGM)
2
| For ILE programs (a *PGM object created by binding one or more *MODULE objects together), the target
| release is determined by examining the target release value for each input *MODULE. If the target release
| values are different, the most recent target release value is used. An ILE program can be created from
| *MODULE objects created by different ILE compilers. The entries in this table for ILE languages under the
| *PGM object type state which target release values are supported by the ILE compiler when creating a
| *MODULE object. The *MODULE object can be used in turn to create an ILE program by using the
| CRTPGM command.
3
| For ILE service programs (a *SRVPGM object created by binding one or more *MODULE objects together),
| the target release is determined by examining the target release value for each input *MODULE. If the
| target release values are different, the most recent target release value is used. An ILE service program can
| be created from *MODULE objects created by different ILE compilers. The entries in this table for ILE
| languages under the *SRVPGM object type state which target release values are supported by the ILE
| compiler when creating a *MODULE object. The *MODULE object can be used in turn to create an ILE
| service program by using the CRTSRVPGM command.
4
| In V4R3, support was added for *STMF sizes up to 4gigabyte - 1byte. A *STMF larger than 2gigabyte - 1
| byte cannot be saved to releases prior to V4R3. In V4R4, support was added for *STMF sizes greater than
| 4gigabyte - 1byte. A *STMF larger than 4gigabyte - 1byte cannot be saved to releases prior to V4R4.
5
| If a journal receiver was attached to a journal when RCVSIZOPT(*MAXOPT1) was in effect, then it cannot
| be saved or restored to a release prior to V4R5M0. Also, it cannot be replicated to any remote journals on
| any systems at release prior to V4R5M0. If a journal receiver was attached to a journal when
| RCVSIZOPT(*MAXOPT2) was in effect, then it cannot be saved or restored to a release prior to V5R1M0.
| Also, it cannot be replicated to any remote journals on any systems at releases prior to V5R1M0. If a
| journal receiver was attached to a journal when any MINENTDTA options were in effect, then it cannot be
| saved or restored to a release prior to V5R1M0. Also, it cannot be replicated to any remote journals an any
| system prior to V5R1M0.
6
| V4R5M0 is the earliest release for *DTAQ if the SIZE and AUTORCL parameters on CRTDTAQ did not
| contain the default values when the data queue was created.
7
| V4R5M0 is the earliest release if *UBIN or *BIN 8 is specified for the format of a message description
| within the message file.
|
|

Testing the Object on the Current Release

Once the object has been created and saved using the TGTRLS parameter it can be
tested on the current-release system. Thus, it is no longer necessary to support and
maintain two development systems (one running the current release and one
running the previous version). Testing this object should be like testing any other
object. Make sure that all the objects that are to be used on the previous-release

330 OS/400 Backup and Recovery V5R1

 system have been saved using the TGTRLS parameter, restored onto the
current-release system, and tested as a group on the current-release system.

Restoring and Using the Object on the Previous Release

Once testing on the current-release system is completed, it is recommended that
the object be distributed on a limited basis to previous-release systems or previous
modification systems. Thus, if problems arise, they can be quickly corrected and
contained with minimal impact to users.

Restrictions for Current Release-to-Previous Release Support

The following restrictions apply when you create and save objects on the current
release, and then restore and use them on the previous release.
v You cannot specify a TGTRLS value earlier than V4R5M0 to save data to optical
media that you initialized with Universal Disk Format (UDF).
v The System/38 environment compilers (CL, COBOL/74, and RPG/III) do not
support the TGTRLS parameter. Programs that are created using these compilers
cannot be saved to, restored, or run on a previous-release system.
v The only way to save an object for a previous-release system is to use the
TGTRLS parameter. If the TGTRLS parameter is not specified on the save
command, and you attempt to restore the object on the previous-release system,
the object is not restored.
v IBM does not support saving IBM-supplied objects (such as system commands,
system programs, IBM spelling aid dictionaries, and so forth) from a
current-release system and restoring and using them on a previous-release
system or previous modification. As a result, the TGTRLS parameter is not
supported on a SAVLIB command that specifies *ALLUSR, *IBM, or *NONSYS
on the LIB parameter.
v IBM does not support new function from the current release to be used on a
previous-release system or previous modification.
v When saving document library objects for a previous release, only folders and
filed documents can be saved. Other items, such as mail or documents that are
not filed cannot be saved using TGTRLS values other than *CURRENT.
v If a current-release program temporary fix (PTF) save file is sent from a current
release-system to a previous-release system for distribution to another
current-release system, object distribution must be used. The Copy PTF
(CPYPTF) command or any save file command, such as DSPSAVF, cannot
process the PTF save file.

Previous Release-to-Current Release Support

Moving All Your Data to the Current Release?

This topic describes considerations when you are moving specific types of
information from an earlier release to the most current release. If you are
upgrading from a PowerPC based system to another PowerPC based system,
refer to the iSeries 940x RISC-to-RISC Road Map book.

Generally, the system to which you are restoring objects must be at the same or a
higher release level than the system from which the objects were saved, unless you
specified a target release value when you saved. When moving data to a higher
level release, you should only move user data. This may include user libraries,
user directories, user profiles, user objects in IBM-supplied libraries, DLOs, and

Chapter 15. Release-to-Release Support 331

 mail. IBM-supplied libraries and IBM-supplied directories should not be restored
to a higher release since these are handled during the licensed program install
process. The target system should have the current-level release installed. This
includes the Licensed Internal Code, OS/400 operating system, IBM-supplied
libraries QGPL and QUSRSYS, OS/400 optional libraries, and any licensed
programs that are purchased. See the Software Installation book to install the current
release.

Considerations when Moving System Customization

Information
Some system customization information that is stored in the QSYS library cannot
be saved. This includes network attributes, system values, the system reply list,
and configuration information. You must manually re-create this information on
your new or upgraded system. In addition, you will not be able to recover your
problem log and your question and answer database. Use the procedure that is
described in “Printing system information” on page 26 to print your current values.

Restoring previous release user data to a new system

The preferred method to restore previous release user data on your new, target
system is to use the iSeries 940x RISC-to-RISC Road Map. The RISC to RISC method
asks you to first install the new, current release onto your old, source system. After
which, you save your old system and then you perform a full system recovery
onto the new, target system. Only use these instructions if it is not possible to
perform the preferred RISC to RISC migration process in the iSeries 940x
RISC-to-RISC Road Map.

This section provides instructions to restore user data from a previous-release

RISC-based system to a later release RISC-based system. Look for supported
releases for software upgrades in the software installation book for your new,
target release. This information explains for which releases you can use these
instructions to restore your previous release user data to your new system.

The restore procedure involves two save steps, and four recovery steps. The save
steps include printing your system information and completely backing up your
old, source system.

The recovery steps on the new, target system involve the following four steps:
1. Install Licensed Internal Code and OS/400 on the target system using new
release distribution media.
2. Restore system and user data to target system using the option 21 save of the
source system.
3. Update the system information on the target system.
4. Install QGPL, QUSRSYS, Base options, and LPPs using new release distribution
media on target system.
This converts restored source data to new target release.

Verify the prerequisites below and proceed to the step-by-step instructions to

restore previous release user data to a new system.

332 OS/400 Backup and Recovery V5R1

Figure 30. Recovery steps for restoring previous release user data to a new system

Prerequisites for the recovery...

These instructions are sometimes used for a system upgrade when you install a
replacement processor.

You must perform the following prerequisite steps before you start the recovery
portion of these instructions:
v If available on your system, you run the RTVSYSINF command on the source
system. Some releases of OS/400 do not support the RTVSYSINF command.
When you run the RTVSYSINF command, the system asks you what library to
use. Typically, you should specify the QUPGRADE library.
v If this is available on your system, you print system information with the
PRTSYSINF command on the source system. Some releases of OS/400 do not
support the PRTSYSINF command. If your release does not support it, refer to
the Backup and Recovery book for the release of your OS/400 for instructions
on how to print the system information.
v If necessary, you save spooled files. For step-by-step instructions on how to save
spooled files, see “Saving spooled files” on page 347.

Note: Job scheduler entries will not be restored. If necessary, make a note of
your current job scheduler entries and recreate them manually on the new
system.

Chapter 15. Release-to-Release Support 333

 v You ran Save menu option 21 of the GO SAVE command on the source system.
Make sure that you specify the following options:
1. Vary off the network servers...*ALL
2. Unmount file systems............Y
v You have a copy of the distribution media for the target system.
v If required, you set up device parity protection and load source mirroring on the
target system. If you have not done so on the new system, consult an iSeries and
AS/400e specialist to configure your DASD for your level of mirroring, and
parity protection before you continue with these instructions.

Restoring previous release user data to a new system:

Step-by-step instructions
Perform the following steps on the new, target system. You will first install only
the Licensed Internal Code and the OS/400 Operating system from the distribution
media from the target release. Do not install any of the base options or LPPs at this
time.
跳转
Check off each item as you complete the task on the target system.
__ 1. IPL the system from the first distribution media.
__ a. Mount the first distribution media on the alternate IPL device. Wait
for the READY status.
__ b. At the CPU control panel, place the system in MANUAL mode.
__ c. Press the Function Select switch (or buttons) to display 02 (IPL) in the
Function display.
__ d. Press Enter.
__ e. Press the Function Select switch (or buttons) to display D (IPL from
tape or CD-ROM) in the Data display.
__ f. Press Enter.
__ g. If the system is powered down, press the power button on the system
to power the system on. Skip to Step 2. Otherwise, continue with Step
1h.
__ h. If the system is powered on, press the Function Select switch (or
buttons) to display 03 (continue the IPL) in the Function display.
__ i. Press Enter.
__ 2. On the Install Licensed Internal Code (LIC) screen, select 2, Install Licensed
Internal Code and Initialize system, to start a ″Scratch Install″ of the
Licensed Internal Code.

334 OS/400 Backup and Recovery V5R1

 Install Licensed Internal Code (LIC)

Disk selected to write the Licensed Internal Code to:

Serial Number Type Model I/O Bus Controller Device

xx-xxxxxxx xxxx xxx x x x

Select one of the following:

1. Restore Licensed Internal Code

=>2. Install Licensed Internal Code and Initialize System
3. Install Licensed Internal Code and Recover Configuration
4. Install Licensed Internal Code and Restore Disk Unit Data
5. Install Licensed Internal Code and Upgrade Load Source

Selection
2

__ 3. On the Install LIC and Initialize System - Confirmation screen, press F10 to
confirm the initialization and to continue the install.

Install LIC and Initialize System - Configuration

Warning:

All data on this system will be destroyed and the Licensed

Internal Code will be written to the selected disk if you
choose to continue the initialize and install.

Return to the install selection screen and choose one of the

other options if you want to perform some type of recovery
after the install of the Licensed Internal Code is complete.

Press F10 to continue the install.

Press F12 (Cancel) to return to the previous screen.
Press F3 (Exit) to return to the install selection screen.

__ a. The Initialize the Disk - Status screen appears.

Initialize the Disk - Status

The load source disk is being initialized.

Estimated time to initialize in minutes: xx

Elapsed time in minutes . . . . . . . .: 0.0

__ b. The Install Licensed Internal Code - Status screen appears.

Install Licensed Internal Code - Status

Install the Licensed Internal Code in progress

Percent Complete . . . . . 0% 8.5 minutes

Chapter 15. Release-to-Release Support 335

 __ 4. On the Disk Configuration Attention Report screen, press F10 to accept any
problems and to continue.

Disk Configuration Attention Report

Type option, press Enter

5=Display Detailed Report

Press F10 to accept all the following problems and continue.

The system will attempt to correct them.

OPT Problem
_ New disk configuration

F3=Exit F10=Accept the problems and continue

__ 5. On the IPL or Install the System screen, select 3, Use Dedicated Service Tools
(DST).

IPL or Install the System

Select one of the following:

1. Perform an IPL
2. Install the operating system
3. Use Dedicated Service Tools (DST)
4. Perform automatic installation of the operating system
5. Save Licensed Internal Code

Selection
3

__ 6. Sign on to DST as DST user, QSECOFR, with the password QSECOFR. The
password is case sensitive; use all capital letters.

Dedicated Service Tools (DST) Sign On

Type choices, press Enter.

DST User . . . . . . . . QSECOFR

DST password . . . . . . QSECOFR

__ a. Select option 4, Work with Disk Units.

__ b. Select option 1, Work with Disk Configuration.
__ c. Select option 3, Work with ASP Configuration.
__ d. Select option 3, Add Units to ASPs.
__ 7. On the Specify ASPs to Add Units to screen, enter ″1″ for each unit that needs
to be in the System ASP (ASP 1).
__ a. If you require more than one ASP, enter the corresponding ASP
number on each selected unit.

336 OS/400 Backup and Recovery V5R1

 Specify ASPs to Add Units to

Specify the ASP to add each unit to.

Specify Serial Resource

ASP Number Type Model Capacity Name

1 00-0103706 6602 030 1031 DD031

1 00-1000341 9337 211 542 DD012
1 00-5000341 9337 211 542 DD015
1 00-7000341 9337 211 542 DD011
1 00-3000341 9337 211 542 DD014
1 00-2000341 9337 211 542 DD013
1 00-61300 6603 074 1475 DD006
1 00-52262 6606 074 1475 DD008
1 00-86978 6606 050 1967 DD009
2 00-95744 6603 074 1475 DD005
2 00-47657 6606 074 1475 DD007
- 00-0238703 6602 074 773 DD051
- 00-0128350 6602 074 773 DD051

__ b. After you complete all units, press Enter.

__ c. If the list of units is correct, press Enter to start initializing the units.
__ 8. On the Problem Report screen, press F10, Ignore problems and continue.

Problem Report

Note: Some action for the problems listed below may need to
be taken. Please select a problem to display more detailed
information about the problem and to see what possible
action may be taken to correct the problem.

Type option, press Enter.

5=Display Detailed Report

OPT Problem
_ Unit possibly configured for Power PC AS

F3=Exit F10=Ignore problems and continue F12=Cancel

__ 9. On the Confirm Add Units screen, press Enter to confirm the selected units.

Chapter 15. Release-to-Release Support 337

 Confirm Add Units

Add will take several minutes for each unit. The system will
have the displayed protection after the unit(s) are added.

Press Enter to confirm your choice for 1=Add units.

Press F9=Capacity Information to display the resulting capacity.
Press F12=Cancel to return and change your choice.

Serial Resource
ASP Unit Number Type Model Name Protection

1 Unprotected
1 00-0103706 6602 030 DD031 Unprotected
2 00-1000341 9337 211 DD012 Unprotected
3 00-5000341 9337 211 DD015 Unprotected
4 00-7000341 9337 211 DD011 Unprotected
5 00-3000341 9337 211 DD014 Device Parity
6 00-2000341 9337 211 DD013 Device Parity
7 00-61300 6603 074 DD006 Device Parity
8 00-52262 6606 074 DD008 Device Parity
9 00-86978 6606 050 DD009 Device Parity
2 Unprotected
10 00-95744 6603 074 DD005 Device Parity
11 00-47657 6606 074 DD007 Device Parity

__ a. The Function Status screen displays the percentage of completion.

__ b. The ″Selected units have been added successfully″ message appears when
the system completes the Add Units process.
__ c. Press F12 to return to the Work with Disk Configuration screen.
__ d. If your system requires mirrored protection, continue to Step 9e. If
your system does not require mirrored protection, press F3 until you
exit the Dedicated Service Tools (DST) screen.
__ e. To start mirrored protection for your system, follow these instructions:
__ 1) On the Work with Disk Configuration screen, select 4, Work with
Mirrored Protection.
__ 2) On the Work with Mirrored Protection screen, select 2, Start
Mirrored Protection.
__ 3) Select an ASP by placing a ″1″ next to it. Press Enter to start
mirrored protection.
__ 4) On the Confirm Continuation screen, press Enter.
__ 5) On the Confirm Start Mirrored Protection screen, press Enter.
__ 6) The Function Status screen displays the completion status of the
Start Mirrored Protection request.
__ 7) The Start mirroring completed successfully message appears on
the Disk Configuration Information Report screen.
__ 8) Press Enter to continue.
__ f. If you use Operations console, follow these instructions to switch from
’local console’ to ’operations console’:
__ 1) On the IPL or Install the System screen, select 3, Use Dedicated
Service Tools (DST). Press Enter to continue.
| __ 2) Sign on to DST as using a service tools user profile that has
| security officer authority and the assigned password.
__ 3) On the Use Dedicated Service Tools (DST) screen, select 5, Work
with DST environment. Press Enter to continue.

338 OS/400 Backup and Recovery V5R1

 __ 4) On the Work with DST Environment screen, select 2, System
Devices. Press Enter to continue.
__ 5) On the Work with System Devices screen, select 6, Console Mode.
Press Enter to continue.
| __ 6) On the Select Console Type screen, select 2, Operations Console
| or option 3 Operations Console (LAN). Press Enter to continue.
__ 7) Press F3 or F12 to get back to the IPL or Install the System
screen.
| __ 10. Load the first volume of installation media that contains OS/400. (This
| volume is labeled B29xx_01, where 29xx is the identifier for the primary
| language.)
| When you are installing from optical media, wait until the In Use indicator
| goes out before you continue.
__ 11. On the IPL or Install the System screen, select 2, Install the Operating
System.

IPL or Install the System

Select one of the following:

1. Perform an IPL
2. Install the Operating System
3. Use Dedicated Service Tools (DST)
4. Perform automatic installation of the Operating System
5. Save Licensed Internal Code

Selection
2

__ a. On the Confirm Install of OS/400 screen, press Enter.

__ b. The Select a Language Group screen displays the primary language
feature that is currently on your save media. Press Enter to accept
this value.

Select a Language Group

Note: The language feature shown is the language feature

installed on the system.

Attention: To keep the same primary language, ensure that the

media you use for installing the operating system
matches the language feature shown. If the operating
system media does not match what is shown, the
installation process will attempt to install the
operating system in a different language feature than
Licensed Internal Code. This is undesirable.
Type Choice, press Enter.
Language Feature . . . . . . . . . 2924

__ c. The Confirm Language Feature Selection screen appears. Press Enter to

continue.
__ 12. On the Add All Disk Units to the System screen, select option 1, Keep the
current disk configuration.

Chapter 15. Release-to-Release Support 339

 Add All Disk Units to the System

Select one of the following:

1. Keep the current disk configuration

2. Perform disk configuration using DST
3. Add all units to the system auxiliary storage pool (ASP)
4. Add all units to the system ASP and balance data

Selection
1

Note: This screen does not appear if you selected all disk units that are
known to the system on Step 7 on page 336.
__ 13. The IPL Step in Progress screen displays the IPL progress.

IPL Step in Progress

IPL step . . . : Storage Management Recovery

Authority Recovery
Journal Recovery
Database Recovery
Journal Synchronization
Start Operating System

__ 14. On the Install the Operating System screen, select option 1, Take defaults.
Make sure that the values for Date and Time are correct. Press Enter to
continue.

Install the Operating System

Type options, press Enter.

Install
option . . . . . 1 1=Take defaults (No other
options are displayed)
2=Change install options

Date
Year . . . . . . 99 00-99
Month. . . . . . 08 01-12
Day. . . . . . . 22 01-31

Time
Hour . . . . . . 16 00-23
Minute . . . . . 45 00-59
Second . . . . . 00 00-59

__ 15. The OS/400 Installation Status screen displays the installation status of the
required OS/400 Installation Profiles and Libraries.

340 OS/400 Backup and Recovery V5R1

 Message ID . . : CPI2070 OS/400 Installation Status

+---------------------------------------------------------------------------------+
Stage 1 |xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx |
+---------------------------------------------------------------------------------+
0 20 40 60 80 100

Installation Objects
Stage Completed Restored

1 Creating needed profiles and libraries . . .:

2 Restoring programs to library QSYS . . . . .:
3 Restoring language objects to library QSYS .:
4 Updating program table . . . . . . . . . . .:
5 Installing database files . . . . . . . . . :
6 Completing OS/400 installation . . . . . . .:

__ 16. The system installs the remaining OS/400 objects.

Message ID . . : CPI2070 OS/400 Installation Status

+---------------------------------------------------------------------------------+
Stage 1 |xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx |
+---------------------------------------------------------------------------------+
0 20 40 60 80 100

Installation Objects
Stage Completed Restored

1 Creating needed profiles and libraries . . .: x

2 Restoring programs to library QSYS . . . . .: x 09
3 Restoring language objects to library QSYS .: x 03
4 Updating program table . . . . . . . . . . .: x
5 Installing database files . . . . . . . . . : x
6 Completing OS/400 installation . . . . . . .:

__ 17. On the Sign On screen, logon as user QSECOFR. You do not need to enter a
password at this time.
__ 18. On the IPL options screen, enter correct values for Date and Time. Only the
following options should be set to Y:
v Start system to restricted state
v Set major system options
v Define or change the system at IPL

IPL Options

Type choices, press Enter.

System date . . . . . . . . . . . . . . . 08/22/99 MM/DD/Y

System time . . . . . . . . . . . . . . . 16:58:00 HH:MM:S
Clear job queues . . . . . . . . . . . . . N Y=Yes, N=No
Clear output queues . . . . . . . . . . . N Y=Yes, N=No
Clear incomplete job logs . . . . . . . . N Y=Yes, N=No
Start print writers . . . . . . . . . . . N Y=Yes, N=No
Start system to restricted state . . . . . Y Y=Yes, N=No

Set major system options . . . . . . . . . Y Y=Yes, N=No

Define or change system at IPL . . . . . . Y Y=Yes, N=No

Chapter 15. Release-to-Release Support 341

| __ a. On the Set Major System Options screen, select N to enable automatic
| configuration.
|
Set Major System Options

Type choices, press Enter.

Enable automatic configuration . . . . N Y=Yes, N=No

Device configuration naming . . . . . NORMAL *NORMAL, *S36

Default special environment . . . . . *NONE *NONE, *S36

__ 19. The Define or Change the System at IPL screen appears.

__ a. Select 3, Commands.
__ b. On the Change System Value Commands screen, select 3, Work with
System Values.
__ c. On the Work with System Values screen, select the System Value that
you plan to change by placing a ″2″ next to it. Press Enter ONLY
after you select all the values.
Update the following System Values. Write down the existing values
so you can update them after the recovery, if necessary.
v Change QALWOBJRST to *ALL
v Change QJOBMSGQFL to *PRTWRAP
v Change QJOBMSGQMX to a minimum of 30
v Change QPFRADJ to 2
v Change QIPLTYPE type to 2
| v Change QVFYOBJRST to 1
__ d. After the system changes the system values, press F3 twice to return
to the Define or Change the System at IPL screen.
__ e. On the Define or Change the System at IPL screen, press F3 to exit and
continue the IPL.
__ 20. On the Change Password screen, type QSECOFR as the current password.
Enter a new password. Re-enter the password to verify and press Enter.
(New password cannot be QSECOFR.)

Change Password

Password last changed . . . . . . xx/xx/xx

Type choices, press Enter.

Current password . . . . . . . . . QSECOFR

New password . . . . . . . . . . . _______
New password (to verify) . . . . . _______

__ 21.
| __ a. To configure 3422, 3430, 3480, or 3490 tape units, follow these
| instructions. If you have a 3490 Model E or F or to configure other
| types of tape units, go to step 21b on page 344.
1) Use the Work with Hardware Resource (WRKHDWRSC)
command to determine the location of the tape controller.
WRKHDWRSC TYPE(*STG)

342 OS/400 Backup and Recovery V5R1

2) Create the controller description for the tape controller by doing
the following:
a) Locate the resource name for the tape controller on the Work
with Storage Resources display. The value 34xx is displayed
in the Type column.
b) Write down the name of the resource.
c) Type a 9 (Work with resource) in the Opt column next to
name of the tape controller and press the Enter key. You see
the Work with Storage Resources display.

Note: If the resource is not listed on the display, you need to

select other resources, such as disk storage controllers.
For some server models, resources are now attached
through combined-function IOPs. Look through the
resources until you find the device you want.
d) Type 5 (Work with controller descriptions) in the Opt column
in front of the tape controller. You see the Work with
Controller Description display.
e) Type 1 (Create) in the Opt column on the top line.
f) Type the controller name (such as TAPCTL01) in the
description field and press the Enter key. You see the Create
Controller Description display.
g) If necessary, type additional information on the display. Then
press the Enter key. You return to the Work with Controller
Descriptions display.
h) If the controller description that you created does not appear,
press F5 (Refresh).
3) To create device descriptions for the tape units that are attached
to the controller, do the following:
a) On the work with Controller Descriptions display, press F23
(More options). The list of options changes.
b) Type 9 (Work with associated descriptions) in the Opt column
in front of the new tape controller. You see the Work with
Associated Descriptions display.
c) Locate the resource for the tape unit. Because no device
description exists, the description says *NONE.
d) Write down the name of the tape resource.
e) Type a 1 (Create) in the Opt column next to the description of
*NONE and press the Enter key. You see the Create Device
Desc (Tape) (CRTDEVTAP).
f) In the Device description field, type a name such as TAP01.
g) In the Resource name prompt, type the name that you wrote
down in step 21a3d. (If you did not write it down, press F12
to return to the display. Repeat steps 21a3d through 21a3g.)
h) Press the Enter key.
i) Additional parameters appear on the display.
j) If necessary, type additional information on the display. Then
press the Enter key. You return to the Work with Associated
Descriptions display.
k) Press F5 (Refresh). The name of the description that you
created should now be associated with the resource.

Chapter 15. Release-to-Release Support 343

 l) Type 8 (Work with configuration status) in front of the
controller description and the device description. You see the
Work with Configuration Status display.
m) Type 1 (Vary on or Make available) in front of the controller
and the devices.
4) Press F3 until your return to your original menu.
__ b. To configure tape units that are not models 34xx, use the following
instructions:
1) Use the Work with Hardware Resource (WRKHDWRSC)
command to determine tape controller name.
WRKHDWRSC TYPE(*STG)
2) Locate the tape controller on the Work with Hardware
Resources display.
3) Type a 9 (Work with resource) next to tape controller name and
press the Enter key.

Note: If the tape controller is not listed on the display, you

need to select other resources, such as disk storage
controllers. For some server models, tape units are now
attached through combined-function IOPs. Look through
the resources until you find the tape unit you want.
4) Locate the resource name for the tape unit (for example,
TAP01).
5) Enter a 5 (Work with Configuration Descriptions) in the Opt
column next to the tape resource name and press the Enter key.
You are shown the Work with Configuration Descriptions
display.
6) Type a 1 (Create) in the Opt field and a tape device description
name (for example, TAP01) in the Description field. Press the
Enter key. You are shown the Create Device Description (Tape)
display.
7) Change any values that you want to change, then press the
Enter key (twice) to create the device description. You are
shown the Work with Configuration Descriptions display again.
The device that you created should appear on the display.
8) Type an 8 (Work with configuration status) in front of the new
device description. You are shown the Work with Configuration
Status display.
9) Type a 1 (Vary on or Make available) in front of the new device.
If the status does not change to Varied On or Available, wait a
few minutes. Then press F5 (Refresh). If the status still does not
change to Varied On or Available, follow normal problem
analysis for the device.
10) Press F3 until you return to the main menu.

344 OS/400 Backup and Recovery V5R1

 AS/400 Main Menu

Select one of the following:

1. User tasks
2. Office tasks
3. General system tasks
4. Files, libraries, and folders
5. Programming
6. Communications
7. Define or change the system
8. Problem handling
9. Display a menu
11. Client Access/400 tasks
90. Sign off

Selection or command
=>

__ 22. Use the save media of the option 21 save from the source system to
perform the following steps to restore the user data and related system
data, and user data to the target system:
__ a. ENDSBS SBS(*ALL) OPTION(*IMMED)
__ b. RSTUSRPRF ALWOBJDIF(*ALL) ENDOPT(*LEAVE)
__ c. RSTCFG OBJTYPE(*ALL) SRM(*NONE) ALWOBJDIF(*ALL) ENDOPT(*LEAVE)
| __ d. RSTLIB SAVLIB(*NONSYS) OPTION(*NEW) ALWOBJDIF(*ALL)
| MBROPT(*ALL) FRCOBJCVN(*NO) ENDOPT(*LEAVE)

Note: This command does not restore the QAUDJRN and

QACGJRN objects and any job scheduler entries.
| __ e. RCLDLO *ALL
__ f. RSTDLO DLO(*ALL) SAVFLR(*ANY) ALWOBJDIF(*ALL) ENDOPT(*LEAVE)

Note: If you have DLOs in any of your user ASPs, you need to use
the following command to restore DLOs to each user ASP:
RSTDLO DLO(*AL) ALWOBJDIF(*ALL) SAVASP(ASP-number)
RSTASP(ASP-number)
| __ g. RST DEV('/QSYS.LIB/media-device-name.DEVD') OBJ(('/*')
| ('/QSYS.LIB' *OMIT) ('/QDLS' *OMIT)) ALWOBJDIF(*ALL)
__ h. To restore spooled files that you saved on your source system, do
the following:
__ 1) In “Saving spooled files” on page 347, you saved spooled files
to database files in a library. If that library has not yet been
restored to your upgraded system, restore it now by using the
RSTLIB command.

Note: Use the RSTLIB command only if you used the SAVLIB
command to save the objects. If you used the SAVOBJ
command, you must use the RSTOBJ command.
__ 2) For each spooled file that you need to restore, do the
following:
__ a) On the printout that you used when you saved the
spooled files, locate the name of the printer file that
was used to create the spooled file. It appears in the
File column on the left side of the printout.

Chapter 15. Release-to-Release Support 345

 __ b) To override the control character value for the
QPRINTS file, type the following command: OVRPRTF
FILE(QPRINTS) CTLCHAR(*FCFC)
__ c) Copy the database file to the printer file by typing the
following: CPYF FROMFILE(LIBSPLF/SPOOLF1)
TOFILE(QGPL/QPRINTS) MBROPT(*REPLACE)
In this example, a spooled file named QPRINTS is
placed on the output queue that is associated with the
definition for the QPRINTS printer file.
__ d) Delete the database file by using the Delete File (DLTF)
command.
__ 3) When you restored all spooled files, delete the override for
the QPRINTS file by typing the following: DLTOVR
FILE(QPRINTS)
__ 23. If you used the RTVSYSINF command on the source system, you must
now run the UPDSYSINF command to update the system information on
the target system.
If your source system does not support the RTVSYSINF command, you
need to use the printed system information to update all system
information such as system values, message reply lists entries, etc. on the
target system.
The RTVSYSINF command does not update all system information. Use the
printed system information to update additional system information such
as access path recovery times, subsystem descriptions, RJE configurations,
etc.
__ 24. Install the base options (including QGPL and QUSRSYS) and other licensed
programs using the distribution media for the target system (current
release) and the GO LICPGM command. Refer to the Software installation
book for your target release.
__ 25. Run the RSTAUT command.
__ 26. Install the cumulative PTF (program temporary fix) media.
From a command line, enter GO PTF to access the PTF menu. Select option
8 (Install program temporary fix package) on the PTF menu. This installs
all of the PTFs in the cumulative PTF package for the licensed programs
that are installed on your system. Refer to the iSeries System PTF Shipping
Information Letter for required special instructions.
If you want to restore individual PTFs, see the System Operation book for
more information about applying individual PTFs.
__ 27. If necessary, change the QALWOBJRST system value back to its original
value by using the WRKSYSVAL command.
__ 28. If necessary, change the QVFYOBJRST system value back to its original
value by using the WRKSYSVAL command.
__ 29. If necessary, change the QJOBMSGQFL system value back to its original
value by using the WRKSYSVAL command.
__ 30. If you do not know the password for the restored QSECOFR profile,
change the password before signing off. Type the following command:
CHGUSRPRF USRPRF(QSECOFR) PASSWORD(new-password)
__ 31. To place scheduled jobs on hold, type WRKJOBSCDE and select option 3 to
hold any scheduled jobs.
__ 32. Type the command, SIGNOFF *LIST or DSPJOBLOG * *PRINT. Check the job
log to ensure that the job restored all objects. To verify if the job restored

346 OS/400 Backup and Recovery V5R1

 all objects, spool the job log for printing along with any remaining spooled
job output. Check for any error messages. Correct the errors and restore
those objects from media.
__ 33. Perform an IPL of the system.
__ a. On the CPU control panel, press Mode Select until the green
indicator is in the NORMAL setting.
__ b. Type the command PWRDWNSYS OPTION(*IMMED) RESTART(*YES *FULL)
IPLSRC(B).
__ 34. If you installed Windows Server for AS/400 on your system and you saved
with your Netfinity® Server in a VARIED OFF setting, perform the
following steps:
__ a. Add the links for the server descriptions. Type the following for
each server description:

ADDNWSSTGL NWSSTG(Storage_Name) NWSD(Server_Description)

__ b. Vary on your Netfinity Servers by typing WRKCFGSTS *NWS and
selecting option 1 to vary on each Netfinity Server.

Saving spooled files

When you save a library that contains output queues, you save the descriptions of
the output queues but not their contents. If you have critical spooled files that you
will need after the upgrade procedure, you can use the following procedure to
save them:
__ 1. Create a library to hold copies of the critical spooled files by using the
CRTLIB command. For example, to create a library called LIBSPLF to hold
your spooled files, type the following: CRTLIB LIB(LIBSPLF)

Note: If the spooled files contain confidential information, specify

AUT(*EXCLUDE) on the CRTLIB command.
__ 2. Use the Work with Output Queue command to locate the spooled files that
you want to save from a designated output queue: WRKOUTQ
OUTQ(lib-name/queue-name) OUTPUT(*PRINT)

Note: If you do not have special output queues that are designated for
critical spooled files, type WRKOUTQ OUTQ(*ALL)
__ 3. Print out and retrieve the listing of the spooled files that you need to save.
__ 4. On the printout, mark the spooled files that you need to save.
__ 5. For each spooled file, do the following:
__ a. Choose a name (8 characters or less) for the spooled file that will help
you to identify it. Each file should have a unique name.
__ b. Create a database file to hold the contents of the spooled file by
typing the following: CRTPF FILE(LIBSPLF/file-name) RCDLEN(133)
Notes:
1) For file-name, substitute the name that you assigned in step 5a.
2) Use an appropriate record length for the spooled file that you are
copying. The record length must be at least 1 character larger than
the spooled data to allow for the control character.
3) If you are copying a large spooled file, specify SIZE(*NOMAX)
when you create the database file.

Chapter 15. Release-to-Release Support 347

 __ c. Copy the contents of the spooled file to the database file that you
created by typing the following: CPYSPLF FILE(spooled-file)
TOFILE(LIBSPLF/file-name) JOB(job-number/user-name/job-name)
SPLNBR(nnn) CTLCHAR(*FCFC)

Note: For spooled-file, substitute the value from the File column on
the listing that you created in step 2 on page 347.
__ d. You may receive message CPA3312 if the spooled file contains special
attributes. Respond with G (GO) to continue saving the contents of
the spooled file.
If you need to save special attributes associated with spooled files
(such as advanced function attributes), see the ’AS/400 Road Map for
Changing to PowerPC Technology’ for more information. You receive
a confirmation message that the data has been copied.
__ 6. Repeat step 5 on page 347, steps 5a on page 347 through step 5d, for each
spooled file that you need to save.
__ 7. If you have additional output queues to process, return to 2 on page 347.
__ 8. Use the SAVLIB command to save the library that contains the copies of
your spooled files.

Restrictions when going from previous release to current

release
To avoid data loss, if you have any of these products installed on your system,
please read this notice before upgrading to OS/400 Version 5, Release 1 or any
later release:
| v 5769SA3 - OS/400 Integration for Novell NetWare
| v 5769XZ1 - OS/2 Warpserver for AS/400
| v 5769AC1 - Cryptographic Access Provider 40–bit
| v 5697ADS - Tivoli® ADSTAR Distributed Storage Management
| v 5769AS1 - WebSphere™ Application Server for AS/400
| v 5769CB1 - ILE Cobol
| v 5769CE1 - AS/400 Client Encryption
| v 5769CF1 - IBM Point-of-Sale Communications Utility for AS/400
| v 5769CL3 - IBM VisualAge® RPG and Cooperative Development Environment for
| AS/400
| v 5769CX2 - IBM Integrated Language Environment® C for AS/400
| v 5769CX5 - IBM VisualAge for C++ for AS/400
| v 5733DME - Domino Migration Engine for iSeries
| v 5769FW1 - IBM Firewall for AS/400
| v TivoITD - Tivoli IT Director
| v 5769MQ2 - IBM MQSeries® for AS/400
| v 5798NC3 - IBM Net.Commerce for AS/400, Version 3.1
| v 5769PM1 - IBM Performance Management/400
| v 5769PW1- IBM Application Development ToolSet for AS/400
| v 5733PY1 - IBM Payment Server™ for AS/400, Version 1.2
| v 5769RD1 - Option 6, OnDemand Client for Windows 3.1
| v 5769RD1- Option 7, OnDemand Client for OS/2
| v 5769RD1- Option 8, OnDemand Client for 32 bit Windows

348 OS/400 Backup and Recovery V5R1

| v 5769RG1- IBM Integrated Language Environment RPG for AS/400
| v 5769SA2 - IBM Integration Services for FSIOP
| v 5769SS1 - Option 15, OS/400 - Common Programming APIs Toolkit
| v 5769SS1 - Option 17, OS/400 - PSF/400 Fax Support
| v 1TDSVR - Tivoli IT Director
| v 5798TBW - IBM Wireless Connection for AS/400
| v 5798TBY - Facsimile Support for AS/400
| v 5639VW5 - Visual Warehouse™ 5.2
| v 5769WP1 - OfficeVision® for AS/400
| v 5763XD1 - Client Access™ for Windows 95/NT
| v 5763XK1 - Client Access Enhanced for Windows 3.1
| v 5649xxx - All Model 150 products
Once you install V5R1M0 or any later release, these products will no longer work.
You cannot use Restore License Programs (RSTLICPGM) to load a previous release
of these products onto a server with V5R1M0 or later version of the operating
system. To avoid losing data that depends on these products, migrate that data
from your server to an accessible location before upgrading to V5R1M0 or any
later release.
v For assistance in migrating your data to AS/400 Netserver, see the redbook The
AS/400 NetServer Advantage, SG24-5196.
v For additional information, read the redpaper: Migration Options for OS/2 Warp
Server for AS/400 and OS/400 Integration for Novell NetWare, REDP0020.
v For alternative ideas to help migrate your data, see information APAR II11689.

Chapter 15. Release-to-Release Support 349

350 OS/400 Backup and Recovery V5R1
Chapter 16. System Synchronization-Planning and Procedures
System synchronization is the second part of moving data between two systems.
You perform system synchronization if you have purchased a new system, you
have moved all of your data to the new system, and one of the following applies:
v The existing system is a production system that needs to stay up and running
while you convert to the new system.
v You want to test the new system before making the switch over.
Since the existing system is a production system, changes occur to the existing
system which do not get applied to the new system. Therefore after you have
loaded the new system, you need to synchronize the new system and the existing
system before you can put the new system into production.

You can perform system synchronization only if one the following is true:
v The new system and the existing system are at the same release. You have fully
reloaded the new system from the existing system either using “Recovering your
entire system after a complete system loss–Checklist 20” on page 96.
v The new system is at a newer release than the existing system. You have fully
reloaded the new system from the source system using one of the procedures in
“Chapter 15. Release-to-Release Support” on page 323.

The method that you use to synchronize the two systems is the side-by-side
method. The underlying principal of the side-by-side method is that you will run
your existing system and your new system in parallel for a test period. During that
test period, you will periodically perform activities to synchronize the data on your
new system with the data on your existing system. At the end of the test period,
you will perform final synchronization activities before moving your production
work to your new system. When you complete your final synchronization, the
software environment on the two systems should be identical.

The topics that follow discuss several different approaches that you can take for
performing synchronization. In all cases, synchronization requires careful planning
and monitoring. It also requires a good understanding of your applications and the
library structure on your system. Running two systems in parallel also requires
strong change-control practices. This chapter focuses primarily on synchronizing
data.

If possible, during the synchronization period you should carefully limit other
changes on your existing system, such as adding or changing user profiles or
changing the system distribution directory. When this type of change to system
customization occurs on your existing system, you need to manually perform the
same updates on your new system.

You might find the security auditing function helpful for keeping track of changes
to system information on your existing system. If you are not familiar with
security auditing see the iSeries Security Reference book. It describes how to set up
security auditing and which values to choose to get the entries that you need.

© Copyright IBM Corp. 1997, 2000, 2001 351

 You can print the entries in the audit journal receiver and enter the same
information on your new system. The Security ToolKit provides a function to
select, format, and print (or display) audit journal entries.

Figure 31. Overview of Synchronization Process

Figure 31 provides an overview of the synchronization process. The process starts

with building an initial test environment that matches a known point on your
existing system (Point 1). Periodically, you establish a new known point
(synchronization point) on your existing system. Point 2 and Point 3 are examples
of synchronization points. After preserving any work that you have performed on
your new system, such as object conversion, you use one of the synchronization
methods to bring your new system to the same level as your existing system.

While you read and plan, consider how the options for synchronization relate to
your current procedures (such as regular save procedures and change control
procedures). By using your existing procedures as a starting point, you can reduce
the level of disruption and build on your existing base of knowledge. For example,
if you currently use object journaling, then object journaling might be a logical part
of your synchronization strategy. If no one in your organization has experience
with object journaling, then it might not be the best solution for you.

Start with a Valid Test Environment

All of the following synchronization methods assume that you start by
loading an entire copy of your existing system to your new system. To create
this initial test environment, you must follow the sequence in your reload
checklist to ensure that the system correctly merges your user data with
licensed program data. The reload checklist is “Recovering your entire system
after a complete system loss–Checklist 20” on page 96 if you are restoring to
the same release or the procedures in “Chapter 15. Release-to-Release
Support” on page 323 if you are restoring to a higher release.

Synchronization Methods: Overview

Following are the common methods for synchronizing test and existing systems:

352 OS/400 Backup and Recovery V5R1

 Moving changed objects
With this method, you periodically save everything on your existing
system that has changed since your last synchronization point. You then
restore these changed objects to the new machine.
Moving libraries
With this method, you periodically copy entire libraries from your existing
system to your new system. This method works best when your programs
are in separate libraries from your database files. You synchronize the
libraries that contain database files.
Moving individual objects
With this method, you periodically copy specific objects, such as database
files, from your existing system to your new system.
Applying journaled changes
With this method, you journal objects on your existing system. You move
the journal receivers from your existing system to your new system. You
apply the journaled changes to the test objects on your new system. This
method is sometimes used in conjunction with moving changed objects.
Refreshing new system
With this method, you periodically refresh your new system by restoring
an entire copy of your existing system.

Table 60 provides a comparison of these synchronization methods. It also shows

where you can read more about each method. You can use these methods
individually or in combination.
Table 60. Comparison of Synchronization Methods
Time Required for Frequency
Method Complexity Risk Synchronization Used Where to Read More About It
Moving changed High Medium Medium High “Moving Changed Objects”
objects to high
Moving libraries Low Low to Medium Medium “Moving Entire Libraries” on
medium to high page 358
Moving objects Medium Low to Low to medium Medium “Moving Individual Objects” on
to high medium page 360
Applying journaled High High Low Low “Applying Journaled Changes”
changes on page 361
Refreshing new Low Low High Low “Refreshing Your new system”
system on page 363

Moving Changed Objects

With this method, you periodically save everything that has changed since your
last synchronization point. You then restore those changed objects to your new
system. The recommended method when you save changed objects is to specify an
exact reference date and time that corresponds to your last synchronization point.
This ensures that the contents of your save tapes are not affected by any
intermediate save operations that might have occurred since your last
synchronization point.

Chapter 16. System Synchronization-Planning and Procedures 353

 Following is an example of the save and restore procedures when you use this
method. You will need to change these sample steps to meet the needs of your
own situation. This example assumes that the last synchronization point was at
1800 hours (6 p.m.) on July 27, 1998.

Steps for Saving Changed Objects

Perform these steps on your existing system:
1. To avoid any problems with inadequate authority, sign on as the security
officer (QSECOFR).
2. Place your system in a restricted state to ensure that you get a stable copy of
the changed objects on your existing system.
3. Use the Save Security Data (SAVSECDTA) command to save all user profiles.
You use this information to correctly synchronize ownership and authority for
any new objects that you move.
4. To save objects that have changed since your last synchronization point, use
the Save Changed Object (SAVCHGOBJ) command. This command example
saves objects in libraries (the QSYS.LIB file system):
SAVCHGOBJ OBJ(*ALL) LIB(*ALLUSR) DEV(tape-device)
OBJJRN(*YES) REFDATE('07/27/xx') REFTIME(180000)
ACCPTH(*YES)

Note: If you are using the SAVCHGOBJ method in conjunction with applying
journaled changes, specify OBJJRN(*NO).
5. If you have user libraries whose names begin with Q, save the changed
objects in those libraries. Repeat step 4 and substitute the name of your Q
library in place of *ALLUSR.

Note: The online information for the LIB parameter tells which Q libraries are
included when you specify *ALLUSR.
6. To save document library objects that have changed since your last
synchronization point, use the Save Document Library Object (SAVDLO)
command:
SAVDLO DLO(*SEARCH) DEV(tape-device)
REFCHGDATE('07/27/xx') REFCHGTIME(180000)
SRCHTYPE(*ALL) OWNER(*ALL)
7. You cannot just save changed mail. You must save all mail, if necessary. To
save mail, use the Save Document Library Object (SAVDLO) command as
follows:
SAVDLO DLO(*MAIL)
8. To save objects in directories that have changed since your last
synchronization point, do the following:

Note: *LANSERVER and *NETWARE are not supported on V4R5M0 or

later.
a. If you have any network server descriptions (NWSDs), you must vary
them off before starting the save procedure. Use the command
WRKCFGSTS CFGTYPE(*NWS) (Work with Configuration Status) to
display the configured NWSDs on your system.
Select option 2 (Vary off) on this display to vary off the NWSDs.

Note: Alternatively, use the Vary Configuration command to vary off a

NWSD:
VRYCFG CFGOBJ(XXX) CFGTYPE(*NWS) STATUS(*OFF)

354 OS/400 Backup and Recovery V5R1

 b. Use the Save (SAV) command to save changed objects:
SAV DEV('/QSYS.LIB/tape-device.DEVD')
OBJ(('/*' *INCLUDE) ('/QSYS.LIB' *OMIT)
('/QDLS' *OMIT)) CHGPERIOD('07/27/xx' 180000)
UPDHST(*YES)
9. To vary on the network servers, use the WRKNWSSTS command and select
option 14. Use the WRKCFGSTS *NWS command to display all network
servers and select option 1 to vary on any that were previously varied off.
10. Use the Start Subsystem (STRSBS) command to restart the controlling
subsystem to return your system to productive use.
11. To display the log of what changed objects the system saved, use the Display
Job Log (DSPJOBLOG) command.
12. Print the job log and highlight each library that was saved. You will need this
information to restore changed objects.

Steps for Restoring Changed Objects

Test Objects to Preserve

This example assumes that you are not making permanent changes to volatile
objects (such as database files) on your new system. When you restore, you
will be overlaying test objects.

It also assumes that after you build your initial new system, you will not
restore programs from the existing system to the new system during
synchronization (because those programs are already converted on your new
system).

If you need to preserve test objects or if programs change on your existing

system, you need to make special plans for your restore procedures.

To restore the changed objects that you saved, perform these steps on your test
system:

For more information about restoring changed objects, refer to “What Happens
When You Restore Objects” on page 42.
1. To avoid any problems with inadequate authority, sign on as the security
officer (QSECOFR).
2. Place your system in a restricted state.
3. To restore the saved user profiles, use the Restore User Profile (RSTUSRPRF)
command:
RSTUSRPRF USRPRF(*ALL) DEV(tape-device)
ENDOPT(*LEAVE)
4. If your new release is V4R3M0 or a newer release, you can omit this step. If
your test machine has a different serial number, use the Change User Profile
(CHGUSRPRF) command to add *ALLOBJ special authority to user profiles, if
necessary.
5. Locate the printout of the job log from your save operation. Use it to
determine which libraries the system saved. If you do not have the job log,
you can use the Display Tape (DSPTAP) command to display the contents of
the save tapes:
DSPTAP DATA(*SAVRST) OUTPUT(*PRINT)

Chapter 16. System Synchronization-Planning and Procedures 355

 6. For each library on the save tapes, type the following:
RSTOBJ OBJ(*ALL) DEV(tape-device) OBJTYPE(*ALL)
SAVLIB(library-name) ENDOPT(*LEAVE) MBROPT(*ALL)
Notes:
a. For both the QGPL library and the QUSRSYS library, you must specify
MBROPT(*NEW). This prevents overlaying new system files with old system
files.
b. Do not specify ALWOBJDIF(*ALL). Normal restore processing with the
default ALWOBJDIF value protects you from accidentally overlaying
critical information. ALWOBJDIF(*ALL) is intended only when you are
initially loading information from one system to another.
c. If your new system has a different ASP configuration from your existing
system, you might need to specify the SAVASP and RSTASP parameters.
7. To restore document library objects that you saved in step 6 on page 354, type
the following:
RSTDLO DLO(*ALL) DEV(tape-device) ENDOPT(*LEAVE)
Notes:
a. Do not use this restore command unless your save tapes contain only
changed DLOs. If you restore all DLOs from your existing system, you
will overlay IBM-supplied objects that are used for Client Access.
b. This command does not restore mail that has changed. Mail gets restored
in step 8.
c. Changes to calendars are restored when you restore the QUSRSYS library.
d. If you have DLOs in more than one ASP, you need to run the RSTDLO
command for each ASP. You specify the SAVASP and RSTASP parameters.
8. To restore mail that you saved in step 7 on page 354, use the Restore
Document Library Object (RSTDLO) command as follows:
RSTDLO DLO(*MAIL)
9. To restore the changed directory objects that you saved in 8b on page 355, type
the following:
RST DEV('/QSYS.LIB/tape-device.DEVD')
OBJ(('/*' *INCLUDE) ('/QSYS.LIB' *OMIT) ('/QDLS' *OMIT))
10. Use the Display Job Log to print your job log:
DSPJOBLOG OUTPUT(*PRINT)

Review it carefully. Whenever you restore changed objects, you are likely to
encounter situations that you will need to recover manually. If you plan to
synchronize your system several times, you might find it useful to create a log
that describes synchronization problems and their resolutions. This will help
to reduce your synchronization time in the future.

“Problems When Restoring Changed Objects” on page 357 describes common

problems and solutions when restoring changed objects.
11. After you resolve any problems that occurred with restored objects, use the
Restore Authority (RSTAUT) command to restore private authorities.

Note: You should wait to restore authority until after you resolve any
problems because some problem resolution steps might require you to
restore new objects.
12. Restart the controlling subsystem and make the system available for more
testing.

356 OS/400 Backup and Recovery V5R1

Problems When Restoring Changed Objects
Because you specify ALWOBJDIF(*NONE) when you restore changed objects, the
system compares heading information on the system copy of an object with
heading information on the media copy. When the system detects discrepancies
that might indicate mismatches, the system does not restore an object. Following
are common cases where this occurs in a test environment and possible solutions:

Problems Restoring Journal Receivers

If you are journaling objects on both your existing system and your new system,
you might have cases where two different journal receivers exist with the same
name. Usually this occurs because you change journal receivers on both systems.
The systems automatically generate the same next-receiver name.

In most environments, you do not need the information that is in the journal
receivers on your new system. Use the Change Journal (CHGJRN) command to
create and attach a new journal receiver with a unique name. Then you can simply
save and delete the journal receivers that you do not need (on your new system).

Note: This strategy applies when you are using a change-object synchronization
method. If you plan to apply journaled changes to synchronize systems, you
need to devise a method for naming and changing journal receivers that
allows you to successfully restore journal receivers.

For more information about the rules to name, attach, and restore journal
receivers, see “Chapter 19. Planning and Setting Up Journaling” on page 383.

Problems Restoring Database File Members

When you delete and re-create a database file, that database file has different
reference information than the previous version. Therefore, the reference
information for the database file that you attempt to restore will not match the
copy on the new system. When this type of mismatch occurs, the system will not
restore the database file. The same problem occurs when you delete and re-create
individual members within a database file.

Following are several options for handling this problem. Choose the correct option
for your situation. Base your decision on your synchronization requirements and
your application architecture. Always make sure that you have a good backup of
your new system.

Recovery Option 1–Restore entire library: A simple solution is to restore the entire
library from your existing system to your new system. To do this, you will need to
first clear the library on your new system. To use this option, you might need to
change your save strategy. For libraries where you regularly delete and re-create
database files or members, you might not be able to use the SAVCHGOBJ
approach.

Recovery Option 2–Delete files or members before restoring: Another option is to

delete (from the new system) the database file or the database file member that
causes the problem. When you use this method, you must make provisions for any
logical files that are dependent on the files or members that you plan to delete. Do
the following:
1. To determine whether dependent logical files exist, use the Display Database
Relationships (DSPDBR) command on your test system:
DSPDBR FILE(library-name/file-name) MBR(member-name)

Chapter 16. System Synchronization-Planning and Procedures 357

 Note: Specify the member only when you need to delete individual members
instead of the whole file.
2. If no database dependencies exist, continue with step 5.
3. On your existing system, use the SAVOBJ command to save each dependent
file to tape.
4. On your test system, use the Delete File (DLTF) command to delete each
dependent file.
5. On your test system, delete the physical files or file members.
6. From your SAVCHGOBJ tape, use the RSTOBJ command to restore the physical
files or physical file members that you could not previously restore.
7. If you saved dependent files in step 3, use the RSTOBJ command to restore
them.

Recovery Option 3–Use a temporary library: A variation of the previous option is

to use a temporary library on your new system. Do the following:
1. On the test system, use the Create Library (CRTLIB) to create a temporary
library.
2. Restore the physical files or file members from your SAVCHGOBJ tape to the
temporary library. (Use the RSTOBJ command with the SAVLIB and RSTLIB
parameters.)
3. To determine whether dependent logical files exist, use the Display Database
Relationships (DSPDBR) command on your test system:
DSPDBR FILE(library-name/file-name) MBR(member-name)

Note: Specify member only when you need to delete individual members
instead of the whole file.
4. If no database dependencies exist, continue with step 7.
5. On the test system, use the Copy File (CPYF) command to copy dependent
files from the original libraries to the temporary library.
6. Delete the dependent files from the original libraries.
7. Delete the physical files from the original libraries.
8. Copy the physical files from the temporary library to the original libraries.
9. If the temporary library contains any dependent files, copy them to the
original libraries.
10. Use the Delete Library (DLTLIB) command to delete the temporary library.

Problems with Object Authority or Ownership

To protect you from someone who might attempt to restore an illicit program to
your system, the system makes authority or ownership changes during some
restore operations. Review the job log to determine whether these changes
occurred. You can use the Change Object Owner (CHGOBJOWN) command to
transfer ownership to the correct user profile. You can use the Edit Object
Authority (EDTOBJAUT) command to make changes to public authority for an
object.

Moving Entire Libraries

When your library structure is organized to separate volatile information (for
example, database files) from static information (for example, programs), this
synchronization method might be simple and effective. You can periodically restore
entire database file libraries from your existing system to your new system.

358 OS/400 Backup and Recovery V5R1

Note: Do not use this method for IBM-supplied (Qxxx) libraries, particularly the
QGPL library and the QUSRSYS library.

Following is an example of the steps for moving a library:

1. On your existing system, sign on with the QSECOFR user profile to avoid
authority problems.
2. Place your existing system in a restricted state to ensure that you get a fixed
(static) copy of the database files.
3. Use the SAVSECDTA command to save all user profiles. You use this
information to correctly synchronize ownership and authority for any new
objects that you move.
4. Use the Save Library (SAVLIB) command to save the libraries to tape:
SAVLIB LIB(library-name) DEV(tape-device)
ENDOPT(*LEAVE) ACCPTH(*YES)
Notes:
a. Specify ENDOPT(*REWIND) when you save the last library.
b. You can specify multiple libraries on the SAVLIB command.
5. Restart the controlling subsystem on your existing system.
6. On your test system, sign on with the QSECOFR user profile to avoid
authority problems.
7. Place your new system in a restricted state to ensure that you do not have
restore problems because of object-locking conflicts.
8. Use the Clear Library (CLRLIB) for each library that you plan to restore. This
eliminates any potential problems with objects not restoring because of
mismatches between the media version and the system version.

Note: If you restore a library that contains Structured Query Language (SQL)
collections that contain *DTADCT objects, for each of these libraries use
the Delete Library (DLTLIB) command. (Use DLTLIB rather than Clear
Library (CLRLIB). SQL collections that contain *DTADCT objects will
fail during the Restore Library (RSTLIB) operation unless you first
delete the library.
9. To restore the saved user profiles, use the RSTUSRPRF command:
RSTUSRPRF USRPRF(*ALL) DEV(tape-device) ENDOPT(*LEAVE)
10. If your new release is V4R3M0 or a newer release, you can omit this step. If
your test machine has a different serial number, use the Change User Profile
(CHGUSRPRF) command to add *ALLOBJ special authority to user profiles, if
necessary.
11. For each library that you saved, use the Restore Library (RSTLIB) command:
RSTLIB SAVLIB(library-name) DEV(tape-device) MBROPT(*NEW)
ENDOPT(*LEAVE) OPTION(*NEW) ALWOBJDIF(*ALL)
Notes:
a. If you have a different ASP organization on your new system, you might
need to specify the SAVASP and RSTASP parameters.
b. You specify ALWOBJDIF(*ALL) because you may be restoring to a system
with a different serial number. ALWOBJDIF(*ALL) links the authorization
lists back with the objects. You should only specify ALWOBJDIF(*ALL)
when you are restoring to an empty library or the library does not exist on
the system.
c. When you restore the last library, specify ENDOPT(*REWIND) unless you have
additional objects to restore from the tape.

Chapter 16. System Synchronization-Planning and Procedures 359

 12. Check your job log to ensure that the system successfully restored the
libraries.
13. Use the RSTAUT command to restore private authorities to objects.

Considerations for Moving Entire Libraries

Following are some considerations when you use this synchronization method:
v You might want to combine this method with the SAVCHGOBJ method. For
example, you might move several large libraries that contain database files. You
can use the SAVCHGOBJ command for other libraries (by using the OMITLIB
parameter on the SAVCHGOBJ command to omit libraries that you are moving
in their entirety). SQL collections containing *DTADCT objects will fail during
the Restore Library (RSTLIB) operation unless you first delete the library.
v When you use this method, you need to decide how to handles DLOs and
objects in directories. You might use a save-changed method for those objects.
Or, you might consider moving entire folders or directories if that method will
work with your folder and directory organization.
v In all cases, it is essential that you have a complete copy of your existing system
on tape before you cut over to your new system. This provides for recovery if
you discover you have neglected to synchronize some critical objects.

Moving Individual Objects

With this method, you periodically copy specific objects (or example, database
files) from your existing system to your new system. This method is most often
used in two situations:
v When you have a short test period, careful change control, and a very
well-defined set of database files that change frequently.
v When you plan to completely rebuild your new system at the end of the test
period. In this case, you might periodically move individual objects to create a
more current set of test data on your new system.

Following is an example of the procedure for moving individual objects:

1. On your existing system, sign on with the QSECOFR user profile to avoid
authority problems.
2. Place your existing system in a restricted state to ensure that you get a fixed
(static) copy of the database files.
3. Use the SAVSECDTA command to save all user profiles. You use this
information to correctly synchronize ownership and authority for any new
objects that you move.
4. Use the SAVOBJ command to save individual objects that you want to
synchronize:
SAVOBJ OBJ(object-name)
LIB(library-name)
OBJTYPE(object-type)
DEV(tape-device) ENDOPT(*LEAVE)
Notes:
a. Specify ENDOPT(*REWIND) for the last object.
b. On the same SAVOBJ command, you can save multiple objects of the same
type from the same library.
5. Restart the controlling subsystem on your existing system.
6. Place your new system in a restricted state.

360 OS/400 Backup and Recovery V5R1

 7. On the new system, use the RSTUSRPRF command:
RSTUSRPRF USRPRF(*ALL) DEV(tape-device)
ENDOPT(*LEAVE)
8. If your new release is V4R3M0 or a newer release, you can omit this step. If
your test machine has a different serial number, use the Change User Profile
(CHGUSRPRF) command to add *ALLOBJ special authority to user profiles, if
necessary.
9. For each library that contains objects that you saved, use the RSTOBJ
command:
RSTOBJ OBJ(*ALL) SAVLIB(library-name)
DEV(tape-device) ENDOPT(*LEAVE) OBJTYPE(*ALL)
Notes:
a. If you have a different ASP organization on your new system, you might
need to specify the SAVASP and RSTASP parameters.
b. When you restore the last object, specify ENDOPT(*REWIND).
10. Check your job log to ensure that the system successfully restored the objects.
11. Use the RSTAUT command to restore private authorities to objects.
12. Restart the controlling subsystem on your new system.

Applying Journaled Changes

With this synchronization method, you use server journaling capabilities to
synchronize the information in journaled objects on your test and existing systems.
This is most commonly used either by installations that already use and
understand journaling or by installations that have very large journaled objects.

The advantage of this method is that you save and restore only the changes that
occur to a journaled object, not the entire object. The disadvantage of this method
is its complexity. See “Chapter 19. Planning and Setting Up Journaling” on
page 383 and “Chapter 20. Working with Journal Entries, Journals, and Journal
Receivers” on page 427 for more information about journaling.

Journal Before Saving

You must set up journaling on your existing system before you perform the
save operation from which you build your new systems. When you start
journaling for an object, the system assigns a journal ID number (JID) to that
object. An object must have a JID before you save it from your existing
system and restore it to your new system or you will not be able to apply
journaled changes to the test version. If a journaled object is an IFS object, the
JID is the same as the object’s file identifier (FID). For more information about
restoring journaled objects see “Restoring Objects That Are Journaled” on
page 235.

Conceptually, when you use journaling to synchronize objects, you do the

following to establish a synchronization point:
1. On your existing system, do the following:
a. Save security data to get a current copy of user profiles and private
authorities.
b. Save new journal receivers (that contain entries that you have not yet
applied on your new system).
c. Save any new objects (that do not yet exist on your new system).

Chapter 16. System Synchronization-Planning and Procedures 361

 2. On your new system, do the following:
a. Restore user profiles (to get any new profiles and current authority
information).
b. Restore any new objects.
c. Restore journal receivers.
d. Apply journaled changes from the journal receivers.
e. Restore authority to any new objects.

Following is an example of the steps for using journaled changes to synchronize

systems:
1. To establish a checkpoint on your existing system, do the following:
a. Use the Work with Journal Attributes (WRKJRNA) command to determine
what journal receivers have been created since your last checkpoint.
b. Write down the names of the new journal receivers.
c. Determine whether any members, files, or data areas have been added to
any journaled objects since your last checkpoint. You can use the DSPJRN
command to do this by typing the following:
DSPJRN JRN(journal-name) JRNCDE((D) (F) (E)) ENTTYP(JF JM EG)
RCVRNG(first-receiver *CURRENT)

For first-receiver, use the name of the first receiver after the most recent
checkpoint.

Note: If you are journaling IFS objects, and your directories are not using
the inherit journaling attribute, look for new IFS objects by adding B
to the JRNCDE parameter, and JT to the ENTTYP parameter.
d. Write the new object names on a list. (You will need to save them later.)
e. If you have other journals on your system, repeat step 1a through step 1c
for each additional journal.
f. For each journal on your system, use the CHGJRN command to detach the
current journal receivers and attach new journal receivers.
g. Use the SAVOBJ command or SAV command to save any newly journaled
objects that you listed in step 1d and step 1c. If the new objects are new file
members, be sure to save only the new members, not the entire files.

Note: The system needs an exclusive lock on an object to save it. You might
need to stop certain application activity on your system to be able to
save the newly journaled objects.
h. Use the SAVOBJ command to save the journal receivers that you listed in
step 1b.
i. If you do not have a current copy of your user profiles on tape, use the
SAVSECDTA command to save them to tape.
j. You have completed establishing a new checkpoint (such as Point 2) on your
existing system.
2. To synchronize the journaled objects on your new system with the existing
versions, do the following:
a. Place your new system in a restricted state.
b. On the new system, use the RSTUSRPRF command:
RSTUSRPRF USRPRF(*ALL) DEV(tape-device)
ENDOPT(*LEAVE)

362 OS/400 Backup and Recovery V5R1

 c. If your new release is V4R3M0 or a newer release, you can omit this step. If
your test machine has a different serial number, use the Change User Profile
(CHGUSRPRF) command to add *ALLOBJ special authority to user profiles,
if necessary.
d. Use the RSTOBJ or RST command to restore any objects that you saved in
step 1g on page 362.
e. Use the RSTOBJ command to restore the journal receivers that you saved in
step 1h on page 362.
f. Use the Apply Journaled Changes (APYJRNCHG) command to bring your
journaled objects to the checkpoint level:
v For the starting receiver, specify the journal receiver that was created and
attached when you established your most recent checkpoint on the
existing system. For the ending receiver, specify *CURRENT.
v For the starting sequence number, specify *FIRST.
v For the ending sequence number, specify *LAST.

Be sure to review the information in “Chapter 20. Working with Journal

Entries, Journals, and Journal Receivers” on page 427 to understand what
journal entries may either be skipped or may stop the apply process.
g. Use the RSTAUT command to restore private authorities for any new
objects that you restored.
h. Restart the controlling subsystem on your new system.

Refreshing Your new system

When you use this method, you periodically rebuild your entire new system from
the current information on your existing system. To ensure that user data and
system data are merged correctly, you must follow the same procedure that you
used when you initially built your new system:
1. Install the Licensed Internal Code from scratch.
2. Install OS/400.
3. Restore user data.
4. Install licensed programs.

Note: The preceding list is a conceptual view of the sequence. Use the checklists
for the complete list of steps.

Before refreshing your new system, be sure to save the work that you have already
performed on your new system. In particular, save any program objects that you
have converted. After you have rebuilt the new system, restore these converted
objects.

Additional Synchronization Tips

Following are some additional considerations when you plan to keep your test and
existing systems synchronized during a test period:
v Synchronization (especially the first few times) can be difficult. You should
always save your entire new system before you begin your synchronization
efforts. With these save tapes available, you have the option of restoring your
entire new system and starting the synchronization again (or changing
synchronization methods).

Chapter 16. System Synchronization-Planning and Procedures 363

 v To synchronize correctly, you need to understand how to save and restore the
authority to objects. When you use the SAVSECDTA command, the system saves
user profiles, authorization lists, and private authorities.
When you restore user profiles (RSTUSRPRF *ALL), the system restores user
profiles and authorization lists. It puts private authority information in work
files on the system. After you restore objects, use the RSTAUT command to
re-establish the private authorities that are in the work files.
v When you are ready to perform your final synchronization before you move
your new system to productive use, be sure to plan both for disaster recovery
and for verification. If possible, keep your former existing system functional
during the verification period in case you discover objects that you did not
properly synchronize.
In addition, make sure that you save an entire copy of your former existing
system to tape before you dismantle it. You might need the objects on these
tapes if you discover errors in your synchronization procedures.
Finally, print listings from your former existing system that you can use as a
basis for verifying the validity of information on your new existing system.
v Good synchronization requires careful change-control procedures. You must plan
for information that you cannot easily synchronize automatically, such as
changes to system information. As much as possible, you should freeze changes
to this type of information during your test period.
v If you do not use the SAVCHGOBJ command as part of your synchronization
strategy, you might need to use special procedures to synchronize calendars and
mail.
Following are the steps for moving calendars from your existing system to your
new system:
1. On your existing system, use the Save Calendar (SAVCAL) command:
SAVCAL CAL(*ALL *ALL) DEV(tape-device)
DATE('07/27/xx' *END)

Note: Option 34 from the Save menu runs the same command.
2. On your new system, use the Restore Calendar (RSTCAL) command:
RSTCAL CAL(*ALL *ALL) DEV(tape-device)
SAVLIB(QTEMP) LABEL(QCAL*ALL*ALL)

Note: Option 33 from the Restore menu runs the same command.

Following are the steps for moving mail from your existing system to your new
system:
1. On your existing system, use the SAVDLO command:
SAVDLO DLO(*MAIL) DEV(tape-device)
2. On your new system, use the RSTDLO command:
RSTDLO DLO(*MAIL) DEV(tape-device)
v To synchronize the BRMS licensed program, do the following:

Note: Use the following tip only for BRMS installations that do not share media
information with other systems.
1. On your existing system, stop all activity that might place locks on objects in
the BRMS libraries. If you have scheduled jobs that use BRMS, you need to
hold them.
2. Mount a tape that is compatible with the tape unit on your new system.
3. Type the following:

364 OS/400 Backup and Recovery V5R1

 SAVLIB LIB(QBRM QUSRBRM) DEV(tape-device)

Note: If you want, you can use save files and transfer the libraries
electronically.
4. On the new system, do the following:
a. Stop all activity that might place locks on objects in the BRMS libraries. If
you have scheduled jobs that use BRMS, you need to hold them.
b. Save a copy of the current BRMS product; enter the following command:
SAVLICPGM LICPGM(57nnBR1) DEV(tape-device)

(Replace nn with the appropriate number for your release,

for example, DSPPTF 5763BR1 for V3R2.)
c. Delete the version of BRMS that has outdated information; enter the
following command:
DLTLICPGM LICPGM(57nnBR1)
d. Mount the tape that you created in step 3 on page 364.
e. Restore the BRMS libraries; enter the following command:
RSTLIB SAVLIB(QBRM QUSRBRM) DEV(tape-device)
f. Load the tape that you created using SAVLICPGM in step 4b.
g. Restore the current version of BRMS; enter the following command:
RSTLICPGM LICPGM(57nnBR1) DEV(tape-device)
h. To set up BRMS again, consult Backup Recovery and Media Services for
iSeries.

Chapter 16. System Synchronization-Planning and Procedures 365

366 OS/400 Backup and Recovery V5R1
Part 5. Considerations for Merging Two or More Systems
Chapter 17. Tips for Merging Two RISC
Systems Into a Single RISC System . . . . . 369
Guidelines for Restoring Information from the
Development System . . . . . . . . . . . 369

© Copyright IBM Corp. 1997, 2000, 2001 367

368 OS/400 Backup and Recovery V5R1
 Chapter 17. Tips for Merging Two RISC Systems Into a Single
RISC System
Merging systems is a complex process. Various factors can affect merging systems
that include the software release, system names, or IBM-supplied objects. The
following information is provided as a guideline to assist you in the planning
process of merging systems. You should consult the documentation for your
additional products to ensure proper migration.

Objects that can migrate from one system only are:

v Directory entries
v Office enrollments
v Distribution lists
v Other IBM-supplied files or libraries
v System values
v Network attributes
v Access path recovery times
v Communications configurations
v System reply list
v Edit descriptions

Determine which system to restore all of these objects from and restore that system
first. If you have a production system and a development system, restore the
production system first, then follow the guidelines below to restore the information
from the development system.

Guidelines for Restoring Information from the Development System

1. Determine which system to restore first. To assist in determining this, you may
need to answer several questions: Which system is more complex? Which has
more users? Which system is more critical to your operations? If you are
choosing between merging a production system and merging a development
system, selecting the production system is recommended.
Restore the production system by following the steps in Table 27 on page 97.
2. User profiles and associated objects that are the same on both systems will not
be merged. However, they are affected in the following manner:
v Object owners, authorization lists, and primary groups will be restored from
the production system.
v Passwords and group connections will be restored from the development
system.
v Merging of private authorities is an AND operation. The object authorities
and data authorities will be added together from both systems. The resulting
authorities on the merged system will be the higher of the matching
authorities from the production and development systems.
| v URSRPRF (*NEW) amd OMITUSRPRF are parameters that may be useful
| when you consolidate systems. They allow you to restore only new user
| profiles or omit certain user profiles. Refer to “Restoring User Profiles” on
| page 214 for more information.

© Copyright IBM Corp. 1997, 2000, 2001 369

 3. Groups of configurations that are needed from the development system may be
restored with the Restore Configuration (RSTCFG) command:
RSTCFG OBJ(workstation) OBJTYPE(*DEVD)
SRM(*NONE)

Automatic configuration may also be enabled to recover the groups of

configurations from the development system.
4. User libraries may be restored with the Restore Library (RSTLIB) command.

Note: Be sure to omit any IBM-supplied libraries such as QGPL and QUSRSYS.
If there are libraries that are the same on both systems, you should consider
using the OPTION(*NEW) parameter to restore only the new objects:
RSTLIB SAVLIB(User library) OPTION(*NEW)

Then determine which objects you want from each system and restore those
objects individually. If there are objects in QGPL or QUSRSYS that are unique
to either system, those objects should be restored individually as well.
5. Documents and folders may be restored with the RSTDLO command. When
saving documents and folders to be restored, any IBM-supplied folders should
be omitted when using the SAVDLO command:
SAVDLO DLO(*ALL) OMITFLR(Q*)

If any IBM-supplied folders are restored, the original information may be

overwritten. Additional considerations will need to be made if any of the DLOs
are from a previous release.
6. The Integrated File System (IFS) may be restored with the following command:
RST OPTION(*NEW)
7. Once the preceding instructions have been completed, run the Restore
Authorities (RSTAUT) command.
8. Once the RSTAUT command completes, perform a normal IPL.

370 OS/400 Backup and Recovery V5R1

 Part 6. Journaling and Commitment Control
Chapter 18. Protecting Access Paths Using Assigning a Journal to an Auxiliary Storage
System-Managed Access-Path Protection . . . 375 Pool . . . . . . . . . . . . . . 403
| Using SMAPP with Independent ASPs . . . . . 375 Associating the Journal with a Journal
Benefits of SMAPP . . . . . . . . . . . 376 Receiver . . . . . . . . . . . . . 404
How SMAPP Works with Access Path Journaling 376 Specifying Where Journal Messages Are Sent 404
How the System Chooses Which Access Paths to Specifying How Journal Receivers for the
Protect . . . . . . . . . . . . . . . 377 Journal Are Changed . . . . . . . . . 405
How SMAPP Affects System Performance and Having the System Delete Journal Receivers
Auxiliary Storage Use . . . . . . . . . . 377 for the Journal . . . . . . . . . . . 406
Working with Recovery Times for Access Paths . . 378 Specifying Receiver Size Options . . . . . 407
How SMAPP Handles Changes in the | Specifying Minimized Entry-Specific Data 409
Configuration of Auxiliary Storage Pools . . . . 380 Creating Journals and Journal
Receivers–Examples . . . . . . . . . . 409
Chapter 19. Planning and Setting Up Journaling 383 Starting Journaling . . . . . . . . . . . 411
Journal Management Commands and APIs . . . 385 How to Start Journaling for Physical Files . . . 411
Commands and APIs for Journals . . . . . 385 How to Start Journaling for DB2® Multisystem
Commands and APIs for Journal Receivers . . 386 Files . . . . . . . . . . . . . . . 412
Commands and APIs for Journal Entries . . . 386 How to Start Journaling for Access Paths . . . 412
Commands and APIs for Starting and Ending | How to Start Journaling for IFS objects . . . . 413
Journaling of an Object . . . . . . . . . 387 | How to Start Journaling for Data Areas and
| Commands and APIs for Getting Information | Data Queues . . . . . . . . . . . . 414
| about Journaled Objects . . . . . . . . . 387 Why You Must Save Objects after Starting
Commands for Recovering an Object by Using Journaling . . . . . . . . . . . . . 414
the Journaled Changes . . . . . . . . . 388 Tasks for Managing Your Journaling Environment 415
Planning Journal Management. . . . . . . . 388 Keep Records of Journaled Objects . . . . . 415
| What Objects Should You Journal? . . . . . 389 Evaluate How System Changes Affect
Should You Journal Objects That the System Journaling . . . . . . . . . . . . . 416
Does Not Journal? . . . . . . . . . . . 390 How to Manage Journals and Journal Receivers 416
Should You Use Access Path Journaling? . . . 391 Changing Journal Receivers . . . . . . 416
How Should Objects Be Assigned to Journals? 391 Changing the Journal State . . . . . . . 418
Should You Journal Before-Images? . . . . . 392 Deleting Journal Receivers . . . . . . . 418
Should Your Journal Receivers Be in a User Deleting a Journal . . . . . . . . . . 419
ASP? . . . . . . . . . . . . . . . 393 Saving Journals and Journal Receivers . . . 420
Journaling with the Save-While-Active Function 394 Correct Order for Restoring Objects
Journaling and System Performance . . . . . . 394 Associated with a Journal . . . . . . . 422
Journaling and Auxiliary Storage . . . . . . . 395 Ending Journaling . . . . . . . . . . . . 423
How to Estimate the Size of a Journal Receiver 396 How to End Journaling for DB2 Multisystem
Estimating the Size of a Journal Files . . . . . . . . . . . . . . . 423
Receiver Method 1. . . . . . . . . . 396 Managing IBM-Supplied Journals. . . . . . . 424
Estimating the Size of a Journal
Receiver–Method 2 . . . . . . . . . 396 Chapter 20. Working with Journal Entries,
Journaling Features That Increase the Journal Journals, and Journal Receivers . . . . . . 427
Receiver Size . . . . . . . . . . . . 397 About Journal Entries . . . . . . . . . . 427
How You Can Reduce the Storage Used by Contents of a Journal Entry. . . . . . . . 428
Journaling . . . . . . . . . . . . 397 Sending Your Own Journal Entries . . . . . . 429
Setting Up Journaling . . . . . . . . . . 398 Displaying and Printing Journal Entries . . . . 429
How Journal Receivers Are Created . . . . . 399 Output for Journal Entries Directed to a Work
Naming Journal Receivers . . . . . . . 399 Station . . . . . . . . . . . . . . 430
Assigning a Journal Receiver to a Library 401 Output for Journal Entries Directed to a
Assigning a Journal Receiver to an Auxiliary Database Output File . . . . . . . . . . 431
Storage Pool . . . . . . . . . . . . 401 Format of Database Output Files . . . . . 431
Specifying a Threshold for a Journal Receiver 401 Processing Journal Entry Data . . . . . . 432
Authority to the Journal Receiver. . . . . 402 Analyzing Your Journal Activity . . . . . 432
How Journals Are Created . . . . . . . . 402 Using the Receive Journal Entry Command . . . 432
Naming Journals . . . . . . . . . . 403 Exit Program for Receiving Journal Entries . . 433
Assigning a Journal to a Library . . . . . 403 Requesting Block Mode . . . . . . . . 435

© Copyright IBM Corp. 1997, 2000, 2001 371

 Using the Retrieve Journal Entry Command . . . 436 Remote Journal Type Descriptions . . . . 476
Using the Retrieve Journal Entries Add Remote Journal Processing . . . . . 477
(QjoRetrieveJournalEntries) API . . . . . . . 437 Journal Attributes for Remote Journals . . . 478
Working With Pointers in Journal Entries . . . . 437 Preventing Journal Receiver Deletion . . . 479
| Working with entries which contain minimized Activating and Inactivating a Remote Journal 479
| entry-specific data . . . . . . . . . . . . 438 Activating the Replication of Journal Entries
| Data Area Considerations . . . . . . . . 439 to a Remote Journal . . . . . . . . . 480
| Database Physical File Considerations . . . . 439 Relational Database Considerations . . . . 485
Using Journaling to Provide an Audit Trail . . . 439 Inactivating the Replication of Journal Entries
Comparing Journal Images . . . . . . . . . 440 to a Remote Journal . . . . . . . . . 485
Journal Receiver Chains . . . . . . . . . . 440 Journal Receivers Associated with a Remote
Using the Work with Journal Attributes Command 442 Journal . . . . . . . . . . . . . 485
Working with the Receiver Directory . . . . 442 Attaching a New Receiver to a Remote
Inoperable Journal Receivers . . . . . . . 444 Journal . . . . . . . . . . . . . 486
Applying and Removing Journaled Changes . . . 444 Activating and Inactivating a Local Journal . . 487
How Applying and Removing Journaled Activating a Local Journal . . . . . . . 487
Changes Works with Referential Constraints . . 445 Inactivating a Local Journal . . . . . . 487
How Applying and Removing Journaled Summary of the Journal Type, Delivery Mode,
Changes Works with Trigger Programs . . . . 446 and Journal State Interactions . . . . . . . 488
Applying Journaled Changes . . . . . . . 446 Removing a Remote Journal . . . . . . . 490
| Integrated File System Considerations . . . 448 Troubleshooting Errors . . . . . . . . . 491
Apply Journaled Changes (APYJRNCHG) Performance Considerations . . . . . . . . 492
Command Examples . . . . . . . . . . 448 Auxiliary Storage Considerations . . . . . . . 494
Removing Journaled Changes . . . . . . . 450 Main Storage Considerations . . . . . . . . 494
Remove Journaled Changes (RMVJRNCHG) Example Remote Journal Function Environments 494
Command Examples . . . . . . . . . . 451 Data Replication Environment . . . . . . . 494
Actions of the APYJRNCHG or RMVJRNCHG Establishing the Data Replication
Command by Journal Code. . . . . . . . 452 Environment . . . . . . . . . . . 495
Work with Journal (WRKJRN) Command Options 455 Normal Run-Time Environment for the Data
Recovery Options . . . . . . . . . . . 456 Replication Environment . . . . . . . 496
Work with Forward Recovery . . . . . . 456 Data Replication Recovery Given a Failure of
Work with Backout Recovery . . . . . . 458 System S1 . . . . . . . . . . . . 497
Display Journal Status . . . . . . . . 459 Hot-Backup Environment . . . . . . . . 497
Recover Damaged Journal . . . . . . . 460 Establishing the Hot-Backup Environment 498
Recover Damaged Journal Receivers. . . . 461 Normal Run-Time Environment for the
Associate Receivers with Journal . . . . . 462 Hot-Backup Environment . . . . . . . 498
Recovery of a Journaled Object Using Journaled Hot-Backup Recovery Given a Failure of
Changes . . . . . . . . . . . . . . 462 System S1 . . . . . . . . . . . . 498
Recovery after Abnormal System End . . . . 462 Displaying Remote Journal Function Information 499
Recovering When a Journal Is Damaged . . . 464 Managing the Remote Journal Function . . . . 499
Recovering When a Journal Receiver Is Keeping Records of Your Remote Journal
Damaged . . . . . . . . . . . . . . 465 Network . . . . . . . . . . . . . . 499
Evaluating How System Changes Affect Your
Chapter 21. Remote Journal Function . . . . 467 Remote Journal Network . . . . . . . . 499
Introduction to the Remote Journal Function . . . 467 Maintaining Journal Receivers . . . . . . . 500
Benefits of the Remote Journal Function . . . 467 Save and Restore Considerations . . . . . . 500
Remote Journal Function Configuration Journal Considerations . . . . . . . . 500
Examples . . . . . . . . . . . . . . 468 Journal Receiver Considerations . . . . . 501
Supported Communications Protocols . . . . 470 Journaled Object Considerations . . . . . 504
Release to Release Considerations . . . . . 471 Save Storage (SAVSTG) Considerations . . . 505
Planning for the Remote Journal Function . . . . 472 Working with Journal Entries and the Remote
Determining Which Journals are Good Journal Function . . . . . . . . . . . . 505
Candidates for Remote Journals . . . . . . 472 Retrieving Journal Entries from a Remote
Determining Which Communications Protocol Journal with Library Redirection . . . . . . 506
and Delivery Mode to Use . . . . . . . . 472 System Dependent Information . . . . . . 506
How Do the Journal Attributes Affect the Retrieving Journal Entries from a Remote
Remote Journal Function? . . . . . . . . 473 Journal During the Catch-up Phase . . . . . 506
Setting up the Remote Journal Function . . . . 473 Retrieving Journal Entries with Commitment
Preparing to use Remote Journal Function . . . 473 Control Considerations . . . . . . . . . 507
Adding a Remote Journal . . . . . . . . 474 Confirmed and Unconfirmed Journal Entries 508
Using Library Redirection . . . . . . . 475 IPL Considerations . . . . . . . . . . . 509

372 OS/400 Backup and Recovery V5R1

 Main Storage Considerations during IPL . . . 510 Commitment Control during Normal Routing Step
Journal Receiver ASP Considerations . . . . . 511 End . . . . . . . . . . . . . . . . 581
Example: Remote Journal Function Recovery . . . 511 Commitment Control during Abnormal System or
Example Scenario Description . . . . . . . 513 Job End . . . . . . . . . . . . . . . 581
Commitment Control during a Save-While-Active
Chapter 22. Commitment Control . . . . . . 525 Operation . . . . . . . . . . . . . . 583
Commitment Control Introduction . . . . . . 525 Commitment Control Considerations and
Terms Used with Commitment Control . . . . . 526 Restrictions . . . . . . . . . . . . . . 584
Commitment Control Startup . . . . . . . . 527 The Size of a Transaction . . . . . . . . 584
Lock-Level Parameter . . . . . . . . . 527 Record Locking. . . . . . . . . . . . 585
Notify Object Parameter . . . . . . . . . 530 Minimizing Locks . . . . . . . . . . . 586
Updates Made to the Notify Object . . . . 531 Commitment Control for Batch Applications . . 588
Commit Scope Parameter . . . . . . . . 533 Performance Considerations for Commitment
Commitment Definition . . . . . . . . 533 Control . . . . . . . . . . . . . . 588
Scope for a Commitment Definition . . . . 534 Miscellaneous Considerations and Restrictions
Commitment Definition Names . . . . . 536 for Commitment Control . . . . . . . . 591
Jobs with Multiple Commitment Definitions Commitment Control Errors . . . . . . . . 592
Example . . . . . . . . . . . . . 537 Error Conditions . . . . . . . . . . . 592
Commit Text Parameter . . . . . . . . . 539 Non-Error Conditions . . . . . . . . . 594
Default Journal Parameter . . . . . . . . 539 Monitoring for Errors after a CALL Command 595
Omit Journal Entries Parameter . . . . . . 540 Error Messages to Monitor for Commitment
Two-Phase Commit–Overview. . . . . . . . 540 Control . . . . . . . . . . . . . . 595
Placing Resources Under Commitment Control . . 541 Normal Commit or Rollback Processing . . . 598
Types of Committable Resources . . . . . . 541 Commit or Rollback Processing During Job End 599
Location of Committable Resources . . . . . 543 Commit or Rollback Processing During IPL . . 599
The Commit Protocol of a Committable Example of Using Commitment Control . . . . 599
Resource . . . . . . . . . . . . . . 544 Example of Transaction Logging File . . . . . 602
The Access Intent of a Committable Resource 544 Starting Application Programs Using a Notify
Files Being Journaled under Commitment Control 545 Object . . . . . . . . . . . . . . . . 608
Sequence of Journal Entries Under Commitment Using a Unique Notify Object for Each Program 609
Control . . . . . . . . . . . . . . . 545 Using a Single Notify Object for All Programs 614
Commit Cycle Identifier . . . . . . . . . . 547 Using a Standard Processing Program . . . . 614
Commit and Rollback Operations . . . . . . 547 Processing Flow . . . . . . . . . . 615
Roles in Commit Processing . . . . . . . 548 Application Program Example. . . . . . 615
How the System Performs a Commit Operation 549 Standard Commit Processing Program . . . 618
How the System Performs a Rollback Operation 550 Deciding If It Is Necessary to Start Again . . 621
Flow of Commit Processing . . . . . . . 551 Commitment Control Practice Problem . . . . . 622
Errors During Commit Processing . . . . . 554 Steps Associated with Logic Flow . . . . . 631
Changing the Commitment Definition to Not
Wait for Outcome . . . . . . . . . . . 554
Changing the Commitment Definition to Not
Select a Last Agent . . . . . . . . . . 557
Changing the Commitment Definition to
Indicate OK to Leave Out . . . . . . . . 558
Changing the Commitment Definition to Allow
Vote Read-Only . . . . . . . . . . . 560
Vote Reliable Affect on Flow of Commit
Processing . . . . . . . . . . . . . 562
Retrieving Commitment Control Information . . . 564
States of the Logical Unit of Work (LUW) . . . . 564
XA Transaction Support . . . . . . . . . . 566
Commitment Control Recovery during Initial
Program Load . . . . . . . . . . . . . 569
Making Heuristic Decisions and Cancelling
Resynchronization . . . . . . . . . . . . 570
Implicit Commit and Rollback Operations . . . . 573
Commitment Control Status . . . . . . . . 576
End Commitment Control (ENDCMTCTL)
Command . . . . . . . . . . . . . . 579
Commitment Control during Activation Group End 580

Part 6. Journaling and Commitment Control 373

374 OS/400 Backup and Recovery V5R1
 Chapter 18. Protecting Access Paths Using System-Managed
Access-Path Protection
This chapter describes system-managed access-path protection (SMAPP):
v Its benefits
v How it relates to explicitly journaling access paths
v How to determine and change the default protection levels

An access path describes the order in which records in a database file are
processed. A file can have multiple access paths, if different programs need to see
the records in different sequences.

When your system ends abnormally because of something like a power

interruption, the next initial program load (IPL) can take much longer than a
| normal IPL. Also, if you are using independent ASPs the next vary on of an
| independent ASP can take much longer than a normal vary on. Rebuilding access
| paths contributes to this long IPL time or the long vary on time for an independent
| ASP. When the system restarts after an abnormal end the server rebuilds access
| paths that were open for updating at the time the server ended. Likewise, when
| you vary on an independent ASP, the server rebuilds access paths that were open
| for updating at the time the independent ASP was varied off. The system does not
| rebuild access paths that are specified as MAINT(*REBLD) when you create them.

| The purpose of system-managed access-path protection (SMAPP) is to reduce the

| amount of time it takes to perform an IPL, or vary on of an independent ASP, after
| an abnormal system end. The system uses information that it has collected to bring
access paths up to date, rather than rebuilding them.

You can specify the target time for rebuilding access paths after the system ends
abnormally. The target time is a goal that the system does its best to achieve. The
actual recovery time for access paths after a specific failure may be somewhat more
or less than this target.

The target recovery time for access paths can be specified for the entire system or
for individual auxiliary storage pools. The system dynamically selects which access
paths to protect to meet this target. It periodically estimates how long it will take
to recover access paths that are open for change.

SMAPP is available beginning with V3R1 of the Operating System/400* (OS/400*)

licensed program. For new systems, the system-wide recovery time for access paths
| is 90 minutes, which is the default. If you move from a release that does not
| provide the SMAPP function to a release that does supports SMAPP, the
| system-wide recovery time for access paths is also set to 90 minutes. Otherwise,
the recovery times remain as you have previously set them.

| Using SMAPP with Independent ASPs

| You can use SMAPP to protect access paths for independent ASPs. When you use
| SMAPP to protect access paths in independent ASPs, you can specify the recovery
| time individually for each independent ASP. This improves the perfomance when
| you vary on your independent ASP after an abnormal vary off. The recovery time
| that you specify moves with the independent ASP if you switch it between

© Copyright IBM Corp. 1997, 2000, 2001 375

| systems. Therefore if you are switching your independent ASPs between systems,
| you only need to specify the recovery time once. The only time the specified
| recovery time is not moved is when the system you are moving the independent
| ASP to has its system recovery time specified as *OFF. In this case the independent
| ASP recovery time is set to *NONE when the independent ASP is varied on.

Benefits of SMAPP
| SMAPP can greatly reduce the amount of time it takes to perform an IPL, or to
| vary on an independent ASP, after an abnormal end. It is an automatic function
that runs without attention. SMAPP determines which access paths to protect
without any intervention by the user. It adjusts to changes in the environment,
such as the addition of new applications or new hardware.

SMAPP does not require any setup. You do not have to change your applications.
You do not have to journal any physical files or even use journaling at all. You
simply need to determine your policy for access path recovery:
| v How long you can afford to spend rebuilding access paths during an IPL, or
| during the vary on of an independent ASP, after a failure.
v How to balance access path protection with other demands on system resources.
v Whether to have different target times for recovering access paths for different
ASPs.
You may need to experiment with different target recovery times for access paths
to achieve the correct balance for your system. If you configure additional user
ASPs, you should also evaluate your access path recovery times.

SMAPP creates and manages its own internal journals and journal receivers. You
cannot use these journals and journal receivers for any other journaling functions.
They do not appear on the Work with Journals display. You do not need to, nor
can you, save them to save media.

SMAPP requires some additional auxiliary storage for journal receivers. However,
SMAPP is designed to keep the additional disk usage to a minimum. SMAPP
manages journal receivers and removes them from the system as soon as they are
no longer needed.

How SMAPP Works with Access Path Journaling

You may choose to journal some access paths yourself by using the Start
Journaling Access Path (STRJRNAP) command. This is called explicit journaling.
“Should You Use Access Path Journaling?” on page 391 describes how to do this.
To journal an access path explicitly, you must first journal all the underlying
physical files. SMAPP does not require that the underlying physical files be
journaled.

The reason for choosing to journal an access path explicitly is that you consider the
access path (and the underlying files) absolutely critical. You want to make sure
that the files are available as soon as possible when the system is started after an
abnormal end.

Under SMAPP, the system looks at all access paths to determine how it can meet
the specified target times for recovering access paths. It may not choose to protect
an access path that you consider critical.

376 OS/400 Backup and Recovery V5R1

 When the system determines how to meet the target times for recovering access
paths, it considers only access paths that are not explicitly journaled.

How the System Chooses Which Access Paths to Protect

An access path is exposed when the access path has changed because records have
been added or deleted or because a key field has changed, and those changes have
not yet been written to the disk. The system periodically examines access path
exposure and estimates how long it would take to rebuild all the exposed access
paths. If the rebuild time exceeds your target recovery times for access paths, the
system selects additional access paths for protection. The system may also remove
access paths from protection if the estimated time for rebuilding access paths
consistently falls below your target recovery times for access paths. The recover
attribute of a file is not used in determining whether to protect access paths.

Some access paths are not eligible for protection by SMAPP:

v A file that specifies MAINT(*REBLD).
v An access path that is already explicitly journaled.
v An access path in the QTEMP library.
v An access path whose underlying physical files are journaled to different
journals.
v An access path for a physical file that was created specifying
FRCACCPTH(YES).
v Any encoded vector access path.

How SMAPP Affects System Performance and Auxiliary Storage Use

SMAPP has some impact on processor performance. The lower the target recovery
time for access paths, the greater this impact is. Usually, the impact on processor
performance is not very noticeable, unless the processor is nearing capacity.

SMAPP causes increased disk activity, which increases the load on disk
input/output processors. Because the disk write operations for SMAPP are
asynchronous, they do not directly affect the response time for a specific
transaction. However, overall response time may be affected because of the
increased disk activity.

Specify target recovery times for access paths either for the entire system or for
individual ASPs, but not for both. If you specify both, the system does extra work
by balancing the overall target with the individual targets.

| The journal receivers that SMAPP uses take additional auxiliary storage. The
| system creates an internal journal and journal receiver for each ASP on your
| system. If the target recovery time for access paths for an ASP is set to *NONE, the
| journal receiver has no entries. The internal journal receivers are spread across all
| the arms in an ASP, up to a maximum of 100 arms.

The system manages the journal receivers automatically to minimize the impact as
much as possible. It regularly discards internal journal receivers that are no longer
needed for recovery and recovers the disk space. The internal journal receivers that
are used by SMAPP require less auxiliary storage than the journal receivers for
explicit journaling of access paths. Internal journal receivers are more condensed
because they are used only for SMAPP entries.

Chapter 18. Protecting Access Paths Using System-Managed Access-Path Protection 377
 If you have already set up journaling for a physical file, the system uses the same
journal to protect any access paths that are associated with that physical file. If the
system chooses to protect additional access paths, your journal receivers will grow
larger more quickly. You will need to change journal receivers more often.

To deal with the increased size of your journal receivers, consider specifying
RCVSIZOPT(*RMVINTENT) on the Create Journal (CRTJRN) or Change Journal
(CHGJRN) command for your journals. If you specify this, the system periodically
removes internal entries from user journal receivers when it no longer needs them
to recover access paths. This prevents your journal receivers from growing
excessively large because of SMAPP. “Specifying Receiver Size Options” on
page 407 provides more information about this option.

The displays for working with recovery for access paths show the current disk
usage for SMAPP. See Figure 32.

If your system cannot support dedicating resources to SMAPP, you can specify
*OFF for the system target recovery time. Before choosing this option, consider
setting the recovery time to *NONE for a normal business cycle, perhaps a week.
During that time, periodically display the estimated recovery time for access paths.
Evaluate whether those times are acceptable or whether you need to dedicate some
system resources to protecting access paths.

If you turn SMAPP off, any disk storage that has already been used will be
recovered shortly thereafter. If you set the SMAPP values to *NONE, any disk
storage that has already been used will be recovered after the next IPL.

Working with Recovery Times for Access Paths

Figure 32 shows the Edit Recovery for Access Paths display. Use the Edit Recovery
for Access Paths (EDTRCYAP) command to access this display:

Edit Recovery for Access Paths

06/19/9x
Estimated system access path recovery time . . . : 70 Minutes
Total disk storage used . . . . . . . . . . . . . : 70 MB
% of disk storage used . . . . . . . . . . . . . : 0.400

Type changes, press Enter.

System access path recovery time . . . *NONE *SYSDFT, *NONE, *MIN,
*OFF, Recovery time

----------Access Path Recovery Time----------- --Disk Storage Used---

ASP Target (Minutes) Estimated (Minutes) Megabytes ASP %
1 30 28 40.000 .700
2 *MIN 4 20.000 10.000
3 *NONE 40 10.000 5.000

Figure 32. Edit Recovery for Access Paths Display

The top part of the display shows the values for the entire system. The bottom part
| of the display shows the values for individual ASPs on the system. If you do not
| have user ASPs that are active, the bottom part of the display says No user ASP
| configured or information not available.

This topic describes some of the information on the display and how to use it. Use
the online help function to get additional information.

378 OS/400 Backup and Recovery V5R1

| The estimated recovery time for access paths is the number of minutes the system
| estimates it will need to recover most of the access paths during an IPL, or during
| a vary on of an independent ASP, after an abnormal end. It is an estimated
maximum, based on most circumstances. It assumes that access paths are being
recovered on a dedicated system (during an IPL) and that all eligible access paths
are being recovered or rebuilt. It does not include time to rebuild access paths that
must be rebuilt for one of the following reasons:
v The access path is damaged.
v The access path was marked as not valid during a previous abnormal system
end and was not successfully rebuilt.
v One of the following commands marked the access path as not valid and was
running when the system failed:
– Copy File (CPYF), if the system chose to rebuild the access path for efficiency.
– Reorganize Physical File Member (RGZPFM)
– Restore Object (RSTOBJ)

If you have user ASPs, the estimated recovery time for access paths for the entire
system (System access path recovery time field) may not equal the total estimated
recovery time for the ASPs (Access Path Recovery Time–Estimated (Minutes)). During
an IPL, or during the vary on of an independent ASP, the system overlaps
processing when recovering access paths to reduce the total time that is required.

The Disk Storage Used field on the display includes only internal system journals
and journal receivers. It does not include any additional space in user-managed
journal receivers for protecting access paths whose underlying physical files are
already journaled.

On the display, you can change the System access path recovery time field or the
Target (Minutes) access path recovery time field for individual ASPs. If you use user
ASPs to separate objects that have different recovery and availability requirements,
you may also want to specify different recovery times for access paths in those
user ASPs.

For example, if you have a large history file that changes infrequently, you may
want to put that file in a separate ASP. You may want to set the access path
| recovery time for that ASP to *NONE. Or, if you have an independent ASP, and
| you want the recovery time to move with the ASP when it is switched to another
| system, you may want to specify a specific time for that ASP.

Note: To change the access path recovery time from *OFF to another value, either
for the entire system or for a specific ASP, your system must be in a
restricted state.

Possible Values for the Access Path Recovery Time:

*SYSDFT Every new system is shipped with an internal default for access
path recovery time. Currently this value is 90 minutes for all
models and configurations. Specifying *SYSDFT sets the value to
the default for your system. You can specify *SYSDFT only for the
entire system, not for individual ASPs.
*NONE No implicit journaling of access paths takes place. The system
monitors current access path exposures. The Recovery for Access
Paths displays will show estimated times.
*MIN The value will be set to the minimum recovery time that can be
achieved for your system.

Chapter 18. Protecting Access Paths Using System-Managed Access-Path Protection 379
 Possible Values for the Access Path Recovery Time:
*OFF No implicit journaling of access paths takes place. The system does
not monitor exposures. System performance is not impacted. The
Recovery for Access Paths displays will not show estimated times.
You can specify *OFF only for the entire system, not for individual
ASPs.
target- recovery- time Specify a value between 10 and 1440 minutes. The system rounds
the value up to the nearest 10-minute boundary. If you specify a
value that is less than the system has calculated is the minimum,
the system protects to the minimum time possible.

You can use two other commands to work with access path recovery:
v Use the Display Recovery for Access Paths (DSPRCYAP) command to display or
print the estimated recovery times and disk usage.
v Use the Change Recovery for Access Paths (CHGRCYAP) command to change
the target recovery times without using an edit display.

The system performance monitor also provides information about access path
recovery times. The Work Management book and the Performance Tools for iSeries
book provide more information about monitoring performance and about what
SMAPP information is available through the tools.

How SMAPP Handles Changes in the Configuration of Auxiliary

Storage Pools
The system considers the size of an auxiliary storage pool (ASP) to determine the
size of the SMAPP journal receiver for that ASP. The system considers the
performance of the disk units assigned to an ASP to determine where to place the
SMAPP journal receiver.

When you perform an IPL, the system checks to see if your ASP configuration has
changed. The system does the following:
v If any disk units have been added or removed from an existing ASP, the system
may change either the size of the SMAPP receiver or the placement of the
receiver.
v If any new ASPs are in the configuration and do not have any access path
recovery times assigned for SMAPP, the system assigns a recovery time of
*NONE for that ASP. If you remove an ASP from your configuration and later
add it back, the access path for that ASP is set to *NONE, even if that ASP
previously had a recovery time for access paths.
v If all user ASPs have been removed from your configuration so that you have
only the system ASP, the system access path recovery time is set to the lower of
the following values:
– The existing system access path recovery time.
– The current access path recovery time for ASP 1. If the current access path
recovery time for ASP 1 is *NONE, the system access path recovery time is
not changed.

| When you vary on an independent ASP, the system checks to see if any disk units
| have been added or removed from the independent ASP. The system may change
| either the size of the SMAPP receiver or the placement of the receiver based on the
| change to the disk units. If this is the first time the independent ASP is varied on,
| then the system assigns a recovery time of *NONE for that independent ASP.

380 OS/400 Backup and Recovery V5R1

| When you add disk units to your disk configuration while your system is active,
| or your independent ASP is varied on, the system does not consider those changes
| in making SMAPP storage decisions until the next time you perform an IPL, or
| vary on your independent ASP. The system uses the size of the ASP to determine
| the threshold size for SMAPP receivers. If you add disk units, the system does not
| increase the threshold size for the receivers until the next IPL, or next vary on of
| the independent ASP. This means that the frequency of changing SMAPP receivers
| will not go down until you perform an IPL, or vary off and on your independent
| ASP.

When you create a new user ASP while your system is active, you should add all
of the planned disks to the ASP at the same time. The system uses the initial size
| of the new ASP to make storage decisions for SMAPP. If you later add more disk
| units to the ASP, those disk units are not considered until the next IPL or vary on
| of the independent ASP. When you create a new user ASP, the access path recovery
time for that ASP is set to *NONE. You can use the EDTRCYAP command to set a
target recovery time for the new ASP, if desired.

Chapter 18. Protecting Access Paths Using System-Managed Access-Path Protection 381
382 OS/400 Backup and Recovery V5R1
 Chapter 19. Planning and Setting Up Journaling
| The main purpose of journal management is to enable you to recover the changes
| to an object that have occurred since the object was last saved. You can also use
| journal management for:
| v An audit trail of activity that occurs for objects on the system.
| v Generate user defined journal entries to record activity, even for objects that do
| not allow journaling.
v Quicker recovery of access paths if your system ends abnormally.
v Quicker recovery when restoring from save-while-active media.
v Assistance in testing application programs.

| You use a journal to define what objects you want to protect with journal
| management. You may have more than one journal on your system. A journal may
| define protection for more than one object.

| You can journal the objects that are listed below:

| v Database physical files
| v Access paths
| v Data areas
| v Data queues
| v Integrated File System (IFS) objects (stream files, directories, and symbolic links).

| The system keeps a record of changes you make to objects that are journaled and
| of other events that occur on the system. These records are called journal entries.
| For example, some journal entries identify activity for a specific database record
| (added, updated 1 or deleted) and for a save, open, or close operation for an object.
| Journal entries can also identify other events that occur, such as security-relevant
| events on the system or changes made by dynamic performance tuning.
| “Appendix D. Journal Entry Information” on page 805 describes all the possible
| journal entry types and their contents.

Each journal entry includes additional control information that identifies the source
| of the activity, including, for example, the user, job, program, time, and date. The
| entries deposited for journaled objects reflect the changes made to that journaled
| object. For example, the entries for changes to database records include the entire
| image of the database record, not just the changed information.

| Journal entries are written to an object called a journal receiver. You can also write
| journal entries for events that you want to record or for objects other than the
| object that you want to protect with journaling. The system sends entries for all the
| objects a particular journal is protecting to the same journal receiver.

Journal receivers are attached to a journal by using the Create Journal (CRTJRN)
and Change Journal (CHGJRN) commands. The system adds journal entries to the
attached receiver. Journal receivers that have been attached to a journal and are

1. If the updated object image after the update is the same as the image before the update, then journal entries are not deposited for
that update.

© Copyright IBM Corp. 1997, 2000, 2001 383

 still known to the system are considered to be associated with that journal. Use the
Work with Journal Attributes (WRKJRNA) command to see a list of receivers
associated with a journal.

| The system adds an entry to the attached journal receiver when an event occurs to
| a journaled object. Each entry is sequentially numbered. For example, it adds an
| entry when you change a record in a journaled database file member and contains
| information that identifies:
v Type of change
v Record that has been changed
v Change that has been made to the record
v Information about the change (such as the job being run and the time of the
change)

| When you are journaling objects, changes to the objects are added to the journal
| receiver. The system does not journal data that you retrieved but did not change. If
| the logical file record format of a database file does not contain all the fields that
| are in the dependent physical file record format, the journal entry still contains all
| the fields of the physical file record format. In addition, if you are journaling access
| paths, entries for those access paths are added to the journal. If the updated
| physical file image after the update is the same as the image before the update,
| and if the file has no variable length fields, then journal entries are not deposited
| for that update. If the updated data area image after the update is the same as the
| image before the update, then journal entries are not deposited for that update.

| Figure 33 on page 385 shows journal processing. Objects A and B are being
| journaled; object C is not. Programs PGMX and PGMY use object B. When you
| make a change to object A or B, the following occurs:
| v The change is added to the attached journal receiver.
| v The journal receiver is written to auxiliary storage.
| v The changes are written to the main storage copy of the object.
| Object C changes are written directly to the main storage copy of the object
| because it is not being journaled. Only the entries added to the journal receiver are
| written immediately to auxiliary storage. Changes against the object may stay in
| main storage until the object is closed.

384 OS/400 Backup and Recovery V5R1

|

|
| Figure 33. Journaling Overview

If you have V4R2M0 or a later release installed on your system, you can take
advantage of the remote journal function. The remote journal function allows you
to associate a journal on a remote system with a journal on a local system. Journal
entries on the local system are replicated to the remote journal receiver.
“Chapter 21. Remote Journal Function” on page 467 provides further information.

Journal Management Commands and APIs

You can use the following commands and APIs for journal management. These
commands are used to create, maintain, and work with objects that are required
for journal management functions. For information about a specific command, refer
to the online command help.

Commands and APIs for Journals

|| Method Description

| Note: For more information about these commands and APIs, click the Programming topic in the Information Center
| at the following Web site: http://www.ibm.com/eserver/iseries/infocenter.
| Add Remote Journal Establishes and associates a remote journal on a target system with a
| (QjoAddRemoteJournal) API or Add journal on a source system. Refer to “Adding a Remote Journal” on
| Remote Journal (ADDRMTJRN) page 474 for more information.

Chapter 19. Planning and Setting Up Journaling 385

 Method Description

Change Journal (CHGJRN) Changes the attributes of a journal and attaches new journal receivers
to the journal. See “Changing Journal Receivers” on page 416.
Change Journal State Changes the journal state for local and remote journals. Refer to
(QjoChangeJournalState) API or Change “Activating and Inactivating a Remote Journal” on page 479 and
Remote Journal (CHGRMTJRN) “Activating and Inactivating a Local Journal” on page 487 for more
information.
Create Journal (CRTJRN) Creates a journal. See “How Journals Are Created” on page 402.
Delete Journal (DLTJRN) Deletes a journal. See “Deleting a Journal” on page 419.
Remove Remote Journal Disassociates a remote journal on a target system from a journal on a
(QjoRemoveRemoteJournal) API or Remove source system. Refer to “Removing a Remote Journal” on page 490 for
Remote Journal (RMVRMTJRN) more information.
| Retrieve Journal Information Retrieves the attributes of a journal including the receiver directory,
| (QjoRetrieveJournalInformation) API journaled objects, and remote journals.
| Work with Journals (WRKJRN) Displays a Work menu for user-selected journals from which the user
| can perform system-assisted recovery of journaled objects, journals, and
| journal receivers. See “Work with Journal (WRKJRN) Command
| Options” on page 455.
| Work with Journal Attributes (WRKJRNA) Displays the attributes of a journal, including the receiver directory,
| journaled objects, and remote journals. See “Using the Work with
| Journal Attributes Command” on page 442.
|

| Commands and APIs for Journal Receivers

Method Description

| Note: For more information about these commands and APIs, click the Programming topic in the Information Center
| at the following Web site: http://www.ibm.com/eserver/iseries/infocenter.
Create Journal Receiver (CRTJRNRCV) Creates a journal receiver. See “How Journal Receivers Are Created” on
page 399.
Delete Journal Receiver (DLTJRNRCV) Deletes a journal receiver. See “Deleting Journal Receivers” on page 418.
Display Journal Receiver (DSPJRNRCVA) Displays the attributes of a journal receiver. The online information for
the display describes its contents.
Retrieve Journal Receiver Information Retrieves the attributes of a journal receiver.
(QjoRtvJrnReceiverInformation) API

Commands and APIs for Journal Entries

Method Description

| Note: For more information about these commands and APIs, click the Programming topic in the Information Center
| at the following Web site: http://www.ibm.com/eserver/iseries/infocenter.
Compare Journal Images (CMPJRNIMG) Compares and lists the difference between the before-image and
after-image of a record or between the current after-image of a record
and the previous after-image of the record. See “Comparing Journal
Images” on page 440 for more information.
Delete Pointer Handle Deletes the specified pointer handle that was generated by the Retrieve
(QjoDeletePointerHandle) API Journal Entries (QjoRetrieveJournalEntries) API, or by the RCVJRNE
command with ENTTYP(*TYPEPTR).
Display Journal (DSPJRN) Displays or prints the journal entries that are in the journal receivers
associated with the specified journal. Writes the journal entries to a
database output file for processing or analysis. See “Displaying and
Printing Journal Entries” on page 429.
| Get Path Name of Object from Its File ID Determines an absolute path name of the object identified by fileid. The
| (Qp0lGetPathFromFileID()) API components of the returned path name are not symbolic links.

386 OS/400 Backup and Recovery V5R1

 Method Description

Receive Journal Entry (RCVJRNE) Allows a specified user program to continuously receive journal entries.
Entries can be received one at a time or in a block. Can be used to help
provide backup on another system. See “Using the Receive Journal
Entry Command” on page 432.
Retrieve Journal Entry (RTVJRNE) Retrieves a journal entry and places it in CL program variables. See
“Using the Retrieve Journal Entry Command” on page 436.
Retrieve Journal Entries Retrieves journal entries into a receiver variable.
(QjoRetrieveJournalEntries) API
Retrieve Journal Identifier Information Retrieves the name of the object that has the JID you specify.
(QJORJIDI) API
| Send Data Queue (QSNDDTAQ) API Sends data to the specified data queue. You can use parameter group
| three to indicate whether the data, in parameter four, came from a
| journal entry.
Send Journal Entry (SNDJRNE) command Adds user entries to a journal receiver. You can use them to record
or Send Journal Entry (QJOSJRNE) API specific events (application checkpoints) that may be important for
| recovery. You can create user entries to perform user-managed
| journaling of object types the system does not journal. See “Sending
Your Own Journal Entries” on page 429.

Commands and APIs for Starting and Ending Journaling of an

Object
Method Description

| Note: For more information about these commands and APIs, click the Programming topic in the Information Center
| at the following Web site: http://www.ibm.com/eserver/iseries/infocenter.
| End Journal (ENDJRN) or End Journal End journaling for IFS objects, data areas, and data queues. See
| (QjoEndJournal) API “Ending Journaling” on page 423.
End Journaling Access Path (ENDJRNAP) Ends journaling access paths. See “Ending Journaling” on page 423.
| End Journal Object (ENDJRNOBJ) End journaling for data areas and data queues. See “Ending Journaling”
on page 423.
| End Journal Physical File (ENDJRNPF) Ends journaling for the database physical files. See “Ending Journaling”
on page 423.
| Start Journal (STRJRN) or Start Journal Start journaling for IFS objects, data areas, and data queues. See “How
| (QjoStartJournal) API to Start Journaling for IFS objects” on page 413 and “How to Start
Journaling for Data Areas and Data Queues” on page 414.
Start Journal Access Path (STRJRNAP) Starts journaling access paths. For more information on journaling
access paths, see the topics “Should You Use Access Path Journaling?”
on page 391 and “How to Start Journaling for Access Paths” on
page 412.
| Start Journal Object (STRJRNOBJ) Start journaling for data areas and data queues. See “How to Start
Journaling for Data Areas and Data Queues” on page 414.
Start Journal Physical File (STRJRNPF) Starts journaling for the database physical files. See “How to Start
Journaling for Physical Files” on page 411.

| Commands and APIs for Getting Information about Journaled

| Objects
|| Method Description

| Note: For more information about these commands and APIs, click the Programming topic in the Information Center
| at the following Web site: http://www.ibm.com/eserver/iseries/infocenter.
| Display File Description (DSPFD) Shows one or more types of information retrieved from the file
| descriptions of one or more database and/or device files.

Chapter 19. Planning and Setting Up Journaling 387

| Method Description

| Display Object Description (DSPOBJD) Shows the names and attributes of specified objects in the specified
| library or in the libraries of the job’s library list.
| Display Object Links (DSPLNK) Shows a list of names of specified objects in directories and options to
| display information about the objects. You can use the Display attributes
| option to get information about journaled objects.
| Get Attributes (Qp0lGetAttr()) API Gets one or more attributes, on a single call, for the object that is
| referred to by the input Path_Name.
| List Objects (QUSLOBJ) API Allows you to generate information about a list of objects. You can use
| this API to find journaling information about those objects.
| Open List of Objects (QGYOLOBJ) API Allows you to generate information about a list of objects. You can use
| this API to find journaling information about those objects.
| Retrieve Object Description (QUSROBJD) Allows you to retrieve information about a specific object. You can use
| API this API to find journaling information about the objects and its
| journaling information.
| Retrieve Object Description (RTVOBJD) Used in a CL program or REXX procedure to retrieve the description of
| a specific object including its journal information.
| Work with Object Links (WRKLNK) Shows a list of names of specified objects in directories and options to
| work with the objects. You can use the Display attributes option to get
| information about journaled objects.
|

| Commands for Recovering an Object by Using the Journaled

Changes
Method Description
| Note: For more information about these commands and APIs, click the Programming topic in the Information Center
| at the following Web site: http://www.ibm.com/eserver/iseries/infocenter.
| Apply Journaled Changes (APYJRNCHG) Uses the journal entries to apply changes that have occurred since a
| journaled object was saved or to a specified point. “Applying Journaled
Changes” on page 446 describes this process.
| Remove Journaled Changes (RMVJRNCHG) Removes changes that have occurred to a journaled object from a
specified point to a previous point (allowed only if before-images and
after-images were journaled for the object). “Removing Journaled
Changes” on page 450 describes this process.

Planning Journal Management

To set your journal management strategy, you must make these decisions:
| v What objects to protect with journaling.
v Whether to journal other objects that the system does not journal.
v Whether to combine journaling with the save-while-active function.
| v How many journals you need and which objects should be assigned to each
| journal.
v Whether to journal after-images only or both before-images and after-images.
v Whether your application programs should write journal entries to assist with
recovery.
v Whether to isolate journal receivers in a user auxiliary storage pool (ASP).
v Whether to use the remote journal function to replicate the journal entries and
receivers to one or more additional systems. Refer to “Chapter 21. Remote
Journal Function” on page 467 for more information.

388 OS/400 Backup and Recovery V5R1

 You also need to make operational decisions about journal management:
v How often should journal receivers be changed and saved?
| v How often should you save journaled objects?
v How should journals and journal receivers be secured?

Finally, you need to balance the benefits of journaling with the impact it may have
on your system performance and auxiliary storage requirements.

| What Objects Should You Journal?

| Consider the following questions when deciding if you should journal all object
| types:

| General questions for all journaled object types

| These questions apply to all object types which can be journaled:

| v How much does the object change? An object with a high volume of transactions
| between save operations is a good candidate for journaling.
| v How difficult would it be to reconstruct the changes made to an object? Are
| many changes made to the object without written records? For example, an
| object used for telephone order entry would be more difficult to reconstruct than
| an object used for orders that arrive in the mail on order forms.
| v How critical is the information in the object? If the object had to be restored
| back to the last save operation, what effect would the delay in reconstructing
| changes have on the business?
| v How does the object relate to other objects on the system? Although the
| information in a particular object may not change often, that object may be
| critical to other, more dynamic objects on the system. For example, many files
| may depend on a customer master file. If you are reconstructing orders, the
| customer master file must include new customers or changes in credit limits.
| v Is there a need to replicate all the actions on the object?
| v After a crash, do you have a requirement to recover the object to a consistent
| state and have a journal entry show what actions completed?
| v What would be the consequence to your operation if a crash damages an object
| that the system is in the process of updating?

| Database physical file unique questions

| v Is the file part of a referential constraint network? If you journal one file that
| participates in a referential constraint, you should journal all the related files.
| Referential constraints are not enforced when you apply or remove journaled
| changes, but the referential integrity of those constraints is verified.
| If you journal all related files, the process for applying and removing journaled
| changes keeps the relationships between your database files valid. If you do not
| journal all related files, your referential constraint may show a status of check
| pending after you apply or remove journaled changes.
| For some types of referential constraints, the system requires that you journal all
| of the related files.
| v Does the file have trigger programs? If the trigger programs only performs
| processing on object types which can be journaled and applied, you should
| journal all such objects processed by the trigger program. If the trigger programs
| do additional work that should be reconstructed during a recovery, consider
| using the API support for sending journal entries.

Chapter 19. Planning and Setting Up Journaling 389

| v In general, database source files should not be journaled. If you use the Start
| Source Entry Utility (STRSEU) command to update a member, every record in
| that member is considered changed and every record is journaled to the journal.
| However, if changes to a source file are critical, you can journal the file in the
| same manner as data files.

| For more information about referential constraints and trigger programs, click the
| Database and file systems topic in the Information Center at the following Web
| site: http://www.ibm.com/eserver/iseries/infocenter.

| Integrated file system unique questions

| v Do you want to protect the actual symbolic links or the object the symbolic link
| points to? When you start journaling on a symbolic link the link is not followed.
| Therefore if you want to protect the actual object with journaling, you have to
| journal the actual object separately.
| v Do you want to automatically protect all objects which are created in a directory
| which itself is journaled? If so, you should consider the use and impacts of the
| inherit journaling attribute that you can associate with a journaled directory.
| v Do you want to protect the structure of the directory tree, or just the data stored
| in stream files within that directory structure? If you just want to protect the
| data stored in stream files, then for performances reasons, it may be best to only
| journal the stream files themselves instead of journaling changes to each
| directory in the directory tree. You should consider this question when you use
| the subtree and inherit journaling attributes options on the start journaling
| interfaces.
| v Do the objects you want to journal reside on an independent ASP? If the objects
| reside on an independent ASP, you cannot journal the objects.

| System objects

| It is recommended that you do not journal changes to IBM-supplied objects. These

| objects are sometimes created and managed differently than user-created objects.
| The system does not assure the recovery of these files even though all recovery
| activity normally succeeds.

Should You Journal Objects That the System Does Not

Journal?
| Some applications depend on information in objects that the system does not
| journal. For example, an application programming interface (API) might use a user
| space to pass data between two jobs.

You can use the Send Journal Entry (SNDJRNE) command or the Send Journal
| Entry (QJOSJRNE) API to write journal entries for these resources. See “Sending
| Your Own Journal Entries” on page 429. If you need to do recovery, you can use a
program to retrieve these journal entries and make sure these application objects
are synchronized with the objects you are journaling.

If you are using commitment control, you can use APIs to register these objects as
committable resources. Chapter 22 provides general information about using APIs
| with commitment control. For information on how to use APIs to send journal
| entries and to register committable resources click the Programming topic in the
| Information Center at the following Web site:
| http://www.ibm.com/eserver/iseries/infocenter.

390 OS/400 Backup and Recovery V5R1

Should You Use Access Path Journaling?
When your system ends abnormally, perhaps because of a power interruption, the
next IPL can take much longer than a normal IPL. Rebuilding access paths
contributes to this long IPL time. When you perform an IPL after an abnormal end,
the system rebuilds access paths that were exposed, except those access paths that
are specified as MAINT(*REBLD) when you create the file. An access path is
exposed if changes have been made to it that have not been written to the disk.

If you journal access paths, the system can use the journal entries to recover access
paths instead of rebuilding them completely. This reduces the time it takes to IPL
after the system ends abnormally. Access path journaling is strictly for the purpose
of system recovery during an IPL. You do not use access path journal entries when
you are applying journal changes to recover a file.

You can allow the system to determine what access paths to protect in order to
meet a target time for recovering access paths. This support, called
system-managed access-path protection, is described in “System-Managed
Access-Path Protection” on page 5.

You can choose to journal some access paths yourself. This is called explicit access
path journaling. For example, if certain access paths and their underlying files are
critical to your operation, you want to ensure that these files are available as soon
as possible after the system ends abnormally. You cannot control which access
paths the system chooses to journal under system-managed access-path protection
(SMAPP). The system may not need to journal the access path that you consider
critical to meet your target recovery times for access paths. In this case, explicitly
journaling an access path is an appropriate decision.

If you choose to journal an access path, remember the following:

v You can journal an access path for a physical file only if the physical file has a
keyed access path or an index created by a referential constraint.
v Before you start journaling an access path, you must journal all the underlying
physical files.
v You can journal only access paths that are defined as MAINT(*IMMED) or
MAINT(*DLY).
v You cannot journal encoded vector access paths.

How SMAPP Is Different from Explicitly Journaling Access Paths:

v SMAPP does not require that underlying physical files be journaled.
v SMAPP determines which access paths to protect based strictly on the target
recovery times for all access paths. You might choose to journal an access path
explicitly because of your requirements for the availability of a specific file.
v SMAPP continually evaluates which access paths to protect and responds to
changes in your system environment.
v SMAPP does not require any user intervention to manage its internal journals
and journal receivers.
v SMAPP uses less disk space for journal receivers because they are detached and
deleted regularly.

How Should Objects Be Assigned to Journals?

You can use one journal to manage all the objects you are journaling. Or, you can
set up several journals if groups of objects have different backup and recovery

Chapter 19. Planning and Setting Up Journaling 391

 requirements. Every journal has a single attached receiver. All journal entries for all
objects being managed by the journal are written to the same journal receiver.

When deciding how many journals you should use and how to assign objects to
journals, consider the following:
v Using one journal (and journal receiver) is the simplest method for managing
both daily operations and recovery.
| v If using a single journal receiver causes a performance bottleneck, you can
| alleviate this by placing the journal receiver in a separate user ASP from the
| objects that you are journaling.
| v To simplify recovery, objects that are used together in the same application
| should be assigned to the same journal.
| v If you are journaling database files, all the physical files underlying a logical file
| should be assigned to the same journal.
v If you have Version 2 Release 3 Modification 0 or earlier of the OS/400 licensed
program, all files opened under the same commitment definition within a job
must be journaled to the same journal. If you have Version 3 Release 1 or later,
this restriction no longer applies in most situations. In commitment control, each
journal is considered a local location. Table 75 on page 544 describes what
conditions are required for more than one local location (journal) to be allowed.
| v If your major applications have completely separate objects and backup
| schedules, separate journals for the applications may simplify operating
| procedures and recovery.
| v If you journal different objects for different reasons; such as recovery, auditing,
| or transferring transactions to another system; you may want to separate these
| functions into separate journals. However, you can assign an object to only one
| journal.
| v If the security of certain objects requires that you exclude their backup and
| recovery procedures from the procedures for other objects, assign them to a
| separate journal, if possible.
v If you have user ASPs with libraries (library user ASP), all objects assigned to a
journal must be in the same user ASP as the journal. The journal receiver may
be in a different ASP. If you place a journal in a user ASP without libraries
(nonlibrary user ASP), objects being journaled must be in the system ASP. The
journal receiver may be in either the system ASP or the nonlibrary user ASP
with the journal. See “Should Your Journal Receivers Be in a User ASP?” on
page 393 for more information about the types of ASPs.
| v If you have independent ASPs and place IFS objects on those independent ASPs,
| then these objects cannot be journaled. All objects assigned to a journal must be
| in the same ASP as the journal and journals cannot be created on independent
| ASPs

Should You Journal Before-Images?

| When you journal an object, the system always writes an after-image for every
| change that is made. You can request that the system write before-image journal
| entries for database files and data areas. This significantly increases the auxiliary
| storage requirements for journaling. All other object types only journal
| after-images.

You may choose to journal before-images for these reasons:

| v Before-images are required for a backout recovery, where you remove journal
| changes (RMVJRNCHG) rather than applying journal changes to a restored copy
| of an object. Backout recovery is often complex, particularly if multiple users

392 OS/400 Backup and Recovery V5R1

| and programs are accessing the same object. It is most commonly used when
| new applications or programs are being tested.
v For database physical files, before-images are required to use the Compare
Journal Images (CMPJRNIMG) command. This command highlights the
differences between the before-images and after-images. It is sometimes used to
audit changes to a database file.
v For database physical files, if you want a copy of the record that is deleted to be
part of the deleted record journal entry information, you must specify
before-images.

| Commitment control requires before-images for the system to roll back

| uncommitted changes. When you open a database file under commitment control,
| the system automatically journals both before-images and after-images while the
| commitment definition is active. If you normally journal only after-images, the
| system writes before-images only for the changes made under commitment control.
| If the system initiates the journaling of before-images, you cannot use them to
| remove journaled changes. Commitment control does not support IFS objects, data
| areas, or data queues. Refer to “Chapter 22. Commitment Control” on page 525 for
| more information on commitment control.

| Access path journaling also requires before-images for the system to use for IPL
| recovery. When you journal access paths, or the system journals an access path to
| provide System-Managed Access-Path Protection, the system will automatically
| journal both before and after-images. If you normally journal only after-images, the
| system also writes before-images if you are journaling the access path.

| Remember that you can select before-images on a object-by-object basis. You

| specify whether you want after-images or both when you start journaling for a
| database file (STRJRNPF command) or a data area (STRJRNOBJ or STRJRN
| command).

Should Your Journal Receivers Be in a User ASP?

You can use user ASPs to control which objects are allocated to which groups of
disk units. If you are journaling many active objects to the same journal, the
journal receiver can become a performance bottleneck. One way to minimize the
performance impact of journaling is to put the journal receiver in a separate user
ASP. This also provides additional protection because your objects are on different
disk units from the journal receiver, which contains a copy of changes to the
objects.

| Note: You cannot store journals or journal receivers on independent ASPs.

| Therefore it is not possible to journal objects that reside on independent
| ASPs.

iSeries and AS/400e servers have three types of ASPs:

v The system ASP contains the operating system. It can also contain user libraries
and objects.
| v A library user ASP contains one or more user libraries or IFS User Defined File
| Systems. It does not contain the operating system. This is the current
recommended method of configuring user ASPs.
| v A nonlibrary user ASP contains no user libraries or IFS User Defined File
| Systems. It may contain journals, journal receivers, and save files. If you place a

Chapter 19. Planning and Setting Up Journaling 393

 journal receiver in a nonlibrary user ASP, the journal must be in either the
system ASP or the same nonlibrary user ASP. The journaled objects must be in
the system ASP.
When ASPs were first introduced, only nonlibrary user ASPs were available.
Many systems still have this type of ASP. However, recovery steps are more
complex for nonlibrary user ASPs. Therefore, for systems implementing
journaling for the first time, library user ASPs are recommended.

Journaling with the Save-While-Active Function

| Journaling can help you with recovery if you use the save-while-active function in
| your backup strategy. If you plan to save an application without ending it for
| checkpoint processing, consider journaling all the objects associated with the
| application. After the save operation is complete, save all the journal receivers for
| the objects you are saving.

If you need to do a recovery, you can restore objects from the save-while-active
media. Then you can apply journal changes to an application boundary.

| For more information about the save-while-active function, click the Backup,
| recovery, and availability topic under Systems Management in the Information
| Center at the following Web site: http://www.ibm.com/eserver/iseries/infocenter.

Journaling and System Performance

| Journaling is designed to prevent transactions from being lost if your system ends
| abnormally or has to be recovered. To do this, changes to a journaled object are
| written immediately to the journal receiver in auxiliary storage. This increases the
| disk activity on your system and may have a noticeable impact on system
| performance. Journaling also increases the overhead associated with opening
| objects and closing objects.

As the number of objects you are journaling increases, the general performance of
the system may be slower. The time it takes to perform an IPL on your system
may also increase, particularly if your system ends abnormally.

The system takes measures to minimize the performance impact of using

journaling features. For example, the system packages before-images and
after-images and any access path changes for a record in a single write operation
to auxiliary storage. Therefore, journaling access paths and before-images, in
addition to after-images, usually does not cause additional performance overhead.
They do, however, add to the auxiliary storage requirements for journaling.

| The system also spreads journal receivers across multiple disk units to improve
| performance. The journal receiver may be placed on up to ten disk units in an ASP.
| If you specify the RCVSIZOPT(*MAXOPT1) or RCVSIZOPT(*MAXOPT2) journal
| option, then the system may place the journal receiver on up to 100 disk units in
| an ASP. Journal entries are written using a “round robin” technique with these
| arms.

You can take measures to minimize the impact of journaling on your system
performance:
v Do not set the force-write ratio (FRCRATIO) parameter for physical files that are
being journaled. You can let the system manage when to write records for the
physical file to disk because the journal receiver has a force-write ratio of 1.

394 OS/400 Backup and Recovery V5R1

 v Isolate journal receivers in a user ASP. This reduces contention when accessing
the disks. See “Should Your Journal Receivers Be in a User ASP?” on page 393.
Your performance may be better if the disk units in the user ASP are
unprotected or mirrored rather than protected through device parity protection.
| If you must use RAID5 disk drives for your journal receiver’s ASP, select the
| ninth and tenth drive of each redundant array of independent disks (RAID) set.
| Only the first eight drives in a RAID set house parity data.
v Consider using record blocking when a program processes a journaled file
sequentially (SEQONLY(*YES)). When you add or insert records to the file, the
records are not written to the journal receiver until the block is filled. You can
specify record blocking with the Override with Database File (OVRDBF)
command or in some high-level language programs.
| If you use the OVRDBF command, do the following:
| – Set the SEQONLY parameter to (*YES).
| – Use a large enough value for the NBRRCDS parameter to make the buffer
| approach the optimal size of 128KB.
v Consider minimizing the fixed length portion of the journal entry using
RCVSIZOPT(*MINFIXLEN) for the journal. When you specify this option, the
job, program, and user profile information is not deposited as part of the journal
entry. Therefore, that information does not have to be retrieved, benefiting
journal performance.
| v Consider omitting information from the journal entry you may not need using
| the OMTJRNE parameter. When you specify the OMTJRNE parameter for
| database physical files you will not deposit the file open and close entries which
| saves processing as well as disk storage space. Similarly, if you specify the
| OMTJRNE paramater for directories and stream files, the object open, close, and
| force entries are not deposited.

Journaling and Auxiliary Storage

| If you are journaling an object, journal management writes a copy of every object
| change to the journal receiver. It writes additional entries for object level activity,
| such as opening and closing the object, adding a member, or changing an object
| attribute. If you have a busy system and journal many objects, your journal
| receivers can quickly become very large. See “Appendix D. Journal Entry
| Information” on page 805 for a complete list of all the journal entries.

The maximum size for a single journal receiver varies. It depends on how the
system allocates the journal receiver across multiple disk arms. The maximum size
ranges from approximately 1.9 GB to 1.0 TB depending on what value you
specified for the associated journal’s Receiver Size Option.

| To avoid possible problems with a journal receiver exceeding the maximum size
| allowed on the system, specify a threshold for the receiver of no more than
| 900 000 000KB if you specified RCVSIZOPT(*MAXOPT1) or
| RCVSIZOPT(*MAXOPT2) for the associated journal. Otherwise, specify a threshold
| of no more than 1 441 000KB. See “Specifying a Threshold for a Journal Receiver”
| on page 401.
Notes:
1. You cannot save or restore any journal receivers that you attach to journals
with RCVSIZOPT(*MAXOPT1) specified to any releases prior to V4R5M0. You
also cannot replicate these receivers to any remote journals on any systems at
releases prior to V4R5M0.

Chapter 19. Planning and Setting Up Journaling 395

| 2. You cannot save or restore any journal receivers that you attach to journals
| with RCVSIZOPT(*MAXOPT2) specified to any releases prior to V5R1M0. You
| also cannot replicate to any remote journals on any systems at releases prior to
| V5R1M0.

How to Estimate the Size of a Journal Receiver

| You can use the methods below to estimate the effect a journal receiver will have
| on auxiliary storage.

| The actual auxiliary storage used will be somewhat larger because the system
| writes additional entries for such actions as opening and closing objects unless you
| specified OMTJRNE(*OPNCLO) when starting journaling for the database physical
| files or you specified OMTJRNE(*OPNCLOSYN) when starting journaling for
| directories or stream files.

Estimating the Size of a Journal Receiver Method 1

This procedure assumes the following:
| v You are journaling database physical files.
v You are journaling after-images only.
v You are using a single journal receiver for an entire day’s transactions.
v You are journaling database physical files only. It does not include estimates for
access path journaling or user-created entries.
v You are not minimizing the fixed length portion of the journal entry
(RCVSIZOPT(*MINFIXLEN) specified for the journal).
| v You are not using the MINENTDTA parameter to minimize entry-specific data
| for the files.

Follow the steps below to estimate the size of a journal receiver:

1. Determine the average record length for all the files that you plan to journal. If
the record lengths vary significantly and the information is available, use a
weighted average based on the relative number of transactions per file.
2. The system-created portion of a journal entry is approximately 155 bytes. (It
varies by the type of journal entry.) Add 155 to the record length from step 1. If
you specify RCVSIZOPT(*MINFIXLEN), the average length of the
system-created portion of a journal entry is approximately 110 bytes. In this
case, add 110 to the record length from step 1.
3. Estimate the number of transactions for journaled files for a day. Multiply the
number of transactions by the record length from step 2. The result is the
estimated bytes of auxiliary storage needed for one day’s transactions in the
journal receiver.

For example:
1. The average record length for journaled files is 115 bytes.
2. The average record length for a journal entry is 270 bytes (115 + 155). In this
example, RCVSIZOPT(*MINFIXLEN) is not specified.
3. The number of journaled transactions per day is 10 000. The total bytes needed
to journal after-images for a day is 2 700 000 (270 x 10 000).

Estimating the Size of a Journal Receiver–Method 2

| Another method for estimating the size of the journal receiver is to run a test. This
| method is more accurate because it includes all journal entries. Additionally, this
| method will work for any object type which can be journaled, not just database

396 OS/400 Backup and Recovery V5R1

| physical files unlike method one. To use this method, you must either have
journaling set up already or you must set it up.

If you are already using journaling, skip steps 1 and 2 below. Instead, issue a
Display Journal Receiver Attributes (DSPJRNRCVA) command before the time
period so you can compare sizes from the beginning of the period to the end.

This method assumes that the same receiver is used during the whole test. If there
is a change journal to attach a new journal receiver during the test, you must
include the sizes of all the receivers.
1. Set up journaling by creating the receiver and journal.
2. Start journaling for all the objects that you plan to journal.
3. Choose a time period (1 hour) with typical transaction rates.
4. After one hour, use the Display Journal Receiver Attributes (DSPJRNRCVA)
command to display the size of the receiver.
5. Multiple the size by the number of hours that your system is active in a day.

Journaling Features That Increase the Journal Receiver Size

If you use some optional features available with journal management, this can
significantly increase auxiliary storage requirements.

| You can select to journal both before-images and after-images. The system uses
| more storage if you select both before-images and after-images, although storage
| use is not necessarily doubled. If you journal access-paths, the before-images and
| after-images are written to the journal receiver when a database file is updated or a
| data area is changed2. Only after-images are written when an object is added
| (write operation) or deleted.

The system requires additional space to journal access paths. The space required
depends on:
v How many access paths are journaled.
| v How often you change the access paths. When you update a record in a
| database file, you cause an access path journal entry only if you update a field
| included in the access path.
v The method used to update access paths. More journal entries are written if you
update access paths randomly than if you update them in ascending or
descending sequence. Doing a mass change to an access path field, such as a
date change, causes the fewest journal entries.

Note: If you have target recovery times for access paths specified (SMAPP) and
you journal database files, the system uses the same journal receiver to
protect access paths for that file. This increases the size of your journal
receivers.

How You Can Reduce the Storage Used by Journaling

You can take measures to reduce the storage needed for journaling:
v Journal after-images only. Unless you are using commitment control,
after-images should be sufficient for your recovery needs. See “Should You
Journal Before-Images?” on page 392.
| v Consider omitting the journal entries for open, close or force operations to
| journaled objects. Use the OMTJRNE parameter on the STRJRNPF or STRJRN

2. Neither the before-image nor after-image is deposited into the journal if the after-image is exactly the same as the before-image.

Chapter 19. Planning and Setting Up Journaling 397

| command. This can have a noticeable affect on both space and performance if an
| application opens, closes, or forces objects frequently.
| However, if you omit the journal entries for opening and closing objects, you
| cannot:
| – Use open and close boundaries when applying or removing journal changes
| (the TOJOBO and TOJOBC parameters).
| – Audit what users opened particular objects.
v Change journal receivers, save them, and free storage more frequently.

Note: Moving journal receivers off-line increases your recovery time because
receivers have to be restored before journal changes can be applied.
v Specify RCVSIZOPT(*RMVINTENT) for journals. This causes the system to
periodically remove internal entries that it no longer needs, such as access path
entries.
v Specify RCVSIZOPT(*MINFIXLEN) for journals. This causes the system to no
longer deposit the job, program, or user profile information in the journal entry,
thus reducing the size of the entries. However, if you require this journal entry
information for audit or other uses, you cannot use this storage saving
mechanism. Additionally, it reduces the options available as selection criteria
used on the DSPJRN, RCVJRNE, RTVJRNE, CMPJRNIMG, APYJRNCHG, and
| RMVJRNCHG commands, or the QjoRetrieveJournalEntries API. See “Using
| RCVSIZOPT(*MINFIXLEN)” on page 408 for more information.
| v Specify the minimize entry-specific data (MINENTDTA) parameter for journals.
| This allows the system to write data to the journal entries in a minimized
| format. See“Specifying Minimized Entry-Specific Data” on page 409 for more
| information.

Setting Up Journaling
After you have decided how you will use journaling, follow these steps to set up
journaling on your system. If you have decided to use more than one journal,
work through all the steps for one journal at a time. Then repeat the steps for the
next journal.

The topics that follow this list of steps describe the steps in more detail.
1. Create the journal receiver using the CRTJRNRCV command.
2. Create the journal using the CRTJRN command.
| 3. Start journaling for each object that you plan to journal:
| v Use the Start Journaling Physical File (STRJRNPF) command for database
| files. Start journaling for only the files that are assigned to the journal that
| you are currently setting up.
| v Use the Start Journal (STRJRN) command for IFS objects. Start journaling for
| only the objects that are assigned to the journal that you are currently setting
| up.
| v Use the Start Journal Object (STRJRNOBJ) command for all other object
| types. Start journaling for only the objects that are assigned to the journal
| that you are currently setting up.
| 4. Start journaling access paths, if you plan to do so, using the STRJRNAP
| command. Start access path journaling for only the access paths that are
| assigned to the journal that you are currently setting up.
| 5. After journaling begins for the object, save the journaled object to preserve its
| journal attribute information. Also, you must save the object because, for

398 OS/400 Backup and Recovery V5R1

| example, you cannot apply journaled changes to a version of the object that
| was saved before journaling was in effect.

Note: If you are journaling distributed physical files, you have to do steps 1, 2, 4,
and 5 on each system in the node group. The system handles distribution of
the start of journaling (step 3, (STRJRNPF command)).

How Journal Receivers Are Created

Use the Create Journal Receiver (CRTJRNRCV) command to create a journal
receiver. Parameters for the command specify where to put the receiver, what its
storage threshold is, and who can access it. The topics that follow describe the
parameters and options. “Creating Journals and Journal Receivers–Examples” on
page 409 shows several examples.

Naming Journal Receivers

When you use the CRTJRNRCV command to create a journal receiver, you assign a
name to the journal receiver. When you use the Change Journal (CHGJRN)
command to detach the current journal receiver and create and attach a new
receiver, you can assign a name or have the system generate one. If you use system
change-journal management, the system generates the name when it detaches a
receiver and creates and attaches a new one.

If you plan to have more than one journal on your system, use a naming
convention that links each journal with its associated receiver.

To simplify recovery and avoid confusion, make each journal receiver name unique
for your entire system, not unique within a library. If you have two journal
receivers with the same name in different libraries and they both become damaged,
the reclaim storage operation renames both journal receivers when they are placed
in the QRCL library. When you use the Rename Object (RNMOBJ) command for a
journal or journal receiver in the QRCL library, you can change the name of the
library back to the original name. You cannot change the name of the journal or
the journal receiver.

When you detach the receiver from the journal and attach a new one, you can
have the system generate the name for the new receiver by incrementing the
previous receiver name. If you use system change-journal management, by
specifying MNGRCV(*SYSTEM) for the journal, the system also generates a new
receiver name when it changes journal receivers.

Table 61 shows the rules the system uses to generate a new receiver name. It
applies these rules in the sequence shown in the table:
Table 61. System-Generated Receiver Names. How the System Generates Journal Receiver
Names When Changing Journal Receivers
Current Name System Action Example
Last 4 characters are numeric. Adds 1 DSTR0001 to DSTR0002
Last character is not numeric. Truncates the name to 6 DSTRCVR to DSTRCV0001
characters, if necessary, and
concatenates 0001
Last character is numeric. Adds 1 DSTR01 to DSTR02
Last non-numeric character is
in position 5 or less.
Last character is numeric. Truncates to 6 characters, if DSTRCVR01 to DSTRCV0001
Last non-numeric character is necessary. Concatenates 0001.
in position 6 or higher.

Chapter 19. Planning and Setting Up Journaling 399

 If you restore a journal to your system, the system creates a new journal receiver
and attaches it to the journal. The system generates a name for the new journal
receiver based on the name of the journal receiver that was attached when the
journal was saved. Table 62 shows the rules the system uses to generate a new
receiver name when you restore a journal:
Table 62. System-Generated Receiver Names. How the System Generates Journal Receiver
Names When Restoring a Journal
Current Name System Action Example
Last 4 or more characters are Adds 1 to the leftmost digit DSTR0001 to DSTR1001
numeric. of the numeric portion.
Last character is not numeric. Truncates to six characters, if DSTRCVR to DSTRCV1000.
necessary. Concatenates 1000.
Ending numeric portion is Pads the left portion of the DSTRCV01 to DSTRCV1001.
less than 4 digits. numeric portion with zeroes
to create a 4-digit suffix.
Adds 1 to the leftmost digit.

If the name generated by the system is the same as the name of a journal receiver
already on the system, the system adds 1 to the name until it creates a name that is
not a duplicate. For example, assume a journal receiver named RCV1 was attached
when the journal was saved. When the journal is restored, the system attempts to
create a new journal receiver named RCV1001. If that name already exists, the
system tries the name RCV1002.

Table 63 shows examples of how the system generates new receiver names:
Table 63. System-Created Journal Receiver Names
Journal Receiver Name
Last Journal Receiver Created by Restoring
Known to the System1 Created by Change Journal2 Journal
A A0001 A1000
ABCDEF ABCDEF0001 ABCDEF1000
ABCDEFG ABCDEF0001 3 ABCDEF1000 3
ABCDEF1234 ABCDEF1235 ABCDEF2234
A0001 A0002 A1001
A1 A2 A1001
A9 A10 A1009
ABCDEF7 ABCDEF0001 3 ABCDEF1007 3
ABCDEF9999 Error 4 ABCDEF0999
A1B15 A1B16 A1B1015
1
If the journal exists on the system, the last journal receiver known to the system is
the journal receiver that is currently attached. If the journal does not exist, the last
journal receiver known to the system is the journal receiver that was attached when
the journal was saved.
2
Either when a user issues the CHGJRN command with JRNRCV(*GEN) or when
the journal is changed by system change-journal management.
3
The last character of the current name is dropped because it exceeds 6 characters.
4
If the journal is set up as MNGRCV(*SYSTEM), the receiver name wraps around to
0’s (ABCDEF0000). If the journal is set up as MNGRCV(*USER), an error occurs
because adding 1 to 9999 causes an overflow condition.

400 OS/400 Backup and Recovery V5R1

 Assigning a Journal Receiver to a Library
When you create a journal receiver, you specify a qualified name that includes the
library for the receiver. The library must exist before you create the journal
receiver.

Possible Values for the Journal Receiver Library:

*CURLIB The receiver is created in the current library for the job. If no
library is specified as the current library for the job, the QGPL
library is used.
library-name: Specify the name of the library where the journal receiver is to
be created.

Assigning a Journal Receiver to an Auxiliary Storage Pool

If you want to place the journal receiver in a library user ASP, you must first create
the library for the journal receiver in the user ASP. If you want to place the journal
receiver in a nonlibrary user ASP, you must first create the library for the journal
| receiver in the system ASP. Placing journal receivers in a different ASP from the
| journaled objects may prevent performance bottlenecks.

Possible Values for the ASP Parameter:

*LIBASP: The storage space for the journal receiver is allocated from the
same ASP as the storage space of the journal receiver’s library.
Use this value if you want the journal receiver in a library user
ASP.
| asp-identifier: Specify a value ranging from 1 through 32 to specify the
| identifier of the ASP from which to have the storage space of
| the journal receiver allocated. Specify an ASP number only if
| you want to place the journal receiver in a nonlibrary user ASP.

Note: The UNIT parameter is no longer supported for the CRTJRNRCV command.
It is still available for compatibility with earlier releases of the OS/400
licensed program.

Specifying a Threshold for a Journal Receiver

When you create a journal receiver, you specify a threshold that indicates when
you want the system to warn you or take action. When the receiver reaches that
threshold, the system takes the action specified in the manage receiver (MNGRCV)
parameter for the journal. See “Specifying How Journal Receivers for the Journal
Are Changed” on page 405.

In specifying a storage threshold, you need to balance the amount of space that
you have available with the system overhead associated with changing journal
receivers frequently. Consider these options:
v Base the size on your available auxiliary storage:
1. Calculate the amount of auxiliary storage space that you have available in
the user ASP for the journal receiver.
2. Assign a receiver threshold that is 75 to 80 percent of that space.
v Base the size on how often you want to change journal receivers:
1. Use the method described in “How to Estimate the Size of a Journal
Receiver” on page 396 to calculate how large your receiver would be for a
day.
2. Determine how many times a day you should detach and save the journal
receiver.

Chapter 19. Planning and Setting Up Journaling 401

 3. Divide the result of step 1 on page 401 by the result of step 2 on page 401.
This is your receiver threshold.

| Do not make the journal receiver size too small, or the system will spend too much
| resource changing journal receivers or sending threshold messages. To avoid
| possible problems with a journal receiver exceeding the maximum size allowed on
| the system, specify a threshold for the receiver of no more than 900 000 000KB if
| you specified RCVSIZOPT(*MAXOPT1) or RCVSIZOPT(*MAXOPT2) for the
| associated journal. Otherwise, specify a threshold of no more than 1 441 000KB.

Possible Values for the THRESHOLD Parameter:

*NONE The journal receiver continues to grow until you detach it or it
reaches the maximum size of a journal receiver.
threshold- value: Specify a value ranging from 1 through 1 000 000 000 that
indicates KB of storage. For example, an entry of 1000 specifies
1 024 000 bytes of storage space. When the size of the space for
the journal receiver is larger than the size you specify, the
system takes the action defined in the MNGRCV parameter for
the associated journal.
Notes:
1. The value should be at least 5000. Using a value less than
5000 may cause operational problems.
| 2. If you plan to use MNGRCV(*SYSTEM), the threshold must
| be a minimum of 100 000 KB if the receiver will be attached
| to a journal with RCVSIZOPT(*MAXOPT1) or
| RCVSIZOPT(*MAXOPT2) specified, otherwise, it must be a
| minimum of 5 000 KB.
| 3. If you plan to attach this journal receiver to a journal that
| does not have RCVSIZOPT(*MAXOPT1) or
| RCVSIZOPT(*MAXOPT2) specified, the maximum you can
| specify is 1 919 999 KB.

Authority to the Journal Receiver

When you create a journal receiver, you specify the authority that all users on the
system have to access it (public authority). The default authority for the
CRTJRNRCV command is *LIBCRTAUT, which means the system uses the value of
the create authority (CRTAUT) parameter for the journal receiver’s library.

| Journal receivers contain copies of changes from all objects being journaled.
| Someone with access to the journal receiver could display confidential data. The
| authority to a journal receiver should be as strict as the authority for the most
| confidential object that is being journaled.

You do not need any authority to the journal or to the journal receiver to use an
object that is journaled. Authority to the journal receiver is checked only when
using commands that operate directly on the receiver. The authority you set for the
journal receiver has no effect on the people using the journaled objects. The iSeries
Security Reference book has more information about the authority required to access
objects and perform commands that use journals and journal receivers.

How Journals Are Created

Use the Create Journal (CRTJRN) command to create a journal. Journals that are
created with the CRTJRN command have a journal type of *LOCAL. The topics

402 OS/400 Backup and Recovery V5R1

 that follow explain the parameters and options available when creating a journal.
“Creating Journals and Journal Receivers–Examples” on page 409 shows several
examples.

Naming Journals
When you create a journal by using the CRTJRN command, you assign a name to
it. If you plan to have more than one journal on your system, use a naming
convention that links each journal with its associated receiver.

To simplify recovery and avoid confusion, make each journal name unique for
your entire system, not unique within a library. If you have two journals with the
same name in different libraries and they both become damaged, the reclaim
storage operation renames both journals when they are placed in the QRCL library.
When you use the RNMOBJ command for a journal in the QRCL library, you can
change the name of the library back to the original library name. You cannot
change the name of the journal itself. In this case, you would not be able to
recovery your journal from QRCL since its name has been changed.

| Using Naming Conventions to Ensure Restore Sequence: To ensure that

| journaling is automatically started again, the journals must be restored before the
| objects being journaled. If the journals and associated objects are in the same
| library, then the system automatically restores the objects in the correct order.

When the journals and associated objects are in different libraries, you must ensure
| that the objects are restored in the correct order. You can name the libraries for the
| journals, objects, and journal receivers to ensure that the objects are restored in the
| correct order. For example, starting the name of the library for the journal with a
| special character, such as #, $, or @, ensures that the library for the journal is
| restored before the library for the objects. In the normal sort sequence, special
characters appear ahead of alphabetic characters.

| Since IFS objects do not exist in a library, your restore processing must ensure the
| objects are restored in the correct order. That is, you must restore your libraries
| which contain the journals before restoring the IFS objects that were journaled to
| that journal.

Assigning a Journal to a Library

When you create a journal, you specify a qualified name that includes the library
for the journal. The library must exist before you create the journal.

Possible Values for the Journal Library:

*CURLIB The journal is created in the current library for the job. If no
library is specified as the current library for the job, the QGPL
library is used.
library-name: Specify the name of the library where the journal is to be
created.

Assigning a Journal to an Auxiliary Storage Pool

| If you want to place the journal in a library user ASP, you must first create the
| library for the journal in the user ASP. If you use a library user ASP, the journal
| and all the objects you are journaling to it must be in the same library user ASP.

| If you want to place the journal in a nonlibrary user ASP, you must first create the
| library for the journal in the system ASP. If the journal is in a nonlibrary user ASP,
| all the objects being journaled to it must be in the system ASP.

Chapter 19. Planning and Setting Up Journaling 403

 Possible Values for the ASP Parameter:
*LIBASP The storage space for the journal is allocated from the same
ASP as the storage space of the journal’s library. Use this value
if you want the journal in a library user ASP.
| asp-identifier Specify a value ranging from 1 through 32 to specify the
| identifier of the ASP from which to have the storage space of
| the journal allocated. Specify an ASP number only if you want
| to place the journal in a nonlibrary user ASP.

Associating the Journal with a Journal Receiver

When you create a journal, you must specify the name of the journal receiver to be
attached to it. The journal receiver must exist before you can create the journal.
The receiver that you attach may not have been previously attached to a different
journal or have been interrupted while being attached to any journal.

Possible Values for the JRNRCV Parameter:

receiver-name: Specify the qualified name of one or two journal receivers to be
associated with this journal.
Note: You can specify up to two journal receivers, but the
system ignores the second journal receiver.

Specifying Where Journal Messages Are Sent

When you create or change a journal, you can specify where messages that are
associated with the journal are sent. If you wish, you can create a program to
monitor this message queue and handle any messages associated with the journal.

Messages that are related to the remote journal function will also be sent to this
message queue. For more information, refer to “Troubleshooting Errors” on
page 491.

A common use for this message queue is to handle threshold messages. When you
create a journal receiver, you can specify a storage threshold. (See “Specifying a
Threshold for a Journal Receiver” on page 401.) The manage receiver (MNGRCV)
parameter for the journal determines what the system does when an attached
receiver reaches its storage threshold.

If you choose to change journal receivers yourself, you can specify where the
system sends messages when the journal receiver exceeds its storage threshold.
You might create a special message queue for this purpose and create a program to
monitor the message queue for the message (CPF7099). When the message is
received, the program can, for example, detach the receiver (CHGJRN) and save it.

If you specify MNGRCV(*SYSTEM) for the journal, the system does not send a
threshold message. Instead, the journal receiver is changed by the system. The
system sends a message (CPF7020), which indicates that the journal receiver for the
journal was successfully detached. The manage receiver (MNGRCV) parameter is
described in “Specifying How Journal Receivers for the Journal Are Changed” on
page 405.

| There are other messages which are sent to this journal message queue related to
| processing for the DLTRCV option. See “Deleting Journal Receivers” on page 418
| for more information.

404 OS/400 Backup and Recovery V5R1

 Possible Values for the MSGQ Parameter:
QSYSOPR If you specify MNGRCV(*USER) for the journal, messages
are sent to the QSYSOPR message queue.
message-queue -name If you specify MNGRCV(*USER) for the journal, messages
| are sent to this message queue. If this message queue is
| not available when the journal operation related messages
| are being sent, the message is sent to the QSYSOPR
| message queue.

Specifying How Journal Receivers for the Journal Are Changed

You can have the system automatically change journal receivers when the
threshold size for the journal receiver is exceeded. Or, you can choose to change
journal receivers using the CHGJRN command.

If you use system change-journal management (MNGRCV(*SYSTEM)), you can

avoid having to do some journal management chores. However, if you are
journaling for recovery purposes, you need to ensure that you save all journal
receivers that have not been saved, not just the currently attached receiver. Also, if
you are journaling for recovery purposes, be sure to specify DLTRCV(*NO).
“Having the System Delete Journal Receivers for the Journal” on page 406
describes the delete receiver (DLTRCV) parameter.

If you use system change-journal management support, you must ensure that your
environment is suitable and that you regularly check the QSYSOPR message queue
and the message queues assigned to your journals.
v If the system cannot complete the CHGJRN command because it cannot obtain
the necessary locks, it retries every 10 minutes. It sends messages (CPI70E5) to
the journal’s message queue and to the QSYSOPR message queue. If this occurs,
you may want to determine why the operation cannot be performed and either
correct the condition or run the CHGJRN command.
v If the system cannot complete the CHGJRN command for any reason other than
lock conflicts, it temporarily discontinues system change-journal management for
that journal and sends a message (CPI70E3) to the message queue assigned to
the journal or to the QSYSOPR message queue. This might occur because a
journal receiver already exists with the name that it would generate. Look at the
messages in the QHST job log to determine the problem. After you correct the
problem, use the CHGJRN command to do the following:
– Detach the current receiver
– Create and attach a new journal receiver
– The system then resumes system change-journal management.
v System change-journal management requires that the storage threshold for the
journal receiver be at least 5000KB.
Notes:
1. The value should be at least 5000. Using a value less than 5000 may cause
operational problems.
| 2. If you plan to use MNGRCV(*SYSTEM), the threshold must be a minimum
| of 100 000 KB if the receiver will be attached to a journal with
| RCVSIZOPT(*MAXOPT1) or RCVSIZOPT(*MAXOPT2) specified, otherwise,
| it must be a minimum of 5 000 KB.
| 3. If you plan to attach this journal receiver to a journal that does not have
| RCVSIZOPT(*MAXOPT1) or RCVSIZOPT(*MAXOPT2) specified, the
| maximum you can specify is 1 919 999 KB.

Chapter 19. Planning and Setting Up Journaling 405

 See “Specifying a Threshold for a Journal Receiver” on page 401 for information
about determining the correct storage threshold for your environment.

Possible Values for the MNGRCV Parameter:

*USER: The user manages the changing of journal receivers by issuing
the Change Journal (CHGJRN) command to attach a new
receiver and to detach an old receiver.
*SYSTEM: When an attached journal receiver reaches its size threshold,
the system detaches the attached journal receiver and creates
and attaches new journal receiver. The new journal receiver has
the same attributes as the detached receiver. During an initial
program load (IPL), the system performs a CHGJRN command
to change the journal receiver and reset the journal sequence
| number. Also, if the journal was attached while
| RCVSIZOPT(*MAXOPT1 or *MAXOPT2) was in effect for the
| journal, the system attempts to perform a CHGJRN command
| to reset the sequence number when the journal receiver’s
| sequence number exceeds 9 900 000 000 000. For all other
| journal receivers, the system attempts this CHGJRN when the
| sequence number exceeds 2 147 000 000.
Notes:
1. To use this option, you must specify a threshold value of at
least 5000KB when you create the journal receivers. If you
plan to use MNGRCV(*SYSTEM), the threshold must be a
minimum of 100 000 KB if the receiver will be attached to a
journal with RCVSIZOPT(*MAXOPT1) or
RCVSIZOPT(*MAXOPT2) specified, otherwise, it must be a
minimum of 5 000 KB. If you plan to attach this journal
receiver to a journal that does not have
RCVSIZOPT(*MAXOPT1) or RCVSIZOPT(*MAXOPT2)
specified, the maximum you can specify is 1 919 999 KB.
2. The system does not reset the journal sequence number
during an IPL if the entries in the receiver may be needed
for commitment control recovery.

Having the System Delete Journal Receivers for the Journal

If you choose system change-journal management, you can also have the system
delete journal receivers that are no longer needed for recovery. The system can
only evaluate whether a receiver is needed for its own recovery functions, such as
recovering access paths or rolling back committed changes. It cannot determine
whether a receiver is needed to apply or remove journaled changes.

The system cannot delete the journal receiver (even if DLTRCV (*YES) is used) if
any of the following conditions is true:
v An exit program that is registered for the Delete Journal Receiver (DLTJRNRCV)
exit point (QIBM_QJO_DLT_JRNRCV) indicates that the receiver is not eligible
for deletion.
v A journal has remote journals associated with it, and one or more of the
associated remote journals does not have a full copy of this receiver.
v The system could not get the appropriate locks that are required to complete the
operation.
v The exit program registration facility was not available to determine if any exit
programs were registered.

406 OS/400 Backup and Recovery V5R1

 If you use system delete-receiver support, you must ensure that your environment
is suitable. You must also regularly check the QSYSOPR message queue and the
message queues that are assigned to your journals.
v If the system cannot complete the DLTJRNRCV command for any of the above
reasons, it retries every 10 minutes. It sends a CPI70E6 message to the journal’s
message queue, and to QSYSOPR message queue. If this occurs, you may want
to determine why the operation cannot be performed and either correct the
condition or run the DLTJRNRCV command.
v If the system cannot complete the command for any other reason, it sends a
CPI70E1 message to the message queue that is assigned to the journal. If you
have not specifically assigned a message queue to the journal, the message will
be sent to the QSYSOPR message queue. Look at the messages in QHST to
determine the problem. After you correct the problem, use the DLTJRNRCV
command on the specific journal receiver.

Do not select to have the detached journal receiver deleted if you may need it for
recovery or if you want to save it before it is deleted. The system does not save the
journal receiver before deleting it. The system does not issue the warning message
(CPA7025) that it sends if a user tries to delete a receiver that has not been saved.

Examples of when you might specify DLTRCV(*YES) include:

v You are journaling for commitment control
v You are journaling for explicit access-path protection
v You are replicating the journal receiver to another system via the remote journal
function, and that system is providing the backup copy of the journal receiver.

Note: You can specify this parameter only if you specify MNGRCV(*SYSTEM) or if
the journal is a remote journal.

Possible Values for the DLTRCV Parameter:

*NO: The detached journal receivers are not deleted by the system.
*YES The detached journal receivers are deleted by the system when
they are no longer needed.

Specifying Receiver Size Options

Following are the possible values for the receiver size option (RCVSIZOPT)
parameter. The topics that follow describe these options.

Possible Values for the RCVSIZOPT Parameter:

*NONE: No receiver size options are selected.
*RMVINTENT Internal entries the system uses to recover during an IPL are
removed periodically when they are no longer needed.
*MINFIXLEN The job, program, and user profile information is not deposited
in the journal entries.
*MAXOPT1 Specifies that the journal receiver can have a maximum size of
approximately 1 TB and a maximum sequence number of
| 9 999 999 999. Additionally, the maximum size of the journal
| entry which can be deposited is 15 761 440 bytes.
| *MAXOPT2 Specifies that the journal receiver can have a maximum size of
| approximately 1 TB and a maximum sequence number of
| 9 999 999 999. Additionally, the largest journal entry can be
| 4 000 000 000 bytes.

Chapter 19. Planning and Setting Up Journaling 407

 Using RCVSIZOPT(*RMVINTENT): A journal receiver holds journal entries that
you might use for recovery and entries that the system might use for recovery. For
example, you might use record level entries, such as database record changes, and
file level entries, such as the entry for opening or closing a file. The system writes
entries that you never see or use, such as entries for explicitly journaled access
paths, for SMAPP, or for commitment control.

You can use the RCVSIZOPT parameter for the journal to have the system
periodically remove internal journal entries from the attached journal receiver
| when it no longer needs them to recover. Specifying RCVSIZOPT(*RMVINTENT)
| may have a very slight impact on system performance, because the system has to
| manage these internal entries separately and periodically remove them. Specifying
RCVSIZOPT(*RMVINTENT) has these benefits:
v It reduces the impact that SMAPP may have on journal receivers for user-created
journals.
v It reduces the size of journal receivers that are on the system.
v It reduces the amount of time and media required to save journal receivers,
because unnecessary entries are not saved.
v It reduces the time that it takes to apply journal entries, because the system does
not have to evaluate unnecessary entries.
| v It reduces the communications overhead if the remote journal function is being
| used because unnecessary entries are not sent.

Using RCVSIZOPT(*MINFIXLEN): The *MINFIXLEN value has the following

effects:
v The job, program, and user profile information is not deposited in the entries.
v This value saves DASD space and some CPU time as well.
v When you view journal entries with this information removed, the displayed
value is *OMITTED.
v To determine if a journal receiver was attached to a journal when *MINFIXLEN
was active, use the DSPJRNRCVA command display.
v You should not use *MINFIXLEN if you require an audit trail.
v *MINFIXLEN limits the selection criteria you can use on the DSPJRN, RCVJRNE,
RTVJRNE, CMPJRNIMG, APYJRNCHG, and RMVJRNCHG commands, and the
QjoRetrieveJournalEntries API.
| v *MINFIXLEN reduces the communications overhead if the remote journal
| function is being used because unnecessary entries are not sent.

Using RCVSIZOPT(*MAXOPT1): Use RCVSIZOPT(*MAXOPT1) to set the

maximum size of a journal receiver attached to your journal to approximately one
terabyte (1 099 511 627 776 bytes) and a maximum sequence number of
| 9 999 999 999. Additionally, the maximum size of the journal entry which can be
| deposited is 15 761 440 bytes. You cannot save or restore these journal receivers to
any releases prior to V4R5M0. Nor can you replicate them to any remote journals
| on any systems at a release prior to V4R5M0.

| Using RCVSIZOPT(*MAXOPT2): Use RCVSIZOPT(*MAXOPT2) to set the

| maximum size of a journal receiver attached to your journal to approximately one
| terabyte (1 099 511 627 776 bytes) and a maximum sequence number of
| 9 999 999 999. However, with RCVSIZOPT(*MAXOPT2), the system can deposit a
| journal entry as large as 4 000 000 000 bytes. You cannot save or restore these
| journal receivers to any releases prior to V5R1M0. Nor can you replicate them to
| any remote journals on any systems at a release prior to V5R1M0.

408 OS/400 Backup and Recovery V5R1

| Specifying Minimized Entry-Specific Data
| On the CRTJRN and CHGJRN commands, you can specify to make minimized
| journal entries. This will decrease the size of your journal entries. When you
| specify the MINENTDTA parameter for an object type, the entry-specific data for
| the entries of those object types may be minimized.
|| Possible Values for the MINENTDTA parameter:
| *NONE No object type allows a journal entry with minimized
| entry-specific data. Journal entries for all journaled objects will
| be deposited in the journal with complete entry-specific data.
| *FILE Journaled physical database files may have journal entries
| deposited with minimized entry-specific data.
| *DTAARA Journaled data areas may have journal entries deposited with
| minimized entry-specific data.
|

| Entries will only be minimized if the minimized entry is smaller in size than a
| complete journal entry deposit would be. Therefore, even if this option is specified,
| not all entries that are deposited will be minimized. An indicator is provided on
| the DSPJRN, RCVJRNE, RTVJRNE commands and QjoRetrieveJournalEntries API
| outputs to indicate whether the entry is actually minimized or not. See
| “Appendix D. Journal Entry Information” on page 805 to see which entries are
| eligible for minimization. See “Working with entries which contain minimized
| entry-specific data” on page 438 for more information and considerations when
| using these entries.

| You cannot save or restore a journal receiver with minimized journal entries to any
| release prior to V5R1M0, nor can they be replicated to any remote journal on a
| system at a release prior to V5R1M0.

Creating Journals and Journal Receivers–Examples

Assume the $DSTJRN library is in the system ASP:
Display Library Description

Library . . . . . . . . . . . . . . . . . : $DSTJRN

Type . . . . . . . . . . . . . . . . . . . : PROD
ASP of library . . . . . . . . . . . . . . : 1
Create authority . . . . . . . . . . . . . : *EXCLUDE
Text description . . . . . . . . . . . . . : Dist Journal Lib

Journal and Receiver in System ASP: The following command creates journal
receiver RCVDST1 in the $DSTJRN library:
CRTJRNRCV JRNRCV($DSTJRN/RCVDST1) THRESHOLD(10000)
TEXT('RECEIVER FOR $DSTJRN JOURNAL')

The journal receiver is placed in the system ASP with the library because *LIBASP
is the default value for the ASP parameter on the CRTJRNRCV command.

Public authority for the journal receiver is *EXCLUDE because the Create authority
value for the library is *EXCLUDE and the default for the authority (AUT)
parameter is *LIBCRTAUT.

This command creates the associated local journal:

Chapter 19. Planning and Setting Up Journaling 409

 CRTJRN JRN($DSTJRN/JRNLA) JRNRCV($DSTJRN/RCVDST1)
MNGRCV(*SYSTEM) DLTRCV(*NO)

The journal is placed in the system ASP with the journal receiver. The system
automatically changes journal receivers when the receiver exceeds 10 240 000 bytes
of storage (the threshold size for the RCVDST1 receiver). The detached receiver is
not deleted.

Journal Receiver in Nonlibrary User ASP: This command creates journal receiver
RCVDST2 in a nonlibrary user ASP:
CRTJRNRCV JRNRCV($DSTJRN/RCVDST2) THRESHOLD(10000)
ASP(2) TEXT('RECEIVER FOR $DSTJRN JOURNAL')

This command creates the local journal in the system ASP:

CRTJRN JRN($DSTJRN/JRNLB) JRNRCVR($DSTJRN/RCVDST2)
MSGQ($DSTJRN/JRNLBMSG)

When the receiver RCVDST2 exceeds 10 240 000 bytes of storage, a message
(CPF7099) is sent to the JRNLBMSG message queue in the $DSTJRN library.

| Note: The objects to be journaled must also be in the system ASP.

Journal and Journal Receiver in User ASPs: Assume the ARLIBR library is in a
user ASP:
Display Library Description

Library . . . . . . . . . . . . . . . . . : ARLIBR

Type . . . . . . . . . . . . . . . . . . . : PROD
ASP of library . . . . . . . . . . . . . . : 3
Create authority . . . . . . . . . . . . . : *USE
Text description . . . . . . . . . . . . . : A/R Receiver Lib

The following command creates journal receiver RCVDST3 in the library user ASP:
CRTJRNRCV JRNRCV(ARLIBR/RCVDST3) THRESHOLD(10000)
TEXT('RECEIVER FOR ARJRN JOURNAL')

Because public authority is not specified, the public authority is set to *USE (the
Create authority value for the ARLIBR library).

The ARLIB library is in a different user ASP:

Display Library Description

Library . . . . . . . . . . . . . . . . . : ARLIB

Type . . . . . . . . . . . . . . . . . . . : PROD
ASP of library . . . . . . . . . . . . . . : 4
Create authority . . . . . . . . . . . . . : *USE
Text description . . . . . . . . . . . . . : A/R Library

This command creates the local journal that is associated with the RCVDST3
journal receiver:
CRTJRN JRN(ARLIB/ARJRN) JRNRCV(ARLIBR/DSTRCV3)

When the RCVDST3 journal receiver exceeds 10 240 000 bytes of storage, a message
is sent to the QSYSOPR message queue (the default).

410 OS/400 Backup and Recovery V5R1

 Notes:
| 1. All objects journaled to the ARJRN journal must be in ASP 4 because the
| journal is in ASP 4.
| 2. In this case, the objects and journal are in the same library. The journal
| receivers are in a library that is saved and restored after the journal library if a
| single command is used, because ARLIBR comes after ARLIB in a normal sort
| sequence.

Starting Journaling
| After you have created the journal, use the appropriate start journal command for
| each object that you want to journal. When journaling has been started for an
| object, the system writes journal entries for all changes to the object.

| The start journal command must obtain an exclusive lock on the object. However,
| for database physical files and integrated file system objects, you can start
| journaling even if an object is open. The recommended procedure for starting
| journaling is:
| 1. Start journaling the object.
| 2. Save the object. If the object is open for changing, this will be a
| save-while-active type save.
| It is highly recommended that you update the history for the object when you
| save it so that APYJRNCHG and RMVJRNCHG processing will have the best
| information for verification. If you saved the object using the SAV command,
| the default value is to not preserve the update history. Therefore, you should
| change the UPDHST value to something other than *NO. For the other save
| related commands, the default value is to preserve the update history.

| How to Start Journaling for Physical Files

| Use the Start Journal Physical File (STRJRNPF) command to start journaling
| physical database files. When you start journaling, you specify whether you want
after-images saved, or both before-images and after-images. See “Should You
Journal Before-Images?” on page 392.

To reduce the number of journal entries, you can omit entries for open operations
and close operations for the file. Specify OMTJRNE(*OPNCLO) on the STRJRNPF
command to do this. If you choose to omit open journal entries and close journal
entries, be aware that:
v You cannot use the journal to audit who has accessed the file.
v You cannot apply or remove journal changes to open boundaries and close
boundaries using the TOJOBO and TOJOBC parameters.

Another way to reduce the number of journal entries for the file is to specify
SHARE(*YES) for the file. You can do this using the Create Physical File (CRTPF)
command or the Change Physical File (CHGPF) command. The system writes a
single open and close entry regardless of how often the shared open data path
(ODP) is opened or closed within a routing step.

For more information on database files, click the Database and file systems topic
in the Information Center at the following web site:
http://www.ibm.com/eserver/iseries/infocenter.

Chapter 19. Planning and Setting Up Journaling 411

 How to Start Journaling for DB2® Multisystem Files
When you successfully start journaling on a distributed file, the system distributes
the start journal request to the other systems in the node group. All systems are
attempted even if there is a failure at any one system. Once journaling is started on
a system in the node group, it stays started even if there is a failure at any of the
other systems.

The journal has to exist with the same name on all systems in the node group. The
journal itself is not distributed, only the STRJRNPF command.

The journal and its receiver are associated only with the changes made to the file
on the one system. If you have two systems in the node group and a file is
updated on both systems, the update on system A is only in system A’s journal
and receiver and the update on system B is only in system B’s journal and receiver.

The journal identifier (JID) will be different on each piece of the distributed file.
Each system piece will have it own JID. This means that the journal entries
deposited on one system cannot be used to APYJRNCHG or RMVJRNCHG those
entries to a different piece of the file on another system.

How to Start Journaling for Access Paths

After you have started journaling for physical files, you can set up explicit
journaling of access paths. You can use the Start Journal Access Path (STRJRNAP)
command to start journaling access paths owned by physical files or logical files.
When you start journaling access paths for a physical file, the system journals any
of these, if they exist:
v Keyed access paths
v Access paths for primary key constraints
v Access paths for unique constraints
v Access paths for referential constraints

All underlying physical files must be journaled to the same journal before you can
start journaling for an access path. The entries created when you journal an access
path are used to recover the access path after the system ends abnormally. They
are not used when you apply or remove journal entries. You can specify
RCVSIZOPT(*RMVINTENT) for the journal to have the system remove these
entries when they are no longer needed for recovery. This reduces the disk storage
requirements for the journal receiver.

You cannot start journaling for an access path that is in use. The STRJRNAP
command must obtain an *EXCL lock on the logical file.

The recommended procedure for starting access path journaling is:

1. Use the STRJRNAP command to start journaling the access path.
2. Save all the underlying physical files, specifying ACCPTH(*YES).

Note: If you have target recovery times for access paths set up on your system,
you may not need to set up explicit journaling for access paths. See
“Chapter 18. Protecting Access Paths Using System-Managed Access-Path
Protection” on page 375 for more information.

412 OS/400 Backup and Recovery V5R1

| How to Start Journaling for IFS objects
| Use the Start Journal (STRJRN) command’or Start Journal (QjoStartJournal) API for
| integrated file system (IFS) objects that you want to journal. You can journal the
| following IFS objects if they are in the Root(’/’), QOpensys, and User-defined file
| systems:
| v stream files(*STMF).
| v directories (*DIR)
| v symbolic link (*SYMLNK)

| For more information on IFS objects, click the Database and file systems topic in
| the Information Center at the following web site:
| http://www.ibm.com/eserver/iseries/infocenter.

| When you use the SAV command to save an IFS object, the default is to not update
| the history information for the object. If you plan to use the APYJRNCHG
| command for the objects you are journaling, you should specify to preserve the
| update history information on the SAV command.

| If you are journaling *DIR or *STMF objects, you can reduce the number of journal
| entries in the journal receiver. If you specify OMTJRNE(*OPNCLOSYN) on the STRJRN
| command you can omit entries for open operations, close operations, and force
| entries for the object. If specify OMTJRNE(*OPNCLOSYN) be aware that:
| v You cannot use the journal to audit who has accessed the object.
| v You cannot apply journal changes to open boundaries and close boundaries
| using the TOJOBO and TOJOBC parameters.
| v The value *OPNCLOSYN is only valid for *DIR and *STMF objects.

| If you are journaling symbolic links, the system does not follow the symbolic link
| to determine what to journal. That is, the system only journals the actual symbolic
| link. If you want to the journal the end object, you must journal the end object
| directly.

| If you are journaling a directory and select INHERIT(*YES) on the STRJRN

| command, then objects created into that directory will be journaled to the same
| journal. Therefore you should use caution because you could be journaling more
| objects than you realize. Also, even if this option is on, if an object is restored to
| the directory, it keeps the journaling attributes it had prior to the restore operation
| (when it was saved). For example, if you restore a stream file that is journaled to
| Journal X, but the directory you restore the stream file to is being journaled to
| Journal Y, the stream file will still be journaled to Journal X, even if the directory
| has the inherit option on.

| If you select SUBTREE(*ALL) on the STRJRN command, journaling only starts on

| objects that exist in the subtree when the STRJRN command is issued. To start
| journaling on objects added to the subtree after this point, you can either issue the
| STRJRN command for each object after it is created, or use the INHERIT option on
| the original STRJRN request.

| If there are object types in the subtree that are not supported for journaling when
| SUBTREE(*ALL) is specified, the unsupported object types are skipped over so that
| only object types that are supported for journaling get journaled.

Chapter 19. Planning and Setting Up Journaling 413

| How to Start Journaling for Data Areas and Data Queues
| After you have created the journal, use the Start Journal (STRJRN) command, the
| Start Journal Object (STRJRNOBJ) command, or the Start Journal (QjoStartJournal)
| API for each data area or data queue you want to journal. When journaling has
| been started for a data area or a data queue, the system writes journal entries for
| all changes to the data area or data queue.

| When you start journaling a data area, you specify whether you want after-images
| saved, or both before-images and after-images. See “Should You Journal
| Before-Images?” on page 392.

| For more information on data queues, see CL Programming. For more information
| on data areas see Work Management.

Why You Must Save Objects after Starting Journaling

Attention: To be able to apply journal changes, you must:
v Save an object after you start journaling it.
v Save a database physical file whenever you add a new member to it.
| v Save an IFS object whenever it is added to a directory and the object
| is automatically journaled because the directory has the inherit
| journaling option on.

| When you start journaling an object, the system assigns a unique journal identifier
| (JID) to that object. If the object is a physical database file, each member is also
| assigned a unique JID. The JID is part of every journal entry added to the journal
| receiver for a given object. The system uses the JID to associate the journal entry
| with the corresponding journaled object. The copy of the object on the save media
| that was saved before it was journaled does not have the journal identifier saved
| with it. Therefore, if this copy of the object is restored to the system, the journal
| entries cannot be associated with the object and cannot be applied. This is why it is
| critical to save the journaled object after journaling is started. Additionally, if the
| object is a physical file, you should save it every time a member has been added to
| it. This ensures that the journal identifiers are saved with the new file members.
| You should also save an IFS object whenever it is added to a directory and the
| object is automatically journaled because the directory has the inherit journaling
| option on.
| Notes:
| 1. The *TYPE4 format for the DSPJRN, RCVJRNE, or RTVJRNE command
| includes the JID for the object. The JID is also included with the *TYPEPTR
| format for the RCVJRNE command, as well as the QjoRetrieveJournalEntries
| API.
| 2. You can use the Retrieve JID Information (QJORJIDI) API to retrieve an object’s
| name (for a non-IFS object) or the file identifier (FID) (for an IFS object), if you
| know its JID. For more information about this API, click the Programming
| topic in the Information Center at the following Web site:
| http://www.ibm.com/eserver/iseries/infocenter.
| 3. If you start journaling on a distributed file, the piece on each system has its
| own unique JID.

414 OS/400 Backup and Recovery V5R1

| Save an object immediately after you have started journaling it, before any changes
| have occurred. Save a database file whenever you add a new member to it. This
| ensures that you can completely recover all the objects by using your saved copy
| and your journal receivers.

Attention:
| Update the history for the object when you save it so that APYJRNCHG and
| RMVJRNCHG processing will have the best information for verification. If
| you save the object using the SAV command, change the UPDHST value to
| something other than *NO. The default value for the SAV command is to not
| preserve the update history. For the other Save related commands, the default
| value is to preserve the update history.

Use one of these methods:

Physical database files, data areas, and data queues

v The Save Changed Objects (SAVCHGOBJ) command and specify
OBJTYPE(*object-type) OBJJRN(*YES)
v The Save Object (SAVOBJ) command
v The Save Library (SAVLIB) command
v The Save (SAV) command

| IFS objects
| v The Save (SAV) command

Save a physical file or a logical file after you start journaling access paths for the
file. This ensures that when you restore the file, journaling access paths is started
automatically.

If you are using distributed files, you must remember to save your file separately
on the systems in the node group after starting journaling for the distributed file.

Tasks for Managing Your Journaling Environment

Managing your journaling environment requires these basic tasks:
| v Keep records of which objects you are journaling.
v Evaluate the impact on journaling when new applications or logical files are
added.
v Regularly detach, save, and delete journal receivers.

Keep Records of Journaled Objects

| You should always have a current list of objects that you are journaling and their
| assigned journals. Print a new list whenever you add or remove objects from the
| journal. Do this to print a list:
1. Type WRKJRN and press the Enter key twice.
2. Write down the names of all the journals or use the PRINT key for each panel
of the display.
| 3. For each journal on the list that is used to journal objects, type WRKJRNA
| JRN(library-name/journal-name) OUTPUT(*PRINT).

Chapter 19. Planning and Setting Up Journaling 415

 Keep the lists with your most recent set of backup media that you used to save the
entire system.

You may need this list for the following reasons:

v You need to recover your journaling environment; for example, if the journal is
damaged or deleted. Although you can recover your journaling environment by
restoring the objects, in many cases starting journaling for the objects is a
quicker and safer method.
v You create new access paths. The system cannot protect access paths, either
explicitly or by using SMAPP, if the underlying physical files are not journaled
to the same journal.
| v You want to move objects to a user ASP. Journaled objects must be in the same
| user ASP as the journal, unless the objects are in the system ASP and the journal
| is in a nonlibrary user ASP.

Evaluate How System Changes Affect Journaling

After you have established your journaling environment, you need to keep up with
changes that occur on your system.

| When you add new applications, evaluate whether the objects should be journaled.

If you use SMAPP, the system automatically considers new access paths when
deciding how to meet your target recovery times for access paths.

Journaling places some limits on what changes you can make. For example:
v You cannot protect a logical file, either explicitly or with SMAPP, if the
underlying physical files are journaled to different journals.
| v You cannot move an object to a different ASP from the ASP of the library that
| contains its journal.

How to Manage Journals and Journal Receivers

| Your journal receivers enable you to recover changes to your important objects.
| They also provide an audit trail of activity that occurs on your system.

Protect your journal receivers by regularly detaching them and saving them; or
you can have the system take over the job of changing journal receivers by
specifying MNGRCV(*SYSTEM) for the journal.

If disk utilization is a problem on your system, you can free storage for journal
receivers when you save them. Freeing storage is preferable to deleting journal
receivers. Journal receivers whose storage has been freed still appear in the receiver
| directory for the journal. If disk utilization is not a problem, leave the journal
| receivers on the system until you have saved all the journaled objects.

Changing Journal Receivers

You can use the Change Journal (CHGJRN) command to change the attributes of
the journal. These attributes are described in “How Journals Are Created” on
page 402. You also use the CHGJRN command to change the receiver for a journal
(detach the current receiver, create and attach a new receiver) and to reset the
sequence number for journal entries.

| You can change the MNGRCV parameter, RCVSIZOPT parameter, or the

| MINENTDTA parameter for the journal only if you are also changing the journal
| receiver.

416 OS/400 Backup and Recovery V5R1

 If you use system change-journal management, the system changes journal
receivers for you.

When you use JRNRCV(*GEN) on the Change Journal (CHGJRN) command, the
system creates the new receiver with the same attributes as the currently attached
receiver, and in the same library. These attributes include the owner, private
authorities, public authority, object auditing, ASP identifier, threshold, and text.

Resetting the Sequence Number for the Journal: Normally, when you change
journal receivers, you continue the sequence number for journal entries by
specifying SEQOPT(*CONT), which is the default. When the sequence number
becomes very large, you can specify SEQOPT(*RESET) on the CHGJRN command
to start the numbering at 1. You can reset the sequence number only when all
changes are forced to auxiliary storage for all journaled objects and commitment
control is not active for the journal. Resetting the sequence number has no effect
on how the new journal receiver is named.

Some conditions prevent you from resetting the sequence number, such as an
active commit cycle. If the system cannot reset the sequence number, you receive
the message CPF7018. If you attempt to use the CHGJRN command with the same
journal receiver name and SEQOPT(*CONT), you may receive the message
CPF701A. To recover, delete the journal receiver and use the CHGJRN command
again.

If you use system change-journal management support (MNGRCV(*SYSTEM)) for

a journal, the sequence number for the journal is reset to 1 whenever you perform
an IPL. When you perform an IPL, the system performs the change journal
operation for every journal on the system that specifies MNGRCV(*SYSTEM). The
operation that the system performs is equivalent to CHGJRN JRN(xxx) JRNRCV(*GEN)
| SEQOPT(*RESET). The sequence number is not reset if journal entries exist that are
| needed for commitment control IPL recovery.

| The maximum sequence number is 2 147 483 136. If you specified

| RCVSIZOPT(*MAXOPT1) or RCVSIZOPT(*MAXOPT2) for the journal that you
| attached the receiver to, then the maximum sequence number is 9 999 999 999. If
| this number is reached, journaling stops for that journal. Whenever you change
| journal receivers, the system tells you what the starting sequence number is
| through message CPF7019. Also, when you are approaching the sequence number
| limit, CPF7019 is additionally sent to the QSYSOPR message queue every time you
| change journal receivers.

| The system sends a warning message (CPI70E7) to the journal’s message queue
| when the sequence number exceeds 2 147 000 000. If you specified
| RCVSIZOPT(*MAXOPT1) or RCVSIZOPT(*MAXOPT2) for the journal that you
| attached the receiver to, the system sends the warning message when the sequence
| number exceeds 9 900 000 000. If you use system change-journal management
| support (MNGRCV(*SYSTEM)) for the journal, the system attempts to change the
| journal and reset the sequence number one time. The message is sent only if the
| attempt is not successful.

When a journal receiver has been detached, it cannot be reattached to any journal.
You can do these things with a detached journal receiver:
v Save or restore it.
v Display entries.
v Retrieve entries.

Chapter 19. Planning and Setting Up Journaling 417

 v Receive entries.
v Use it to apply or remove journaled changes.
v Use it to compare journaled images.
v Display its status or position in a receiver chain.
v Delete it.
v Replicate it with the remote journal function.

Changing the Journal State

You can change the journal state of the local journal to allow or disallow the
deposit of journal entries. Refer to “Activating and Inactivating a Local Journal” on
page 487 for more information.

You can also change the journal state of a remote journal to inactivate the
replication of journal entries to that remote journal. Refer to “Inactivating the
Replication of Journal Entries to a Remote Journal” on page 485 for more
information.

Deleting Journal Receivers

You can use the Delete Journal Receiver (DLTJRNRCV) command to delete a
journal receiver if you no longer need it. You can also use an option from the Work
with Journal Receiver Directory display to delete a journal receiver. Use the Work
with Journal Attributes (WRKJRNA) command and press F15 to access the Work
with Journal Receiver Directory display.

If you have space on your system, wait to delete journal receivers until they are
| unlikely to be needed for a recovery. (The journaled objects have been saved.)
Restoring journal receivers before applying or removing journaled changes may
significantly increase your recovery time.

If you are journaling only for access path protection or for commitment control and
do not need the journal receivers to recover journaled changes, you can have the
system delete journal receivers that are no longer needed. Specify
MNGRCV(*SYSTEM) and DLTRCV(*YES) for the journal.

You cannot delete a journal receiver that is attached to a local journal. Journal
receivers must be deleted in the same order they were attached to a journal except
that a damaged or inoperable receiver can be deleted regardless of this restriction.
If an attached receiver is damaged, you must detach it (CHGJRN command) before
you delete it. You can also delete a journal receiver in the middle of a receiver
directory chain if one of the following conditions are true:
v The journal has DLTRCV(*YES) specified
v The journal is a remote journal

If you need the journal receivers for a recovery, you should ensure that they are
not deleted without first being saved. Display the journal receiver directory using
the WRKJRNA command and pressing F15 from the Work with Journal Attributes
display. The receiver directory shows which receivers have been saved. A date of
00/00/00 in the Save date column indicates that a journal receiver has not been
saved.

If you attempt to delete a receiver that has not been saved, the system sends the
inquiry message CPA7025. If the reply to the message is ignore (I), the journal
receiver is deleted. If the reply is cancel (C), the operation is not completed. You
can use the system reply list to specify the reply that the system sends for a
particular job.

418 OS/400 Backup and Recovery V5R1

 Note: No message is sent when the system deletes a journal receiver if you have
specified DLTRCV(*YES) for the journal. By specifying DLTRCV(*YES), you
have indicated that you do not need the journal receivers for user recovery.

If you attempt to delete a receiver that is attached to a remote journal, the system
sends the inquiry message CPA705E. The results of the reply to the message are the
same as those that occur with message CPA7025. You cannot delete a journal
receiver that is attached to a remote journal if the remote journal has a journal state
of *ACTIVE. Refer to “Chapter 21. Remote Journal Function” on page 467 for more
information.

An exit point is available for the Delete Journal Receiver (DLTJRNRCV) command
for systems that are running V4R2M0 or a later release. One example of using this
exit point is a situation where your application is using the data in the journal
receiver. The application is dependent on the journal receiver being present until
your application processing is complete. By registering an exit program with the
QIBM_QJO_DLT_JRNRCV exit point, the program will be called every time a
journal receiver is deleted from the system. If your program determines that your
application is not yet done with the receiver, it can indicate that the journal
receiver is not eligible for deletion.

If you must delete the receiver regardless of what an exit program indicates, you
can specify *IGNEXITPGM for the DLTOPT parameter on the DLTJRNRCV
command. This parameter value requests that any user exit programs that are
registered for QIBM_QJO_DLT_JRNRCV exit point be ignored.

You can also use the following values for the DLTOPT parameter:
*IGNTGTRCV
Ignore target receiver. If you specify this value, the system does not verify
that all remote journals that are associated with this journal, and are
immediately downstream on a target system, have full copies of this
journal receiver. The delete operation will continue, even if a remote
journal does not have a full copy.
*IGNINQMSG
Ignore inquiry message. Inquiry message CPA7025 will not be presented,
even if this receiver has not been fully saved. Also, inquiry message
CPA705E is not presented to the user even if the receiver is attached to a
remote journal. The delete operation continues.

Deleting a Journal
Each journal on the system causes additional time and resource to be used at each
IPL after an abnormal end. If you no longer need a journal, you should delete it
using the Delete Journal (DLTJRN) command. The system does not allow a journal
to be deleted if any of the following conditions exist:
| v You are journaling objects to it.
v Commitment control is active, and the journal is associated with a commitment
definition.

Note: If you have certain types of referential constraints defined, the system
starts commitment control if it is not already started. For example, if you
have defined a cascaded delete constraint for an object, the system starts
commitment control if you open the object for a delete operation. The
default commitment definition that is created is active until the job ends.
v Any of the associated remote journals have a journal state *ACTIVE.

Chapter 19. Planning and Setting Up Journaling 419

| If you need to recover objects that were journaled to a journal that you deleted,
| restore the journal from a saved copy or create a new journal with the same name
| in the same library.

Then restore all the needed receivers before using the APYJRNCHG or
RMVJRNCHG command to apply or remove changes with that journal. You can
use an option on the Work with Journals display to reassociate any journal
receivers that are still on the system (if they were created on Version 3 Release 1 or
later). Use the Work with Journals (WRKJRN) command. If the receivers were
created on earlier versions, you must restore them from newest to oldest. See
“Journal Receiver Chains” on page 440.

If you no longer need a journal and its associated receivers, perform the following
steps:
1. If commitment control is active and the journal is associated with it, end
commitment control with the ENDCMTCTL command.
2. End journaling of all logical files associated with the journal using the
ENDJRNAP command.
| 3. End journaling of all database files associated with the journal using the
| ENDJRNPF command.
| 4. End journaling of all IFS objects associated with the journal using the ENDJRN
| command.
| 5. End journaling of all other object types associated with the journal using the
| ENDJRNOBJ command.
6. If any commitment definitions are active that use this journal as the default
journal, use the ENDJOB command to end the jobs that are using the
commitment definitions. This includes commitment control that is started
because of a referential constraint.
7. If any remote journals have a journal state of *ACTIVE, deactivate them by
using the Change Journal State (QjoChangeJournalState) API or the Change
Remote Journal (CHGRMTJRN) command. Refer to “Inactivating the
Replication of Journal Entries to a Remote Journal” on page 485 for more
information.
8. Delete the journal using the DLTJRN command.
9. Delete the journal receiver using the DLTJRNRCV command.

Saving Journals and Journal Receivers

You can use these commands to save journals and journal receivers:
v Save Object (SAVOBJ)
v Save Changed Objects (SAVCHGOBJ)
v Save Library (SAVLIB)
| v Save (SAV)

You can save a journal receiver when it is attached to the journal. You should save
the journal receiver again when it is no longer attached, so that you have all the
journal entries saved.

When you save a journal receiver that is no longer attached, you can free storage.
However, a journal receiver whose storage has been freed must be restored before
it can be used for recovery.

Following are several examples of approaches you might take in detaching and
saving journal receivers.

420 OS/400 Backup and Recovery V5R1

Using SAVCHGOBJ to Save Journal Receivers: One technique for saving journal
receivers is to use the SAVCHGOBJ command. For example, if all your journal
receivers are in a library called RCVLIB, use this command:
SAVCHGOBJ OBJ(*ALL) LIB(RCVLIB) OBJTYPE(*JRNRCV)
DEV(media-device-name) ENDOPT(*LEAVE)

This saves all journal receivers that have any new entries since the entire library
was saved. The advantage to this method is that you can completely automate the
saving of journal receivers. You can leave a save media volume mounted and
schedule a job to run periodically. If you are managing journal receivers yourself,
the job can run the CHGJRN command for each journal and then run the
SAVCHGOBJ command. If you are using system change-journal management, the
job needs to run only the SAVCHGOBJ command.

Also, you can specify a threshold for the journal receiver and message queue for
the journal. If you have specified MNGRCV(*USER) for the journal, you can create
a CL program to do the following:
1. Monitor the journal message queue for the message (CPF7099).
2. When the message is received, run the CHGJRN command.
3. Run the SAVCHGOBJ command to save all journal receivers that have changed
since the entire library was saved.

A possible disadvantage to using the SAVCHGOBJ command to save journal

receivers is that you save the journal receivers that are currently attached. They are
saved as partial receivers. If you need to do a recovery, you may need to handle
the error condition that occurs when you attempt to restore the partial receiver
over the receiver that is currently on the system and has not yet been saved.

Saving Journal Receivers by Name–Manual Method: Another technique is to

save journal receivers individually. You can use the WRKJRNA command to
display the receiver directory for each journal. The receiver directory tells which
journal receivers have not yet been saved. Then use the SAVOBJ command to save
them.

The advantage to using this technique is that each journal receiver is saved only
once. You will not have problems with duplicate names and partial receivers if you
need to restore. The disadvantage to this technique is that it requires manual effort
to determine the names of the journal receivers to be saved.

Saving Journal Receivers by Name–Automated Method 1: You can use a

combination of system change-journal management and a CL program to automate
most journal management tasks. Do the following:
v Specify a threshold size for the journal receiver.
v Specify MNGRCV(*SYSTEM), DLTRCV(*NO), and a message queue for the
journal.
v Use a CL program to monitor the journal message queue for the message
(CPF7020) that indicates that the system has successfully detached the journal
receiver.
v Your CL program can then save the receiver that was detached and optionally
delete it.

Saving Journal Receivers by Name–Automated Method 2: An alternate method

of automatically saving journal receivers is to use a program that utilizes the
Retrieve Journal Information (QjoRetrieveJournalInformation) API. You can use this

Chapter 19. Planning and Setting Up Journaling 421

 API to determine the journal receiver directory and which receivers are not saved.
The program could then save the journal receivers that are not marked as saved.
This program could be set up to run on a regular basis or as part of normal
processing.

| Keeping Records of Your Journal Receivers: Choose the method for saving
| journal receivers that works best for your organization. Then be sure to keep track
| of what you do. Label your save media so that you know which journal receiver
| media volumes are required to apply journal changes to the last complete saved
| copy of the journaled objects.

Think through possible recovery scenarios. For example, assume this is your save
procedure:
v You save all user libraries on Sunday evening.
| v You save changed objects every evening.
v You save journal receivers every 2 hours during normal business hours.

| What are your recovery steps if you lose a journaled object at 3 p.m. on Thursday?

Correct Order for Restoring Objects Associated with a Journal

For the system to automatically reestablish your journaling environment, restore
objects in this sequence:
1. Journals
2. Based-on physical files
3. Dependent logical files
| 4. Other journaled object types
5. Journal receivers.

Note: The journal receivers can actually be restored at any point after the journals
are restored. They do not have to be restored after the journaled objects.

When these objects are in the same library, the system restores them in the correct
| sequence. When these objects are in different libraries or directories, you must
| restore them in the correct sequence, or you must manually reestablish your
| journaling environment after the restore operation.

If all of the journal receivers were created on V3R1 or later, you can restore them
in any sequence. After restoring them, use option 9 (Associate receivers with
journal) from the Work with Journal (WRKJRN) command display to build the
receiver chain in the correct sequence. Option 9 can also be used to build the
receiver chain if you restore the journal after the journal receivers.

If any journal receivers were created before V3R1, you must restore the journal
receivers from newest to oldest to build the receiver chain correctly. The journal
must be on the system for the receiver chain to be built. See “Journal Receiver
Chains” on page 440 for more information.

| If you restore journaled objects before restoring the journal, you must use one of
| the commands to start journaling again:
| v The Start Journal Physical File (STRJRNPF) command
| v The Start Journal Access Path (STRJRNAP) command
| v The Start Journal (STRJRN) command
| v The Start Journal Object (STRJRNOBJ) command

422 OS/400 Backup and Recovery V5R1

 Note: Your journals and journal receivers may be in different libraries. If this is
true, you should ensure that the library that will contain the journal
receivers is on the system before restoring the journal. Ensuring this will
also ensure that the journal receiver is created in the desired library, since a
journal receiver is created when the journal is restored. Only the library
needs to be on the system, not the journal receivers in that library. If you do
not ensure this, you may need to create a journal receiver in the desired
journal receiver library. You would then have to run the Change Journal
(CHGJRN) command to attach the new receiver to your journal.

Ending Journaling
You may need to end journaling for several reasons:
v If a journal is damaged and you need to delete it, you must first end journaling
for all objects assigned to the journal.
v In some situations, you may want to end journaling before running a large batch
application, if that application has exclusive use of the object. This is done either
to improve the speed of the batch application or to reduce the auxiliary storage
needed for the journal receiver. If you do this, use this method:
| 1. End journaling for the objects.
2. If journaling physical files save them specifying ACCPTH(*YES).
| 3. If journaling other object types, save them.
4. Run the batch application.
| 5. Start journaling for the objects.
| 6. Save the physical files, specifying ACCPTH(*YES).
| 7. Save the other journaled objects.

Use the following commands to end journaling :

v End Journal Access Path (ENDJRNAP) command
v End Journal Physical File (ENDJRNPF) command
| v End Journal (ENDJRN) command
| v End Journal Object (ENDJRNOBJ) command

| You must end journaling for any access paths based on a physical file or data area
| before you can end journaling for the physical file.

In the following cases, the system implicitly ends journaling:

| v When you delete an object, journaling is ended for the object.
v When you remove a physical file member, journaling is ended for the member.
v When you remove a physical file member, journaling is ended for any access
paths associated with the member unless an access path is shared and journaled
by another file member.
v When you delete a file, journaling is ended for any access paths associated with
the file unless an access path is shared and journaled by another file.

How to End Journaling for DB2 Multisystem Files

When you successfully end journaling on a distributed file, the system distributes
the end journal request to the other systems in the node group. All systems are
attempted even if there is a failure at any one system. Once journaling is ended on
a system in the node group, it stays ended even if there is a failure at any of the
other systems.

Chapter 19. Planning and Setting Up Journaling 423

 Even if a distributed file is not locally journaled, and if you specify the file name
and the journal name on the ENDJRNPF command, the system will still attempt to
distribute the end-journal request to the other systems in the file node group.

Managing IBM-Supplied Journals

The operating system and some licensed programs use journals to provide audit
trails and assist with recovery. Table 64 lists some of the IBM-supplied journals:
Table 64. IBM-Supplied Journals
Journal Name Library Name Description
QACGJRN QSYS Keeps job accounting information. The Work Management
book describes using this optional journal.
QAOSDIAJRN QUSRSYS Provides recovery for the document library files and the
distribution files. Used by Integrated xSeries Server for
iSeries and OfficeVision.
QAUDJRN QSYS Keeps an audit record of security-relevant activity on the
system. The iSeries Security Reference book describes
using this optional journal.
QCQJMJRN QUSRSYS Provides an audit trail for Managed System Services.
QDSNX QUSRSYS Provides an audit trail for DSNX activity.
QLYJRN QUSRSYS Keeps a log of transactions made to the Application
Development Manager datastore files. Used by the
system if recovery is necessary. The ADTS/400:
Application Development Manager User’s Guide provides
more information about this journal.
QLYPRJLOG QUSRSYS Keeps the project logs for the Application Development
Manager licensed program. Used by the system if
recovery is necessary. The ADTS/400: Application
Development Manager User’s Guide book provides more
information about this journal.
QLZALOG QUSRSYS Used by the licensed management program to log
requests that exceed the usage limit of a license.
QPFRADJ QSYS Keeps a log of dynamic performance tuning information.
The Work Management book describes using this optional
journal.
QSNADS QUSRSYS Provides an audit trail for SNADS activity.
QSNMP QUSRSYS Provides an audit trail for network management
information. The Simple Network Management Protocol
(SNMP) Support book describes using this journal.
QSXJRN QUSRSYS Provides a log of the activity that occurs in the database
files for service-related activity. The information in this
journal should be kept for 30 days.
QVPN0001 QUSRSYS Provides an audit trail for Virtual Private Networking
(VPN) connections. The TCP/IP Configuration and
Reference describes this journal.
QZCAJRN QUSRSYS Contains a record for each SNMP PDU in and out of the
SNMP agent, by PDU type (SNMP GET, SNMP
GETNEXT, SNMP SET, SNMP TRAP). The TCP/IP
Configuration and Reference book provides more
information about this journal.
QZMF QUSRSYS Provides an audit trail for the mail server framework.
The AnyMail/400 Mail Server Framework Support book
provides more information about this journal.

424 OS/400 Backup and Recovery V5R1

If you are using licensed programs or system functions that require these journals,
you should consult the documentation for those functions for instructions on how
to manage the journals and journal receivers.

In general, you should use the Change Journal (CHGJRN) command to detach the
journal receiver and create and attach a new receiver on a regular basis. You may
need to save detached receivers before deleting them, or you may be able to delete
them without saving them. This depends on how the journal receivers are being
used and whether the journal is defined as MNGRCV(*SYSTEM).

In some cases, you can use the automatic cleanup function of Operational Assistant
to remove detached journal receivers that are no longer needed. The System
Operation book describes using the automatic cleanup function.

Following is a description of the procedure for changing journal receivers.

1. Type CHGJRN(library-name/journal-name) JRNRCV(*GEN). This command:
a. Detaches the currently attached receiver.
b. Creates a new receiver with a name and attributes that follow the
conventions established by the function that uses the journal.
c. Attaches the new receiver to the journal.
2. Use the Save Object (SAVOBJ) command to save the detached journal receiver.
Specify object type *JRNRCV.
3. Use the Delete Journal Receiver (DLTJRNRCV) command to delete the receiver.
If you try to delete the detached receiver without saving it, you receive a
warning message.

Chapter 19. Planning and Setting Up Journaling 425

426 OS/400 Backup and Recovery V5R1
 Chapter 20. Working with Journal Entries, Journals, and
Journal Receivers
This chapter describes working with journaling information, including the
following:
v Contents of journal entries
v Sending your own journal entries
v Displaying and retrieving journal entries
v Receiving journal entries
v Working with the receiver chain
v Using the Work with Journal Attributes command
v Applying and removing journaled changes
v Using the Work with Journal command

About Journal Entries

Every journal entry is stored internally in a compressed format and must be
converted by the operating system to an external form before it can be shown to
the user. You cannot modify or access the journal entries directly. Not even the
| security officer can remove or change journal entries in a journal receiver. You can
| use these journal entries to help you recover your objects or analyze changes that
| were made to the objects.

Table 65 shows the methods available to work with journal entries:

Table 65. Methods for Working with Journal Entries
Method How It Is Used
| Note: For more information about these commands and APIs, click the Programming topic in the Information Center
| at the following Web site: http://www.ibm.com/eserver/iseries/infocenter.
| Apply Journaled Changes Uses the journal entries to apply changes that have occurred since an object was
| (APYJRNCHG) command saved or to a specified point. “Applying Journaled Changes” on page 446
| describes this process.
Compare Journal Images Compares and lists the difference between the before- and after-image of a
(CMPJRNIMG) command record or between the current after-image of a record and the previous
after-image of the record. See “Comparing Journal Images” on page 440 for more
information.
Delete Pointer Handle Deletes the specified pointer handle that was generated by the Retrieve Journal
(QjoDeletePointerHandle) API Entries (QjoRetrieveJournalEntries) API, or by the RCVJRNE command with
ENTTYP(*TYPEPTR).
Display Journal (DSPJRN) Displays or prints the journal entries that are in the journal receivers associated
command with the specified journal. Writes the journal entries to a database output file for
processing or analysis. See “Displaying and Printing Journal Entries” on
page 429.
Get Path Name of Object from Its Determines an absolute path name of the file identified by fileid. The
File ID (Qp0lGetPathFromFileID()) components of the returned path name are not symbolic links.
API
Receive Journal Entry (RCVJRNE) Allows a specified user program to continuously receive journal entries. Entries
command can be received one at a time or in a block. Can be used to help provide backup
on another system. See “Using the Receive Journal Entry Command” on
page 432.
Retrieve Journal Entry (RTVJRNE) Retrieves a journal entry and places it in CL program variables. See “Using the
command Retrieve Journal Entry Command” on page 436.

© Copyright IBM Corp. 1997, 2000, 2001 427

 Table 65. Methods for Working with Journal Entries (continued)
Method How It Is Used
Retrieve Journal Entries Retrieves journal entries into a receiver variable.
(QjoRetrieveJournalEntries) API
Remove Journaled Changes Removes changes that have occurred to an object from a specified point to a
(RMVJRNCHG) command previous point (allowed only if before-images and after-images were journaled
for the object). “Removing Journaled Changes” on page 450 describes this
process.
| Send Data Queue (QSNDDTAQ) Sends data to the specified data queue. You can use parameter group three to
| API indicate whether the data, in parameter four, came from a journal entry.
Send Journal Entry (SNDJRNE) Adds user entries to a journal receiver. You can use them to record specific
command or Send Journal Entry events (application checkpoints) that may be important for recovery. You can
(QJOSJRNE) API create user entries to perform user-managed journaling of object types the
system does not journal. See “Sending Your Own Journal Entries” on page 429.

Contents of a Journal Entry

| The entries that are created when there is a change that is made to a journaled
| object contain:
| v Information that identifies the type of change.
| v Information that identifies the data that was changed.
| v The after-image of the data.
| v Optionally, the before-image of the data (this is a separate entry in the journal).
| v Information that identifies the job, the user, the time of change, and so on.
| v The journal identifier of the object. This is used to present the name and library
| of the object. For the IFS objects, the name is presented as a file identifier (FID).
| You can use the Qp0IGetPathFromFileID to determine the IFS object path name
| v Information that indicates if the entry-specific data is minimized.

| The system also places entries in the journal that are not for a particular journaled
| object. These entries contain information about the operation of the system and the
| control of the journal receivers.

Each journal entry is sequentially numbered without any missing numbers until
| the Change Journal (CHGJRN) command resets the sequence number. However,
| sequence numbers may be missing when journal entries are converted and shown
| to the user because the system uses some entries only internally.

| When the system exceeds the largest sequence number, a message is sent to the
| system operator identifying the condition and requesting action. No other journal
| entries can be added to the journal until the journal receivers are changed and the
| sequence number is reset. “Changing Journal Receivers” on page 416 provides
| more information about what the system does when you approach the maximum
| sequence number and how to reset it.

A journal entry that is converted for displaying or processing contains a

fixed-length prefix portion that is followed by a variable-length portion. The
variable-length portion contains entry-specific data and, in some cases, null-values
indicator data. The format of the converted entry depends on the command that
you use and the format that you specify. The entry-specific data varies by entry
type. The SNDJRNE command or the QJOSJRNE API specifies the entry-specific
data for user-created journal entries.

428 OS/400 Backup and Recovery V5R1

| See “Appendix D. Journal Entry Information” on page 805 for all details on all
| possible journal entries and the information associated with them.

Sending Your Own Journal Entries

Use the Send Journal Entry (SNDJRNE) command or the Send Journal Entry
(QJOSJRNE) API to add your own entries to a journal. The system places these
entries in the journal’s attached journal receiver along with the system-created
journal entries.

| To help identify your entries, you can associate each entry with a particular
| journaled object. If you use the QJOSJRNE API, you can include the commit cycle
| identifier with the journal entry and send a larger amount of entry-specific data.

You may add entries to the journal to identify significant events (such as a
checkpoint) or to help in the recovery of your applications. On the SNDJRNE
command, the data specified on the ENTDTA parameter becomes the Entry-Specific
Data field in the journal entry, and the TYPE parameter value becomes the entry
type field. On the QJOSJRNE API, you use the entry data parameter to specify the
entry-specific data and the journal entry type parameter to specify the entry type.
For both the command and API deposits, the entries journal code is ’U’.

For more information about using the QJOSJRNE API, click the Programming topic
in the the Information Center at the following Web site:
http://www.ibm.com/eserver/iseries/infocenter.

Displaying and Printing Journal Entries

Use the Display Journal (DSPJRN) command to display journal entries. The entries
are displayed at a work station, printed, or written to an output file. You cannot
directly access the journal entries in the form in which they are contained in the
journal receivers.

“Appendix D. Journal Entry Information” on page 805 describes each type of

journal entry and the information that it contains. It shows the layouts for the
fixed-length portion and the variable-length portion of the journal entry. Look
under the Programming topic on the Information Center at the following Web site:
http://www.ibm.com/eserver/iseries for complete layouts for the model database
output files that are provided by the system.

Often, to prepare for a recovery, you display or print the journal entries first.
“Appendix D. Journal Entry Information” on page 805 lists all the possible journal
entry types. Use this list to help you analyze the journal entries and to do the
following:
| v Prepare for the recovery of a particular object. The list contains the information
| you need to specify the starting and ending points for applying and removing
| journaled changes.
| v Determine the functions that have been performed on the objects that are being
| journaled (such as save and restore, clear, reorganize).
v Determine the functions that have been performed on the journal (such as
attaching new journal receivers).
v Determine the functions that have been performed on the associated journal
receivers (such as save and restore).
| v Review the activity that has occurred on an object.

Chapter 20. Working with Journal Entries, Journals, and Journal Receivers 429
 v Analyze journal entries for debugging or problem analysis.
v Analyze journal entries for an audit trail.

The DSPJRN command either can selectively list journal entries for a particular
member of a file or list the entries for all files within a particular library. You can
further identify journal entries by specifying other selection criteria such as:
v Journal entries for specific entry types or journal codes, such as U (user-created
entries)
v Journal entries for a particular job, program, or file
v Commit cycle identifier
v Date and time
| v Dependent entries (referential integrity, triggers, and entries that will be ignored
| during an Apply Journaled Changes (APYJRNCHG) or Remove Journaled
| Changes (RMVJRNCHG) operation)
v Any combination of these:

The online help describes all the parameters for the DSPJRN command. To view
the help, type DSPJRN on a command line and press F1.

Specifying Journal Codes

You can display entries that have specific journal codes, such as all
file-member-level entries (F), all record-level entries (R), or all security entries (T).
You specify journal codes in paired values. The first value in the pair is the journal
code. The second value indicates whether the file selections you have specified
should apply when deciding to display entries with the journal code.

Following is an example:
DSPJRN JRN($JRNLIB/JRNA) FILE(CUSTLIB/FILEA)
JRNCDE((F *ALLSLT) (R *ALLSLT)
(U *IGNFLSLT))...

In this example, entries for the FILEA file with journal codes F and R are displayed
if the entries meet all other selection criteria, such as date and time. Entries with
journal code U are displayed regardless of whether they are for file FILEA, because
ignore file selection (*IGNFLSLT) is specified for journal code U. Entries with
journal code U must meet all other selection criteria, such as date and time, to be
displayed. See Table 110 on page 807 for a list of the possible valid journal codes.

Output for Journal Entries Directed to a Work Station

If you direct the output from the DSPJRN command to the requesting work
station, basic information about the journal entries appears. Use the roll key to
display the next sequential set of entries. When the receiver range includes an
attached journal receiver, TOENT(*LAST) is specified on the command, and the last
journal entries in the journal are displayed. Press the Page Down key to display
any new journal entries that are added to the attached receiver since the last time
the Page Down key was pressed.

Note: The attached journal receiver in receiver range refers to the journal receiver
that was currently attached when the DSPJRN command was first issued.
That journal receiver could be detached while you are looking at the data
on-line. If that occurs, paging down does not display any entries added after
that receiver was detached.

430 OS/400 Backup and Recovery V5R1

 Output for Journal Entries Directed to a Database Output File
If you direct the output from the DSPJRN command to a database output file, you
can further restrict the journal entries you want to process by creating logical files
over the database output file.

Each journal entry occupies one record in the output file. Each has a fixed length
portion for standard files. Before-images and after-images occupy separate records.
The ENTDTALEN parameter controls the length of the field that is used to contain
the record image. The ENTDTALEN parameter also controls whether the field is a
fixed or variable length field. If the journal entry is smaller than the output file
record, the journal entry is padded with blanks. If the journal entry is larger than
the output file record, the remainder of the journal entry is truncated, and the
system issues a warning message. To avoid truncation, specify the maximum
record length in your files for the ENTDTALEN parameter on the DSPJRN
command or specify *CALC for the ENTDTALEN parameter to allow the system
to calculate the length of the specific data field so no entry is truncated.

If you write journal entries to a database output file, you can write application
programs that will process the data to:
v Write your own apply program.
v Correct data that has been incorrectly updated.
v Remove or review all changes that were made by a particular program.
However, if you remove all changes that were made by a particular program,
| you could remove some valid updates. For example, assume that two work
| station users are using the same program to update an object, and one user
| enters some data that is not valid. If you remove all invalid data changes that
are made by that program, you also remove the valid data that is entered by the
other work station user.

Format of Database Output Files

When you direct the output of the DSPJRN command to a database file, the system
creates the output file records in a standard format. The system creates the
database file in one of these standard formats that are determined by the value
that is specified for the OUTFILFMT parameter:
Type Format
*TYPE1
The system uses the QJORDJE format in the model output file QADSPJRN
to create the output file for the converted journal entries. See Table 111 on
page 815.
*TYPE2
The system uses the QJORDJE2 format in the model output file QADSPJR2
to create the output file for the converted journal entries. In addition to all
the fields that are included in the *TYPE1 format, this format includes two
additional fields: one for the user profile which the job was running under
when the journal entry was added, and one for the name of the system on
which the journal entry was added. See Table 112 on page 818.
*TYPE3
The system uses the QJORDJE3 format in the model output file QADSPJR3
to create the output file for the converted journal entries. In addition to all
the fields included in the *TYPE2 format, this format includes a different
format for the date field and a field for the null value indicators that

Chapter 20. Working with Journal Entries, Journals, and Journal Receivers 431
 correspond to the record image in the journal entry. “Variable-Length
Portion of a Journal Entry” on page 821 describes the null value indicator.
See Table 113 on page 819.
*TYPE4
The system uses the QJORDJE4 format in the model output file QADSPJR4
to create the output file for the converted journal entries. In addition to all
the fields that are included in the *TYPE3 format, this format includes
information about referential constraints and trigger programs. This format
includes the journal identifier (JID) that is associated with the entry, if there
| is one. This format also has a field that indicates whether or not the journal
| entry will be ignored by the execution of APYJRNCHG or RMVJRNCHG
| commands. See Table 114 on page 820.

You can create an output file to hold the output from the DSPJRN command, but
the format has to match the format of one of the IBM-supplied output files.
“Appendix D. Journal Entry Information” on page 805 provides information about
these output files.

Processing Journal Entry Data

There are many ways to work with the journal entry data, including the
entry-specific data, depending on the command that you use to process the journal
entry data.
v Use your high-level language (HLL) to subdivide the fields into subfields.
v Use the Retrieve Journal Entry (RTVJRNE) command and the substring built-in
function.
v Use the Receive Journal Entry (RCVJRNE) command and the substring built-in
function.
v Use the Retrieve Journal Entries (QjoRetrieveJournalEntries) and map out the
data that is returned.

Analyzing Your Journal Activity

You can use the DSPJRN command to help analyze your journal entries. For
example, you could determine how many of each type of entry (such as add or
update) was done for a specific file or by a specific user.

Using the Receive Journal Entry Command

You can write a program to receive journal entries as they are written to the
journal receiver. When you use the Receive Journal Entry (RCVJRNE) command,
you can specify a user-defined program, called an exit program, to receive journal
entries. The program can, for example, write the entries to tape or to an OS/400
intersystem communications function (ICF) file that sends them to a backup
| system. You can use the received entries to update a backup copy of the primary
| object on the backup system. You cannot use these received entries with
| system-supplied recovery commands (APYJRNCHG and RMVJRNCHG) to update
| your objects because the RCVJRNE command converts the entries to their external
| form. You must write your own program to apply the changes that are contained
| in the entries to the objects.

The RCVJRNE command supports the same selection criteria as the DSPJRN
command. You can specify which entries go to the exit program.

For example, you can choose not to receive journal entries that are generated by
the action of trigger programs or referential constraints. If you have a user-written
program that updates the files on a second system with the journal entries, you

432 OS/400 Backup and Recovery V5R1

 probably want to specify DEPENT(*NONE). The actions performed by trigger
programs or referential constraints are duplicated automatically on the second
system if your database definitions are the same and you replay the original file
operations.

You can specify DELAY(*NEXTENT) to have journal entries sent to your program
as soon as they are written to the journal receiver. You can also specify a time
interval. When that interval ends, the exit program is called. Either new entries are
sent or an indicator is sent that there are no new entries.

Exit Program for Receiving Journal Entries

You use two parameters to communicate between your exit program and the
system when you are receiving journal entries. The system uses the first parameter
for the contents of one or more journal entries that it is passing to the exit
program. The exit program uses the first parameter to indicate the block length if
the exit program requests block mode.

The system and the exit program use the second parameter to communicate about
status changes, such as requesting block mode or ending the RCVJRNE command.
The second parameter is a character field that is three bytes long. Following are the
possible values for the first byte of the second parameter:

Possible Values for the First Byte of the Second Parameter:

0 This value is passed from the system to the exit program. It indicates that no
journal entry is being passed on this call of the exit program.
1 This value is passed from the system to the exit program. It indicates that a
single journal entry is being passed on this call of the exit program. If the
specified entry format is not *TYPEPTR, then Figure 34 on page 435 shows
the layout of the first parameter. Otherwise, the layout is the same as
returned to the QjoRetrieveJournalEntries API interface. For more
information, look in the Information Center at ath following Web site:
http://www.ibm.com/eserver/iseries/infocenter.
2 This value is passed from the system to the exit program. It indicates that
block mode is in effect. One or more journal entries are being passed on this
call of the exit program. If the specified entry format is not *TYPEPTR, then
Figure 35 on page 435 shows the layout of the first parameter.Otherwise, the
layout is the same as returned to the QjoRetrieveJournalEntries API interface.
For more information, see the Information Center at the following Web site:
http://www.ibm.com/eserver/iseries/infocenter.
3 This value is passed from the system to the exit program. It indicates that no
journal entry is being passed on this call of the exit program because the
journal receiver that was attached when the RCVJRNE command was started
is no longer attached. The system ends the RCVJRNE command after
returning this value to the exit program.
4 No journal entry is passed on this call to the exit program, and no more
entries can be passed unless the local or remote journal is activated.
Note: This value can only be passed to the exit program when receiving
journal entries from the attached receiver of a local or remote journal. The
journal state for the journal must be *INACTIVE.
8 This value is passed from the exit program to the system. It indicates that the
system should begin block mode and pass multiple entries to the exit
program. See “Requesting Block Mode” on page 435 for more information.
9 This value is passed from the exit program to the system. It indicates that the
RCVJRNE command should be ended.

Chapter 20. Working with Journal Entries, Journals, and Journal Receivers 433
 Possible Values for the Second Byte of the Second Parameter:
N This value is passed from the system to the exit program. Additional journal
entries are not currently available to be passed after this call of the exit
program, or the RCVJRNE command will end after this call of the exit
program.
Y This value is passed from the system to the exit program. Additional journal
entries are currently available to be passed after this call of the exit program.

Possible Values for the Third Byte of the Second Parameter:

’00’ x One or more journal entries are being passed to the exit program and the
object names in the fixed length portion of each journal entry do not
necessarily reflect the name of the object at the time the journal entry was
deposited into the journal.
Note: This value is only returned when receiving journal entries from a
journal receiver that was attached to a journal prior to V4R2M0.
0 No journal entries are currently being passed, so the information that is
normally returned in this byte is not applicable.
1 One or more journal entries are being passed to the exit program. The object
names in the fixed length portion of each journal entry reflect the name of
the object at the time the journal entry was deposited into the journal.
2 One or more journal entries are being passed to the exit program. The object
names in the fixed length portion of each journal entry do not necessarily
reflect the name of the object at the time the journal entry was deposited into
the journal. The object name in the fixed length portion of the journal entry
may be returned as a known name for the object prior to the journal entry
being deposited into the journal. The object name in the fixed length portion
of the journal entry may also be returned as *UNKNOWN.
Note: This value will only be returned when receiving journal entries from a
remote journal and the remote journal is currently being caught up from its
source journal. A remote journal is being caught up from its source journal
when the QjoChangeJournalState API or CHGRMTJRN command is invoked
and is currently replicating journal entries to the remote journal. After the
call to the QjoChangeJournalState API or CHGRMTJRN command returns,
the remote journal is maintained with a synchronous or asynchronous
delivery mode, and the remote journal is no longer being caught up.

Refer to “Retrieving Journal Entries from a Remote Journal During the

Catch-up Phase” on page 506 in the Chapter 21. Remote Journal Function
chapter for more information.

Any information that is passed from the exit program to the system in the second
byte or third byte is ignored.

The second byte of the second exit program parameter is provided whether journal
entries are being processed as a single journal entry per call of the exit program, or
as a block of journal entries per call.

Note: When an N is passed to the exit program in the second byte of the second
parameter indicated that no additional journal entries are currently
available, it does not necessarily mean that when the exit program returns,
that the RCVJRNE command will have to wait for additional journal entries
to be deposited into the journal. By the time the exit program returns,
additional journal entries may already be available and depending upon
what was specified on the DELAY parameter, may or may not be
immediately passed to the exit program. If DELAY(N) was specified the

434 OS/400 Backup and Recovery V5R1

 system will wait N seconds before passing the journal entries to the exit
program. If DELAY(*NEXTENT) was specified, the journal entries will
immediately be passed to the exit program.

Requesting Block Mode

You can request block mode at any time. When you specify 8 for the value of the
first byte of the second parameter, you must specify the block length in the first 5
bytes of the first parameter as a zoned decimal (Zoned (5,0)) field. 99999 bytes is
the maximum block size. After you have requested block mode, the system
remains in block mode until the RCVJRNE processing is ended.

If you request block mode and the system is already using block mode, your
request is ignored. You cannot change the size of the block from the size you
specified when you first requested block mode.

Format of the First Parameter: If the specified entry format is not *TYPEPTR, and
if you are using single-entry mode, the format of the first parameter looks like
Figure 34:

│W─LnA─S │
│W─────────A─────────────S│W00000S│

Legend: LnA = Length of Entry (Zoned 5,0)

A = Entry (including length of entry)
00000 = End of Record

Figure 34. First Parameter of RCVJRNE Command–Single-Entry Mode

The first 5 bytes contains the length of the entry. The last 5 bytes contains all
zeroes. The length of the entry does not include the 5 bytes of zeroes at the end of
the record.

If the specified entry format is not *TYPEPTR, and if you are using block mode,
the format of the first parameter looks like Figure 35:

LnBLK
│W─────S│

│W─LnA─S │
│W───────────A───────────────S│

│W─LnB─S │
│W───────────B─────────────────────S│

│W─LnC─S │
│W───────────C───────────────────S│W00000S│

Legend: LnBlk = Length of the Block (Zoned 5,0)

LnA, LnB, LnC = Length of the Entry (Zoned 5,0)
A, B, C = Entry (including length of entry)
00000 = End of Record

Figure 35. First Parameter of RCVJRNE Command–Block Mode

The first 5 bytes contains the total length of the block. This length includes the 5
bytes for the total block length, the 5 bytes of the End of Record field at the end of

Chapter 20. Working with Journal Entries, Journals, and Journal Receivers 435
 the block, and all of the length and data fields in between. If no entry is being
passed, this Block Length field contains zeroes. The block always ends with a
5-byte End of Record field containing zeroes.

The system fills the block with as many complete entries as it can fit within the
block size that you specified. The system does not send a partial entry to fill the
block size. If the specified entry format is not *TYPEPTR, the maximum number of
bytes that are available for the journal entries is 99989 bytes. 10 bytes in each block
are reserved for the Block Length field and for the End of Record field. If the
specified entry format is *TYPEPTR, the maximum number of bytes that are
available is 99999 bytes.

If you specify a block size that is not valid, the system begins block mode but it
sends only one journal entry per block. The system sends message CPD7095 to
indicate that you have specified a block size that is not valid. If you specify a block
size that is not valid or too small for a single journal entry, the system still returns
at least one journal entry to the exit program. If the specified entry format is
*TYPEPTR, the block size must be at least 13 bytes to be considered valid.

When the System Sends a Record: When block mode is in effect, the system uses
the following rules to determine when to call the exit program:
v If the block does not contain any entries but the next entry would exceed the
maximum size for the block, then the entry is placed into the block. The exit
program is called. The system always passes at least one complete journal entry
to the exit program.
v If the next entry to be put into the block would exceed the maximum size for
the block and the current block has entries in it, then the current block of entries
is passed to the exit program.
v If the current block has one or more entries in it and no additional entries in the
journal meet the selection criteria, the current block of entries is passed to the
exit program.

When in block mode, the specification for the DELAY parameter is used only when
the current block is empty and no entries are currently available to be returned to
the exit program.

Using ENTFMT(*TYPEPTR) With the RCVJRNE Command: If the specified

entry format is *TYPEPTR, the layout of the journal entry data is the same as the
layout that is described in the QjoRetrieveJournalEntries API interface. The layout
is the same for both single entry and block entry mode when you specify
*TYPEPTR. For more information, see the Information Center at the following Web
site: http://www.ibm.com/eserver/iseries/infocenter.

When you specify *TYPEPTR, the journal entry data may have pointers that will
point to additional entry-specific data. See “Working With Pointers in Journal
Entries” on page 437 for more information.

Using the Retrieve Journal Entry Command

Use the Retrieve Journal Entry (RTVJRNE) command in a CL program to retrieve a
journal entry and place it in variables in the program. You can retrieve the
following:
v Sequence number
v Journal code
v Entry type

436 OS/400 Backup and Recovery V5R1

 v Journal receiver name
v Library name for the journal receiver
| v Journal entry-specific data

You can use this method to create programs to automate recovery. “Appendix D.
Journal Entry Information” on page 805 provides the layout of the fixed-length
portion and the variable-length portion of the journal entry. For the format of the
| record for the RTVJRNE command, click the Programming topic in the Information
| Center at the following Web site: http://www.ibm.com/eserver/iseries/infocenter.
“Using the Retrieve Journal Entry (RTVJRNE) Command in a Program” on
page 766 shows an example program.

Using the Retrieve Journal Entries (QjoRetrieveJournalEntries) API

The Retrieve Journal Entries (QjoRetrieveJournalEntries) API allows you to retrieve
journal entries into a receiver variable. The available journal entry information is
similar to what is provided by using the Display Journal (DSPJRN), Receive
Journal Entry (RCVJRNE), and Retrieve Journal Entry (RTVJRNE) commands. This
API also provides additional journal entry data that cannot be retrieved with these
commands. This additional data is accessed using pointers. Refer to “Working With
Pointers in Journal Entries” for more information.

For further details, including the layout of the information that is returned, look in
the Information Center under the Programming topic at the following Web site:
http://www.ibm.com/eserver/iseries/infocenter.

Working With Pointers in Journal Entries

Under certain conditions, not all of the journal entry data will be immediately
retrievable from a journal entry. Instead, part of the journal entry information will
include pointers to additional journal entry-specific data. These pointers will only
be retrieved if the QjoRetrieveJournalEntries API or the *TYPEPTR format is used
on the RCVJRNE command. In all other retrievals of journal entry data, *POINTER
would be in the field where a pointer could exist. An incomplete data indicator has
been added to indicate if the journal entry-specific data has data missing which
can only be retrieved thru a pointer.

If the QjoRetrieveJournalEntries API or the *TYPEPTR format on RCVJRNE

command is used and the incomplete data indicator field is 1, the journal
entry-specific data will contain pointers. For all other interfaces, if the incomplete
data indicator is 1, the journal entry-specific data will have the character string
*POINTER in the field where an actual pointer would be placed if the API or
*TYPEPTR interfaces were used. The incomplete data indicator field could be set to
1 if the journal entry-specific data exceeds 32766 bytes, or if the journal entry is
associated with a database file which has one or more fields of data type BLOB
(binary large object), CLOB (character large object), or DBCLOB (double-byte
character large object). See Table 110 on page 807 for which journal entry types can
set the incomplete data indicator on.

These pointers can only be used with the V4R4M0 and later versions of the
following languages:
v ILE/COBOL
v ILE/RPG
v ILE/C if the TERASPACE parameter is used when compiling the program. See
ILE C/400® Programmer’s Guide, SC09–2069, for more information.

Chapter 20. Working with Journal Entries, Journals, and Journal Receivers 437
 There are some considerations you need to be aware of when using the pointer
data:
v The pointer can only be used by the process or job which retrieved or received
the journal entry which contained the pointer. The pointer cannot be passed on
to another job, nor can it be stored to use at a later date by a different job or
process.
v The pointer will only give you read access to the additional data. Write
operations to that pointer are not allowed.
v The data that is being pointed to actually resides in the journal receiver.
Therefore, you should ensure that you protect the journal receiver from deletion
until you use the data. To prevent a journal receiver from being deleted before
the data is used, you can register an exit point for the DLTJRNRCV command.
For more information, refer to “Deleting Journal Receivers” on page 418.
v For files with fields of data type BLOB (binary large object), CLOB (character), or
DBCLOB (double-byte character large object), use SQL to update the files. For
more information on the layout of the database records when LOB fields are
included, refer to the Information Center under the Database and File Systems
category.
| If you are using this journal for replication purposes, the journal entry can be
| used with the appropriate corresponding database operations if they are done
| using the ILE/C language. Contact your service representative for access to the
| information on how to activate this support.

If any journal entries are returned with pointers, the journal entry will also contain
a pointer handle. This pointer handle must be used to free up any allocations
associated with the pointer data once the pointer data has been used. The
considerations for this pointer handle are as follows:
v Using the pointer data means any of the following:
– Addressing the information and copying the addressed data to another object
– Using the journal entry-specific data directly to modify another object. For
example, using the data to update a database file with the journal entry
which represents a database record update for a file which included LOBs.
– Ignoring the additional data that is pointed to
v If you used the QjoRetrieveJournalEntries API, use the Delete Pointer Handle
(QjoDeletePointerHandle) API to delete the pointer handle when you are done
using it. For further information on the QjoDeletePointerHandle API,
http://www.ibm.com/eserver/iser
v If you used the RCVJRNE command, you must use the associated pointer prior
to returning control from the exit program. The system will delete all pointer
handles after the return from the exit program call. There is no problem if the
exit program were to also use the QjoDeletePointerHandle API to delete the
pointer handle prior to returning control from the exit program.

| Working with entries which contain minimized entry-specific data

| If you have selected to use the MINENTDTA parameter for the journal, then some
| of your journal entries entry-specific data will be minimized. The entries will only
| be minimized if the minimization technique will deposit a journal entry which is
| smaller in size than the complete entry would be. See Table 110 on page 807 for
| which specific journal entry types can possibly be minimized. When the entry is
| minimized, the fixed-length portion of the journal entry will have the minimized
| entry-specific data indicator on. Currently, only data areas and database physical
| files can have their entry-specific data minimized.

438 OS/400 Backup and Recovery V5R1

| Data Area Considerations
| The layout of the data area entries which are minimized is exactly the same as the
| layout if the entry was not minimized. The only difference is that only the bytes
| which actually changed are deposited rather than depositing all the bytes on the
| change request. See Table 143 on page 853 for more information about the entry
| layout of the change data area entries.

| Database Physical File Considerations

| The layout of the minimized record changes entries is completely different than the
| layout when the entry is not minimized. The data it not even recognizable nor
| readable as sophisticated hash techniques are used in addition to only operating on
| actual changed bytes. Additionally, the Null-value-indicators field will be used,
| even if the file is not null capable, to provide additional information that can be
| used by database operations. Therefore, if you want to use the journal as an audit
| mechanism, you may not want to choose this option for database physical files
| since you will not be able to read the actual change made.

| If you are using this journal for replication purposes, the journal entry can be used
| with the appropriate corresponding database operations if they are done using the
| ILE/C language. Contact you service provider for access to the information on
| how to activate this support.

Using Journaling to Provide an Audit Trail

| You can use journal management to provide an audit trail of changes that were
| made to your objects. You can determine which program or user made changes to
| objects by using the journal entries. For database physical files, you can determine
| what changes were made to specific records by using the CMPJRNIMG command.

Journal management can be used to provide an audit trail because of the following
reasons:
v The security officer can not even remove or change the journal entries.
v The journal entries represent a chronological sequence of events.
v Each journal entry in the system is sequentially numbered without gaps until the
CHGJRN command resets the sequence number. A journal entry is written if the
sequence number is reset. When you display the journal entries, there can be
gaps in the sequence numbers because some journal entries are only used
| internally by the system. These gaps occur if you are using commitment control,
| database file journaling, or access-path journaling.
v The journal contains entries that indicate when each journal receiver was
changed and the name of the next journal receiver in the chain.
| v Whenever journaling for an object is ended or whenever an object is restored an
| entry is written.

Remember that the date and time recorded in the journal entries depends on the
date and time entered during an IPL and therefore, may not represent the actual
date and time. Also, if you use shared files, the program name that appears in the
journal entry is the name of the program that first opened the shared file.

A special journal, that is called the audit (QAUDJRN) journal, can provide a record
of many security-relevant events that occur on the system. The iSeries Security
Reference book provides more information about using the QAUDJRN journal.

Chapter 20. Working with Journal Entries, Journals, and Journal Receivers 439
 For a complete audit trail, you should not specify RCVSIZOPT(*MINFIXLEN). If
this parameter is specified, the job, program, and user profile information are not
journaled.

Comparing Journal Images

Use the Compare Journal Images (CMPJRNIMG) command to compare and list the
differences between the before-image of a record and the after-image of that record,
or the after-image of a record with the previous after-image of that record.

| You can only use the CMPJRNIMG command for journaled physical database files.
You cannot use the CMPJRNIMG command for journal entries that have
minimized entry-specific data. If you specified the MINENTDTA parameter on the
CRTJRN or CHGJRN commands, the journal entries have minimized entry-specific
| data.

If the journaled files have null-capable fields, the null value indicators
corresponding to the fields in the before-image of the record are compared with
the null value indicators corresponding to the fields in the after-image of the
record. A field-by-field basis does this.

The printed output from the CMPJRNIMG command shows the before-images and
after-images of a record followed by a line that indicates (with asterisks) the
specific change in the record on a character-by-character basis. If you compare the
after-images, the output shows the previous after-image of the record and the
current after-image of the record, followed by a line indicating the changes.

Note: If this command is used to compare journal images for a file that contains
any fields of data type BLOB (binary large object), CLOB (character large
object), or DBCLOB (double-byte character large object), these fields are not
included in the comparison. All other fields in the file are compared.

The online help provides more information about using the CMPJRNIMG
command. To view the help, type CMPJRNIMG on a command line, and press F1.

Journal Receiver Chains

Journal receivers that are associated with a journal (that is presently or previously
attached to the journal) are linked in one or more receiver chains. Each journal
receiver, except the first one, has a previous receiver that was detached when the
current receiver was attached. Each journal receiver, except the one that is currently
attached, also has a next receiver.

Figure 36 illustrates the process by which journal receiver chains are created. If you
leave the previously attached receivers RCVA7 through RCVA9 on-line, you can
use them to apply changes, to remove changes, or to display journal entries
without restoring them first.

440 OS/400 Backup and Recovery V5R1

 Journal
┌──────────┐
│ JRNA │
└──────────┘
O
│ ┌─────────────┐ Currently
└─────────┤ RCVA10 │ Attached
┌──S│ ├───┐ Receiver
│ └─────────────┘ │
Next │ ┌─────────────┐ │ Previous
Receiver └───┤ RCVA9 │W──┘ Journal
┌──S│ ├───┐ Receiver
│ └─────────────┘ │
│ ┬─────────────┐ │
└───┤ RCVA8 │W──┘
┌──S│ ├───┐
│ └─────────────┘ │
│ ┌─────────────┐ │
│ │ RCVA7 │ │
└───┤ │W──┘
└─────┬───────┘
N Receivers RCVA1
┌─────────────┐ through RCVA6 have
│ Backup │ been saved and
│ Media │ deleted from the
└─────────────┘ system.

Figure 36. Creating a Journal Receiver

If a complete copy of a receiver is missing in a chain of journal receivers linked

together in the previously described relationship, the result is a chain break.
Receiver chain breaks should be avoided. A receiver chain break indicates that any
changes made between the last entry in the last receiver in one chain and the first
entry in the first receiver in the next chain are not available in any journal receiver
on the system.

A set of receivers for a journal that has one or more receiver chain breaks has
multiple receiver chains. Receiver chain breaks result from the following:
v You restored an old journal receiver and its next receiver is not on the system.
v A journal receiver was saved while it was attached, a partial receiver is restored,
and no complete copy of the receiver is on the system or restored.
v A receiver that has not had its storage freed by a save operation is restored, and
the next receiver has had its storage freed by a save operation.
v The journal is restored. All journal receivers associated with the previous copy of
the journal (before the journal was deleted and restored) will not be in the same
receiver chain as the currently attached journal receiver.
v The user or the system deleted a damaged or destroyed journal receiver from
the middle of a chain.
v A journal receiver from another system is restored. The journal receiver will be
associated with a journal at restore time if the associated library and journal on
the source system had the same library name and journal name as the library
and journal on the target system.
v You chose to replicate specific receivers instead of all receivers in the receiver
directory chain. This occurred while replicating journal receivers from a source
system to a target system. For more information, refer to “Chapter 21. Remote
Journal Function” on page 467.

Chapter 20. Working with Journal Entries, Journals, and Journal Receivers 441
 The APYJRNCHG, RMVJRNCHG, RCVJRNE, DSPJRN, RTVJRNE, and
CMPJRNIMG commands, and the QjoRetrieveJournalEntries API, cannot be used
across multiple receiver chains. If multiple receiver chains exist, you need to
determine:
v Whether any journal entries are missing.
v Whether your data will be valid if you use more than one receiver chain.
If you decide to proceed, you must run a separate command for each receiver
chain.

You can use the Work with Journal Attributes (WRKJRNA) command to display
the receiver chain (F15) and work with journal receivers. See “Using the Work with
Journal Attributes Command”.

Using the Work with Journal Attributes Command

The Work with Journal Attributes display lets you display information about the
journal and related journal receivers. Use the WRKJRNA command. This display
identifies:
v The journal receivers currently attached to the journal
v A directory of the journal receivers still on the system that are associated with
the journal.
| v The names of all of the objects that are being journaled through the journal.
v The attributes of the journal
v Information about all remote journals that are associated with the journal

Working with the Receiver Directory

Figure 37 on page 443 shows an example of the Work with Receiver Directory
display. To reach this display, use the WRKJRNA command. Press F15 (Receiver
directory) from the Work with Journal Attributes display.

442 OS/400 Backup and Recovery V5R1

 Work with Receiver Directory

Journal . . . . . . : JRNLA Library . . . . . . : DSTJRN

Total size of receiver (in kilobytes) . . . . . . . . . . . . : 155648

Type options, press Enter.

4=Delete 8=Display attributes

Attach Save
Opt Receiver Library Number Date Status Date
_ RCVDST1 $DSTJRN 00001 05/18/9x SAVED 05/19/9x
_ RCVA0001 $DSTJRNA 00002 05/19/9x SAVED 05/20/9x
_ RCVA0002 $DSTJRNA 00003 05/20/9x PARTIAL 05/21/9x
_ RCVA1002 $DSTJRNA 01001 05/21/9x ATTACHED 00/00/00

Bottom

Parameters or command
===>____________________________________________________________________________
F3=Exit F4=Prompt F5=Refresh F9=Retrieve F11=Display size
F12=Cancel

Figure 37. Work with Receiver Directory

The partial status of a journal receiver on this display indicates the following:
v A journal receiver was saved while it was attached to the journal. This means
that additional entries will be recorded in the journal that is attached to this
receiver after the save operation has occurred. The receiver was later restored,
and no complete version is available.
v The journal receiver is associated with a remote journal. It does not contain all
the journal entries that are in the associated journal receiver that is attached to
the source journal.
v A partial receiver does not contain all the entries that are recorded in the journal
while this receiver was attached. It does contain entries that are recorded up to
the last save operation.

| You can use partial receivers to apply or remove changes from an object. If you
| attempt to restore a saved receiver while a more current version of the receiver is
| on the system, an escape message is sent to prevent you from restoring the
| receiver. The system makes sure that the most complete version is preserved.

You can use a partial receiver as the last receiver in the receiver chain for an
APYJRNCHG command only if you specify a sequence number for the TOENT
parameter. You can use a partial receiver as the first receiver in the receiver chain
for a RMVJRNCHG command only if you specify a sequence number for the
FROMENT parameter.

The system does not allow you to delete a journal receiver from the middle of the
receiver chain unless one of the following conditions exists:
v The journal has DLTRCV(*YES) specified
v The journal is a remote journal

Chapter 20. Working with Journal Entries, Journals, and Journal Receivers 443
 This ensures logical recovery. However, if a journal receiver is damaged, you can
delete it from the middle of the chain. If an attached journal receiver is damaged,
you must detach the damaged receiver (CHGJRN command) before you can delete
it.

The system does not prevent you from deleting a receiver that was once attached
and is not saved or that is required to provide adequate recovery. If you try to
delete a journal receiver that was once attached but has not been saved, the system
issues an inquiry message. You can then continue or cancel the delete operation.
You may use the system reply list to specify the reply the system is to send for this
inquiry message (rather than explicitly responding to each inquiry message).

You must ensure that the journal receivers are not deleted until you no longer need
them. Use the Work with Receiver Directory display to determine which journal
receivers have been saved. A date of 00/00/00 in the Save Date column indicates
that a journal receiver has not been saved. A status of PARTIAL indicates that the
receiver was saved when it was attached and then restored. More entries may have
been added after the receiver was saved. The most complete version of the journal
receiver is no longer on the system because it was destroyed during a failure. You
have restored an older, partial version. Refer to “Deleting Journal Receivers” on
page 418 for more information.

Inoperable Journal Receivers

| If you have specified journaling for any objects, the system ensures that you have
| corrected problems that affect journaling before continuing with operations on
| those objects. If the attached journal receiver becomes inoperable, the operation
that writes a journal entry is interrupted and the system sends an inquiry message
that notifies the system operator. The operator can use the CHGJRN command to
change the journal receiver. You can then respond to the inquiry message. A
receiver can become inoperable if the receiver is damaged, the maximum sequence
number has been reached, or there is no more space.

Applying and Removing Journaled Changes

| Use the Apply Journaled Changes (APYJRNCHG) command to apply changes to
| these object types:
| v Database file member
| v IFS object
| v Data area
| Use the Remove Journaled Changes (RMVJRNCHG) command to remove changes
| that were made to these object types:
| v Database file member
| v Data area

| To use the APYJRNCHG or RMVJRNCHG command to recover an object, the

| object must be currently journaled. The journal entries must have the same journal
| identifier (JID) as the object. To ensure the journal identifiers are the same, save the
| object immediately after journaling is started for the object and each time a
| member is added to the database file or an IFS object is added to a directory with
| the inherit journaling option on. “Why You Must Save Objects after Starting
| Journaling” on page 414 has more information about saving journaled objects and
| about JIDs.

444 OS/400 Backup and Recovery V5R1

 Some types of entries in the journal receiver cause the apply or remove process to
stop. These entries are written by events that the system cannot reconstruct. When
one of these events is encountered, the process ends and a message is sent
indicating the sequence number of the last journal entry that was successfully
applied or removed and the reason the process ends. Certain illogical conditions,
such as a duplicate key in a database file defined as unique, can also cause
processing to end.

Table 66 on page 452 shows how the APYJRNCHG and RMVJRNCHG commands
| handle journal entry types. It shows which entry types cause processing to end
| and what processing is done when the entry is applied or removed.

How Applying and Removing Journaled Changes Works with

Referential Constraints
When you apply or remove journaled changes, referential constraints are not
enforced. In the following cases, files may be in CHECK PENDING status after
you have applied or removed journaled changes:
v When you restore a file that already exists, the referential constraints for the
system copy of the file are used. Some of the journaled changes that you apply
may have been valid with the referential constraints that were associated with
the saved copy. However, they are not necessarily valid with the current
referential constraints. If you have changed the referential constraints on the file,
considering doing one of the following before applying or removing journaled
changes:
– Deleting the system copy and then restoring the file
– Recreating the changes to the referential constraints

When you apply or remove journaled changes, the system attempts to verify the
referential constraints at the end of the command, before returning control to
you. This may result in a CHECK PENDING status.
v Some referential constraints cause an action to another file. You may define a
constraint so that deleting a record in one file causes a related record to be
deleted in another file. Because referential constraints are not enforced when you
apply journaled changes, the second delete operation does not happen
automatically. However, if you are journaling both files and applying journaled
changes to both files, the system applies the journal entry for the second file
when it encounters it.
If one of the files in a referential constraint was not journaled or is not included
when you apply or remove journaled changes, the referential constraint will
probably be put in CHECK PENDING status.

The *TYPE4 and *TYPEPTR output format for journal entries and the
QjoRetrieveJournalEntries API interface includes information about whether a
journal entry was created because of changes that occurred to a record that was
part of a referential constraint.

Look on the Information Center under the Database and File System categories for
more information about working with referential constraints.

Chapter 20. Working with Journal Entries, Journals, and Journal Receivers 445
 How Applying and Removing Journaled Changes Works with
Trigger Programs
The system does not call trigger programs when it is applying or removing journal
entries. If an event occurs that would normally cause a trigger program to run, it is
up to you to ensure that the processing performed by the trigger program is
recovered correctly.

Normal recovery processing should work correctly if all of the following are true:
| v The trigger program only performs processing on object types which can be
| journaled and applied
| v The processed object types are journaled
| v Journaled changes are applied to or are removed from all the objects that are
| affected by the trigger program

| If additional work is performed by the trigger program or objects other than object
| types which can be journaled and applied are updated, you must use user-written
| programs to recover the work performed by the trigger program.

| If you use trigger programs to perform these actions, consider using the QJOSJRNE
| API to send journal entries when trigger programs are called. See “Sending Your
Own Journal Entries” on page 429. To help with recovery, you can develop a
program to retrieve these entries and perform the same operations.

The *TYPE4 and *TYPEPTR output format for journal entries and the
QjoRetrieveJournalEntries API interface includes information about whether a
journal entry was created because of actions that were performed when a trigger
program was called.

Look in the Information Center under the Database and File System topics for
more information about working with trigger programs.

Applying Journaled Changes

| If an object becomes damaged or is not usable. You can recover the object using
| the Apply Journaled Changes (APYJRNCHG) command directly, or by using the
| Work Journal (WRKJRN) command and following the prompts. For a description
| of the options, see the on-line information or refer to the topic “Work with Journal
| (WRKJRN) Command Options” on page 455. You must first reestablish the object to
| a condition that you know is undamaged.
| v To reestablish the object, restore the last saved copy of the object. The object
| must have been saved while it was being journaled.
| v If you saved a database physical file by using the CPYF command, use the CPYF
| command to restore the member.
| v If the member of a database physical file was just initialized, initialize the
| member again using the Initialize Member (INZPFM) command or a
| user-created application program.
v If a member of a database physical file was just reorganized, reorganize the
member again using the Reorganize Physical File Member (RGZPFM) command.

For more information about the CPYF, INZPFM, and RGZPFM commands, refer to
the online help. To view the help, type one of the commands on a command line,
and press F1.

446 OS/400 Backup and Recovery V5R1

 If the journal receivers are deleted or saved with their storage that is freed since
the object was last saved (or since some other point). You must restore the needed
journal receivers, from the newest to oldest.

Note: Journal receivers do not need to be restored in a particular sequence if they

are created on Version 3 Release 1 or later. The system establishes the
receiver chains correctly when they are restored.

| The system applies the changes to the object in the same order as they were
| originally made. When you use the APYJRNCHG command, the object cannot be
| in use by anyone else.

| When the condition of the object has been established, use the APYJRNCHG
| command to apply the changes that are recorded in the journal to the object. On
| the APYJRNCHG command, specify the first journal entry to be applied to the
| object. This entry can be selected from any of the following points:
| v After the last save of the object
v From the first journal entry
v From an identified sequence number that corresponds to a date and time stamp
| v From an identified sequence number that corresponds to the start or end of a
| particular job’s use of the object provided that you did not specify one of the
| following:
| – OMTJRNE(*OPNCLO) when starting journaling for the object
| – OMTJRNE(*OPNCLOSYN) when starting journaling for a directory or stream
| file
| – RCVSIZOPT(*MINFIXLEN) for the journal at any time while the object was
| journaled
v A specific sequence number.

You can stop applying the journal entries at:

v The end of the data in the last journal receiver in the receiver range
v A particular entry in the journal
v A date and time stamp
v A commitment boundary
| v The start or end of a particular job’s use of the data in the object provided you
| did not specify the following:
| – OMTJRNE(*OPNCLO) when starting journaling for the object
| – OMTJRNE(*OPNCLOSYN) when starting journaling for a directory or stream
| file
| – RCVSIZOPT(*MINFIXLEN) for the journal at any time while the object was
| journaled
| v The journal entry that indicates when the object was last restored
v A specific sequence number

You can ensure that commitment transaction boundaries are honored on the apply
journaled changes operations by using the commit boundary (CMTBDY) parameter
on these commands.

Note: If the system encounters a journal entry that causes the apply or remove
process to stop, the commitment boundary may not be honored. Table 66 on
page 452 shows which entry types cause processing to end.

Chapter 20. Working with Journal Entries, Journals, and Journal Receivers 447
 Use the Display Journal (DSPJRN) command to identify the desired starting and
ending points. If you use a control language (CL) program for your recovery
procedures, use the Retrieve Journal Entry (RTVJRNE) command to retrieve a
journal entry and place it in program variables. For information about specifying
the program variables, look under the Programming topic on the Information
Center at the following Web site: http://www.ibm.com/eserver/iseries.You can
also use the QjoRetrieveJournalEntries API to retrieve the information into a High
Level Language (HLL) program.

| Integrated File System Considerations

| Sometimes, if there is a create or delete entry in range of journal entries being
| applied, an apply of journaled changes to a directory could cause the creation or
| deletion of an object. This is different than what occurs for database physical files.

| If you are journaling a directory with the INHERIT(*YES) option, and an object is
| created into that directory, the system will automatically start journaling that object
| and deposit associated create and start journal object journal entries. The apply of
| these create and start journal entries during the apply operation on the directory
| will then create the objects and start journaling for them during they apply. For
| any subsequent journaled entries for that object, the apply operation will apply
| any entries that it encounters for that object as well. Similarly, if an entry is
| encountered which deletes an IFS object, that object is actually deleted as part of
| the apply operation.

| Additionally, the apply operation can include applying any IFS journal entry that
| adds a link to the journaled directory, such as moving a non-journaled object into
| the journaled directory, or adding a new hard link to a non-journaled object into
| this journaled directory.

| Note that as objects are created, they are included in the maximum number of
| objects which can be applied as part of one APYJRNCHG request. But, even
| though objects are deleted, they are still included in the maximum number of
| objects which can be applied limit.

| See Table 66 on page 452 for what operations are applied for IFS related journal
| entries.

| Many journaled IFS operations use system initiated commitment control for the
| duration of the operation. These operations are not considered completed
| successfully unless the commitment control cycle is committed.

| Note: Commitment control, here, refers to commitment control that the system
| initiates. IFS operations cannot be included in a user initiated commitment
| control cycle.
| For IFS journal entries that are part of a commitment control cycle, you should not
| apply individual entries from within the cycle without applying the entire commit
| cycle. Using CMTBDY(*YES) on the APYJRNCHG command can help enforce this.
| If you do not use this option, and choose a specific starting point, you choose to
| start applying entries in a commit cycle, from the Start of commit cycle (C SC)
| entry for that cycle. Likewise, if you choose to end applying at a specific point, end
| on the Commit (C CM) or Rollback (C RB) entry for that cycle.

Apply Journaled Changes (APYJRNCHG) Command Examples

| The following are examples of the APYJRNCHG command applied to a database
| physical file, integrated file system (IFS) object, and data area.

448 OS/400 Backup and Recovery V5R1

| The following examples show database physical files, data areas, and IFS objects
| being processed separately. However, you can use one APYJRNCHG command if
| you use the OBJ parameter for files and data areas, and the OBJPATH parameter
| for the IFS objects on one command invocation

| Database physical file

The following command applies the changes in journal JRNA to the first member
of all files in the library DSTPRODLIB that are being journaled to journal JRNA:
APYJRNCHG JRN(JRNLIB/JRNA) FILE((DSTPRODLIB/*ALL))

Because the RCVRNG parameter is not specified, the system determines the range
of journal receivers to use as a result of the save information for the files. Because
the FROMENT parameter is not specified, the system applies the changes that
begin with the first journal entry.

If the file was last saved with the save-while-active function, the saved copy of
each file member includes all record-level changes in the journal entries up to the
corresponding F SS journal entry. In this case, the system applies changes that
begin with the first journal entry that follows the F SS entry.

If the file was last saved when it was not in use (normal save), the saved copy of
each member includes all record-level changes in the journal entries up to the
corresponding F MS member saved journal entry. In this case, the system applies
changes that begin with the first journal entry that follows the F MS entry.

The following command applies the changes to the file from the journal receivers
that are currently attached to the journal:
APYJRNCHG JRN(JRNLIB/JRNA) FILE((LIBA/FILEA MBR1))
RCVRNG(*CURRENT) FROMENT(*FIRST)
TOENT(*LAST)

The *CURRENT journal receiver is the journal receiver that is attached to journal
JRNA at the beginning of the operation. The system applies the changes from the
first journal entry in this receiver to the last journal entry in this receiver. Changes
are applied to member MBR1 of the file FILEA.

The following command applies the changes in the journal JRNA to all members of
the file FILEA beginning with the first journal entry after the file member was last
saved:
APYJRNCHG JRN(JRNLIB/JRNA) FILE((LIBA/FILEA *ALL))
TOJOBC(000741/USERP/WORKSTP)

The operation continues until the specified job closes any of the members in the
file that it opened. The operation is not restricted only to those journal entries that
are recorded by the specified job.

Note: This example works only if you do not specify OMTJRNE (*OPNCLO) when
starting journaling for the file or you did not specify
RCVSIZOPT(*MINFIXLEN) for the journal at any time while the file was
journaled).

| IFS object

Chapter 20. Working with Journal Entries, Journals, and Journal Receivers 449
| The following command applies the changes in journal JRNA to the objects in the
| directory MyDirectory, and its subdirectories, that are being journaled to journal
| JRNA:
| APYJRNCHG JRN(JRNLIB/JRNA) OBJPATH(('/MyDirectory')) SUBTREE(*ALL)

| Because the RCVRNG parameter is not specified, the system determines the range
| of journal receivers to use as a result of the save information for the objects.
| Because the FROMENT parameter is not specified, the system applies the changes
| that begin with the journal entry for the last save of each of the objects.

| If the object was last saved with the save-while-active function, the saved copy of
| each object includes all record-level changes in the journal entries up to the
| corresponding B FW journal entry. In this case, the system applies changes that
| begin with the first journal entry that follows the B FW entry.

| If the object was last saved when it was not in use (normal save), the saved copy
| of each object includes all changes in the journal entries up to the corresponding B
| FS saved journal entry. In this case, the system applies changes that begin with the
| first journal entry that follows the B FS entry.

| Data area

| The following command applies the changes to the data area DATA1 from the
| journal receiver that is currently attached to the journal:
| APYJRNCHG JRN(JRNLIB/JRNA) OBJ((LIBA/DATA1 *DTAARA))
| RCVRNG(*CURRENT) FROMENT(*FIRST)
| TOENT(*LAST)

| The *CURRENT journal receiver is the journal receiver that is attached to journal
| JRNA at the beginning of the operation. The system applies the changes from the
| first journal entry in this receiver to the last journal entry in this receiver. Changes
| are applied to data area DATA1.

Removing Journaled Changes

| Depending on the type of damage to the journaled object and the amount of
| activity since the object was last saved, removing changes from the object can be
| easier than applying changes to the object. Use the Remove Journaled Changes
| (RMVJRNCHG) command directly or the Work with Journal (WRKJRN) command
| and follow the prompts to remove (or back out) changes from an object if
| before-images were journaled.

For a description of the journal options, see the topic “Work with Journal
(WRKJRN) Command Options” on page 455, or the on-line information for the
WRKJRN command. The changes are removed in reverse chronological order from
the order in which they were originally made to the object.

| On the RMVJRNCHG command, you identify the first journal entry to be removed
| from the object. This entry can be from:
v The last journal entry that is contained within the range of journal receivers
specified
| v The entry that corresponds to the last save of the object
v An identified sequence number

450 OS/400 Backup and Recovery V5R1

| You can control the changes that are removed from the object. For example,
| assume that an application updated data incorrectly for a period of time. In this
| case, you could remove the changes from the object until that application first
| opened the object.

You can stop removing journaled changes at:

v The end of data in the journal receivers. (This corresponds to the first journal
entry that was recorded on the range of journal receivers that are specified.)
v An identified sequence number that corresponds to a particular entry in the
journal.
| v The start of a particular job’s use of the object (if you did not specify
| OMTJRNE(*OPNCLO) when starting journaling for the file or you did not
| specify RCVSIZOPT(*MINFIXLEN) for the journal at any time while the object
| was journaled).

You can ensure that commitment transaction boundaries are honored on the
remove journaled changes operations by using the CMTBDY parameter on these
commands.

Note: If the system encounters a journal entry that causes the apply or remove
process to stop, the commitment boundary may not be honored. Table 66 on
page 452 shows which entry types cause processing to end.

Use the Display Journal (DSPJRN) command to identify the desired starting and
ending points for removing the changes. If you use a control language (CL)
program for your recovery procedures, use the Retrieve Journal Entry (RTVJRNE)
command to retrieve a journal entry and place it in program variables. For
information about the program variables for the RTVJRNE command, look under
the Programming topic on the Information Center at the following Web site:
http://www.ibm./eserver/iseries/infocenter. You can also use the
QjoRetrieveJournalEntries API to retrieve the information into a High Level
Language (HLL) program.

Remove Journaled Changes (RMVJRNCHG) Command

Examples
Even though the following examples show database physical files and data areas
being processed separately, you can do them with one RMVJRNCHG command if
you use the OBJ parameter for both object types.

Database physical file

The following command removes the changes in journal JRNA from the first
member of FILEA:
RMVJRNCHG JRN(JRNLIB/JRNA) FILE(DSTPRODLIB/FILEA)
RCVRNG(*CURRENT)

The *CURRENT journal receiver is the journal receiver that is attached to journal
JRNA at the beginning of the operation. The system starts removing the changes
beginning with the latest entry for that member in this receiver and continues to
the earliest entry for that member in this receiver.

The following command removes the changes in journal JRNA from the first
member of FILEA:

Chapter 20. Working with Journal Entries, Journals, and Journal Receivers 451
 RMVJRNCHG JRN(JRNLIB/JRNA) FILE(DSTPRODLIB/FILEA)
RCVRNG(JRNLIB/RCVA10 JRNLIB/RCVA8)

The system starts removing the changes beginning with the last entry (the latest
entry) for that member in journal receiver RCVA10 and continues to the first entry
(the earliest entry) for that member on journal receiver RCVA8.

| Data area

| The following removes the changes in JRNA from data area DATA1 from the last
| save entry to entry number 1003.
| RMVJRNCHG JRN(JRNLIB/JRNA) OBJ((LIBA/DATA1 *DTAARA))
| RCVRNG(*CURRENT) FROMENT(*LASTSAVE) TOENT(1003)

| If the last save operation used the save-while-active function, the system starts by
| removing changes from the entry preceding the last E EW start of save entry. If the
| last save operation was a normal save operation, the system starts by removing
| changes from the entry that precedes the last E EW member saved entry. In the
| example, journaled changes are removed back to entry 1003.

Actions of the APYJRNCHG or RMVJRNCHG Command by

Journal Code
Table 66 on page 452 shows the actions that are taken by the APYJRNCHG or
RMVJRNCHG command by journal code and entry type. If All is specified for the
Entry Type, it indicates that all entry types for that journal code have the specified
actions taken by the APYJRNCHG or RMVJRNCHG command.
Table 66. Actions by Journal Code and Entry Type
Journal Entry
Code Type Operation APYJRNCHG RMVJRNCHG
A All Ignores Ignores
| B AA Change audit attribute Attribute is changed Ignores
| B AJ Start of apply Ends Ignores
| B AT End of apply Ends Ignores
| B BD IFS object deleted Ignores Ignores
| B B0 Begin create Ignores Ignores
| B B1 Create summary Object is created and linked Ignores
| B B2 Link to existing object Object is linked Ignores
| B B3 Rename, move object Object is moved or renamed Ignores
| B B4 Remove link (parent directory) Object link is removed Ignores
| B B5 Remove link (link) Object link is removed Ignores
| B CS IFS object closed Ignores Ignores
| B ET End journaling for object Ends Ignores
| B FA IFS object attribute changed Attribute is changed Ignores
| B FC IFS object forced Ignores Ignores
| B FF Storage for object freed Ignores Ignores
| B FR IFS object restored Ends Ignores
| B FS IFS object saved Ignores Ignores
| B FW Start of save Ignores Ignores
| B JT Start journaling for object Ignores Ignores
| B OA Change object Authority is changed Ignores
| B OF IFS object opened Ignores Ignores
| B OG Change primary group Primary group is changed Ignores
| B OI Object in use at abnormal end, Ignores Ignores
| object is synchronized1

452 OS/400 Backup and Recovery V5R1

 Table 66. Actions by Journal Code and Entry Type (continued)
Journal Entry
Code Type Operation APYJRNCHG RMVJRNCHG
| B OI Object in use at abnormal end, Ends Ignores
| object is not synchronized1
| B OO Change Object Owner Owner is changed Ignores
| B TR IFS object truncated Object is truncated Ignores
| B WA Write, after-image Object is updated Ignores
C All Ignores Ignores
D All Ignores Ignores
| E EA Update data area, after image Data area modified Ignores
| E EB Update data area, before image Ignores Data area modified
| E ED Data area deleted Ends Ends
| E EG Start journal for data area Ignores Ends
| E EH End journal for data area Ends Ignores
| E EI Data area in use, object Ignores Ignores
| synchronized1
| E EI Data area in use, object not Ends Ends
| synchronized1
| E EL Data area restored Ends Ends
| E EM Data area moved Ignores Ignores
| E EN Data area renamed Ignores Ignores
| E ES Data area saved Ignores Ignores
| E EQ Data area changes applied Ends Ends
| E EU RMVJRNCHG command started Ends Ends
| E EW Start of save for data area Ignores Ignores
| E EX Data area changes removed Ends Ends
| E EY APYJRNCHG command started Ends Ends
F AY Journaled changes applied Ends Ends
| F CB Change File member Ignores Ignores
F CE Change end of data Member end of data Ends
changed2
F CL Member closed Ignores Ignores
F CH File changed Ignores Ignores
F CR Member cleared Member cleared of all Ends
records2
F DE Member deleted record count Ignores Ignores
| F DM Delete member Ignores Ignores
F EJ End journaling Ends Ignores
F EP End journaling access paths Ignores Ignores
F FD Member forced to auxiliary Ignores Ignores
storage
F FI Internal format information Ignores Ignores
F IU Object synchronized1 Ignores Ignores
F IU Object not synchronized1 Ends Ends
F IZ Member initialized Initialized records inserted in Initialized records
member deleted from member
F JM Start journaling member Ignores Ends
F JP Start journaling access paths Ignores Ignores
| F MC Create member Ignores Ignores
F MD Member deleted Ends Ends
F MF Member saved with storage freed Ends Ends
F MM Member moved Ignores Ignores
F MN Member renamed Ignores Ignores
F MR Member restored Ends Ends
F MS Member saved Ignores Ignores

Chapter 20. Working with Journal Entries, Journals, and Journal Receivers 453
 Table 66. Actions by Journal Code and Entry Type (continued)
Journal Entry
Code Type Operation APYJRNCHG RMVJRNCHG
F OP Member opened Ignores Ignores
F PD Access path deleted Ignores Ignores
F PM Logical owning member of access Ignores Ignores
path moved
F PN Logical owning member of access Ignores Ignores
path renamed
F RC Journaled changes removed Ends Ends
F RG Member reorganized Ends Ends
| F RM Member reorganized Ignores Ignores
F SA Start of APYJRNCHG Ends Ends
F SR Start of RMVJRNCHG Ends Ends
F SS Start of save active Ignores Ignores
I All Ignores Ignores
J All Ignores Ignores
L All Ignores Ignores
M All Ignores Ignores
O All Ignores Ignores
P All Ignores Ignores
| Q All Ignores Ignores
R BR Before-image updated for Ignores Record updated with
rollback operation before-image
R DL Record deleted Record deleted Record updated with
before-image
R DR Record deleted for rollback Record deleted Record updated
operation
| R IL Increment record limit Ignores Ignores
R PT Record written to member Record written to member Record deleted from
member
R PX Record added directly to member Record added Record deleted from
member
R UB Record updated (before-image) Ignores Record updated with
before-image
R UP Record updated (after-image) Record updated with Ignores
after-image
R UR After-image updated for rollback Record updated with Ignores
operation after-image
S All Ignores Ignores
T All Ignores Ignores
U User- User entry Ignores Ignores
specified
1
The Flag field in the journal entry indicates whether the object is synchronized (0 = object was synchronized;
1 = object was not synchronized).
2
Applying journaled changes stops at this entry if referential constraints that this entry violates are active
during the apply operation.

In addition to the entries that cause the command to end, the system ends the
Apply Journaled Changes (APYJRNCHG) or Remove Journaled Changes
(RMVJRNCHG) command if any format error (such as an undefined entry for that
file member) or logical error (such as updating a record that has not been inserted
or a duplicate key exception) is encountered when the command is run.

454 OS/400 Backup and Recovery V5R1

| For entries that end the APYJRNCHG or RMVJRNCHG command, a message
| identifying the reason for the end is placed in the job log, and the corresponding
| change is not made to the object. The message contains the sequence number of
the journal entry on which the failing condition was detected. Analyze the error,
make the necessary correction, and then start applying or removing journal
changes again using the appropriate sequence number.

For example, if the entry that caused the APYJRNCHG command to end is entry
code F of type RG, you must reorganize the physical file member referred to in the
journal entry. Use the same options that were originally specified on the reorganize
request when the journal entry was recorded in the journal receiver. Resume
applying journal changes by starting with the journal entry that follows the ’F RG’
reorganize physical file member journal entry.

| The APYJRNCHG and RMVJRNCHG commands send an escape message and

| ends the operation if any required journal receiver defined by the RCVRNG
| parameter is not on the system and associated with the journal. Use the WRKJRNA
command to select the journal receiver directory display, to see which journal
receivers are on the system and associated with the journal. The escape message
contains the name of the required journal receiver if the reason code of message
CPF7053 is 1 or if message CPF9801 is sent.

When the processing of the APYJRNCHG or RMVJRNCHG command ends with

an escape message, the objects can be partially changed. To determine how many
changes were applied or removed for each object, review the diagnostic messages
in the job log prior to the final escape message for each object, or use the DSPJRN
command to display the journal entries indicating completion of the command.

| The command completion journal entries by object type are as follows :

| Database physical file members
| F journal code and an entry type of AY or RC
| Integrated file system objects
| B journal code and entry type of AJ
| Data area objects
| E journal code and entry type of EQ or EX.

| The Count field in the journal entry contains the number of journal entries that are
applied or removed.

| The system puts out a maximum of 8192 diagnostic messages from Apply or
| Remove Journaled Changes. If you have more than 8192 objects with which you
| are working, then looking in the journal at the journal entries is probably the best
| way to determine how many changes were applied to the objects.

For more information on the journal codes, entry types, and journal entries, see
“Appendix D. Journal Entry Information” on page 805.

Work with Journal (WRKJRN) Command Options

The Work with Journal (WRKJRN) command shows a display that provides you
with options to display journal status and guides you through various types of
recovery involving journals, journal receivers, and journaled files. The options are
shown on the following display:

Chapter 20. Working with Journal Entries, Journals, and Journal Receivers 455
| Note: You cannot use this interface to assist with recovery if you are journaling
| object types other than database physical files and access paths.

Work with Journals

Type options, press Enter.

2=Forward recovery 3=Backout recovery 5=Display journal status
6=Recover damaged journal 7=Recover damaged journal receivers
9=Associate receivers with journal

Opt Journal Library Text

JRNACC DSTA1 JOURNAL FOR ACCOUNTS

Bottom
Selection or command
===>

F3=Exit F4=Prompt F9=Retrieve F12=Cancel

All of the options are valid for local journals. Only option 9 is valid for remote
journals.

Recovery Options
When it is necessary to restore an object during the following recovery procedures,
the system prompts for the values that are needed with a prompt of the RSTOBJ
command. The values that are known for OBJ, SAVLIB, and OBJTYPE parameters
appear, but they cannot be changed. If the values for DEV, VOL, SAVF, SAVDATE,
SAVTIME, and OPTFILE can be determined by the system (depending on the
amount of damage), they are also filled in. If the values are not shown, you must
enter the values.

If more than one object is restored, the system groups the objects according to the
library specified in the SAVLIB parameter. Within each group, the system further
orders the objects according to the values that are specified for the DEV, VOL,
SAVF, SAVDATE, SAVTIME, OPTFILE parameters. This grouping results in a
minimum number of restore object prompts as a result of information known to
the system.

Note: If the object was saved to an optical device, only the first 6 characters of the
volume ID and the first 17 characters of the optical file name can be
determined.

Work with Forward Recovery

Option 2 on the Work with Journals display shows the Work with Forward
Recovery display and a list of the file members that are being journaled.

456 OS/400 Backup and Recovery V5R1

 Work with Forward Recovery

Journal . . . . . . : JRNACC Library . . . . . . : DSTA1

Position to . . . . .
Library . . . . . .

Type options, press Enter.

1=Add member to list 2=Apply journaled changes 3=Restore
4=Remove member from list

Opt File Library Member Status

_ __________ __________ __________
_ PAYFILE1 PAYLIB QTR192

Bottom
F3=Exit F12=Cancel

The Work with Forward Recovery display contains a status field for each file
member. The status field for each member indicates the following:
v NOT FOUND. The system cannot locate the specified database file.
v DAMAGED. The member is damaged and needs to be restored.
v NOT SYNCHRONIZED. The journal is inoperable, or the journal receiver that is
used for this member is damaged and needs to be restored.
v RESTORE COMPLETE. The restore of the member is complete.
v RECOVERED. Recovery has completed successfully.
v NOT JOURNALED. The member is not journaled to any journal and cannot be
recovered.
v DIFFERENT JOURNAL. The member is not journaled to the journal you are
working with.
v Blank. The member and the journal are usable and everything is synchronized.

Option 1 (Add member to list) on the display allows you to add a member to the
list. You may want to do this if you want to restore those members.

Option 2 (Apply journaled changes) shows the APYJRNCHG command prompt,

which applies the journaled changes to the file members, and changes the status to
RECOVERED (if the apply operation was successful). If the apply operation was
not successful, messages appear indicating why, and the status remains the same.

If any required receivers are missing or damaged while running the APYJRNCHG
command, the system displays prompts for the restore procedures for the missing
or damaged receivers.

If any of the members in the list have a status of DAMAGED when option 2 is
selected, you are prompted with the command necessary to recover the file
member. For files that are damaged, recovery involves the restore of the last save
that is followed by the Apply Journaled Changes (APYJRNCHG) command. The
system guides you through recovery as follows:
1. The system identifies all the logical files dependent on the specified damaged
file. The Dependent Logical Files display appears identifying these files.

Chapter 20. Working with Journal Entries, Journals, and Journal Receivers 457
 2. The dependent logical files are deleted.
3. The files to be recovered (or restored) are deleted by the system.
4. The system displays prompts for the restore of files to be recovered. After all
restores are completed successfully, the files to be recovered are allocated
exclusively to prevent any other processing. This allocation is maintained until
the recovery procedures are complete.
5. The system displays prompts for the restores of the dependent logical files.
6. An APYJRNCHG command is prompted with FROMENT(*LASTSAVE) and
TOENT(*LASTRST).
7. If the APYJRNCHG command encounters a required journal receiver that is not
on-line, the system prompts for the restore of the required receiver and again
starts the APYJRNCHG command.

When the recovery process is complete, the status field for the member indicates
RECOVERED (if the operation was successful). If the operation failed, the status
field remains unchanged, and messages appear indicating why the operation
failed.

Option 3 (Restore) prompts you for the files to restore. Use this option if any
members have a status of NOT FOUND. Members that are restored successfully
have a status of RESTORE COMPLETE. Members that are not restored keep their
old status. A message is sent indicating that the restore did not complete
successfully. All members that are restored are included in the list of members to
recover.

Note: The last save information is provided for the restore operation. If either of
the following are true, the RSTOBJ command must be used instead of option
3 (Restore):
v The device provided is tape, diskette, or optical and you choose to restore from
a save file (*SAVF).
v The device provided is a save file (*SAVF) and you choose to restore from tape,
diskette, or optical media.

Option 4 (Remove member from list) on the display removes file members from
the list of members to be recovered.

Work with Backout Recovery

The same options on the Work with Forward Recovery display are available on the
Work with Backout Recovery display (option 3 on the Work with Journals display).
In addition, the option to restore the file is not valid for backout recovery. The
status field that is shown on the Work with Backout Recovery display is either
blank or it indicates the same status as for forward recovery, except for RESTORE
COMPLETE.

458 OS/400 Backup and Recovery V5R1

 Work with Backout Recovery

Journal . . . . . . : JRNACC Library . . . . . . : DSTA1

Position to . . . . .
Library . . . . . .

Type options, press Enter.

1=Add member to list 2=Remove journaled changes
4=Remove member from list

Opt File Library Member Status

_ __________ __________ __________
_ PAYFILE1 PAYLIB QTR192

Bottom
F3=Exit F12=Cancel

Option 1 (Add member to list) on the display allows you to add a member to the
list.

Option 2 (Remove journaled changes) on the display shows the Remove

Journaled Changes (RMVJRNCHG) command prompt, removes the journaled
changes, and changes the status to RECOVERED (if the operation was successful).
If any required journal receivers are missing or damaged while the RMVJRNCHG
command is running, the system displays prompts for the necessary restore
procedures for the missing or damaged receivers. If the remove operation was not
successful, messages appear indicating why the status remains the same.

If any members in the list have a status of NOT FOUND or DAMAGED when on
the Work with Backout Recovery display, the operation is not allowed. These
members must be recovered in a forward fashion after they have been restored.
Forward recovery of specific files must be used for this type of recovery.

Option 4 (Remove member from list) on this display allows you to remove file
members from the list.

Display Journal Status

Option 5 on the Work with Journal display shows the current status of the journal.
It shows if the last system end was NORMAL or ABNORMAL, and if the journal
is damaged. The damage status is NONE or FULL.

Chapter 20. Working with Journal Entries, Journals, and Journal Receivers 459
 Display Journal Status

Journal . . . . . . : PAYJRN Library . . . . . . : QSYS

Last system end status . . . . . . . . . . . . . : NORMAL

Journal damage status . . . . . . . . . . . . . . : NONE
All objects synchronized . . . . . . . . . . . . : YES

Attached Damage
Receiver Library Status
PAYJRNRCV2 PAYRCVLIB NONE

Bottom
Press Enter to continue.

F3=Exit F12=Cancel

If the last system end was abnormal, this display indicates whether the system
synchronized the journaled objects or not. This indicates if the system
synchronized each object in use during the abnormal end to match the entries in
the attached journal receiver during the previous initial program load (IPL).

If the last system end was normal, the display indicates that all objects are
synchronized with the journal. If the journal is damaged, the display indicates that
the system was unable to determine whether or not all objects are synchronized.

The display also presents information about the currently attached receiver and its
damage status. The damage status of the receiver can be NONE, PARTIAL, or
FULL. If the journal damage is such that the system cannot determine the status of
the attached journal receiver, no attached receiver shows on the display.

If some objects are not synchronized or damage has been detected, a message
appears indicating the form of recovery that you should perform.

Recover Damaged Journal

Option 6 on the Work with Journals display verifies that the journal is damaged
before proceeding with recovery. If the journal is not damaged, an information
message appears.

Recovery for a damaged journal guides you through the following steps:
1. The system attempts to determine which files are currently being journaled to
the indicated journal. If the system cannot successfully build this list, a message
appears before the recovery operation begins.
2. Journaling, for all access paths that are currently being journaled to the
specified journal, is ended.
3. Journaling, for all files that are currently being journaled to the specified
journal, is ended .
4. The system deletes the journal.
5. The system presents the Recover Damaged Journal display, which asks you
whether to restore or create the journal:

460 OS/400 Backup and Recovery V5R1

 a. If the journal will be restored, the system prompts for the values that are
needed for the restore operation.
b. If the journal will be created, the system prompts for the receiver names
and attributes with the CRTJRNRCV command prompt. The system
prompts for values needed to create the journal with the CRTJRN command
prompt, with known values that are shown.
6. The list of files for which journaling is to be started again is shown. When you
press the Enter key, journaling is started for all files that are listed.
7. The list of files that contains access paths for which journaling is to be started
again appears. When you press the Enter key, journaling for the access paths is
started for the files that are listed.
8. The system associates all applicable receivers with the re-created or restored
journal so that a restore of these receivers is not necessary.
A journal receiver is associated with a journal if the journal receiver appears in
the journal receiver directory. A receiver that was previously attached to a
journal, but is not currently associated with a journal, cannot be used with the
journal commands such as DSPJRN, APYJRNCHG, and RMVJRNCHG.

As the recovery of a damaged journal proceeds, the Display Journal Recovery

Status display appears. The information on this display is updated as the operation
progresses to indicate which steps have been completed, which steps have been
bypassed, and which step will be run next. Whenever a user action is required, the
status display is replaced by the appropriate prompt display.

The status field indicates the following operation status:

v PENDING. The step has not been started.
v NEXT. The step will be performed next (after the Enter key is pressed).
v BYPASSED. The step was not performed. (It was not necessary).
v COMPLETE. The step has been performed.

The first display you usually see after the first status display is the Recover
Damaged Journal display. Use this display to choose whether the journal is to be
created or restored.

When the last step of the recovery process is complete, a message appears
indicating that all files for which journaling was started should be saved to
establish a new recovery point.

Note: If the damaged journal had any remote journals associated with it, use the
Add Remote Journal (QjoAddRemoteJournal) API or ADDRMTJRN
command to reassociate those remote journals. See “Adding a Remote
Journal” on page 474 for more information.

Recover Damaged Journal Receivers

Option 7 on the Work with Journals display checks to determine which journal
receivers that are associated with the specified journal are damaged. If none are
damaged, a message appears.

If there are damaged journal receivers associated with the specified journal, the
Recover Damaged Journal Receivers display appears and lists those receivers.

The status fields initially show a value of DAMAGED. After recovery has been
successfully completed, the status shows a value of RECOVERED (receiver
recovered).

Chapter 20. Working with Journal Entries, Journals, and Journal Receivers 461
 Recovery for a damaged journal receiver guides you through the following steps:
1. If the attached receiver is damaged, you must run a CHGJRN to attach a new
receiver.
Indicate that you want create a new receiver. The system presents the
CRTJRNRCV command prompt for receiver name and attributes. After you
create the new receiver, the system shows the CHGJRN command prompt.
If the attached receiver is not damaged, the preceding step is omitted.
2. The damaged journal receiver is deleted.
3. Prompts for the restore of the damaged journal receiver is shown. Any of the
values on the prompt can be changed except the receiver name. Save
information in the prompt is provided by the system.

Associate Receivers with Journal

Option 9 on the Work with Journals display should be used if the journal was
restored or created again. The system associates all applicable receivers with the
restored or recreated journal so that a restore of these receivers is not necessary.

A journal receiver is associated with a journal if the journal receiver appears in the
journal receiver directory. A receiver that was previously attached to a journal but
is not currently associated with a journal cannot be used with the journal
commands, such as DSPJRN, APYJRNCHG, and RMVJRNCHG.

Recovery of a Journaled Object Using Journaled Changes

| You can recover from many types of damage to journaled objects by using
| journaled changes. For example, an object is damaged and becomes unusable, an
| error in an application program caused records to be improperly updated, or
| incorrect data was used to update an object. In each of these instances, simply
restoring a saved version of the object may result in the loss of a significant
amount of data. If you use the Apply Journaled Changes (APYJRNCHG) command
to apply journaled changes, significantly less data may be lost. You can use the
Remove Journaled Changes (RMVJRNCHG) command to recover from improperly
| updated records or incorrect data if before-images have been journaled. This
| command removes (or backs out) changes that were made to an object. See
“Applying and Removing Journaled Changes” on page 444.

| Note: To apply or remove journaled changes to or from a restored copy of the

| object, the object must have been saved while it was being journaled. For an
| explanation of why objects must be saved, see “Why You Must Save Objects
| after Starting Journaling” on page 414.

Recovery after Abnormal System End

| If the system abnormally ends while you are journaling objects, the system does
| the following:
| 1. Brings all journals, journal receivers, and objects you are journaling to a usable
| and predictable condition during the IPL, including any access paths being
| journaled and in use at the time the system abnormally ended.
2. Checks all recently recorded entries in the journal receivers that were attached
to a journal.
3. Places an entry in the journal to indicate that an abnormal system end
occurred. When the system completes the IPL, all entries are available for
processing.
4. Checks that the journal receivers attached to journals can be used for normal
| processing of the journal entries. If some of the objects you are journaling could

462 OS/400 Backup and Recovery V5R1

 not be synchronized with the journal, the system sends a message (CPF3172) to
the history log (QHST) that identifies the journals that could not be
synchronized. If a journal or a journal receiver is damaged, the system sends a
message to the history log identifying the damage that occurred (CPF3171
indicates that the journal is damaged, and the message CPF3173 or CPF3174
indicates that the journal receiver is damaged).
| 5. Recovers each object that was in use at the time the system ended abnormally,
| using the normal system recovery procedures for objects.
| In addition, if an object being journaled was opened for output, update, or
| delete operations, the system performs the following functions so changes to
| that object will not be lost:
| a. Ensures that the changes that appear in the object. Changes that do not
| appear in the journal receiver are not in the object.
| b. Places an entry in the journal receiver that indicate whether the object was
| synchronized with the journal. For database files, if the file could not be
| synchronized with the journal, the system places a message (CPF3175) in
| the history log identifying the failure, and you must correct the problem.
| For other journaled objects, the system places message CPF700C in the
| history log identifying the failure, and you must correct the problem.

| A synchronization failure can occur if the data portion of the object is damaged,
| a journal receiver required to perform the synchronization is damaged, or the
| journal is inoperable.

Procedure for Abnormal System End Recovery: After an abnormal system end,
perform the following:
1. Perform a manual IPL.
| 2. Check the history log to determine if there are any damaged object, objects that
| are not synchronized, or any damaged journals or journal receivers.
3. If necessary, recover the damaged journals or journal receivers as described in
“Recovering When a Journal Is Damaged” on page 464 and “Recovering When
a Journal Receiver Is Damaged” on page 465.
| 4. If there is a damaged object:
| a. Delete the object.
| b. Restore the object from the latest saved version.
| c. Allocate the object so no one else can access it.
d. Restore the needed journal receivers from newest to oldest, if they are not
on-line.

Note: Journal receivers that are created on Version 3 Release 1 or later do

not need to be restored in a particular sequence. The system
establishes the receiver chains correctly when they are restored.
| e. Use the APYJRNCHG command to apply the changes to the object.
| f. Deallocate the object.
| 5. If an object could not be synchronized, use the information in the history log
| and in the journal to determine why the object could not be synchronized and
| how to proceed with recovery. For example, you may need to use the DFU or a
| user-written program to bring a database file to a usable condition.
6. Determine which applications or programs were active, and determine where
to restart the applications from the information in the history log and in the
journal.

Chapter 20. Working with Journal Entries, Journals, and Journal Receivers 463
 If a journaled access path is in use during an abnormal system end, that access
path does not appear on the Edit Rebuild Access Path display.

If the maintenance for the access path is immediate or delayed, the system
automatically recovers the access path during IPL. A status message appears for
each access path whose maintenance is immediate or delayed as it is being
recovered during an IPL. The system places a message (CPF3123) in the system
history log for each access path that is recovered through the journal during the
IPL. This message appears for access paths that are explicitly journaled and for
access paths that are protected by SMAPP.

Recovering When a Journal Is Damaged

If a journal becomes damaged, the system sends message CPF8135 to the system
operator and to the job log. Use the Work with Journal (WRKJRN) command to
help you in the recovery of a damaged journal. Select option 6 (Recover damaged
journal) on the Work with Journals display to recover a damaged journal.

For a description of the Work with Journals display, see “Work with Journal
(WRKJRN) Command Options” on page 455, or the WRKJRN command in the
online command help. To view the help, type WRKJRN on a command line, and
press F1.

| It is recommended that you use the Work with Journals (WRKJRN) command to
| recover a damaged journal if you are only journaling physical files and access
| paths to this journal. The WRKJRN command performs all the steps that are
described below except for saving the physical files and logical files. The WRKJRN
command associates the receivers with the recovered journals without you having
to delete and restore the receivers.

Use the following steps to recover a damaged journal without using the WRKJRN
command:
1. End journaling for all access paths associated with the journal by using the
ENDJRNAP command.
2. End journaling for all physical files associated with the journal by using the
ENDJRNPF command.
| 3. End journaling for all IFS objects by using the ENDJRN command.
| 4. End journaling for all other object types by using the ENDJRNOBJ command.
5. Delete the damaged journal by using the DLTJRN command.
6. Create a journal receiver (CRTJRNRCV command) and create a journal with the
same name and in the same library as the damaged journal (CRTJRN
command), or restore the journal from a previously saved version.
7. Start journaling the physical files that were journaled by using the Start Journal
Physical File (STRJRNPF) command.
8. Start journaling the access paths that were journaled by using the Start Journal
Access Path (STRJRNAP) command.
| 9. Start journaling IFS objects with the Start Journal (STRJRN) command.
| 10. Start journaling other object types with the Start Journal Object (STRJRNOBJ)
| command.

| Note: You can also restore your journaling environment by deleting and restoring
| all the objects that were being journaled. Objects that were journaled at the
| time of their save automatically begin journaling at restore time if the
| journal is on-line.

464 OS/400 Backup and Recovery V5R1

| 11. Save the journaled objects to allow for later recovery.
12. Associate the old journal receivers with the new journal. Do the following:
a. Type WRKJRN and press the Enter key.
b. On the prompt display, enter the name of the journal.
c. From the Work with Journal display, select option 9 (Associate receivers).
d. Press F12 to cancel the display.
e. Type WRKJRNA JRN(library-name/journal-name) and press the Enter key.
f. From the Work with Journal Attributes display, press F15 to display the
receiver directory. If the journal receivers have not been reassociated
correctly, perform the following steps. These steps are usually required only
if you have journal receivers created before V3R1.
1) Save the journal receiver that was attached to the damaged journal.
2) Delete it and restore it and any previously attached journal receiver you
need. You must delete and then restore the receiver after the journal is
restored or re-created in order to associate the journal receiver with the
journal. The journal receiver must be restored newest to oldest.
3) Use the WRKJRNA command to display the receiver directory again.

Each time a journal is restored, a new receiver chain is started because the last
journal receiver on the chain that existed prior to the restore process did not have
the newly created receivers as its next receivers.

Note: If the damaged journal had any remote journals associated with it, use the
Add Remote Journal (QjoAddRemoteJournal) API or ADDRMTJRN
command to reassociate those remote journals. See “Adding a Remote
Journal” on page 474 for more information.

Recovering When a Journal Receiver Is Damaged

If a journal receiver becomes damaged, the system sends message CPF8136 or
message CPF8137 to the system operator and the job log. To recover from a
damaged receiver, do the following:
v If the damaged receiver is currently attached to a journal, use the Change
Journal (CHGJRN) command to attach a new receiver and detach the damaged
receiver.
v If the journal receiver is not currently attached to a journal, delete the journal
receiver by using the Delete Journal Receiver (DLTJRNRCV) command and
restore a previously saved copy.
v If the journal receiver was never attached to a journal, delete the receiver and
create it again or restore it.

| If the journal receiver is partially damaged, all journal entries except those in the
| damaged portion of the journal receiver can be viewed using the Display Journal
| (DSPJRN) command. Using this list, you can determine what you need to do to
| recover your objects. Applying or removing journal changes cannot be done with a
| partially damaged journal receiver.

It is recommended that you use the Work with Journals (WRKJRN) command to
recover a damaged journal receiver. For a description of the Work with Journals
display, see “Work with Journal (WRKJRN) Command Options” on page 455, or
the WRKJRN command in the online command help. To view the online help, type
WRKJRN at a command line, and press F1. The online help also contains a
description of the journal menus.

Chapter 20. Working with Journal Entries, Journals, and Journal Receivers 465
466 OS/400 Backup and Recovery V5R1
 Chapter 21. Remote Journal Function
Introduction to the Remote Journal Function
| The remote journal function allows you to establish journals and journal receivers
| on a remote system that are associated with specific journals and journal receivers
| on a local system. The remote journal function can replicate journal entries from
the local system to the journals and journal receivers that are located on the remote
system after they have been established.

Benefits of the Remote Journal Function

The remote journal function helps to efficiently replicate journal entries to one or
more systems. You can use the remote journal function with application programs
| to maintain a data replica. A data replica is a copy of the original data that resides
on another iSeries or AS/400e server. The original data resides on a primary
| system. Applications make changes to the original data during normal operations.
Previous to V4R2M0, you could have accomplished a similar function by using the
Receive Journal Entry (RCVJRNE) command. (“Using the Receive Journal Entry
Command” on page 432 provides information on the RCVJRNE command.) In that
environment, the RCVJRNE exit program receives journal entries from a journal,
and then sends the journal entries to the remote system by using whatever
communications method is available. All of this processing occurs asynchronously
to the operation that is causing the journal entry deposit and takes place at an
application layer of the system. The remote journal function replicates journal
entries to the remote system at the Licensed Internal Code layer. Moving the
replication to this lower layer provides the following:
v The remote system handles more of the replication overhead
v Improved overall system performance and journal entry replication performance
v An option is available to have replication occur synchronously to the operation
that is causing the journal entry deposit
v Journal receiver save operations can be moved to the remote system

Figure 38 on page 468 and Figure 39 on page 468 illustrate a comparison of a

hot-backup environment with and without the remote journal function.
Hot-backup is the function of replicating an application’s dependent data from a
| primary system to a backup system. The primary system is the system where the
| original data resides. The backup system is the system where a replica of the
| original data is being maintained. In the event of a primary system failure, you can
| perform a switch-over to the backup system.

© Copyright IBM Corp. 1997, 2000, 2001 467

|

|
| Figure 38. Hot-Backup Environment without Remote Journal Function, and Application-Code Based Apply

|
| Figure 39. Hot-Backup Environment with Remote Journal Function, and Application-Code Based Apply

Remote Journal Function Configuration Examples

The following figure shows the two basic remote journal function configurations.

468 OS/400 Backup and Recovery V5R1

 Broadcast Configuration
-----------------------

┌────S(T)
│ ┌───┐
│ │RJ1│
│ └───┘ Cascade Configuration
│ ---------------------
│
│
(S)─────┼────S(T) (S)──────────S(T) (S)──S ... ──S(T)
┌───┐ │ ┌───┐ ┌───┐ ┌───┐ ┌───┐ ┌───┐
│LJ │ │ │RJ2│ │LJ │ │RJ1│ │RJ2│... ...│RJn│
└───┘ │ └───┘ └───┘ └───┘ └───┘ └───┘
│ . (S)──────────S(T)
│ .
│ .
│ Remote journals that cascade to other remote journals.
│
│ ┌───┐
│ │RJn│
│ └───┘
└────S(T)

A local journal replicating journal entries

to multiple remote journals.

Figure 40. Broadcast and Cascade Remote Journal Configurations

A broadcast configuration is a journal that replicates its journal entries to one or

more remote journals. A cascade configuration is a remote journal that replicates its
journal entries to an additional remote journal. The additional remote journal can
replicate the entries to yet another remote journal, and so on. The remote journal
function configurations can stand alone or can be combined with one another. For
example, one or more of the remote journals in the broadcast configuration could
cascade down to several additional remote journals. Likewise, one or more remote
journals in the cascade configuration could broadcast out to one or more remote
journals.

A local journal is populated by applications that are depositing journal entries. A

remote journal is populated by receiving its journal entries from either a local or
another remote journal. The journals are paired, as depicted in Figure 40 where (S)
represents a journal on a source system, and (T) represents a journal on a target
system. In the cascade configuration, a remote journal can be a recipient of journal
entries (a target), and a replicator of journal entries (a source) at the same time.

| A source system is a system where a journal resides and is having its journal
| entries replicated to a remote journal on a target system.

Note: A source system is not necessarily the primary system. For example, a
remote journal that is cascading its journal entries to another remote journal
is said to reside on a source system.

| A target system is a system where a remote journal resides and is receiving journal
| entries from a journal on a source system.

A remote journal network includes the local journal and all of the remote journals
that are downstream from that local journal. You can set up the remote journal
network in broadcast configuration, cascade configuration, or a combination of the
two configurations.

Chapter 21. Remote Journal Function 469

| In many environments, users attempt to minimize the amount of processing that is
| performed on the local or primary system. As much of the processing as possible is
shifted to other systems in the network. A combination of the broadcast and
cascade configurations allows for this when replicating the journal entries from a
single system to multiple other systems. For example, replicating a local journal to
a single remote journal on a target system will minimize the replication cost on the
primary system. Then, from the target system, the replicated journal can be
asynchronously replicated by either a broadcast or cascade configuration to other
remote journals on other systems. This allows all of the journal entries to be
known to all desired systems, while requiring a minimal amount of processing on
the primary system.

The following characteristics apply to local journals and to any journal receivers
that were attached to local journals:
v Objects can be journaled to local journals.
v Journal entries can be directly deposited to local journals. For example, the Send
Journal Entry (SNDJRNE) command or the Send Journal Entry (QJOSJRNE) API
can be used to send journal entries directly to a local journal.

The following characteristics apply to remote journals and to any journal receivers
that were attached to remote journals:
v Objects cannot be journaled to remote journals.
v Journal entries cannot be directly deposited to remote journals. For example, the
Send Journal Entry (SNDJRNE) command or API (QJOSJRNE) cannot be used to
send journal entries directly to a remote journal.
v Journal entries are only replicated to remote journals from an associated source
journal. A source journal is the journal on the source system to which a remote
journal has been added. A source journal can be either a local or a remote
journal.
v The information in the journal entries such as time stamps, system name, and
qualified journal receiver names reflect information as deposited in the local
journal for this remote journal network.
v The information in the journal receiver such as attach time and detach time
reflect the information as it is for the local journal for the remote journal
network.
| v Certain attributes of the remote journal are fixed and determined based on the
| source journal, such as the values for receiver size options and the values for
| minimize entry-specific data. These attributes for the remote journal cannot be
| changed except by changing the attributes for the source journal.
v Remote journals cannot be saved and restored to any release prior to V4R2M0.

Supported Communications Protocols

The remote journal function supports the following communications protocols for
replicating the journal entries to the remote systems:
v OptiConnect for OS/400. If you want to use the OptiConnect for OS/400
support, you must purchase and install the required hardware and software for
that support. Refer to OptiConnect for OS/400, SC41-5414-02 for more
information.
v Systems Network Architecture (SNA). If you want to use SNA for the transport,
there are no additional software considerations. The software support is in the
base operating system. You must purchase whatever hardware is appropriate for
your configuration. Refer to SNA Distribution Services, SC41-5410-01 for more
information.

470 OS/400 Backup and Recovery V5R1

 v Transmission Control Protocol/Internet Protocol(TCP/IP). If you want to use
TCP/IP for the transport, there are no additional software considerations. The
software support is in the base operating system. You must purchase whatever
hardware is appropriate for your configuration. Refer to TCP/IP Configuration and
Reference, SC41-5420-03 for more information.

Specifying a relational database (RDB) directory entry will identify the

communications protocol that the remote journal function will use. The RDB that is
specified must meet the following rules:
v The communications protocol must be one of the remote journal function
supported protocols.
v The remote location name in the RDB cannot refer to the *LOCAL database.
v The RDB cannot use an application requester driver program (*ARDPGM) to
locate the target system.
| For more information on creating relational databases, look in the Information
| Center at the following Web site: http://www.ibm.com/eserver/iseries/infocenter.

Security of the remote journal function is dependent on the communications

protocol security. The remote journal function does not alter the security
characteristics that are available.

The communications function that is identified by the RDB can be shared by other
activity. However, you may consider isolating the remote journal function activity
in order to have the best performance.

Release to Release Considerations

All systems that use the remote journal function must be at V4R2M0 or a later
release.

Only those receivers which have been attached to a journal, after V4R2M0 is
installed, will be allowed to be replicated by the remote journal function.

All journals that were created prior to V4R2M0 are considered local journals.

Information APAR II11476 contains a list of PTFs that enable other releases to work
with V4R4 in a remote journal network. Information APAR II12001 contains the
| same list of program temporary fixes (PTF) for V4R5. Information APAR II12556
| contains the same list of program temporary fixes (PTF) for V5R1.
Notes:
1. If you specify RCVSIZOPT(*MAXOPT1) on the journal that you attach a journal
receiver to, you cannot replicate the journal receivers to any remote journals on
any systems at a release prior to V4R5M0.
| 2. If you specify RCVSIZOPT(*MAXOPT2) on the journal that you attach a journal
| receiver to, you cannot replicate the journal receivers to any remote journals on
| any systems at a release prior to V5R1M0.
| 3. If you specified MINENTDTA for either *FILE or *DTAARA on the journal to
| which you attached a journal receiver, you cannot replicate the journal receivers
| to any remote journals on any systems at a release prior to V5R1M0.

Chapter 21. Remote Journal Function 471

 Planning for the Remote Journal Function

Determining Which Journals are Good Candidates for Remote

Journals
| Journals that you are currently replicating, or that you plan to replicate, in their
| entirety to one or more systems, are excellent candidates for the remote journal
| function.

Journals with high activity that require frequent saves and deletes of the associated
journal receivers during the day are also good candidates for the remote journal
function. The backup system can take over the journal receiver save processing,
and the primary system could specify the MNGRCV(*SYSTEM) and
DLTRCV(*YES) attributes. This would free up disk space on the primary system as
| quickly as possible. The backup system is the system where a replica of the
| original data is being maintained. The primary system is the system where the
| original data resides.

You may have applications that are so critical to your business that any downtime
will impact your operations. The application dependent data is a good candidate
to protect with the remote journal function. Application dependent data is any data
that a particular application would depend upon if that application were to be
interrupted and had to be restarted.

| For example, you may have a database that has a lot of query activity that impacts
| your system performance. That database is a good candidate to replicate to another
system so that the query activity moves to that system. The remote journal
function can assist in that replication.

Determining Which Communications Protocol and Delivery

Mode to Use
The greater the volume of traffic, that is the higher the rate of journal entry
deposits, the faster communications method you should choose. If your traffic is
minimal, then a slower communications method can be adequate.

The delivery mode defines how journal entries are replicated to a remote journal.
The delivery mode only applies when actively replicating the journal entries from
a journal on a source system to a remote journal on a target system. The delivery
mode can be either synchronous or asynchronous.

If the application dependent data is critical and the loss of journal entries would
impact your business, then you should use the synchronous delivery mode.
Synchronous delivery mode is only valid when activating a remote journal that is
associated with a local journal.

It may be acceptable that the remote system does not have all the journal entries as
they are being deposited or replicated into the source journal. If this is true, the
asynchronous delivery mode is a good choice to minimize the impact to the source
journaling throughput.

The choice of delivery mode and communications protocol are closely linked. Since
the synchronous delivery mode will affect the interactive users response time, the
faster the communications protocol the better. This again will be dependent on the
journal entry deposit rate.

472 OS/400 Backup and Recovery V5R1

 See “Performance Considerations” on page 492 for more details.

How Do the Journal Attributes Affect the Remote Journal

Function?
Reducing the size of the journal receivers on the source system will reduce the
| communications overhead of the remote journal function. Therefore, you may want
| to consider only journaling after images, not journaling open, close, or force
| entries. You may also want to consider using the various receiver size options, or
| using the minimized entry-specific data values. See “How You Can Reduce the
Storage Used by Journaling” on page 397 for more information.

You can also refer to “Journal Attributes for Remote Journals” on page 478 and
“Auxiliary Storage Considerations” on page 494 for more details.

Setting up the Remote Journal Function

This section describes the steps you would use to create and work with a remote
journal network or environment. It discusses how to establish and maintain one
remote journal that is associated with one local journal.

Note: If you want to make a more complicated broadcast or cascade configuration,

use the following steps for each of the remote journals in the configuration.

Preparing to use Remote Journal Function

Before establishing the remote journal environment, do the following steps.
1. Determine the extent of your remote journal network or environment.
See “Planning for the Remote Journal Function” on page 472.
2. Determine what library redirection, if any, you will be using for the remote
journals and associated journal receivers. Library redirection is the ability to
allow the remote journal and associated journal receivers to reside in different
libraries on the target system from the corresponding source journal and its
associated journal receivers.
See “Using Library Redirection” on page 475.
3. Ensure that all selected libraries exist on the target systems. You will need to
consider whether or not library redirection will be used when adding the
remote journal.
4. Create the appropriate local journal if it does not already exist.
See “How Journals Are Created” on page 402 for more information on creating
local journals.
5. Configure and activate the communications protocol you have chosen to use.
See “Supported Communications Protocols” on page 470 for more information.
After you have configured the communications protocol, it must be active
while you are using the remote journal function. For example, if you are using
the OptiConnect for OS/400 bus transport method, then the OptiConnect for
OS/400 subsystem, QSOC, must be active. QSOC must be active for both the
source system and the target system, and the appropriate controllers and
devices must be varied on. If you are using a SNA communications transport,
vary on the appropriate line, controller, and devices and ensure subsystem
QCMN is active on both systems. If you are using TCP/IP, you must start
TCP/IP by using the Start TCP/IP (STRTCP) command, including the
distributed data management (DDM) servers.

Chapter 21. Remote Journal Function 473

 Note: This chapter is not intended to explain communications configuration
details. Refer to the appropriate documentation for more details.
6. If one does not already exist, create the appropriate relational database (RDB)
directory entry that will be used to define the communications protocol for the
remote journal environment.

| For more information on the APIs and commands that are described in the
| following sections, click the Programming topic in the Information Center at the
| following Web site: http://www.ibm.com/eserver/iseries/infocenter.

Adding a Remote Journal

Adding a remote journal creates a remote journal on a target system and associates
that remote journal with the journal on the source system. This occurs if this is the
first time the remote journal is being established for a journal.

If a remote journal environment has previously been established, adding a remote

journal re-associates the remote journal on the target system with the journal on
the source system.

Note: The journal on the source system can be either a local or remote journal.

You can establish and associate a remote journal on a target system with a journal
on the source system by one of the following methods:
v Use the Add Remote Journal (QjoAddRemoteJournal) API on the source system.
v Use the Add Remote Journal (ADDRMTJRN) command on the source system.
See “Add Remote Journal Processing” on page 477 for additional details.

The following guidelines apply when adding a remote journal:

v You can only associate a remote journal with a single source journal.

Note: The same remote journal can then have additional remote journals that
are associated with it that are located on other target systems. This is the
cascade configuration that is shown in Figure 40 on page 469.
v A maximum of 255 remote journals can be associated with a single journal on a
source system. This can be any combination of asynchronously maintained or
synchronously maintained remote journals.
The terms asynchronously maintained and synchronously maintained both
describe a remote journal function delivery mode for journal entry replication. If
a journal is asynchronously maintained, control is returned to the application
generating the journal entry on the source system without waiting for the
journal entry to be replicated to the remote journal. An asynchronously
maintained remote journal may lag several journal entries behind the total
number of journal entries in the journal on the source system.
If a journal is synchronously maintained, control is not returned to the
application generating the journal entry on the local system until the journal
entry is replicated to the remote journal.
v The remote journal will only have its attached receiver populated with journal
entries that are replicated from the corresponding journal receiver on the source
system. No journal entries can be directly deposited to a remote journal.
v The remote journal function is only provided for local journals with a single
attached receiver. Therefore, all remote journals will also only have a single
receiver attached.

474 OS/400 Backup and Recovery V5R1

Using Library Redirection
Library redirection provides a means for remote journals and any of their
associated journal receivers to optionally reside in differently named libraries on
the target system from the corresponding local journal and journal receivers on the
local system. You can specify library redirection by using the Add Remote Journal
(QjoAddRemoteJournal) API or the ADDRMTJRN command. When using the
QjoAddRemoteJournal API, specify a different name in the Remote Journal Library
name field or the Remote Journal Receiver Library field. When using the
ADDRMTJRN command, specify a different name for the Target Journal Library
parameter or the Remote Receiver Library parameter. When a remote journal is
added, its journal type specification influences how much redirection you can
specify.

“Remote Journal Type Descriptions” on page 476 describes the various types of
remote journals that can be added, as well as a description of their redirection
characteristics.

If redirection is not specified, then the remote journal will reside in a library that
has the same name as the library that contains the source journal.

Note: Library redirection for the journal object must be specified when replicating
the journal entries to a target system for any journal starting with the letter
Q in a library starting with Q. This does not apply to the QGPL library. This
restriction prevents collisions between local and remote journals that are
used for system functions. One example of this is journal QAUDJRN in
library QSYS which is used for security auditing.

If no redirection is specified for the journal receiver, then the remote journal
receiver will reside in a library whose name is the same as the library for the
source journal receiver. For example, the source journal has two receivers that are
associated with it, receiver RCV0001 in library LIBA, and receiver RCV0002 in
library LIBB. If no journal receiver library redirection is specified, then the journal
entries in RCV0001 in library LIBA on the source will be replicated to RCV0001 in
library LIBA on the target system. The journal entries in RCV0002 in library LIBB
on the source will be replicated to RCV0002 in library LIBB on the target system.
Therefore, both libraries, LIBA and LIBB, will need to exist on the target system
prior to the invocation of the remote journal function. If journal receiver library
redirection is specified with a redirected receiver library specification of RMTLIB,
then both RCV0001 and RCV0002 would be in library RMTLIB on the target
system.

For *TYPE1 remote journals, the library redirection or the selection of no library
redirection for the journal and journal receivers can only be modified by doing the
following:
v Remove all *TYPE1 remote journals.
v Change the local journal and attach a new journal receiver.
v Delete the remote journal from the target system.
v Add the *TYPE1 remote journal, specifying the new library redirection, if any.

For *TYPE2 remote journals, the library redirection or the selection of no library
redirection for the journal and journal receivers can only be modified by doing the
following:
v Remove the *TYPE2 remote journal.
v Delete the remote journal from the target system.

Chapter 21. Remote Journal Function 475

 v Add the *TYPE2 remote journal, specifying the new library redirection, if any.

Remote Journal Type Descriptions

The two types of remote journals are *TYPE1 and *TYPE2. The two types identify
operational characteristics of a remote journal and its associated journal receivers.
Table 67 is an overview of the different remote journal types and their
characteristics.

Note: There are no performance differences between the types of remote journals.
Table 67. Journal Types and Associated Characteristics
Journal Type
Local Remote
Remote Journal Type N/A *TYPE1 *TYPE2
Remote Journal Types That *TYPE1 *TYPE2 *TYPE1 *TYPE2 *TYPE2
Can Be Added
Remote Journal Name N/A Journal name must be the Journal name may be
same as the local journal. different from the source
journal.
Journal Library Redirection N/A Journal library name may A given redirected library
be redirected to a single may be specified when
different library from that adding a remote journal.
of the local journal. All Subsequent adds of *TYPE2
*TYPE1 remote journals remote journals may
associated with a given specify a different library
local journal must reside in redirection than was
the same named library. specified on any previously
added remote journal.
Journal Receiver Library N/A Receiver library name may A given redirected library
Redirection be redirected to a single may be specified when
different library from that adding a remote journal.
of the receivers associated Subsequent adds of *TYPE2
with the local journal. remote journals may
specify a different library
redirection than was
specified on any previously
added remote journal.
Journal Receiver Library N/A The target library used The target library used
Redirection Used on when replicating a receiver when replicating a receiver
Activate from the source journal to from the source journal to
this remote journal will this remote journal will
reflect the library reflect the library
redirection that was in redirection that is currently
place for the receiver, if defined for the target
any, at the time the receiver journal.
was attached to the source
journal.1

476 OS/400 Backup and Recovery V5R1

Table 67. Journal Types and Associated Characteristics (continued)
Journal Type
Local Remote
Receiver Restore Receivers associated with Receivers associated with Receivers associated with a
Characteristics2 the local journal can be the local journal or any of given *TYPE2 remote
saved and restored to the the *TYPE1 remote journals journal can be saved and
local system or to any of can be saved and restored restored to the local system
the systems for the *TYPE1 to the local system or to or to the same system
remote journals and be any of the systems where where the *TYPE2 remote
linked into the correct the *TYPE1 remote journals journal resides and be
receiver chain of the local reside and be linked into linked into the correct
journal or the *TYPE1 the correct receiver chain of receiver chain of the
remote journal. the journal. journal.

1
If the journal receiver was attached to a journal when no remote journals were added, then no library
redirection is assumed for that journal receiver if that receiver is specified during activation. Therefore, the
journal receiver will be created in the same library on the target system as it is on the local system.
2
A journal receiver from any system in the remote journal network may always be restored to any system if
the receiver is being restored into the original or redirected receiver library. Otherwise, receivers can always
be restored to any system and associated with a local journal if a local journal by the same name as the
original local journal is found residing in the same named original local journal library.
See “Save and Restore Considerations” on page 500 for more information.

Add Remote Journal Processing

Note: In the following section, you should be aware whether library redirection is
in effect for the remote journal. If it is in effect, any library name processing
will substitute the redirected library name for the library name that is used
for the operation on the target system.

The following is the input that you must provide to add a remote journal to a
source journal:
1. The journal name and library on the source system to which the remote journal
is being added.
2. The remote journal name and library on the target system that is being added.
3. A relational database directory entry, which identifies the target system and
other necessary communications information.
4. The type of remote journal to be added.
5. Optionally, the journal or journal receiver library redirection.
6. Optionally, the values for the journal message queue, text, and delete receivers
attributes to be applied to any newly created remote journal.

Some of the processing which takes place as part of adding a remote journal is as
follows:
v A user profile with the same name as the user profile which is adding a remote
journal must exist on the target system. A check is performed on the target
system to verify that the user profile exists. If the profile does not exist on the
target system, then an exception is signaled, and the processing ends.
v A check is performed to verify that the target system has a library by the same
name as the library for the journal on the source system. If the library does not
exist on the target system, then an exception is signaled, and the processing
ends.

Chapter 21. Remote Journal Function 477

 v A check is performed on the target system to determine if a journal by the same
qualified name as the journal on the source system already exists. If a journal
already exists, it can be used for the remainder of the add remote journal
processing if it meets the following conditions:
1. It is a remote journal.
2. It was previously associated with this same source journal or part of the
same remote journal network.
3. The type of the remote journal matches the specified remote journal type.

If a journal was found, but does not meet the above criteria, then an exception is
signaled, and the processing ends. Otherwise, the remote journal is used for the
rest of the add remote journal processing.
v If no journal is found on the specified target system, then a remote journal is
created on the target system. The new remote journal has the same
configuration, authority, and audit characteristics of the source journal. The
journal that is created has a journal type of *REMOTE.
The creation of the journal on the target system is performed as though the
journal was being saved and restored to the target system. Therefore, the
ownership of the journal on a target system will follow the same rules as with
the existing save and restore functions. If the user profile which owns the
journal on the source system is on the target system, then that profile will own
the created journal on the target system. If the user profile does not exist on the
target system, then the profile QDFTOWN will own the journal on the target
system.
Additionally, if the remote journal is created, the values for the journal attributes
of text, journal message queue and delete receivers will be taken from what is
specified on the API invocation. After the remote journal has been created, these
values can be changed by using the Change Journal (CHGJRN) command for the
remote journal on the remote system. After the remote journal is created, any
changes to these attributes on the source journal will not cause equivalent
changes to the remote journal. See “Journal Attributes for Remote Journals” for
more information.
v When adding the remote journal, you must specify the type of remote journal to
add. The remote journal type influences the library redirection rules and other
operational characteristics for the journal. See “Remote Journal Type
Descriptions” on page 476 for more information.

The remote journal does not have an attached journal receiver after the add remote
journal processing completes. In addition, the journal state for the remote journal is
set to *INACTIVE. A journal state of *INACTIVE means that the remote journal is
not ready to receive any journal entries from the journal on the source system.
During this time, journal entries can continue to be deposited or replicated into the
journal on the source system. However, no entries are replicated to the newly
added remote journal until you activate that remote journal. (You activate the
remote journal by using the Change Journal State (QjoChangeJournalState) API or
the Change Remote Journal (CHGRMTJRN) command.) Refer to “Activating the
Replication of Journal Entries to a Remote Journal” on page 480 for further details.

Journal Attributes for Remote Journals

When a remote journal is created by the add remote journal processing, the remote
journal’s initial attributes are defined by the add request and the source journal.
Various journal attributes for a remote journal are treated as follows:

478 OS/400 Backup and Recovery V5R1

 ASP
If the library for the remote journal resides in an ASP, the remote journal will
be created in that ASP.
Journal message queue
Defined on add request. Once the remote journal is created, the journal
message queue, can be modified by using the Change Journal (CHGJRN)
command on the remote journal on the remote system.
Delete receivers
Defined on add request. Once the remote journal is created, the delete receivers
attribute can be modified by using the Change Journal (CHGJRN) command
on the remote journal on the remote system.
Manage receivers
Does not apply. The managing of the receivers for the remote journal is driven
by the management of the source journal.
| Minimize entry-specific data options
| Does not apply. The minimize entry-specific data options in effect for the
| remote journal are driven by the minimize entry-specific data options in effect
| for the source journal.
Receiver size options
Does not apply. The receiver size options in effect for the remote journal are
driven by the receiver size options in effect for the source journal.
Text
Defined on add request. Once the remote journal is created, the text can be
modified by using the Change Journal (CHGJRN) command on the remote
journal on the remote system.

Preventing Journal Receiver Deletion

Once a remote journal has been added to a source journal, the system provides
additional protection of the journal receivers attached to the source journal. This
protection prevents the user from deleting the journal receiver until that receiver
has been fully replicated to the remote journal. (“Deleting Journal Receivers” on
page 418 provides information on deleting journal receivers.) Once a remote
journal has been added to a journal, the receiver that was currently attached to the
source journal at the time the remote journal was added is protected. Any journal
receivers that are attached subsequently are also protected. The protection is
eliminated when you remove the remote journal. (You use the Remove Remote
Journal (QjoRemoveRemoteJournal) API or the Remove Remote Journal
(RMVRMTJRN) command to remove the remote journal.)

Activating and Inactivating a Remote Journal

Note: In the following section, you should be aware whether library redirection is
in effect for the remote journal. If it is in effect, any library name processing
will substitute the redirected library name for the library name that is used
for the operation on the target system.

Activating a remote journal means starting and then maintaining the replication of
journal entries from a source journal to a remote journal. Activating a remote
journal always occurs from the source system.

Chapter 21. Remote Journal Function 479

 Inactivating a remote journal means ending the replication of journal entries from
the source journal to the remote journal. Inactivating a remote journal can be
performed from the source or target systems. However, the preferred method is to
inactivate from the source system.

If this is the first time the remote journal is being activated, activating a remote
journal creates one or more journal receivers on the target system. Activating the
remote journal also establishes the connection between the source and remote
journal so that journal entry replication can begin.

If the remote journal has previously been activated, the creation of additional
journal receivers may or may not be required on the target system. This would
occur prior to establishing the connection between the source and remote journal
so that journal entry replication can resume.

Use the Change Journal State (QjoChangeJournalState) API, Change Remote

Journal (CHGRMTJRN) command, or CHGJRN command to activate and inactivate
a remote journal.

Activating the Replication of Journal Entries to a Remote Journal

You can activate the replication of journal entries from a journal on a source
system to a remote journal on a target system by using one of the following
methods:
v The Change Journal State (QjoChangeJournalState) API
v The Change Remote Journal (CHGRMTJRN) command
Both the QjoChangeJournalState API and the CHGRMTJRN command must be
issued from the source system.

In order to activate the replication of journal entries to a given remote journal, the
following must be true:
v The remote journal that you wish to activate must not have a journal state of
*ACTIVE. For instance, this might seem to be a reasonable request if you wanted
to simply change the delivery mode from synchronous to asynchronous.
However, the remote journal must be inactive before you can activate it.
v The remote journal that you wish to activate must not be actively replicating
journal entries to other remote journals, as in a cascade configuration. You must
inactivate the remote journals that are immediately downstream before
activating the remote journal.

You need to provide the following input in order to activate the replication of
journal entries to a remote journal on a target system:
v The journal name and library on the source system from which journal entries
will be replicated.
v The remote journal name and library on the target system to which journal
entries will be replicated.
v A relational database directory entry, which identifies the target system and
other necessary communications information.
v The delivery mode to be used.
Specify either synchronous or asynchronous delivery mode.
Journal entries can be replicated to the target system either synchronously or
asynchronously. Synchronous delivery means that the journal entry is replicated
to the target system concurrently with the entry being written to the local
receiver on the source system. The entry is known on the target system, in main

480 OS/400 Backup and Recovery V5R1

 storage, prior to returning control to the user application that deposited the
journal entry on the source system. Therefore, the target system knows of all
journal entries as they are being made in real-time on the source system. Using
this mode allows for recovery without losing journal entries on the target system
if the source system fails. Providing journal entries synchronously to a target
system will have some impact to the journaling throughput on the local system.
Synchronous delivery mode is only supported when a remote journal is
associated with a local journal.

Note: There are certain circumstances, when using synchronous mode, where
journal entries may not be immediately available for retrieval from the
target system.
– Some entries that are not required for data recovery may not be
| immediately available for retrieval from the target system. For example,
| journal entries for a file close (journal code ’F’, entry type ’CL’) or a
| stream file open, (journal code ’B’, entry type ’OF’).
– User-generated journal entries that use the Send Journal Entry
(SNDJRNE) command or the Send Journal Entry API (QJOSJRNE) may
not be immediately available for retrieval. If either you, or your
application, do not request to force these user-generated entries they
will only be replicated to the remote journal when some other action
forces them. Therefore, you should periodically specify FORCE(*YES)
when using the send journal entry functions.
– Journal entries that are associated with commitment control
transactions may not be immediately available for retrieval from the
remote system. These entries will be retrievable after the following
journal entries have been deposited into the source journal:
- Journal code ’C’, journal entry type ’CM’ (Commit)
- Journal code ’C’, journal entry type ’RB’ (Rollback)
Refer to “Retrieving Journal Entries with Commitment Control
Considerations” on page 507 for more information.

Replicating a journal entry asynchronously means that the journal entry is

replicated to the target system after control is returned to the application
depositing the journal entry on the source system. Using this mode allows for
recovery that may lose some journal entries if the source system fails. However,
this mode has less impact to the journal throughput on the local system in
comparison with the synchronous mode.

Journal entry latency may occur when remote journals are asynchronously
maintained. Journal entry latency is the difference between the journal entries
that exist in the remote journal on the target system from those residing in the
journal on the source system. From a recovery standpoint, the source system
may be some number of journal entries ahead of what journal entries are known
on the target system.
v The point in the chain of journal receivers where the replication of journal
entries should start.

Note: The activation of the remote journal can be a long running process. This
may occur if there are a large number of journal receivers and entries that
initially must be caught-up in the remote journal.

Caught-up is a conceptual state when all journal entries known to the

journal on the source system are also known to reside in the remote

Chapter 21. Remote Journal Function 481

 journal on the target system. The journal entries that reside in the
attached journal receiver of the remote journal could be confirmed or
unconfirmed journal entries.

Catch-up refers to the process of replicating journal entries that existed in

the journal receivers of the source journal before the remote journal was
activated. The catch-up phase is the most efficient method of replicating
journal entries to a remote journal. Control does not return to the
requester of the activation of the remote journal until this catch-up
processing has completed. You will want to consider this when making
the following selection.

Select the starting point for journal entry replication:

*ATTACHED
The replication of journal entries starts with the journal receiver that is
currently attached to the remote journal on the target system. The
journal entries are replicated from the corresponding journal receiver
that is associated with the journal on the source system. The replication
starts with the journal entries that follow the last journal entry that
currently exists in the attached journal receiver on the target system.
The remote journal on the target system might not have an attached
journal receiver. If this occurs, the journal receiver that is currently
attached to the journal on the source system is created on the target
system. That journal receiver is then attached to the remote journal on
the target system. Then journal entries are replicated starting with the
first journal entry in the journal receiver that is currently attached to the
journal on the source system.
If the journal on the source system does not have an attached journal
receiver, no journal entries can be replicated, and an error is returned.
This is only possible in the case of a remote journal that is associated
with another remote journal.
*SRCSYS
The replication of journal entries starts with the journal receiver that is
currently attached to the journal on the source system.
If the corresponding journal receiver exists and is attached to the remote
journal on the target system, journal entries are replicated. Replication
starts with the journal entries that follow the last journal entry that
currently exists in the attached journal receiver on the target system.
Otherwise, if the corresponding journal receiver exists but is not
attached to the remote journal on the target system, no journal entries
can be replicated. The system returns an error.
If the corresponding journal receiver does not exist on the target system,
the journal receiver is created and attached to the remote journal on the
target system. Journal entries then are replicated starting with the first
journal entry in the journal receiver that is currently attached to the
journal on the source system.
If the journal on the source system does not have an attached journal
receiver, journal entries cannot be replicated, and the system returns an
error. This is only possible in the case of a remote journal that is
associated with another remote journal.

482 OS/400 Backup and Recovery V5R1

 Qualified journal receiver name
The replication of journal entries starts with the specified journal
receiver name for the journal on the source system.
If the corresponding journal receiver exists and is attached to the remote
journal on the target system, journal entries are replicated. Replication
starts with the journal entries that follow the last journal entry that
currently exists in the attached journal receiver on the target system.
Otherwise, if the corresponding journal receiver exists but is not
attached to the remote journal on the target system, no journal entries
can be replicated. The system returns an error.
If the corresponding journal receiver does not exist on the target system,
the journal receiver is created and attached to the remote journal on the
target system. Journal entries then are replicated starting with the first
journal entry in the specified journal receiver.
If the journal on the source system is not associated with the specified
journal receiver, no journal entries can be replicated, and an error is
returned.

The creation of any receiver on the target system by the change journal state
processing is performed as though the receiver was being saved and restored to
the target system. Therefore, the ownership of the receiver on a target system
will follow the same rules as with the existing save and restore functions. If the
user profile which owns the receiver on the source system is on the target
system, then that profile will own the created receiver on the target system. If
the user profile does not exist on the target system, then the profile QDFTOWN
will own the receiver on the target system.

Additionally, information such as the audit attributes of the source journal

receiver at the time it was attached to the source journal will be incorporated
into the created journal receiver on the target system.

If the library for the journal receiver resides in an ASP, the journal receiver will
be created in that ASP. See “Journal Receiver ASP Considerations” on page 511
for more information.

Note: The remote journal function does not support nonlibrary ASPs for the
ASP of the remote journal receiver.
v If an asynchronous delivery mode was specified, then the sending task priority
may also be specified. If a value is not specified, the system selects a default
priority, which is higher than what the user can specify for this value.

Note: Setting this value too large may cause a greater journal entry latency or
lag.

Catch-up phase is initiated after the following actions occur:

v A request has been issued on the source system to activate a remote journal
v The system has determined which journal receivers and journal entries to
replicate to the target system
There is a difference between the catch-up phase processing and the run-time
synchronous or asynchronous processing. Catch-up processing replicates the
following to the target system:
v Those journal entries that already exist in the journal on the source system

Chapter 21. Remote Journal Function 483

 v Those journal entries that are deposited or replicated to the source journal
during the catch-up processing
Run-time synchronous or asynchronous processing occurs as part of the actual
deposit or replication of journal entries into the currently attached receiver on the
source system. While in the catch-up phase, the journal delivery mode will be
either *ASYNCPEND or *SYNCPEND, depending on the delivery mode that was
specified.

The catch-up phase is the most efficient method of transporting journal entries to a
remote journal in bulk.

The following is a high-level overview of the catch-up phase and related

processing:
1. Determine the starting point in the journal receiver on the source system.
2. If necessary, create a receiver on the target system and attach it to the remote
journal.
3. Replicate or complete replication for all of the journal entries that are contained
in the receiver on the source system to the corresponding receiver on the target
system.
4. If the receiver on the source system is the currently attached receiver, complete
the catch-up processing by transitioning into synchronous or asynchronous
remote journal function mode. Catch-up phase is complete, and control returns
to the requestor of the remote journal activation.
The remote journal will now be maintained synchronously or asynchronously
as additional journal entries are deposited, or replicated, into the attached
receiver on the source system.
5. If the receiver on the source system is not the currently attached receiver for
the journal on the source system, one of the following steps is performed:
v If there is a next receiver within the source journal’s chain of receivers, go
back to step 2. Replicate journal entries by starting with the first entry in the
next receiver.
For more information on receiver directory chains, see “Journal Receiver
Chains” on page 440.
v If there is no next receiver, (which indicates that a receiver chain break
exists), the catch-up phase is complete. Processing does not transition into
synchronous or asynchronous mode and the change journal state processing
ends. A final escape message is sent indicating that processing has ended.

After the system transitions a given remote journal to either the synchronous or
asynchronous remote journal function mode, the system continues to maintain that
mode. This continues until the remote journal function is inactivated for that
remote journal by using the Change Journal State (QjoChangeJournalState) API or
Change Remote Journal (CHGRMTJRN) command, or a failure occurs. See
“Inactivating the Replication of Journal Entries to a Remote Journal” on page 485
for more information on inactivating remote journals. See “Troubleshooting Errors”
on page 491 for more information on the possible failure conditions.

The replication of journal entries to an individual remote journal is performed

independently from the replication of journal entries to any other defined remote
journal. This is important if a given target system fails or if communications to a
target system fails from a particular source system. If either one occurs, the remote
journal function will end to those affected remote journals that reside on that target
system and are maintained from the source system. All other remote journals that

484 OS/400 Backup and Recovery V5R1

are being maintained from the source system will continue to function normally.
For example, a source journal could have two remote journals on two different
systems. In this situation, if the replication of entries from the source journal to the
second journal ended, the replication of entries from the source journal to the first
remote journal would not necessarily end. If a given remote journal has any type
of failure, the system ends the remote journal function. Appropriate messages are
signaled to either system or both systems involved, but the remote journal function
for other remote journals would not be affected. Likewise, the communications line
speed for a given asynchronously maintained remote journal will not affect the
speed for another asynchronously maintained remote journal using a different
physical transport.

Relational Database Considerations

Once a remote journal is activated, the remote journal function will work with the
communications configuration defined by the specified relational database (RDB)
entry as long as the remote journal is active. However, the information will be
taken from the RDB at the point in time when the remote journal was activated.
Therefore, even if the definition of the RDB entry is changed while a remote
journal has a journal state of *ACTIVE, none of those changes will take effect
immediately. If the remote journal is inactivated, and then activated again, the new
RDB entry definition will take effect. When you view the remote journal
information, the RDB entry information that is displayed represents the state of the
RDB entry information when the remote journal was last activated. See
“Displaying Remote Journal Function Information” on page 499.

Inactivating the Replication of Journal Entries to a Remote

Journal
You can also use the Change Journal State (QjoChangeJournalState) API and
Change Remote Journal (CHGRMTJRN) command to inactivate the replication of
journal entries to a remote journal. For this purpose, the API can be initiated from
either the source system or the target system.The CHGRMTJRN command can only
be initiated from the source system. The Change Journal (CHGJRN) command can
be used on the target system to inactivate the remote journal. It is recommended
that the replication of entries be ended from the source system whenever possible,
rather than from the target system. Usually, ending replication from the target
system for a remote journal is only necessary when the source system has failed,
and the system has not ended the remote journal function.

If you are inactivating an asynchronously maintained remote journal, you can

request that the remote journal function be ended immediately or in a controlled
fashion. For an immediate end, any journal entries which have already been
queued for replication will not be sent to the remote journal. For a controlled end,
any journal entries which have already been queued for replication will be sent to
the remote journal. When all queued entries have been sent to the target system,
the system sends message CPF70D3 to the journal message queue. The message
indicates that the remote journal function has been ended. If you are inactivating a
synchronously maintained journal, the remote journal function is ended
immediately, regardless of the whether an immediate or controlled end was
requested. Similarly, if the remote journal is in the catch-up phase of processing,
the remote journal function is ended immediately. This is also regardless of the
whether an immediate or controlled end was requested.

Journal Receivers Associated with a Remote Journal

Journal receivers that are associated with a remote journal are exact replicas of the
corresponding journal receivers that are associated with the journal on the source
system. The receiver directory for a remote journal is maintained in the same way
as the receiver directory is maintained for the related source journal. Consecutive

Chapter 21. Remote Journal Function 485

 receivers associated with a remote journal are linked together to form a receiver
chain. This is just like the corresponding receivers for the source journal are placed
into a similar receiver directory chain. Receiver chain breaks are forced and
maintained in a similar manner for local and remote journals.

For more information on journal receiver directory chains, see “Journal Receiver
Chains” on page 440.

However, the following are some other differences for remote journals and the
journal receivers that were attached to remote journals:
v A remote journal does not have to have a currently attached journal receiver. If
the remote journal is ready to receive journal entries, then it must have an
attached receiver. All the journal entries will be replicated to that attached
receiver.
v The receiver that is currently attached to a remote journal that is in the catch-up
phase can be a different journal receiver than is currently attached to the source
journal.
The receiver that is currently attached to an asynchronously maintained remote
journal can be a different journal receiver than is currently attached to the source
journal.
The receiver that is currently attached to a synchronously maintained remote
journal will be the same journal receiver as is currently attached to the source
journal.
v The journal receiver that is attached to a remote journal can be deleted if the
journal state of that journal is not *ACTIVE.
v You can delete the journal receivers that are associated with a remote journal in
any order, regardless of their position within the receiver directory chain.
v The creation date and time stamps for remote journals are always those of the
system on which the journals were created by the remote journal function. This
is also true for journal receivers that were attached to remote journals.
v The save and restore date and time stamps for remote journals are always those
of the system on which the save or restore operation took place. This is also true
for the journal receivers that are associated with the remote journals.
v The attach and detach time stamps for a journal receiver that was attached to a
remote journal are always those of the attach and detach time stamps of the
local journal receiver.
v When a journal receiver that is associated with a remote journal is saved, deleted
or restored, the following journal entries are not deposited:
– J RD - Journal receiver deleted
– J RF - Journal receiver saved, storage freed
– J RR - Journal receiver restored
– J RS - Journal receiver saved

Attaching a New Receiver to a Remote Journal

A change journal operation may occur on the source system to attach a new
receiver to a local journal. When this happens, the remote journal function
automatically attaches a new receiver to those remote journals that are currently
being maintained synchronously or asynchronously. If the journal sequence
numbers were reset as part of the change journal operation performed for the local
journal, the remote journal function will also reset the journal sequence number for
each remote journal. This keeps the journal sequence numbers synchronized
between the local journal and the remote journal. For remote journals that are
being synchronously maintained, a coordinated change journal operation is

486 OS/400 Backup and Recovery V5R1

 performed for the local journal on the source system and the remote journals on
the target systems. For asynchronously maintained remote journals, the new
receiver is attached when the target system receives the journal entry with journal
code ’J’ and entry type ’PR’ (previous receiver).

If the change journal operation fails on the target system, the remote journal
function ends for that remote journal, and processing continues on the source
system. The system sends a message to the journal message queue that indicates
that the remote journal function failed. When applicable, the system sends remote
journal failure type messages to the related journal message queues on both the
affected source and target systems. See “Troubleshooting Errors” on page 491 for
more information.

A change journal operation to attach a new receiver cannot be initiated directly for
a remote journal. New journal receivers are always attached to the remote journal
by the remote journal function as new receivers are attached to the local journal.
However, a change journal operation can be performed on a remote journal to
change several other attributes for the remote journal such as the journal message
queue or delete receivers value.

A change journal operation to attach a new receiver to a local journal that has an
associated remote journal in the catch-up phase can be performed. This is
regardless of whether the remote journal is currently being caught-up from a
detached or the currently attached receiver on the local system. The catch-up phase
of processing will not transition into synchronous or asynchronous delivery mode
until the end of the currently attached receiver for the local journal is reached.

Activating and Inactivating a Local Journal

Journal entries with journal code ’J’ and entry types ’LA’ and ’LI’ are deposited
when the local journal is activated and inactivated respectively.

Activating a Local Journal

When a local journal is created, the journal state of that journal is *ACTIVE. This
means that journal entries can be deposited to the local journal. If a local journal
has been inactivated, you can use the Change Journal State
(QjoChangeJournalState) API or Change Journal (CHGJRN) command to activate
the local journal.

Inactivating a Local Journal

Inactivating a local journal changes the journal state for the local journal to
*INACTIVE which prevents additional journal entry deposits. You can use the
Change Journal State (QjoChangeJournalState) API or Change Journal (CHGJRN)
command to inactivate a local journal. This is useful when performing a planned
switch-over3 of a primary system to a backup system or a switch-back4 from a
backup system to a primary system. The function is intended to prevent additional
journal entry deposits into the local journal of those journal entries that would
| affect the data content for journaled objects or for user-generated journal entries. A
| local journal can be inactivated with journaled objects open and in-use. If this
occurs, additional operations that attempt to generate journal entries will result in
an Entry Not Journaled exception (CPF7003) being signaled to the application. The
reason code for the exception will be code 10. An Entry Not Journaled exception

3. Switch-over describes the processing that is performed by a hot-backup application to logically promote a backup system to
assume the role of a primary system.
4. Switch-back describes the processing performed by a hot-backup application to allow the primary system to reassume its role
from a previously promoted backup system.

Chapter 21. Remote Journal Function 487

 will also be signaled if you attempt to send user journal entries to an inactivated
local journal. You would attempt to send the user journal entries by using the Send
Journal Entry (SNDJRNE) command or the Send Journal Entry (QJOSJRNE) API
interfaces.

Note: There are some exceptions to the rule that no journal entries may be
deposited in a journal which has a journal state of *INACTIVE. Table 110 on
page 807 lists which journal entries will still be deposited even though the
journal is inactive.

You can change the journal state for a local journal to *INACTIVE at any time,
with the following exceptions:
1. If any commitment control transaction that is associated with the journal has
any pending changes. This would include database files that are journaled to
the local journal that was opened under commitment control with pending
changes. This would also include API commitment control resources that use
the journal.
2. If any restore operation is in progress on the system.

Summary of the Journal Type, Delivery Mode, and Journal

State Interactions
The following table summarizes the remote journal function with respect to the
various journal types, delivery modes, and journal states.

The journal state describes an attribute for a journal. The attribute value can be
either *ACTIVE or *INACTIVE. For a local journal, *ACTIVE indicates that journal
entries are currently allowed to be deposited into the journal. *INACTIVE indicates
that journal entries are not allowed to be deposited.

The journal state for a remote journal on a target system that is associated with a
journal on a source system can be viewed in one of two ways.

When viewed from the source system, *ACTIVE indicates that journal entries are
currently being replicated to that remote journal on the target system. *INACTIVE
indicates that journal entries are not currently being replicated.

When viewed from the target system, *ACTIVE indicates that journal entries are
currently being received from the journal on the source system. *INACTIVE
indicates that the target journal is not ready to receive journal entries from the
source journal.
Table 68. Summary of the Journal Type, Delivery Mode and Journal State Interactions
Journal Type Delivery Mode Journal State Comments
| *LOCAL N/A *ACTIVE Objects journaled to the local journal can be changed,
| and entries can also be deposited into the local journal
| using the Send Journal Entry (SNDJRNE) command or
| the Send Journal Entry (QJOSJRNE) API interfaces. The
currently attached journal receiver may or may not be
currently replicated to one or more remote journals. This
depends upon whether any remote journals have been
added to the local journal’s definition, and if so, the
current journal state for each of those remote journals.

488 OS/400 Backup and Recovery V5R1

 Table 68. Summary of the Journal Type, Delivery Mode and Journal State Interactions (continued)
Journal Type Delivery Mode Journal State Comments
*LOCAL N/A *INACTIVE This is the state of a local journal after using the Change
Journal State (QjoChangeJournalState) API or Change
Journal (CHGJRN) command specifying
JRNSTATE(*INACTIVE) to not allow deposits into the
local journal, or after an IPL and the journal state of the
local journal was *INACTIVE when the system ended.

| Objects journaled to the local journal cannot be restored

| or changed, and entries can not be deposited into the
| local journal using the Send Journal Entry (SNDJRNE)
| command or the Send Journal Entry (QJOSJRNE) API
| interfaces, until the journal state for the local journal is
| again changed to *ACTIVE. This can be performed by
using the Change Journal State (QjoChangeJournalState)
API or Change Journal (CHGJRN) command specifying
JRNSTATE(*ACTIVE).
*REMOTE *SYNCPEND *ACTIVE This is the state after a remote journal has been activated
using the Change Journal State (QjoChangeJournalState)
API or CHGRMTJRN command and the processing is
still in the catch-up phase of remote journal activation.
Synchronous delivery mode was requested on the API
invocation.
*REMOTE *SYNC *ACTIVE This is the state after a remote journal has been activated
using the Change Journal State (QjoChangeJournalState)
API or CHGRMTJRN command, after catch-up has
completed, and changes to the currently attached journal
receiver for the journal on the source system are being
replicated synchronously to the remote journal on the
target system.
*REMOTE *ASYNCPEND *ACTIVE This is the state after a remote journal has been activated
using the Change Journal State (QjoChangeJournalState)
API or CHGRMTJRN command and the processing is
still in the catch-up phase of remote journal activation.
Asynchronous delivery mode was requested on the API
invocation.
*REMOTE *ASYNC *ACTIVE This is the state after a remote journal has been activated
using the Change Journal State (QjoChangeJournalState)
API or CHGRMTJRN command, after catch-up has
completed, and changes to the currently attached journal
receiver for the journal on the source system are being
replicated asynchronously to the remote journal on the
target system.
*REMOTE *SYNC *INACTPEND This is the state of a remote journal, viewed from the
target system where some failure has occurred and
either the system is in the process of inactivating the
remote journal, or unconfirmed journal entries exist in
the remote journal. See “Confirmed and Unconfirmed
Journal Entries” on page 508 for more information.
*REMOTE *ASYNC *INACTPEND This is the state of a remote journal, viewed from the
target system where some failure has occurred and the
system is in the process of inactivating the remote
journal.

Chapter 21. Remote Journal Function 489

Table 68. Summary of the Journal Type, Delivery Mode and Journal State Interactions (continued)
Journal Type Delivery Mode Journal State Comments
*REMOTE *ASYNC *CTLINACT This is the state after a remote journal has been
inactivated using the Change Journal State
(QjoChangeJournalState) API or CHGRMTJRN
command, a controlled inactivate was requested on that
invocation and that controlled inactivate has not yet
completed.
*REMOTE N/A *INACTIVE This is the state after a remote journal has been added
and associated with a journal on a source system.
However, the journal state for the added remote journal
has yet to be activated or has been inactivated using the
Change Journal State (QjoChangeJournalState) API,
CHGRMTJRN command, or by an IPL. No delivery
mode is in effect for an inactivated remote journal.

Removing a Remote Journal

Note: In the following section, you should be aware whether library redirection is
in effect for the remote journal. If it is in effect, any library name processing
will substitute the redirected library name for the library name that is used
for the operation on the target system.

You can disassociate a remote journal on a target system from a journal on a

source system by using one of the following methods:
v The Remove Remote Journal (QjoRemoveRemoteJournal) API.
v The Remove Remote Journal (RMVRMTJRN) command.
Both the QjoRemoveRemoteJournal API and the RMVRMTJRN command must be
started on the source system for the journal on the source system identifying which
remote journal to remove.

When using either of these methods, the replication of journal entries to the remote
journal to be removed cannot be currently active. If the remote journal state is
*ACTIVE, use the Change Journal State (QjoChangeJournalState) API or Change
Remote Journal (CHGRMTJRN) command to inactivate the replication of journal
entries to the remote journal.

The remote journal, and any associated journal receivers, are not deleted from the
target system when you use the Remove Remote Journal
(QjoRemoveRemoteJournal) API or RMVRMTJRN command. Neither method
initiates any processing on the target system. Once the remote journal is removed
from the journal on the source system, you are responsible for deleting the remote
journal and associated journal receivers, if desired.

You can add this remote journal back to the remote journal function definition for
the journal on the source system. Use the Add Remote Journal
(QjoAddRemoteJournal) API or ADDRMTRJN command.

Once a remote journal is removed, the journal receivers are no longer protected
from deletion. See “Preventing Journal Receiver Deletion” on page 479 for more
information.

The following is the input that you must provide to remove a remote journal on a
target system:

490 OS/400 Backup and Recovery V5R1

 1. The journal name and library on the source system from which the remote
journal is being removed.
2. The remote journal name and library on the target system that is being
removed.
3. A relational database directory entry, which identifies the target system and
other necessary communications information.

Troubleshooting Errors
Several different error conditions can occur when the remote journal function is
active. When an error condition is encountered, the system automatically ends the
remote journal function on the source system to that remote journal. The system
notifies you that a failure occurred. Failure notification is made on both the source
system and the target system. Notification is made by sending a message to the
journal message queues associated with the source and target journals as
appropriate.

In some cases where the source system fails, you may have to inactivate the remote
journal on the target system. (You inactivate the remote journal on the target
system by using the Change Journal State (QjoChangeJournalState) API or the
CHGJRN command.) For a synchronously maintained remote journal, inactivating
the remote journal on the target system will discard all unconfirmed journal
entries. See “Confirmed and Unconfirmed Journal Entries” on page 508 for more
information.

Note: The source system may not detect that an error has occurred on the target
system until the next time a journal entry is replicated to that failing target
system.

Additional messages can be sent to the journal message queue for normal remote
journal processing. For example, if you requested a controlled inactivate of the
remote journal, a message will be sent to the message queue when the inactivate
processing has completed.

Note: Even though the remote journal function has been ended, the local journal is
not automatically inactivated. Therefore the local system journal entry
deposits will continue normally.

The remote journal function messages that are sent to the journal message queue
are listed below.
CPF70D3
A controlled inactivate of a remote journal has completed.
CPF70D4
The remote journal function is no longer active due to various reasons. For
a synchronously maintained remote journal, there may be unconfirmed
entries which may need to be processed prior to the remote journal being
inactivated.
CPF70D5
The remote journal function is no longer active and has been ended due to
various reasons. There are no unconfirmed entries.
CPF70D6
The remote journal function was ended due to storage constraints.

Chapter 21. Remote Journal Function 491

 CPF70D7
There was a problem on the target system while attempting to execute a
change journal.
CPF70DB
A severe error has occurred with the remote journal function, and service
should be notified.
CPF70DC
There was a problem on the target system while attempting to execute a
change journal.

Display the messages on your system for more information.

Performance Considerations
There are two main performance objectives for the remote journal function. To
provide a timely delivery of journal entries to a target system and to minimize
impacts to the journaling throughput on the source system. Even though both
aspects are very important for both synchronous and asynchronous delivery
modes, each mode prioritizes the two in a different order. The top priority for
synchronous delivery mode is a timely delivery of journal entries. For
asynchronous delivery mode, the top priority is to minimize impacts to journaling
throughput.

All performance considerations that are currently used for a local journal still
apply and should continue to be employed. See “Journaling and System
Performance” on page 394 for information on local journaling performance. The
following are additional factors that may affect the performance of the remote
journal function. The factors are listed in the order of importance.
1. Transport method
Either the OptiConnect for OS/400 bus transport or a communications
transport could be considered when using a synchronous delivery. However,
depending on the rate of the journal activity, the OptiConnect for OS/400 bus
transport may be the most suitable method and asynchronous transfer mode
(ATM) may be a good alternate. You will have to weigh the response time
impacts of the synchronous delivery mode in your environment against the
communications overhead of the transport methods.
Either the OptiConnect for OS/400 bus transport or a communications
transport could be considered when using an asynchronous delivery. A
communications transport will have to be used when replicating journal entries
over a long distance. The most important performance factors regarding a
communications transport method are the overall rated speed of the
communications resource and any existing traffic already using the
communications resource.
2. Number of remote journals that are being maintained
With respect to the job performing the journal entry deposit, the impact of the
remote journal function increases by an equal factor for each remote journal
that is added. For example, if you have three synchronously maintained
journals, the impact to the job is three times that of one synchronously
maintained journal.
However, the impact to the job performing the journal entry deposit for an
asynchronously maintained journal is significantly less than that for a
synchronously maintained journal.

492 OS/400 Backup and Recovery V5R1

 IBM recommends that only one synchronous remote journal be maintained for
a given local journal. This is because the application cannot continue until the
journal entry has been replicated to the remote journal.
With respect to the system performance impacts, the overhead typically
increases by less than an equal factor for each additional remote journal.
3. Arrival rate of journal entries that are being deposited on the local system
The higher the arrival rate of journal entries being deposited on the local
system, the greater the chance of affecting journaling throughput for
synchronous or asynchronous delivery. This may cause asynchronous
journaling to fall further behind.
4. Batch versus Interactive
In general, higher remote journal throughput can be maintained when many
interactive jobs generate the journal throughput rather than a single-threaded
batch job. This also requires less journal and remote journal overhead.
5. CPU utilization on the source system
The higher the CPU utilization of the source system, the greater the chance of
affecting journaling throughput for synchronous or asynchronous delivery. This
may cause asynchronous journaling to fall further behind.
6. CPU utilization on the target system
The higher the CPU utilization of the target system, the greater the chance of
affecting journaling throughput for synchronous or asynchronous delivery. This
may cause asynchronous journaling to fall further behind.
7. The value set for the sending task priority when using the asynchronous
delivery mode
The larger the value, the smaller effect the remote journal function will have on
the system, but the further the target system may lag behind the source system.

Performance considerations regarding the catch-up phase when activating the

remote journal function include the following:

Note: The catch-up processing that is performed by the remote journal function is
the most efficient method of replicating the journal entries with the remote
journal function.
1. Total number of bytes for all of the journal entries that need to be caught up
The larger the total size, the longer the catch-up phase will run.
2. Transport method
In general, an OptiConnect for OS/400 bus transport method will outperform a
communications transport method. The difference depends on the exact
configuration and type of communications method to be used.
3. Disk protection on the target system
At high data transfer rates, disk units with device parity protection in the ASP
on the target system can limit the performance of the catch-up phase. One
example of this is when you use the OptiConnect for OS/400 bus transport
method. Having mirrored or unprotected disk units in the ASP on the target
system would eliminate this effect.
4. CPU utilization on the source system
The higher the CPU utilization of the source system, the greater the chance of
affecting the performance for the catch-up phase.
5. CPU utilization on the target system
The higher the CPU utilization of the target system, the greater the chance of
affecting the performance for the catch-up phase.

Chapter 21. Remote Journal Function 493

 6. Delivery mode
The performance of the catch-up phase does not depend on the delivery mode
that was specified, synchronous or asynchronous.

For additional information see the Redbook AS/400 Remote Journal Function for
High Availability and Data Replication (SG245189-00).

Auxiliary Storage Considerations

Auxiliary storage will be required on both the source and target systems. The
amount that is required will be about the same on both systems. Anything that is
done to minimize the amount of auxiliary storage required on the source system
will reduce the amount of auxiliary storage required on the target system.
Additionally, the less auxiliary storage used, or smaller the journal receivers are,
the less data is transmitted on the communications links. Therefore, the
communications overhead will be reduced.

See “How You Can Reduce the Storage Used by Journaling” on page 397 for more
information on ways to reduce the auxiliary storage usage.

If the target system is not working for any extended period of time, enough
auxiliary storage on the source system is needed to keep the journal receivers
on-line. This will be required until the target system becomes available.

Main Storage Considerations

Providing greater amounts of main storage in the *BASE main storage pool on the
source system will improve remote journal performance. This is especially true in
environments with multiple asynchronously maintained remote journals.

Providing greater amounts of main storage in the *BASE main storage pool on the
target system will improve remote journal performance. This is especially true in a
remote journal network with a high volume of activity. The additional storage will
keep the number of page faults to a minimum, and reduce the impacts to the
target system.

Example Remote Journal Function Environments

This section describes two typical environments that the remote journal function
can effectively be used with:
v A data replication environment
v A hot-backup environment

Data Replication Environment

Figure 41 on page 495 describes a typical remote journal environment that is used
for data replication purposes only. Data replication is the function of maintaining a
separate copy of data from an original copy, keeping the two copies consistent
with each other.

494 OS/400 Backup and Recovery V5R1

|

|
| Figure 41. Typical Data Replication Environment with Remote Journal Function

| Local objects on system S1 are journaled to local journal JRN in library JLB1. A
remote journal is defined on system S2, where JRN has been redirected to library
JLB2. This remote journal is receiving journal entries from the local journal on
| system S1. A hot-backup application apply is replaying the changes to the data
| replica on system S2. The data replica is journaled to a local journal, JRN in library
| JLB1, for system recovery purposes only. If system S2 was to fail, the system would
| perform recovery for the objects by using this local journal.

| A hot-backup application is used to assist in replicating data from one system to

| another. A hot-backup application typically performs the following:
1. If the primary system fails, it performs a switch-over to the backup system.
This function then logically promotes the backup system to assume the role of
the primary system
2. After the failed primary system is restarted, it performs a switch-back operation
so that the primary system can again assume the role of the primary system.

| A hot-backup application apply defines the part of a hot-backup application that

| actually performs the replay operations to the data replica. This usually occurs on
| the backup system when maintaining a data replica.

| Note: In the following sections, a hot-backup application apply is performing the

| replay of operations to the data replica on the target system. This is for
simplification purposes, even though this example is for the data replication
environment. The Apply Journaled Changes (APYJRNCHG) and Remove
Journaled Changes (RMVJRNCHG) commands cannot be used for this apply
processing.

Establishing the Data Replication Environment

Establishing the data replication environment in Figure 41 requires the following:
| 1. The objects and local journal on system S1 are already assumed to exist. The
journal state for the local journal is also assumed to be *ACTIVE. The
communications environment and associated RDB entries already exist and are
established.
2. Use the Add Remote Journal (QjoAddRemoteJournal) API or ADDRMTJRN
command on system S1 to create the remote journal on system S2, and to
specify library redirection. This indicates that the journal’s library, JLB1 on

Chapter 21. Remote Journal Function 495

 system S1, will be redirected to library JLB2 on system S2. The journal
receiver’s library, RLB1 on system S1, will be redirected to library RLB2 on
system S2.
Note that after this step, the remote journal exists, but no receiver is currently
attached.
3. To establish a clean breakpoint, you may want to perform a change journal
operation to attach a new journal receiver at this time.
Notes:
a. Step 4 restores local journal JRN in library JLB1 and attaches receiver X1002 in
library RLB1. It then restores the objects, and starts journaling for the objects to
the restored local journal.
b. Since you cannot save and restore the contents of data queues, you will want to
take that into consideration when priming the data replica for any data queue
objects.
| 4. Save the local journal and objects from system S1 and restore them to system
| S2. This primes the data replica and establishes the local journaling
| environment on system S2.
5. Use the Change Journal State (QjoChangeJournalState) API or Change Remote
Journal (CHGRMTJRN) command on system S1 to activate the remote journal
on system S2. Specify that the remote journal should start with the
*ATTACHED receiver. Since no receiver is attached to the remote journal, the
receiver that is currently attached to the local journal on system S1 (X2) is
created on the target system. This receiver is then attached to the remote
journal. Journal entries are replicated, starting with the first journal entry in
receiver X2.
An additional parameter on the Change Journal State (QjoChangeJournalState)
API and CHGRMTJRN command indicates whether the remote journal function
is to be maintained synchronously or asynchronously. Depending on how the
remote journal is maintained, other parameters may also apply.
| 6. The hot-backup application apply process receives or retrieves journal entries
| from the remote journal, starting with the entries that were deposited after the
| data was saved and primed into the data replica. The process then starts
| replaying the changes that are contained in these journal entries to the data
| replica.

Normal Run-Time Environment for the Data Replication

Environment
Normal run-time environment for the data replication environment in Figure 41 on
page 495 requires the following steps:
1. You can stop and start the replication of journal entries to the remote journal as
desired. Use the Change Journal State (QjoChangeJournalState) API or
CHGRMTJRN command to activate and inactivate the remote journal. Each
time the remote journal is activated, *ATTACHED is specified as the point in
the receiver chain to start receiving journal entries. The system checks the
currently attached remote journal receiver for journal entries and replicates the
next journal entry in sequence.
You must specify the delivery mode when activating the remote journal. If
desired, it can be different on each activation of the remote journal.
2. Change journal operations that attach a new receiver to the local journal on
system S1 are performed by the remote journal function as required on the
target system. The remote journal function attaches the associated receivers to
the remote journal automatically. If the remote journal is being maintained
synchronously, the change journal operation to attach a new receiver is

496 OS/400 Backup and Recovery V5R1

 essentially a coordinated operation between the source and target systems. If
the remote journal is being maintained asynchronously, the change journal
operation to attach a new receiver on the target system is performed differently.
In this case, it is triggered when the journal entry with journal code ’J’ and
entry type ’PR’ is received by the remote journal on the target system.
| 3. The hot-backup application apply continues to replay changes to the data
| replica as received or retrieved from the receivers associated with the remote
| journal.
4. If desired, the receivers that are associated with the local journal on system S1
can be deleted when each receiver is replicated to system S2. This can be
accomplished by using the DLTRCV(*YES) support or manually deleting the
receivers on system S1.
| You can save the receivers from system S2. If necessary, you can use the
| receivers for recovery of the original data on system S1 at a later time.
See “Deleting Journal Receivers” on page 418 for more information.

Data Replication Recovery Given a Failure of System S1

Recovery for Figure 41 on page 495 is simpler than those involving hot-backup.
| Cases where the logical ownership of the data is given to another system are more
| complicated. (These are cases where a backup system is logically promoted to
assume the role of the primary system.) This is because after system S1 has
completed its IPL, the remote journal function catch-up phase from the local
journal on system S1 to the remote journal on system S2 will always allow a
| resynchronization of the two sets of data. Data resynchronization is recovery
processing that is performed during switch-back processing by a hot-backup
| application apply. This processing ensures that the original data is consistent with
| the data replica, and contains all the correct changes. The main objective of this,
| besides assuring data consistency, is to eliminate re-priming the original data from
| the data replica.

Recovery complications can be prevented in the cases where a user does not
| switch-over to the backup system. What prevents them is an assumption that the
| hot-backup application apply logic will not receive and replay unconfirmed journal
| entries to the data replica if system S2 loses communications with system S1. In all
| cases where system S1 fails, a switch-over to S2 is not performed, and the
| unconfirmed journal entries are not replayed to the data replica. Therefore, the
| data replica on system S2 can never ’get ahead’ of the data on system S1. This
| greatly simplifies data synchronization. See “Confirmed and Unconfirmed Journal
Entries” on page 508 for a discussion on unconfirmed entries.

Hot-Backup Environment
Figure 42 on page 498 describes a typical remote journal environment that is used
for hot-backup purposes.

Chapter 21. Remote Journal Function 497

 Figure 42. Typical Hot-Backup Environment with Remote Journal Function

Establishing the Hot-Backup Environment

The same steps that are used for the data replication environment (see
“Establishing the Data Replication Environment” on page 495) are used for the
hot-backup environment except for this additional last step:
1. A remote journal is also established on the local system S1 and associated with
the local journal that is created on system S2. This remote journal will receive
or retrieve the journaled changes that are made when S2 assumes the role of
| the primary system. Note that this local journal and remote journal pair will
| only be used when replicating changes back to the original data. During
normal run-time processing, the remote journal, JLB2/JRN, that is defined on
system S1 is not active. When it is not active, it is not receiving or retrieving
journal entries from the local journal, JLB1/JRN, on system S2.

Normal Run-Time Environment for the Hot-Backup Environment

The same steps that are used for the data replication environment are used for the
hot-backup environment. See “Normal Run-Time Environment for the Data
Replication Environment” on page 496.

Hot-Backup Recovery Given a Failure of System S1

See “Example: Remote Journal Function Recovery” on page 511.

498 OS/400 Backup and Recovery V5R1

Displaying Remote Journal Function Information
When you are working with the remote journal function, you will want to be able
to view the remote journal network. You may also want to view the various
attributes, journal states, or delivery modes. The Work with Journal Attributes
(WRKJRNA) displays include the list of all remote journals that are associated with
a given journal. When looking at a specific journal, you can see information about
the journal’s source journal, if any. Additionally, you can see all remote journals
which are immediately downstream from the specified journal. If those remote
journals are cascaded to other remote journals, you will not be able to see any
cascaded remote journal information. To see that information, you must invoke the
WRKJRNA command for that remote journal on its own system.

This information is also available through the Retrieve Journal Information

(QjoRetrieveJournalInformation) API.

Additionally, the Display Journal Receiver Attributes (DSPJRNRCVA) displays

provide additional information about the remote journal characteristics of the
journal receivers. The DSPJRNRCVA command also has an API counterpart to
allow program retrieval of the journal receiver information, the Retrieve Journal
Receiver Information (QjoRtvJrnReceiverInformation) API.

Managing the Remote Journal Function

Managing the remote journal function requires these basic tasks:
v Keeping records of your remote journal network.
v Evaluating the impact on the remote journal network as new applications are
added or the system workload grows.
v Considering the ramifications of journal receivers on two systems which require
regular save and delete processing.
v Considering the save and restore implications of the remote journal network.

Keeping Records of Your Remote Journal Network

You should always have a current list of the remote journals that are associated
with local journals, and their associated communications information.

For each journal which has remote journals associated with it, use the following
command: WRKJRNA JRN(library-name/journal-name) OUTPUT(*PRINT).

Alternatively, you could use the Retrieve Journal Information

(QjoRetrieveJournalInformation) API to retrieve the information and place it in a
file.

Additionally, to get the related relational database information, use the following
command: WRKRDDIRE RDB(*ALL) OUTPUT(*PRINT).

Remember to do this for all cascaded remote journals as well, not just the local (or
primary) system.

Evaluating How System Changes Affect Your Remote Journal

Network
After you have initially established your remote journal network, you need to keep
up with changes that occur on the system.

Chapter 21. Remote Journal Function 499

 If the amount of work that is going to the journals which you are replicating
increases, you may need to consider upgrading the communications method.

The traffic rate for work other than the remote journal function may increase on a
communications method that is shared. If this occurs, you may need to consider
separating the various pieces of communications traffic so that the remote journal
function is not impaired. This is especially important if you are using the
synchronous delivery mode.

An application that is being protected may become more critical to your business,
where any time that the system is not working is considered disastrous. If this
occurs, you may need to consider upgrading that application’s remote journal to
using the synchronous delivery mode so that no journal entries are lost.

Maintaining Journal Receivers

As with local journals, you should regularly save and delete your journal receivers
to minimize the amount of on-line auxiliary storage which is used by the journal
receivers.

If you plan to off-load the save overhead to the backup system, the journal
receivers can be quickly deleted from the primary system after they have been
replicated to the backup system. Consider using the DLTRCV(*YES) attribute to
have the system manage this delete processing. On your backup system you could
use the DLTRCV(*NO) attribute on the remote journal and manage the receiver
| save processing as you did before. Remember that the source journal receiver
| cannot be deleted until it has been replicated to all associated remote journals. If
you had cascaded remote journals, you may want to use the DLTRCV(*YES)
attribute on the local journal, and on the lowest level remote journal. You would
use DLTRCV(*NO) on the cascaded remote journal since you plan to do your save
processing on that system.

The Delete Journal Receiver exit point, QIBM_QJO_DLT_JRNRCV, may be of

assistance as well. For example, you may want to add an exit program to
QIBM_QJO_DLT_JRNRCV which verifies that the journal receiver is no longer
needed for any hot-backup application apply processing before it can be deleted.
Refer to “Deleting Journal Receivers” on page 418.

Note: The change journal processing for a remote journal is driven by the source
journals change journals. See “Attaching a New Receiver to a Remote
Journal” on page 486 for more information.

Save and Restore Considerations

Journal Considerations
IBM recommends that you save the remote journal network after the addition of
any and all remote journals that will be associated with the journal. This includes
saving the local journal and any associated remote journals, as well as the journal
receivers that are associated with the local journal.

Follow the basic save and restore rules for journals that are listed here:
1. A saved local journal is always restored as a local journal.
2. A saved remote journal is always restored as a remote journal.
3. As with all prior save and restore support for journals, the support will not
allow a restore-over operation for a journal. This is true for both local and
remote journals.

500 OS/400 Backup and Recovery V5R1

 4. When restored, a local or remote journal is always restored to the library from
which it was saved. For a local journal, this library is referred to as the original
journal library. For a remote journal, this library is referred to as the redirected
journal library.
For a remote journal, library redirection may not have been specified when
adding the remote journal to the local journal’s definition. If this occurs, then
the redirected journal library name is the same name as the original journal
library name.

Note: This is always true except in the case where the journal was saved from
library QRCL. (The journal could reside in library QRCL due to prior
Reclaim Storage processing.) In that case, the RSTLIB parameter must be
specified on the restore request, and you must specify the library where
the journal originally resided. For a local journal, this is existing support
and is not new. For a local journal, the library that must be explicitly
specified is the original library.

This support logically extends to remote journals. For a remote journal, the
redirected library must be explicitly specified on the RSTLIB parameter of the
restore request.
5. If remote journals are associated with a journal when a journal is saved, the
information that is related to the added remote journals is also saved.
When the journal is restored, the information that is saved about its remote
journals is also restored. This information is included as part of that journal’s
definition. This is true whether the journal being saved is a local or a remote
journal. When restored, the restored journal’s definition will only include the
saved, immediately downstream remote journal definitions.

Note: None of the actual downstream remote journals are actually verified as
part of the restore operation. Any necessary validation of the remote
journal information occurs when you activate that particular remote
journal by using the Change Journal State (QjoChangeJournalState) API
or Change Remote Journal (CHGRMTJRN) command.
6. Local journals are always restored in the active state. The following are the
reasons behind this design point:
| a. When you restore journals and journaled objects at the same time, it allows
| the system to successfully start journaling for those restored objects. The
| system cannot successfully start journaling for journals and journaled
| objects that are restored in the inactive state.
b. It also follows the logic that a restore of a journal is similar to the create of
a journal. Local journals are always created in the active state.

N to N-1 Save and Restore Considerations for Journals: The following are
considerations when performing a save of a journal and the target release is prior
to V4R2M0:
v If the journal is a remote journal, it cannot be saved.
v If the journal is a local journal, it will be saved, but the remote journal related
information will be removed. This includes information about downstream
remote journals.

Journal Receiver Considerations

Figure 43 on page 502 illustrates the restore relationships for journal receivers that
are associated with remote journals, based on the remote journal type.

Chapter 21. Remote Journal Function 501

 Note: Receivers can always be saved from a journal, and then restored to another
local journal of the same name. However, they will be placed in their own
separate receiver chain.

Figure 43. Journal Receiver Restore Characteristics and Remote Journal Types

There are several unique rules which govern where the journal receivers that are
associated with a remote journal can be restored. The rules also address the
placement of the journal receivers in the receiver directory chain of a local or
remote journal. These rules are influenced by the remote journal type of the journal
to which the journal receiver was attached. These rules are also influenced by the
library redirection that was in effect when that receiver was attached. See “Remote
Journal Type Descriptions” on page 476.

The following items describe the rules that are used by the system when restoring
journal receivers:
1. The system first attempts to find an appropriate remote journal. When
searching for a remote journal, the following rules are used:
a. If the saved receiver was originally associated with a local or *TYPE1
remote journal, then search for a *TYPE1 remote journal.
1) If a *TYPE1 remote journal was defined at the time this receiver was
attached, then use the journal and receiver library redirection that was
in effect and saved with the receiver. If no *TYPE1 remote journal was
defined at the time this receiver was attached, then the original journal
library and receiver library names will be used when searching for the
*TYPE1 remote journal.

502 OS/400 Backup and Recovery V5R1

 2) If a *TYPE1 remote journal is found, and the current receiver library
redirection for the found *TYPE1 remote journal matches the library
name where the receiver is being restored, the journal receiver will be
associated with the found *TYPE1 remote journal.
b. If the receiver was originally associated with a *TYPE2 remote journal, then
search for a *TYPE2 remote journal. When searching for the *TYPE2 remote
journal, a journal with the same name as the name that was saved with the
receiver will be used.
1) The journal receiver will be associated with a found *TYPE2 remote
journal if the following conditions are met:
v A *TYPE2 remote journal is found with the correct name in the correct
library
v The found journal is in the exact same remote journal network as that
of the saved receiver
v The receiver is being restored to the same named system as the name
of the system at the time the receiver was saved
2. If a remote journal was not found, then the system searches for a local journal.
When searching for a local journal, the original journal and journal library
names are used.
The journal receiver will be associated with a found local journal if the
following conditions are met:
v A local journal is found by the correct name in the correct library
v The original journal receiver library name for the found journal matches the
library name where the receiver is being restored
3. If a local journal cannot be found, the restore operation will be allowed to
proceed. The journal receiver will not be associated with any journal, if the
receiver is being restored to the original or redirected receiver library.
4. Still honoring the previous receiver restore rules, the following must also be
true if the receiver is being restored over an existing receiver:
a. If the receiver is not being associated with any journal (as previously
determined from the prior receiver restored rules), then:
1) The receiver creation time stamps must match.
2) If the saved receiver was ever associated with a journal, then it must
have been previously associated with a journal of the same type as that
of the existing receiver.
3) If the saved receiver was ever associated with a remote journal network,
then it must have been previously associated with the same remote
journal network as that of the existing receiver.
4) The saved receiver must have at least as many entries as the existing
receiver.
b. If the receiver is being associated with a local journal, then:
1) If the saved receiver was originally associated with a local journal, then
the receiver creation time stamps must match.
2) If the saved receiver was not originally associated with a local journal,
then the saved receiver must have been originally associated with the
same remote journal network as that of the existing receiver.
3) The saved receiver must have at least as many entries as the existing
receiver.
c. If the receiver is being associated with a *TYPE1 remote journal, then:

Chapter 21. Remote Journal Function 503

 1) The receiver creation time stamps must match, and the saved receiver
must have been originally associated with a local or *TYPE1 remote
journal.
d. If the receiver is being associated with a *TYPE2 remote journal, then:
1) The receiver creation time stamps must match, and the saved receiver
must have been originally associated with the same *TYPE2 remote
journal.

Note: When receivers are saved from or restored to a target system and associated
with a remote journal, no journal entries are deposited to indicate that the
save or restore occurred. However, the object save and restored date and
time stamps are updated accordingly.

Non-Replicated Journal Receiver Protection Considerations: The protection

provided, which prevents journal receivers that are not fully replicated to all
associated remote journals from being deleted, is removed when the journal
receiver is restored.

Unconfirmed Journal Entries Save Considerations: When a journal receiver that

is associated with a remote journal is saved, only those journal entries which have
been confirmed are saved to the media. Therefore, no unconfirmed journal entries,
nor any journal entries that would not survive any IPL journal recovery processing,
will be saved.

Journal Receivers Saved with STG(*FREE) Considerations: Even if a journal

receiver has not been fully sent to all known remote journals, such a journal
receiver can be saved with STG(*FREE). However, a diagnostic message is left in
the joblog indicating the freeing of the journal receiver storage without the journal
receiver first being fully replicated to all downstream remote journals. This is in
contrast to the default action taken when attempting to delete a receiver that has
not been fully replicated to all downstream remote journals.

N to N-1 Save and Restore Considerations for Journal Receivers: The following
are considerations when performing a save of a journal receiver and the target
release is prior to V4R2M0:
v If the journal receiver is associated with a remote journal, it will be saved, but
the remote journal related information will be removed.

Journaled Object Considerations

| For an object that is restored and associated with an inactive local journal, the
| restore of the object completes normally. However, an informational message is
| sent to the joblog. This message indicates that journaling could not be started for
| the object. Existing restore logic will send a final escape message for the restore
operation.

| For an object that is being restored-over and it is currently journaled to an inactive

| local journal, the restore for the object is prevented. A diagnostic message is left in
the joblog. The diagnostic message is followed by the appropriate message that is
currently sent by restore. The current message indicates that the restore for the
particular object failed.

| The system will send a diagnostic message for any object in which the ’object
| restored’ journal entry could not be sent due to a problem with the journal or
| attached journal receiver. The system always attempts to start journaling for an
| object that was journaled at save time to the same named journal, in the same
| named library, during a restore operation. This is still true, and there are no

504 OS/400 Backup and Recovery V5R1

 processing changes to note if a local journal is found by the restore processing.
| However, if a remote journal is found by the restore processing, the restore
| completes successfully, but journaling is not started for the restored object. A
diagnostic message is sent that indicates that a remote journal was found by the
restore processing. This message is followed by the message that is already sent
that indicates journaling was not started.

| In a hot-backup configuration, a local journal is used on the backup system to

| capture the changes that are made to the objects on the remote system. This occurs
when the remote system is logically promoted to assume the role of the primary
| system. The local journal that is being used on a backup system may not be in the
| exact same-named library as the journal that is being used for the object at save
| time. If this occurs, you are responsible for starting journaling for the restored
| objects.

Note: This is a fundamental reason to use library redirection for all defined remote
journals.

Save Storage (SAVSTG) Considerations

If a system is restored from SAVSTG media, the primary remote journal function
concerns have to do with configuration changes involving additionally defined
remote journals. These remote journals were established after the SAVSTG media
was produced. If a primary system is restored from SAVSTG media, journal
receivers can be restored back to the primary system from versions saved from any
of the associated remote journals in the remote journal environment. If a backup
system is restored from SAVSTG media, then the catch-up phase for activating the
remote journal can replicate all necessary journal receivers that are still on-line
from the primary system to the restored backup system. Those journal receivers
that are not on-line, and were attached to a *TYPE1 remote journal, can be restored
back to the backup system. They can be restored from any saved versions of the
journal receivers that were previously taken from one of the following:
v The primary system
v Any of the associated remote journals in the remote journal environment
See “Journal Receiver Considerations” on page 501 for the journal receiver restore
rules which will be used for such a restore.

Another consideration occurs as part of the processing that is performed by the

system when restoring journal receivers. Before associating a journal receiver with
a local journal and retaining any remote journal information, both the journal
library name and the system name must be correct. This allows the system to
differentiate between a local journal that was originally created and one that was
restored to a different physical system using SAVSTG media. This case assumes
that the user assigns a new system name as part of the SAVSTG procedure.

A related case can potentially cause trouble. In this case, the system was also
restored using SAVSTG media. However, the system it was restored to was not the
same physical system but it still had the same name as the system from where the
media was produced. You should avoid duplicating this situation.

Working with Journal Entries and the Remote Journal Function

Working with the journal entries in a remote journal is essentially the same as
working with the journal entries in a local journal. See “Chapter 20. Working with
Journal Entries, Journals, and Journal Receivers” on page 427 for more information.
The exceptions are as follows:

Chapter 21. Remote Journal Function 505

 v Retrieving journal entries with library redirection
v System dependent information
v Retrieving the journal entry object names during catch-up processing
v Retrieving journal entries with commitment control
v Confirmed and unconfirmed journal entries

Retrieving Journal Entries from a Remote Journal with Library

Redirection
All journal entries that are retrieved from a remote journal will have the object
names as they exist on the local system.

The following journal entries will show the name of the journal receiver as it was
on the local system even if the entry is displayed on a remote system. This is
because these entries really apply to the version of the journal receiver that existed
on the local system.
v J PR - Previous Receiver entry
v J NR - Next Receiver entry
v J RD - Receiver Deleted
v J RR - Receiver Restored
v J RS - Receiver Saved
v J RF - Receiver Saved with storage Freed
v Object saved entries - See Table 137 on page 850 for a list of the possible entry
types.
v Journal changes applied entries - See Table 115 on page 822 for a list of the
possible entry types.
v Journal changes removed entries - See Table 115 on page 822 for a list of the
possible entry types.

System Dependent Information

The following information in the journal entries is based on the system of the
original local journal. It is not based on the system of the remote journal where the
entries are viewed.
v System name
v Date
v Time stamp

Retrieving Journal Entries from a Remote Journal During the

Catch-up Phase
During the catch-up phase, journal entries that have been replicated to the target
system can be retrieved from the remote journal.

You can activate and inactivate the remote journal function while concurrently
running the following commands to view journal entries on the target system:
v Display Journal (DSPJRN)
v Retrieve Journal Entry (RTVJRNE)
v Receive Journal Entry (RCVJRNE)
v Retrieve Journal Entries (QjoRetrieveJournalEntries) API

506 OS/400 Backup and Recovery V5R1

 When the remote journal is in the process of being caught-up from the attached
journal receiver on the source system, two things can happen with respect to
objects and their names in the journal entries.
1. First, if journaling is started for any objects on the source system, the object
name that is given on the target system in the start journal entry may be
*UNKNOWN.
2. Second, if any move or rename operations take place, the last object name that
was known before the catch-up phase started is what will be given. The actual
new name may not be available until the catch-up phase is complete.

If you are using the DSPJRN or RTVJRNE command, additional informational

messages will indicate that this situation occurred. If you are using the RCVJRNE
command, additional information is provided on the exit program interface to help
distinguish these situations as well. Refer to “Using the Receive Journal Entry
Command” on page 432 for more information. If you are using the
QjoRetrieveJournalEntries API, additional information is provided in the returned
| data to help distinguish these situations. For more information, look in the
| Information Center at the following Web site:
| http://www.ibm.com/eserver/iseries/infocenter. When necessary, the system
attempts to minimize the possibility of showing these inconsistencies by
temporarily delaying the processing performed by these commands.

Once the catch-up phase is completed, these inconsistencies will be resolved, and
complete information will again be available.

Retrieving Journal Entries with Commitment Control

Considerations
Special performance related processing is done by the system when depositing
entries that are associated with commitment control transactions to a local journal.
When a job deposits a journal entry that is not associated with a commitment
control transaction, that job waits for the local journal I/O to auxiliary storage to
complete. After completion, control is given back to the application. A different
technique is used for those journal entries that are associated with a commitment
control transaction which results in the application being given control back before
the local journal I/O is complete. This special processing has some ramifications
when retrieving journal entries from a remote journal.

For journal entries deposited related to a commitment control transaction, a job

only waits for the local journal I/O to complete when the following journal entries
are being deposited into the local journal:
v Journal code ’C’, journal entry type ’CM’ (Commit)
v Journal code ’C’, journal entry type ’RB’ (Rollback)
For remote journals, those journal entries that the job that is making the deposit
does not wait for are not immediately replicated or scheduled to be replicated to
the remote journal. Prior to the CM (Commit) or RB (Rollback) entry being
deposited, there is no guarantee as to when the journal entries for open
commitment control transactions will be retrievable from the remote journal.

After the commit or rollback operation is complete for a particular commitment

control transaction, all journal entries associated with that transaction are
immediately retrievable from an asynchronously maintained remote journal.
However, there may be some journal entry delivery latency due to the transport
method that is being used.

Chapter 21. Remote Journal Function 507

 For a synchronously maintained remote journal, all journal entries associated with
the commitment control transaction are assured to be retrievable after the CM
(Commit) or RB (Rollback) entry is deposited.

Interspersed local journal I/O, for journal entries not associated with a
commitment control transaction, can also affect when the journal entries associated
with a commitment control transaction can be retrieved from the remote journal. In
this I/O a job actually waits for the local journal I/O to complete. This
interspersed local journal I/O will also cause the journal entries related to the
commitment control transaction to be replicated to the remote journal. Once in the
remote journal, and when later remote journal I/O makes them confirmed, the
journal entries that are related to the commitment control transaction are
retrievable.

Note: These considerations would also apply to any user generated entries that
use the Send Journal Entry (SNDJRNE) command or API (QJOSJRNE). If the
application or user never requests to force these user generated entries, they
will only be replicated to the remote journal when some other action forces
the journal entries. Therefore, you will wish to periodically specify
FORCE(*YES) when using these send journal entry functions.

| These considerations also apply to any database physical file open or close journal
| entries; or directory or stream file open, close, or force entries.

Confirmed and Unconfirmed Journal Entries

For a local journal, all entries are confirmed entries. There is no concept of
unconfirmed entries.

For a remote journal that is maintained asynchronously, all entries are confirmed
entries. For a remote journal that is maintained synchronously, there are both
confirmed and unconfirmed entries. Unconfirmed entries will only become
important if you are using the remote journal support for a hot-backup or data
replication environment, and the source system has a failure such that the target
system will take over processing.

Confirmed journal entries are journal entries replicated to a target system, and the
state of the I/O to auxiliary storage for the same journal entries on the primary
system is known to have completed.

Unconfirmed journal entries are entries replicated to a target system, but the state
of the I/O to auxiliary storage for the same journal entries on the primary system
is not known. Unconfirmed entries only pertain to remote journals that are
maintained synchronously. The remote I/O to the remote journal is overlapped
with the local I/O to the local journal for better performance. Such journal entries
on the target system are held in the data portion of the journal receiver. However,
the journal entries are not officially included with the remainder of the journal
entries until the confirmation of the I/O for the same entries is received from the
primary system. For performance reasons, confirmation of these entries is not
typically sent to the target system until some later delivery of journal data to the
target system.

While the journal entries are unconfirmed on a target system, the entries typically
cannot be retrieved from the remote journal. You can retrieve the journal entries by
using the INCENT(*ALL) parameter on the DSPJRN, RTVJRNE, or RCVJRNE
commands. You can also retrieve the journal entries by specifying *ALL for the

508 OS/400 Backup and Recovery V5R1

 include entries key for the QjoRetrieveJournalEntries API. The INCENT(*ALL)
parameter, or include entries key specification of *ALL, requests that all confirmed
and unconfirmed entries are included. This means that for synchronous remote
journal function, the last few journal entries are not immediately retrievable from
the remote journal by using the default command invocations. This is true even
though all journal entries physically reside in both the local journal and the remote
journal. This is done so that application programs do not make decisions on the
target system by using journal entries that may not end up being deposited into
| the local journal. This is because those journal entries would not cause a change to
| the original data.

With respect to a hot-backup application apply, in most circumstances only the

confirmed journal entries in the remote journal are of interest. In the data
replication environment, a hot-backup application apply would probably never
want to apply any unconfirmed journal changes. This is because any subsequent
activation of the remote journal will ensure that the journal entries in the remote
journal will match the journal entries in the source journal. However, as described
in “Example: Remote Journal Function Recovery” on page 511, knowledge of the
unconfirmed journal entries is essential during the switch-over and switch-back
processing for a hot-backup environment.

When a remote journal is inactivated, all unconfirmed entries are removed from
the remote journal. It is important that those entries are retrieved prior to the
remote journal being inactivated, if those entries are desired for additional
processing on the backup system. The message that is sent to the journal message
queue when the remote journal is inactivated by the system will indicate if the
remote journal has any unconfirmed journal entries. See “Troubleshooting Errors”
on page 491.

IPL Considerations
When a system that has a local journal ends either normally or abnormally, the
subsequent IPL processing ensures that the journal state for the journal is the same
as it was when the machine ended.

Regardless of whether the journal state for a local journal is *ACTIVE or

*INACTIVE, the system will always deposit an IPL entry into a local journal. The
journal entry type will be ’IA’ or ’IN’ with a journal code of ’J’.

| Likewise, if the system ends with the data being actively changed, the system will
| also deposit the object in-use entry into any local journal regardless of the journal
| state. See Table 123 on page 824 for a list of possible object in-use journal entries.
| The flag indicator in the in-use journal entry also indicates if the system was able
| to successfully reconcile the local journal with the object during the IPL processing.

An Entry Not Journaled exception (CPF7003), with reason code 10, will be signaled
to any program performing an operation that attempts to generate a journal entry
into an inactivated local journal during IPL processing. Journal entry deposits will
not be possible until the local journal is activated after the IPL by the Change
Journal (CHGJRN) command or Change Journal State (QjoChangeJournalState)
API. This restriction does not cause a problem for commitment control recovery
processing that normally occurs during an IPL. A local journal cannot be
inactivated if there are any open commitment control transactions that are
associated with the journal. Therefore, no commitment control recovery processing
will ever be necessary during an IPL for an inactivated local journal.

Chapter 21. Remote Journal Function 509

 Note: It might seem reasonable for all local journals that have associated remote
journals to be put in the *INACTIVE journal state during IPL processing if
the system last ended abnormally. However, if a hot-backup application
apply is performing any necessary recovery, the local journals must always
be *ACTIVE to allow for journal entries to be deposited into the local
journals. Therefore, the journal state is preserved across the IPL.

The replication of journal entries to each of the associated remote journals is

implicitly ended when the local system ends. The remote journal on the target
system must be inactivated before restarting the replicating of journal entries to a
particular remote journal. Use the Change Journal State (QjoChangeJournalState)
API or Change Journal (CHGJRN) command to inactivate the remote journals on
the target system. The replicating of journal entries to any remote journal is not
started again until each desired remote journal is activated. Reissuing the Change
Journal State (QjoChangeJournalState) API or Change Remote Journal
(CHGRMTJRN) command from the source system activates each desired remote
journal. After an IPL, you are not required to reassociate the desired remote
journals with the journal on the source system. (You reassociate the remote journals
by using the Add Remote Journal (QjoAddRemoteJournal) API or ADDRMTJRN
command.)

Main Storage Considerations during IPL

In addition to unconfirmed I/O for journal entries, the preservation of main
storage for a failed system also needs to be considered during recovery
0processing. Given certain system failures, main storage may or may not be
preserved during the ensuing IPL to recover from the system failure.

Therefore, it is possible for journal entries to survive in a local journal after a

system failure even if the local, or remote, I/O was never performed for those
journal entries.

| This effectively means that IPL recovery on a primary system may preserve
| changes that were not replicated to any of the remote journals, even the
| synchronously maintained remote journals. “Example: Remote Journal Function
Recovery” on page 511 demonstrates that journal entries that survive a system
failure in this manner can be accounted for using the remote journal function.
| These journal entries will not cause a total re-priming of the original data when
| switching-back from a backup system which took over the role of the primary
| system.

Application programs that were in the process of generating such surviving journal
entries never had control returned to them. From the application’s perspective, it
truly is not known if the operation completed or not. No dependencies or
decisions could have been made on such operations. This includes dependencies or
decisions by the application performing the operation, or any other application that
could be possibly dependent upon the data affected by the operation.

This means that the strategy that is discussed in “Example: Remote Journal
Function Recovery” on page 511, of backing out such operations is acceptable.

| Because of these main storage preservation considerations, it is recommended that

| both the before and after images be journaled for any objects, if possible. With the
| before images, the work can then be backed out after the IPL. If the data activity is

510 OS/400 Backup and Recovery V5R1

| not backed out after the IPL, the alternative would be to re-prime the primary
| system data completely from the backup data which had assumed the role of the
| primary.

Note: Commitment control, especially two-phase commitment control, can cause

some additional considerations and potential complications. For example, if
any of the entries that were preserved but not yet confirmed were a commit
or a rollback operation, then the transaction will have to be reconciled
accordingly between the primary system, and the backup system.

Journal Receiver ASP Considerations

The receiver configuration is the ASP the receiver resides in, and how the data for
the receiver is spread across the disk arms within that ASP. A remote journal
receiver will have the same receiver configuration as its corresponding source
receiver. If the source receiver is in an ASP that is spread across multiple disk
units, then the remote journal receiver will also be configured to use the same
number of disk units. The remote journal receiver may be in an ASP that has fewer
disk units than the ASP that contains the journal receiver on the source system. If
this occurs, the remote journal receiver will still be configured as if it still had that
same number of disk units as the source journal receiver. However, the data may
physically be going to a fewer number of disk units.

Note: If the remote journal receiver is in an ASP with fewer disk arms then the
source journal receiver, then performance may be impacted. This is because
the disk arms for the remote receiver will be moving considerably more
than the disk arms will be moving for the source receiver. Therefore, we
recommend that the number of disk arms is the same on the source and
remote journal receivers ASPs.

Likewise, the journal receiver on the source system may be in an ASP that has
fewer disk units than the ASP that contains the remote journal receiver. If this
occurs, the remote journal receiver will not take advantage of all possible disk
units on the target system.

Example: Remote Journal Function Recovery

This section gives an overview of a simple hot-backup recovery scenario that uses
the remote journal function support.

| Note: The following examples only discuss database physical files. All the
| concepts, however, apply to any journaled object type.

Chapter 21. Remote Journal Function 511

 Figure 44. Example Remote Journal Environment for Hot-Backup Recovery

The following considerations apply to Figure 44:

v Journal receivers are not specifically called out in the diagram. They have been
omitted in an attempt to simplify the scenario and to focus on the recovery steps
for the database. Where necessary, processing specific to journal receivers is
referred to in the scenario.
v Likewise, library redirection for the journals and journal receivers is not
specifically called out in the diagram. Again, this is omitted in an attempt to
simplify the scenario. In the scenario, the libraries for any of the journals or
journal receivers could be redirected to a library that is different from that being
used for the corresponding objects on the other system.
| v The diagram simply refers to the original data in the diagram as DB on the
| primary system S1 and DB’ as the data replica on the backup system S2. DB
| could be one or more journaled objects, and DB’ contains a replica for each of
| the journaled objects in DB.

For simplicity, the scenario below treats DB as a single database file and DB’ as its
replica.

In normal run mode, the following is assumed:

v System S1 is the primary system.
| v The original data that is denoted by DB is journaled to an active local journal
| PJ1.
v Remote journal BJ1 on backup system S2 is active, and unless otherwise noted,
is synchronously receiving journal entries from journal PJ1.
| v A hot-backup application apply, not shown in the diagram, is asynchronously
| replaying, or applying, the changes to the data replica, DB’.
| v The data replica DB’ is journaled to local journal PJ2 on system S2.
| The journal state for journal PJ2 is *ACTIVE so that the hot-backup application
| apply can replay the necessary changes to the data replica.

512 OS/400 Backup and Recovery V5R1

 v Remote journal BJ2 has a journal state of *INACTIVE (journal entries are not
| replicated to it). Remote journal BJ2 is only active when accepting the data
| changes back from system S2. This occurs after system S2 had been promoted to
assume the role of the primary system due to a planned or unplanned outage of
system S1, and after system S1 has resumed operations.
v The primary system, S1, has failed.
v The decision has been made to switch-over to the backup system, S2.

Note: The following hot-backup recovery strategy does not require that both the
before- and after-journal images are journaled to the local journal. However,
if any hot-backup recovery strategy has the hot-backup application apply
resynchronization processing back out in-flight changes during the
switch-back to the primary system, then they may be required. This will
become more clear after working through the following recovery scenario.

Example Scenario Description

More unconfirmed journal entries are present in BJ1 than are known to journal PJ1
during the IPL of system S1. (Main storage was not preserved as part of the IPL
recovery for system S1.)

Figure 45 on page 514 shows the state of the journals and data at the time of the
failure:

Chapter 21. Remote Journal Function 513

 Figure 45. More unconfirmed journal entries present in BJ1 than known in PJ1.

| At the time of the system failure, journal entries 12-19 had already been deposited
| into PJ1 and confirmed in BJ1. The corresponding data changes had also already
| been reflected in the data replica, DB’, on system S2. Journal entries 20-25 had been
built and validated in main storage on S1 and sent to BJ1, and then system S1
failed. Main storage was not preserved when S1 failed. So at the time of the
failure, the last known confirmed sequence number in BJ1 was 19; unconfirmed
was 25. The last known sequence number in PJ1 will be 19 during the ensuing IPL
on system S1.
1. On system S2, the first step to be performed as part of the switch-over
processing is to allow the hot-backup application apply processing to
complete the replay of confirmed operations as identified in journal BJ1. This
would include replaying all journal entries up through and including
sequence number 19.

Note: Sequence numbers 20-25 would not yet be replayed because the I/O for
those journal entries is not yet confirmed from the local journal PJ1. The
Receive Journal Entry (RCVJRNE) command or
QjoRetrieveJournalEntries API that is being used to retrieve the entries
from the remote journal will not return sequence numbers 20-25 to the
exit program, unless specifically requested to do so by using the

514 OS/400 Backup and Recovery V5R1

 INCENT(*ALL) parameter on the command. You can also request this
by specifying *ALL for the include entries key on the API.
2. After all confirmed journal entries have been replayed, then a change journal
operation is performed to attach a new journal receiver to local journal PJ2 on
system S2 to establish a clean recovery point.

| Note: The change journal operation is not absolutely essential, but attaching a
| new journal receiver to PJ2 at this time makes clear what information
| will need to be sent back to system S1 later to replay back to the
| original data. Performing the change journal operation also prevents the
remote journal function from having to re-replicate all of the journal
entries that were previously generated into the currently attached
| journal receiver of PJ2. (The journal entries were generated into the
| receiver as part of replaying the database changes to the data replica on
| system S2.)

Chapter 21. Remote Journal Function 515

 Figure 46. Switch-over processing. System S2 is now ready to allow applications to run.

| 3. The unconfirmed journal entries are then read from BJ1 and replayed to the
| data replica. They are retrieved from BJ1 by using the Receive Journal Entry
(RCVJRNE) command or QjoRetrieveJournalEntries API, specifically
| requesting that unconfirmed journal entries be returned. Journal entries
| 140-145 are generated into journal PJ2 when replaying these changes to the
| data replica.

516 OS/400 Backup and Recovery V5R1

 4. The QjoChangeJournalState API or CHGJRN command has inactivated the
remote journal BJ1. During this operation, the system physically removes the
unconfirmed journal entries from BJ1. The last known sequence number in BJ1
is now 19.
5. The replay processing on S2 sends a user entry that indicates the point in time
when the database was switched-over. The user entry is sequence number 146
in Figure 46 on page 516, Journal code ’U’, entry type ’SW’.
Figure 46 on page 516 shows the state of BJ1, PJ2, and DB’ when system S2 is
ready to assume the role of the primary system.
6. After these steps are performed on system S2, applications can now be started
on S2 and use DB’ as the database to be updated. Applications continue to
work and deposit journal entries 147-200.
| 7. System S1 is IPLed and normal IPL recovery finds the end of the journal for
| PJ1 to be sequence number 19. IPL recovery ensures that all changes up to
| sequence number 19 are reflected in the original data. The IPL for S1
completes with journal PJ1 being left in the *ACTIVE state, as this was the
state of the journal when the system failed.
Figure 47 shows the state of both systems after system S1 has completed its
IPL. This is after system S2 has been running as the primary system, but
before database DB is resynchronized with DB’. (The database changes
represented in PJ2 by journal sequence numbers 147-200 are not shown in DB’
for simplicity.)

Chapter 21. Remote Journal Function 517

Figure 47. System S2 has assumed the role of the primary, and DB’ is now being updated. IPL processing has
completed on system S1.

8. The first step in resynchronizing DB with DB’ after the IPL completes is to
activate the remote journal BJ2 by using the QjoChangeJournalState API or
Change Remote Journal (CHGRMTJRN) command. Specify that the process
should start with the attached journal receiver on the source system (S2). This
starts the transport of journal entries representing the changes made on S2 as
part of replaying the unconfirmed journal entries plus all changes made to
DB’ while S1 was unavailable. While this transfer is progressing (during

518 OS/400 Backup and Recovery V5R1

 catch-up processing, which then transitions into synchronous or asynchronous
remote journal function mode), changes are still being made by applications to
DB’.
9. Either before or during the transport of journal entries to BJ2, the last known
sequence number in BJ1 (19) must be sent over and made known to the
hot-backup application apply.

Note: This could be included as information in the SW user journal entry. See
step 5 on page 517.
10. Changes that are known to PJ1 (after the last known sequence number in BJ1)
| are then backed-out of the original data DB on system S1. For this particular
| scenario, no changes need to be backed out of the original data.

Note: For those scenarios which do require this back-out processing, both the
before and after image journal entries would be required.

| Figure 48 on page 520 shows the state of both systems before changes are
| replayed back to the original data on system S1. (Again, the data changes
| represented in PJ2 by journal sequence numbers 147-300 are not shown in DB’
| for simplicity.)

Chapter 21. Remote Journal Function 519

 Figure 48. State of the journals and databases just before starting to replay the changes back to the original data, DB.

| 11. The process of replaying the changes back to the original data on system S1
| begins. The changes that are replayed include those changes that were made
to DB’ as part of the switch-over processing. (The switch-over processing

520 OS/400 Backup and Recovery V5R1

 replayed the data changes for the unconfirmed journal entries (sequence
numbers 140-145)). Additional changes include those data changes that were
deposited while system S2 had assumed the role of the primary system
(sequence numbers 147-300). Note that changes are still being made to DB’ on
system S2 and journal entries are still being generated into local journal PJ2 on
system S2.
12. When the user decides that S1 should again assume the role of the primary
system, the applications on S2 are ended. The local journal PJ2 on system S2 is
inactivated by using the Change Journal (CHGJRN) command or
QjoChangeJournalState API. This prevents any additional changes from being
made to DB’. Figure 49 on page 522 shows the state of both systems just before
the local journal PJ2 is inactivated making way for system S1 to again assume
the role of the primary system.
13. The remaining changes are then allowed to be replicated to BJ2. After all
changes have been sent to BJ2, you can inactivate BJ2 by using the
QjoChangeJournalState API or the CHGRMTJRN command.
| 14. After all of the journal entries have been replayed to the original data on S1, a
| new journal receiver is attached to PJ1 to clearly denote a new recovery point.

| Note: The change journal operation is not absolutely essential. However,

| attaching a new journal receiver to PJ1 at this time makes clear where
| to start replaying changes back to the data replica on system S2.
Performing the change journal operation also prevents the remote
journal function from having to send back all of the journal entries that
were previously generated into the currently attached journal receiver
| of PJ1. (The journal entries were generated in the receiver as part of
| replaying the data changes back to the original data on system S1.)

Chapter 21. Remote Journal Function 521

Figure 49. Preparing to let S1 again assume the role of the primary system.

522 OS/400 Backup and Recovery V5R1

 Figure 50. Preparing to let S1 again assume the role of the primary system (continued).

| 15. Application programs can now make changes to the original data DB on
| system S1.
16. When you determine that it is time to start replicating the changes made on
the primary system to the backup system, you can activate the remote journal
BJ1. (You activate the remote journal B1 by using the QjoChangeJournalState
API or CHGRMTJRN command.)
When activating the remote journal, you can indicate to send journal entries
| starting with the attached journal receiver on the source system. If this occurs,
| then only those journal entries that are required to be replayed to the data
| replica will be sent to system S2.

Note: You can start with the attached receiver, only if you did the change
journal to attach a new receiver that was mentioned in step 14 on
page 521.

| The hot-backup application apply processing on system S2 will need to start

| replaying data changes. Data changes are replayed starting with the first entry
| in the journal receiver that is being sent over.

Note: If the complete chain of journal receivers is desired from system S1, the
remote journal could be activated indicating to start with the attached
journal receiver as known to the remote journal, BJ1. This will complete
the sending of the journal receiver that contains the IPL entry (sequence
| number 20). The process will then move on to the next journal receiver
| that contains the journal entries where the hot-backup application apply
| will start replaying changes to the data replica. An alternative to that
approach would be to save and restore the detached journal receiver to
system S2.
17. You can now activate local journal PJ2 on system S2 by using the
QjoChangeJournalState API or CHGJRN command.

Chapter 21. Remote Journal Function 523

 18. After local journal PJ2 has been activated, a change journal operation for PJ2
to attach a new journal receiver needs to be performed.

| Note: The change journal operation is not absolutely essential. However,

| attaching a new journal receiver to PJ2 at this time makes clear where
| the replaying of changes back to the data replica started on system S2.
Performing the change journal operation also avoids the remote journal
function from having to later send all of these hot-backup application
apply generated journal entries back to system S1.

The newly attached journal receiver will contain journal entries that will not
have to be sent back to system S1.
| 19. After the operation is performed, the hot-backup application apply can be
| started on system S2 to start replaying changes to the data replica. The
hot-backup application apply starts with the source system sending the newly
attached journal receiver.

524 OS/400 Backup and Recovery V5R1

 Chapter 22. Commitment Control
Prerequisite: To understand the information in the following topic, it is assumed
that you have a basic understanding of the journal management function. For more
information about journal management, see Chapter 19 and Chapter 20.

Commitment Control Introduction

Commitment control is a function that allows you to define and process a group
of changes to resources, such as database files or tables, as a logical unit of work.

Throughout this topic, the term logical unit of work (LUW) is used. A logical unit
of work (LUW) is defined as a group of individual changes to objects on the
system that should appear as a single atomic change to the user. End users and
| application programmers might think of an LUW as a transaction.If you are
| familiar with the term Unit of Work (UOW), an LUW has the same meaning.

Commitment control ensures that either the entire group of individual changes
occur on all systems that participate or that none of the changes occur. The transfer
of funds from a savings to a checking account is an example of changes that can be
grouped together. To the user, this is a single transaction. However, more than one
change occurs to the database because both savings and checking accounts are
updated.

You can use commitment control to design an application so that it can be started
again if a job, an activation group within a job, or the system ends abnormally. The
application can be started again with assurance that no partial updates are in the
database due to incomplete logical units of work from a prior failure.

Commitment control is started and ended using the Start Commitment Control
(STRCMTCTL) and End Commitment Control (ENDCMTCTL) commands.

Commitment control allows you to:

v Ensure that all changes within an LUW are completed for all resources that are
affected.
v Ensure that all changes within an LUW are removed if processing is interrupted.
v Remove changes that are made during an LUW when the workstation user
application determines that an LUW is in error.

LUWs can be any of the following:

v Inquiries in which no database file changes occur.
v Simple transactions in which one database file is changed each time the Enter
Key is pressed.
v Complex transactions in which one or more database files are changed each time
the Enter key is pressed.
v Complex transactions in which each time the Enter key is pressed one or more
database files are changed, but these changes represent only a part of a logical
group of transactions.
v Simple or complex transactions that involve database files at more than one
location. The database files can be:

© Copyright IBM Corp. 1997, 2000, 2001 525

 – On a single remote system
– On the local system and one or more remote systems
– Assigned to more than one journal on the local system. Each journal can be
thought of as a local location.
v Simple or complex transactions on the local system that involve objects other
than database files.

Terms Used with Commitment Control

Figure 51 on page 527 shows a simple overview of the commitment control process.

This topic briefly defines the commitment control terms that are shown in
Figure 51 on page 527. Later in the chapter, each term is explained in more detail.

A commitment definition contains information about the resources that are under
commitment control for a job or an activation group.

A commitment boundary is established by starting commitment control or by

performing a commit operation. For a particular commitment definition, a
commitment boundary is defined to be any time there are no outstanding
uncommitted changes for the objects that are being changed under commitment
control. When a committable change is made to a committable resource, the
current LUW is no longer at a commitment boundary.

A commit operation establishes a new commitment boundary. A rollback operation

brings the objects that are being changed for the current LUW back to the previous
commitment boundary.

A resource manager is a program or function that is responsible for making

changes to committable resources. For local database files, it is the database
management function. For remote resources, the communications program is the
resource manager.

A transaction manager is a program or function that is responsible for managing

all the resource managers on a system. It communicates with transaction managers
on other systems. On the system, the transaction manager function is part of the
operating system, unless you are using XA transaction support. Refer to “XA
Transaction Support” on page 566 for more information.

A committable resource is a local or remote object that can be placed under

commitment control.

A committable change is a change to a committable resource that can be

committed or rolled back. A committable change is referred to as pending or
uncommitted until the change is committed or rolled back.

The initiator is the system where the program is running that requests a commit or
rollback operation for an LUW.

An agent is any system where a resource manager, other than the initiator, is
participating in an LUW. “Roles in Commit Processing” on page 548 provides more
information about initiators and agents.

526 OS/400 Backup and Recovery V5R1

 ┌──────────────────────────┐
│ 1.STRCMTCTL command │ The system creates a
│ ├─────S commitment definition.
│ │ A commitment boundary
│ │ is established.
│ │
│ │
├──────────────────────────┤
│ 2.Call PGMX │
│ │
│ │
│ │
│ │
├──────────────────────────┤
│ 3.Open files F1 and F2 │ The files become
│ ├─────S committable resources.
│ │ The operating system
│ │ function responsible for
│ │ changing the files becomes
│ │ the resource manager.
├──────────────────────────┤
│ 4.Update files F1 and F2 │ These changes are called
│ ├─────S committable changes.
│ │ They are pending.
├──────────────────────────┤
│ 5.COMMIT ├─────S The pending committable
│ │ changes associated with
│ │ the commitment definition
│ │ are committed. A new
│ │ commitment boundary is
│ │ established.
├──────────────────────────┤
│ 6.End PGMX │
├──────────────────────────┤
│ 7.ENDCMTCTL ├─────S The commitment definition
│ │ is ended, if it was
│ │ started by this program.
└──────────────────────────┘

Figure 51. Overview of Commitment Control Process

Commitment Control Startup

When you use the Start Commitment Control (STRCMTCTL) command you can
specify these things:
v The default lock level.
v The notify object.
v The scope for the commitment definition.
v Text for the commitment definition.
v A default journal.
v Whether to omit certain journal entries.

These parameters are explained in the topics that follow.

Lock-Level Parameter
The value you specify for the LCKLVL parameter on the Start Commitment
Control (STRCMTCTL) command becomes the default level of record locking for

Chapter 22. Commitment Control 527

 database files that are opened and placed under commitment control for the
commitment definition. The default level of record locking cannot be overridden
when opening local database files. However, database files accessed by SQL can be
opened with a different level of record locking than what is specified as the default
lock level on the STRCMTCTL command.

The lock level should be specified according to your needs, the wait periods
allowed, and the release procedures used most often. The following descriptions
apply only to files that are opened under commitment control.

*CHG Lock Level: Use this value if you want to protect changed records from
changes by other jobs running at the same time. For files that are opened under
commitment control, the lock is held for the duration of the LUW. For files not
opened under commitment control, the lock on the record is held only from the
time the record is read until the update operation is complete.

*CS Lock Level: Use this value to protect both changed and retrieved records from
changes by other jobs running at the same time. Retrieved records that are not
changed are protected only until they are released, or a different record is
retrieved.

The *CS lock level ensures that other jobs are not able to read a record for update
that this job has read. In addition, the program cannot read records for update that
have been locked with a record lock type of *UPDATE in another job until that job
accesses a different record.

*ALL Lock Level: Use this value to protect changed records and retrieved records
that are under commitment control from changes by other jobs running under
commitment control at the same time. Records that are retrieved or changed are
protected until the next commit or rollback operation.

The *ALL lock level ensures that other jobs are not able to access a record for
update that this job has read. This is different from normal locking protocol. When
the lock level is specified as *ALL, even a record that is not read for update cannot
be accessed if it is locked with a record lock type of *UPDATE in another job.

Table 69 shows the duration of record locks for files under and not under
commitment control.
Table 69. Lock Duration by Lock-Level Parameter
Request LCKLVL Parameter Duration of Lock Lock Type
Read only No commitment control No lock None
*CHG No lock None
*CS From read to next read, *READ
commit, or rollback
*ALL From read to commit or *READ
rollback

528 OS/400 Backup and Recovery V5R1

Table 69. Lock Duration by Lock-Level Parameter (continued)
Request LCKLVL Parameter Duration of Lock Lock Type
Read for update then update No commitment control From read to update or *UPDATE
or delete1 delete
*CHG From read to update or *UPDATE
delete
Then from update or delete *UPDATE
to next commit or rollback2
*CS From read to update or *UPDATE
delete
Then from update or delete *UPDATE
to next commit or rollback2
*ALL From read to update or *UPDATE
delete
Then from update or delete
to next commit or rollback2
Read for update then release1 No commitment control From read to release *UPDATE
*CHG From read to release *UPDATE
*CS From read to release, *UPDATE
commit, or rollback
Then from release to next *UPDATE
read, commit, or rollback
From read to release, *UPDATE Then from release to next
commit, or rollback commit or rollback
*UPDATE
Add No commitment control No lock None
*CHG From add to commit or *UPDATE
rollback
*CS From add to commit or *UPDATE
rollback
*ALL From add to commit or *UPDATE
rollback
Write direct No commitment control For duration of write direct *UPDATE
*CHG From write direct to *UPDATE
commit or rollback
*CS From write direct to *UPDATE
commit or rollback
*ALL From write direct to *UPDATE
commit or rollback

1
If a commit or rollback operation is performed after a read-for-update operation, but before the record is
updated, deleted, or released, the record is unlocked during the commit or rollback operation. The
protection on the record is lost as soon as the commit or rollback completes.

Chapter 22. Commitment Control 529

Table 69. Lock Duration by Lock-Level Parameter (continued)
Request LCKLVL Parameter Duration of Lock Lock Type

2
If a record is deleted but the commit or rollback has not yet been issued for the transaction, the deleted
record does not remain locked. If the same or a different job attempts to read the deleted record by key, the
job receives a record not found indication. However, if a unique keyed access path exists over the file,
another job is prevented from inserting or updating a record with the same unique key value as that of the
deleted record until the transaction is committed.

A record lock type of *READ is obtained on records that are not read for update
when the lock level is *CS or *ALL. This type of lock prevents other jobs from
reading the records for update but does not prevent the records from being
accessed from a read-only operation.

A record lock type of *UPDATE is obtained on records that are updated, deleted,
added, or read for update. This type of lock prevents other jobs from reading the
records for update, and prevents jobs running under commitment control with a
record lock level of *CS or *ALL from accessing the records for even a read-only
operation.

Programs that are not using commitment control can read records locked by
another job, but cannot read records for update, regardless of the value specified
for the LCKLVL parameter.

The lock level specified for a commitment definition when commitment control is
started for an activation group or for the job applies only to opens associated with
that particular commitment definition.

Note: The *CS and *ALL lock-level values protect you from retrieving a record that
currently has a pending change from a different job. However, the *CS and
*ALL lock-level values do not protect you from retrieving a record using a
program running in one activation group that currently has a pending
change from a program running in a different activation group within the
same job.

Within the same job, a program can change a record that has already been changed
within the current LUW as long as the record is accessed again using the same
commitment definition. When using the job-level commitment definition, the access
to the changed record can be made from a program running within any activation
group that is using the job-level commitment definition.

Notify Object Parameter

A notify object is a message queue, data area, or database file that contains
information identifying the last successful LUW completed for a particular
commitment definition if that commitment definition did not end normally. The
information used to identify the last successful LUW for a commitment definition
is given by the commit identification that associates a commit operation with a
specific set of committable resource changes.

The commit identification of the last successful LUW for a commitment definition
is placed in the notify object only if the commitment definition does not end
normally. This information can be used to help determine where processing for an
application ended so that the application can be started again.

530 OS/400 Backup and Recovery V5R1

If journaled resources participate in the current LUW and a commit operation is
performed with a commit identification, the commit identification is placed in the
commit journal entry (journal code and entry type of ’C CM’) that identifies that
particular LUW as being committed. A commit journal entry containing the
commit identification is sent to each journal associated with resources that
participated in the LUW.

Table 70 shows how you specify the commit identification and its maximum size. If
the commit identification exceeds its maximum size, it is truncated when it is
written to the notify object.
Table 70. Specifying a Commit Identification
Maximum Characters in Commit
Language Operation Identification

CL COMMIT command 30001

ILE RPG* COMIT operation code 40001
PLI. PLICOMMIT subroutine 40001
ILE C* _Rcommit function 40001
ILE COBOL* COMMIT verb Not supported
SQL COMMIT statement Not supported
1
If the notify object is a data area, the maximum size is 2000 characters.

When a notify object is updated with the commit identification, it is updated as

follows:
v Database file
If a database file is used as the notify object, the commit identification is added
to the end of the file. Any existing records will be left in the file. Because several
users or jobs can be changing records at the same time, each commit
identification in the file should contain unique information to associate the data
with the job and commitment definition that failed. The file that serves can be
journaled.
v Data area
If a data area is used as the notify object, the entire content of the data area is
replaced when the commit identification is placed in the data area. If more than
one user or job is using the same program, only the commit identification from
the last commitment definition that did not end normally will be in the data
area. Consequently, a single data area notify object may not produce the correct
information for starting the application programs again. To solve this problem,
use a separate data area for each commitment definition for each workstation
user or job.
v Message queue
If a message queue is used as a notify object, message CPI8399 is sent to the
message queue. The commit identification is placed in the second-level text for
message CPI8399. As with using a database file for the notify object, the contents
of each commit identification should uniquely identify a particular commitment
definition for a job so that an application program can be started again.

Updates Made to the Notify Object

For purposes of the notify object, the following are considered uncommitted
changes:
v An update to a record that is made under commitment control.
v A record that is deleted under commitment control.

Chapter 22. Commitment Control 531

 v An object level change that is made to a local DDL object under commitment
control.
v A read operation performed for a database file that was opened under
commitment control. This is because file position is brought back to the last
commitment boundary when a rollback operation is performed. If you perform a
read operation under commitment control, the file position is changed and
therefore, an uncommitted change then exists for the commitment definition.
v A commitment definition with one of the following resources that are added is
always considered to have uncommitted changes:
– An API commitment resource
– A remote Distributed Relational Database Architecture (DRDA*) resource
– A Distributed Database Management Architecture (DDM) resource
– An LU 6.2 resource
This is because the system does not know when a real change is made to the
object or objects that are associated with these types of resources. “Types of
Committable Resources” on page 541 has more information about how you add
and work with these types of resources.

The system makes updates to the notify object and are based on the following
ways that a commitment definition can end:
v If a job ends normally and no uncommitted changes exist, the system does not
place the commit identification of the last successful commit operation in the
notify object.
v If an implicit commit operation is performed for an activation-group-level
commitment definition when the activation group is ended, the system does not
place the commit identification of the last successful commit operation in the
notify object.

Note: Implicit commit operations are never performed for the *DFTACTGRP or
*JOB commitment definitions.
v If the system, job, or an activation group ends abnormally before the first
successful commit operation for a commitment definition, the system does not
update the notify object because there is no last commit identification. To
differentiate between this condition and a normal program completion, your
program must update the notify object with a specific entry prior to completing
the first successful commit operation for the commitment definition.
v If an abnormal job end or an abnormal system end occurs after at least one
successful commit operation, the system places the commit identification of that
commit operation in the notify object. If the last successful commit operation did
not specify a commit identification, then the notify object is not updated. For an
abnormal job end, this notify object processing is performed for each
commitment definition that was active for the job. For an abnormal system end,
this notify object processing is performed for each commitment definition that
was active for all jobs on the system.
v The system updates the notify object with the commit identification of the last
successful commit operation for that commitment definition if all of the
following occur:
– A non-default activation group ends.
– An implicit rollback operation is performed for the activation-group-level
commitment definition.
– At least one successful commit operation has been performed for that
commitment definition.

532 OS/400 Backup and Recovery V5R1

 If the last successful commit operation did not specify a commit identification,
then the notify object is not updated. An implicit rollback operation is performed
for an activation-group-level commitment definition if the activation group is
ending abnormally or errors occurred when closing the files opened under
commitment control that were scoped to that activation group. For more
information about scoping database files to activation groups and how activation
groups can be ended, see the reference book for the ILE language that you are
using.
v If uncommitted changes exist when a job ends normally and at least one
successful commit operation has been performed, the commit identification of
the last successful commit operation is placed in the notify object and the
uncommitted changes are rolled back. If the last successful commit operation did
not specify a commit identification, then the notify object is not updated. This
notify object processing is performed for each commitment definition that was
active for the job when the job ended. For more information on the functions
that are done during the normal ending of a job, see the topic “Commitment
Control during Normal Routing Step End” on page 581.
v If uncommitted changes exist when the ENDCMTCTL command is run, the
notify object is updated only if the last successful commit operation specified a
commit identification:
– For a batch job, the uncommitted changes are rolled back and the commit
identification of the last successful commit operation is placed in the notify
object.
– For an interactive job, if the response to inquiry message CPA8350 is to
rollback the changes, the uncommitted changes are rolled back and the
commit identification of the last successful commit operation is placed in the
notify object.
– For an interactive job, if the response to inquiry message CPA8350 is to
commit the changes, the system prompts for a commit identification to use
and the changes are committed. The commit identification that is entered on
the prompt display is placed in the notify object.
– For an interactive job, if the response to inquiry message CPA8350 is to cancel
the ENDCMTCTL request, the pending changes remain and the notify object
is not updated.

Commit Scope Parameter

When commitment control is started, the system creates a commitment definition.
The commit scope parameter identifies the scope for the commitment definition.
The default is to scope the commitment definition to the activation group of the
program making the start commitment control request. The alternative scope is to
the job.

Commitment Definition
Each commitment definition is known only to the job that started commitment
control and is not known by any other job. The commitment definition contains
information that pertains to the resources that are being changed under
commitment control within that job. The system maintains the commitment control
information in the commitment definition as the commitment resources change,
until the commitment definition is ended.

A commitment definition generally includes:

v The parameters on the Start Commitment Control (STRCMTCTL) command.
v The current status of the commitment definition.

Chapter 22. Commitment Control 533

 v Information about database files and other committable resources that contain
changes that are made during the current LUW.

Programs can start and use multiple commitment definitions . Each commitment
definition for a job identifies a separate LUW that has committable resources
associated with it. These LUWs can be committed or rolled back independently
from LUWs that are associated with other commitment definitions that are started
for the job.

Scope for a Commitment Definition

The scope for a commitment definition is used to indicate which programs that
run within the job will use that commitment definition. The default scope for a
commitment definition is to the activation group of the program that starts
commitment control. Only programs that run within that activation group will use
that commitment definition. Commitment definitions that are scoped to an
activation group are referred to as activation-group-level commitment definitions.
Many activation-group-level commitment definitions can be active for a job at one
time. However, each activation-group-level commitment definition can be
associated only with a single activation group. The programs that run within that
activation group can associate their committable changes only with that
activation-group-level commitment definition.

The alternative scope for a commitment definition is to the job. A commitment

definition with this scope value is referred to as the job-level (*JOB) commitment
definition. Any program running in an activation group that does not have an
activation-group-level commitment definition started uses the job-level
commitment definition, if it has already been started by another program for the
job. Only a single job-level commitment definition can be started for a job.

In addition to the job-level commitment definition and commitment definitions

that are scoped to a particular activation group, a job can have other commitment
definitions active that are started by some system functions. These types of
commitment definitions do the following:
v Function independently from any other commitment definition active for a job.
v Do not have any particular scope.
v Are used only by the system function that started the commitment definition.

| An example of a function that starts these types of commitment definitions is the

| Problem Log.

For a given activation group, only a single commitment definition can be used by
the programs that run within that activation group. Therefore, programs that run
within an activation group can either use the job-level or the activation-group-level
commitment definition, but not both at the same time.

Even when the job-level commitment definition is active for the job, a program can
still start the activation-group-level commitment definition if no commitment
control requests or operations have been performed for the job-level commitment
definition by any program running within that activation group. Otherwise, the
job-level commitment definition would first have to be ended before the
activation-group-level commitment definition could be started. Commitment
control requests or operations for the job-level commitment definition that would
prevent the activation-group-level commitment definition from being started
include:

534 OS/400 Backup and Recovery V5R1

v An open, full or shared, of a database file under commitment control.
v Adding an API commitment resource by using the QTNADDCR API.
v A commit operation.
v A rollback operation.
v Adding a remote resource under commitment control.
v Changing commitment options by using the QTNCHGCO API.
v Bringing the commitment definition to a rollback required state by using the
QTNRBRQD API.
v Sending a user journal entry that includes the current commit cycle identifier by
using the QJOSJRNE API with the Include Commit Cycle Identifier parameter.

Likewise, if the programs within an activation group are currently using the
activation-group-level commitment definition, it must first be ended before the
job-level commitment definition could be used by the programs running within
that same activation group.

When opening a database file, the open scope for the opened file may be either to
the activation group or to the job with one restriction. If the file is being opened
under commitment control and is being scoped to the job, then the program
making the open request must be using the job-level commitment definition.

Thread Scoped Transactions: As described earlier, commitment definitions are

normally scoped to an activation group. If a job is multi-threaded, all threads in
the job have access to the commitment definition and changes made for a
particular transaction may be spread across multiple threads. That is to say, all
threads whose programs run in the same activation group participate in a single
logical unit of work.

There are cases where it is desirable for transactional work to be scoped to the
thread, rather than an activation group. In other words, each thread would have its
own commitment definition and transactional work for each commitment
definition would be independent of work performed in other threads.

This is supported by DB2 UDB for iSeries by using the Change Job (QWTCHGJB)
API to change the job to run in SQL server mode. When an SQL connection is
requested in SQL server mode, it is routed to a separate job. All subsequent SQL
operations that are performed for that connection are also routed to that job. When
the connection is made, completion message SQL7908 is sent to the SQL server
mode job’s job log indicating which job the SQL requests are being routed to. The
commitment definition is owned by the job that is indicated in this message. If
errors occur, it may be necessary to look at the job logs for both jobs to understand
the source of the problem since no real work is being done in the job performing
the SQL statements.

When running in SQL server mode, only SQL interfaces may be used to perform
work under commitment control. Embedded SQL or Call Level Interface (CLI) may
be used. All connections made through embedded SQL in a single thread are
routed to the same back-end job. This allows a single commit request to commit
the work for all the connections, just as it would in a job that is not running in
SQL server mode. Each connection made through the CLI is routed to a separate
job. The CLI requires work that is performed for each connection to be committed
or rolled back independently.

Chapter 22. Commitment Control 535

 You cannot perform the following operations under commitment control when
running in SQL server mode:
v Record changes that are made via non-SQL interfaces
v Changes to DDM files
v Changes to API commitment resources
You cannot start commitment control directly in a job running in SQL server mode.
For more information on SQL server mode, refer to the Information Center under
the Database and File Systems categories at the following Web site:
http://www.ibm.com/eserver/iseries/infocenter..

Commitment Definition Names

The system gives names to all commitment definitions that are started for a job.
Table 71 shows various commitment definitions and their associated names for a
particular job.
Table 71. Commitment Definition Names for a Job
Commitment Definition
Activation Group Commit Scope Name
Any Job *JOB
Default activation group Activation group *DFTACTGRP
User-named activation group Activation group Activation group name (for
example, PAYROLL)
System-named activation Activation group Activation group number (for
group example. 0000000145)
None None QDIR001 (example of a
system-defined commitment
definition for system use
only). System-defined
commitment definition names
begin with Q.

Note: Only Integrated Language Environment* (ILE*)-compiled programs can start

commitment control for activation groups other than the default activation
group. Therefore, a job can use multiple commitment definitions only if the
job is running one or more ILE-compiled programs. For more information
about the Integrated Language Environment, see the ILE Concepts book.

Original Program Model (OPM) programs that are run in the default activation
group, and by default use the *DFTACTGRP commitment definition. In a mixed
OPM and ILE environment, the job-level commitment definition should be used if
all committable changes made by all programs are to be committed or rolled back
together.

An opened database file scoped to an activation group can be associated with

either an activation-group-level or job-level commitment definition. An opened
database file scoped to the job can be associated only with the job-level
commitment definition. Therefore, any program, OPM or ILE, that opens a
database file under commitment control scoped to the job needs to use the
job-level commitment definition.

Application programs do not use the commitment definition name to identify a

particular commitment definition when making a commitment control request.
Commitment definition names are primarily used in messages to identify a
particular commitment definition for a job. Instead, the system determines which
commitment definition to use based on which activation group the requesting

536 OS/400 Backup and Recovery V5R1

program is running in. This is possible because the programs that run within an
activation group at any point in time can only use a single commitment definition.

Jobs with Multiple Commitment Definitions Example

Figure 52 on page 538 shows an example of a job that uses multiple commitment
definitions. It indicates which file updates are committed or rolled back at each
activation group level. The example assumes that all of the updates that are made
to the database files by all of the programs are made under commitment control.

Chapter 22. Commitment Control 537

┌──────────────────────────────────────────────────────────────────────┐
│ *JOB Commitment Definition │
└──────────────────────────────────────────────────────────────────────┘
O O
┌─────────────┐ │ ┌──────────────┐ │
│*DFTACTGRP │ │ │Commitment │ │
│Commitment │ │ │Definition Y │ │
│Definition │ │ └──────────────┘ │
└─────────────┘ │ O │
O │ │ │
┌─────┴───────┐ ┌───────┴──────┐ ┌──────┴───────┐ ┌──────┴──────┐
│ Default │ │ Activation │ │ Activation │ │ Activation │
│ Activation │ │ Group X │ │ Group Y │ │ Group Z │
│ Group │ │ │ │ │ │ │
├─────────────┤ ├──────────────┤ ├──────────────┤ ├─────────────┤
│Program MAIN │ ┌S│Program PGMX │ ┌S│Program PGMY │ ┌S│Program PGMZ │
├─────────────┤ │ ├──────────────┤ │ ├──────────────┤ │ ├─────────────┤
│STRCMTCTL │ │ │STRCMTCTL │ │ │STRCMTCTL │ │ │ │
│ LCKLVL(*ALL)│ │ │ LCKLVL(*CHG) │ │ │ LCKLVL(*CHG) │ │ │ │
│ │ │ │CMTSCOPE(*JOB)│ │ │ │ │ │ │
│ │ │ │ │ │ │ │ │ │ │
│Update files │ │ │ Update files │ │ │ Update files │ │ │ │
│ F1 and F2 │ │ │ F3 and F4 │ │ │ F5 and F6 │ │ │ │
│ │ │ │ │ │ │ │ │ │ │
│ Call PGMX ──┼─┘ │ Call PGMY ───┼──┘ │Error detected│ │ │ │
│ │ │ │ │ │ │ │ │
│ │ │ │ │ 1--ROLLBACK │ │ │ │
│ │ │ │ │ (Files F5 │ │ │ │
│ │ │ │ │ and F6) │ │ │ │
│ │ │ │ │ │ │ │ │
│ │ │ │ │ Update files │ │ │ │
│ │ │ │ │ F5 and F6 │ │ │ │
│ │ │ │ │ │ │ │ │
│ │ │ │ │ 2--COMMIT │ │ │ │
│ │ │ │ │ (Files F5 │ │ │ │
│ │ │ │ │ and F6) │ │ │ │
│ │ │ │ │ │ │ │ │
│ │ │ W───────────┼────┤ RETURN │ │ │ │
│ │ │ │ └──────────────┘ │ │ │
│ │ │ Call PGMZ ───┼─────────────────────┘ │ Update file │
│ │ │ │ │ F7 │
│ │ │ │ │ │
│ │ │ W───────────┼───────────────────────┤ RETURN │
│ │ │ │ └─────────────┘
│ │ │ 3--COMMIT │
│ │ │ (Files F3, │
│ │ │ F4 and F7) │
│ │ │ │
│ W───────────┼───┤ RETURN │
│ │ └──────────────┘
│ 4--COMMIT │
│ (Files F1 │
│ and F2) │
└─────────────┘

Figure 52. Using multiple commitment definitions in a job

Table 72 on page 539 shows how files would be committed or rolled back if the
scenario that are shown in Figure 52 changes:

538 OS/400 Backup and Recovery V5R1

Table 72. Additional Examples of Multiple Commitment Definitions in a Job
Effect on Changes to These Files:
Change in Scenario F1 and F2 F3 and F4 F5 and F6 F7
PGMX performs a Still pending Rolled back Already committed Rolled back
rollback operation
instead of a commit
operation
(3--COMMIT becomes
ROLLBACK).
PGMZ performs a Still pending Committed by PGMZ Already committed Committed
commit operation
before returning to
PGMX.
PGMZ attempts to Still pending Still pending Already committed Still pending
start commitment
control specifying
CMTSCOPE(*ACTGRP)
after updating file F7.
The attempt fails
because changes are
pending using the
job-level commitment
definition.
PGMX does not start Still pending Not under Already committed File F7 cannot be
commitment control commitment control opened because no
and does not open *JOB commitment
files F3 and F4 with definition would exist
COMMIT(*YES). (PGMX did not create
PGMZ attempts to it).
open file F7 with
COMMIT(*YES).

Commit Text Parameter

The commit text parameter identifies the specific text to be associated with a
commitment definition when displaying information about the commitment
definitions started for a job. If no text is specified, the system provides a default
text description.

Default Journal Parameter

You can specify a default journal when you start commitment control. You might
use a default journal for these reasons:
v You want to capture LUW journal entries. These entries can assist you in
analyzing the history of what resources are associated with an LUW. They are
not used for applying and removing journaled changes. The omit journal entries
(OMTJRNE) parameter determines whether the system writes LUW entries.
v You want to improve performance for jobs that close files and open them again
within a routing step. If you close all the files assigned to a journal that is not
the default journal, all the system information about the journal is removed from
the routing step. If a file that is assigned to that journal is opened later, all the
information about the journal must be created again. The system keeps
information about the default journal with the commitment definition, whether
or not any resources that are assigned to the journal are active.

Chapter 22. Commitment Control 539

 Omit Journal Entries Parameter
If you specify a default journal to improve performance, you can use the
OMTJRNE parameter to prevent the system from writing LUW journal entries.
Having the system write LUW entries significantly increases the size of your
journal receiver and degrades performance during commit and rollback operations.

LUW entries can be useful when you are setting up and testing either your
commitment control environment or a new application.

LUW entries are written to the default journal regardless of the value of the
OMTJRNE parameter under these conditions:
v A system error occurs during a commit or rollback operation.
v A manual change is made to a resource that participated in an LUW, and the
change caused a heuristic mixed condition. See Table 79 on page 565 for a
description of the heuristic mixed condition. This type of manual change is
called a heuristic decision.

You can use the information about what resources participated in the LUW to
determine what action to take in these situations.

Table 124 on page 824 through Table 130 on page 845 show the layouts for the
entry-specific data for LUW journal entries.

Two-Phase Commit–Overview
For systems using Version 3 Release 1 and later of the OS/400 licensed program,
the system supports two-phase commit operations. Two-phase commit is intended
to ensure that committable resources on multiple systems remain synchronized.

OS/400 supports two-phase commit in accordance with the SNA LU 6.2

architecture. For more detailed information about the internal protocols used by
the system for two-phase commit, refer to the SNA Transaction Programmer’s
Reference for LU Type 6.2.

For systems using releases prior to V3R7 of OS/400, the system supports the
Presumed Nothing protocols of SNA LU 6.2. For systems using V3R7 and later
releases of OS/400, the system supports the Presumed Abort protocols of SNA LU
6.2. When connections are made between pre-V3R7 systems and V3R7 and later
systems, Presumed Nothing protocols are used between those systems.

| Starting with V5R1M0, two-phase commit is also supported using TCP/IP as a

| Distributed Unit of Work (DUW) DRDA protocol. To use TCP/IP DUW
| connections, all of the systems (both the application requester and the application
| server) must be at V5R1M0. For more information about DRDA see the Open
| Group Technical Standard, DRDA V2 Vol. 1: Distributed Relational Database
| Architecture at the following Web site:
| http://www.opengroup.org/

| Under two-phase commit, the system performs the commit operation in two
waves:
v During the prepare wave, a resource manager issues a commit request to its
transaction manager. The transaction manager informs any other resources it

540 OS/400 Backup and Recovery V5R1

 manages and the other transaction managers that the LUW is ready to be
committed. All the resource managers must respond that they are ready to
commit. This is called the vote.
v During the committed wave, the transaction manager that initiated the commit
request decides what to do, based on the outcome of the prepare wave. If the
prepare wave completes successfully and all participants vote ready, the
transaction manager instructs all the resources it manages and the other
transaction managers to commit the LUW. If the prepare wave does not
complete successfully, all the transaction managers and resource managers are
instructed to roll back the LUW.

Placing Resources Under Commitment Control

When you place an object under commitment control, it becomes a committable
resource. It is registered with the commitment definition. It participates in each
commit operation and rollback operation that occurs for that commitment
definition.

The topics that follow describe these attributes of a committable resource:

v Resource type
v Location
v Commit protocol
v Access intent

Types of Committable Resources

Table 73 shows:
v The types of committable resources.
v How they are placed under commitment control.
v How they are removed from commitment control.
v Restrictions that apply to the resource type.
Table 73. Types of Committable Resources
What Kinds of
How It Is Placed Under How It Is Removed From Changes Are
Resource Type Commitment Control Commitment Control Committable Restrictions
FILE– local Opening under v Closing the file, if no Record level changes No more than
database files commitment control1 changes are pending. 4 000 000 records can
be locked for a single
v If changes are pending
LUW.
when the file is closed,
after performing the
next commit or rollback
operation.
DDL– Running SQL under v Performing a commit or Object level changes, Only object-level
object-level commitment control rollback operation after such as: changes made using
changes to the object-level change. v Create SQL SQL are under
local SQL Package commitment control.
tables and SQL
v Create SQL Table
collections.
v Drop SQL Table

Chapter 22. Commitment Control 541

 Table 73. Types of Committable Resources (continued)
What Kinds of
How It Is Placed Under How It Is Removed From Changes Are
Resource Type Commitment Control Commitment Control Committable Restrictions
DDM– remote Opening under v Closing the file, if no Record level changes
distributed commitment control3 changes are pending.
data
v If changes are pending
management
when the file is closed,
(DDM) file
after performing the
next commit or rollback
operation.
LU 6.2– Starting the v Ending the
protected conversation4 conversation
conversation6
®
| DRDA – Using SQL CONNECT v Ending the connection
distributed statement2
relational
database
API– local API QTNADDCR API v QTNRMVCR API The user program The application must
commitment determines this. provide an exit
resource5 Journal entries may program to be called
be written by the user during commit,
program using the rollback, or
QJOSJRNE API to resynchronization
assist with tracking operations.
these changes.
|| TCP–TCP/IP Using SQL CONNECT v Ending the SQL
|| connection statement to an RDB connection, or closing
|| defined to use TCP/IP the DDM file if no
|| connections, or opening changes are pending. If
|| a DDM file defined with the DDM file is closed
|| a TCP/IP location with changes pending,
| the connection is closed
| after performing the
| next commit or rollback
| operation.

542 OS/400 Backup and Recovery V5R1

 Table 73. Types of Committable Resources (continued)
What Kinds of
How It Is Placed Under How It Is Removed From Changes Are
Resource Type Commitment Control Commitment Control Committable Restrictions

1
For details on how to place a database file under commitment control, refer to:
v The appropriate language reference manual.
| v The Information Center at the following Web site: http://www.ibm.com/eserver/iseries/infocenter
2
For more information about using SQL with commitment control, see the Information Center at the
following Web site: http://www.ibm.com/eserver/iseries/infocenter.
3
For details on how to place a remote database file under commitment control, look in the Information
Center at the following Web site: http://www.ibm.com/eserver/iseries/infocenter.
4
When a DDM connection is started, the DDM file specifies PTCCNV(*YES), and the DDM file is defined
with an SNA remote location, an LU62 resource is added with the DDM resource.
When a DRDA connection is started, an LU62 resource is added with the DRDA resource if both of the
following are true:
v The program is using the distributed unit of work connection protocols.
v The connection is to an RDB defined with an SNA remote location.
For more information about starting protected conversions, see the APPC Programming book.
5
For more information about API commitment resources and these API interfaces, see the Information Center
at the following Web site: http://www.ibm.com/eserver/iseries/infocenter
6
The presumed abort level of LU 6.2 two-phase commit is implemented in V3R7. For information about LU
6.2, see the SNA Transaction Programmer’s Reference for LU Type 6.2

Location of Committable Resources

A committable resource can be either a local resource or a remote resource.

Local Committable Resources: A local committable resource resides on the same

system as the application. Each journal associated with resources under
| commitment control can be thought of as a local location. All the resources that are
| registered without a journal (optionally for both DDL resources and API resources)
| can be thought of as a separate local location.

Remote Committable Resources: A remote committable resource resides on a

different system from the application. A remote location exists for each unique
conversation to a remote system. A commitment definition may have one or more
remote locations on one or more remote systems.

“Location of Committable Resources” shows the types of committable resources

and their locations:
Table 74. Local and Remote Committable Resources
Resource Type Location
FILE Local
DDL Local
API Local
DDM Remote
LU62 Remote
DRDA Local or remote
| TCP Remote

Chapter 22. Commitment Control 543

 The Commit Protocol of a Committable Resource
Commit protocol is the capability a resource has to participate in one-phase or
two-phase commit processing. Local resources, except API committable resources,
are always two-phase resources. A two-phase resource is also called a protected
resource. Remote resources and API committable resources must be registered as
one-phase resources or two-phase resources when they are placed under
commitment control. Table 75 shows what types of committable resources can
coexist in a commitment definition with a one-phase resource:
Table 75. Coexistence Rules for One-Phase Resources and Two-Phase Resources
Resource Type Can Coexist With
One-phase API resource Other local resources. No remote resources.
One-phase remote resource Other one-phase resources at the same
location. No local resources.

The Access Intent of a Committable Resource

When a resource is placed under commitment control, the resource manager
indicates how the resource will be accessed:
v Update
v Read only
v Undetermined

The access intent determines how the resources participate together in an LUW.
“The Access Intent of a Committable Resource” shows what access intents are
possible for a particular type of resource and how the system determines the
access intent for a resource when it is registered:
Table 76. Access Intents for Resource Types
How the Access Intent Is
Resource Type Possible Access Intents Determined

FILE Update, read only Based on how the file was opened
DDL Update Always update
API Update Always update
DDM Update, read only Based on how the file was opened
LU62 Undetermined Always undetermined
DRDA Update, read only, For DRDA Level 1, the access
undetermined intent is update if no other remote
resources are registered. Otherwise,
the access intent is read only. For
DRDA Level 2, the access intent is
always undetermined.
| TCP Undetermined Always undetermined

The access intent of resources that are already registered determines whether a
new resource can be registered. The following rules apply:
v A one-phase resource whose access intent is update cannot be registered when
any of the following is true:
– Resources whose access intent is update are already registered at other
locations.
– Resources whose access intent is undetermined are already registered at other
locations.

544 OS/400 Backup and Recovery V5R1

 – Resources whose access intent is undetermined are already registered at the
same location and the resources have been changed during the current LUW.
v A two-phase resource whose access intent is update cannot be registered when a
one-phase resource whose access intent is update is already registered.

Files Being Journaled under Commitment Control

A database file (resource type FILE or DDM) must be journaled before it can be
opened for output under commitment control. A file does not need to be journaled
in order to open it for input only under commitment control. An error occurs if:
v An attempt is made to open a database file for output under commitment
control, but the file is not currently journaled.
v No commitment definition is started that can be used by the file being opened
under commitment control.

If only the after-images are being journaled for a database file when that file is
opened under commitment control, the system automatically starts journaling both
the before- and after-images. The before-images are written only for changes to the
file that occur under commitment control. If other changes that are not not under
commitment control occur to the file at the same time, only after-images are
written for those changes.

Only record-level committable changes are automatically written to a journal by

the system. The journal is then used by the system, if necessary, for recovery
purposes. Object-level committable changes and committable changes for API
commitment resources are not written to the journal automatically by the system.
The exit program for the API resource can use the Send Journal Entry (QJOSJRNE)
API to write journal entries to provide an audit trail or to assist with recovery. The
content of these entries is controlled by the user exit program.

Recovery for object-level commitment resources is performed by the system using

a mechanism other than a journal. Recovery for API commitment resources is
accomplished by calling the commit and rollback exit program associated with
each particular API commitment resource. The exit program has the responsibility
for performing the actual recovery that is necessary for the situation.

Sequence of Journal Entries Under Commitment Control

Table 77 on page 546 shows the sequence of entries that are typically written while
a commitment definition is active. The tables in Appendix D have more
information about the contents of the journal entries. Notice that commitment
control entries are written to a journal (local or remote) only if at least one of the
following is true:
v The journal is specified as the default journal on the STRCMTCTL command.
v At least one file assigned to the journal is opened under commitment control.
v At least one API commitment resource associated with the journal is registered
under commitment control.

Chapter 22. Commitment Control 545

 Table 77. Sequence of Commitment Control Journal Entries
Entry Type Description Where It Is Written When It is Written
’C BC’ Begin To the default journal, if one is specified on When the STRCMTCTL command is
commitment the STRCMTCTL command. issued
control
To the journal for each local location. When the first file assigned to a journal
is opened or when an API resource is
registered for a journal.
’C SC’ Start commit To the journal for each local location. When the first record change occurs for
cycle the LUW for a file assigned to this
journal.
When a commit operation is
performed, if cursor position has
changed for a file journaled to that
journal for the LUW, or a commit
identifier is specified on the commit
request, and an SC record has not
already been written to the journal for
the LUW. In this case, both an SC
journal entry and a CM journal entry
are written.
To the journal for an API resource. When the QJOSJRNE API is first used
with the Include Commit Cycle Identifier
key.
| Journal DDL object level To the journal associated with the object When updates occur.
| codes D entries being updated. Only journal entries that
| and F contain a commit cycle identifier represent a
| DDL object level change that is part of the
| LUW.
Journal Record level To the journal associated with the file being When the updates occur.
code R entries updated.
Journal User-created To the journal associated with an API If the application program uses the
code U entries resource. SNDJRNE command or the QJOSJRNE
API.
’C CM’ Commit To the journal for each location. When the commit has completed
successfully.
To the default journal. If any committable resources are
associated with the journal.
’C RB’ Rollback To the journal for each local location. After the rollback operation has
completed.
To the default journal. If any committable resources are
associated with the journal.
’C LW’ End LUW To the default journal, if one is specified on When the commit or rollback operation
the STRCMTCTL command. The system has completed.
writes an LW header record and one or more
detail records. These entries are written only
if OMTJRNE(*NONE) is specified on the
STRCMTCTL command or if a system error
occurs.

546 OS/400 Backup and Recovery V5R1

 Table 77. Sequence of Commitment Control Journal Entries (continued)
Entry Type Description Where It Is Written When It is Written
’C EC’ End To the journal for each local location. When the ENDCMTCTL command
commitment completes.
control
To a local journal that is not the default When a commit boundary is
journal. established, following the point when
all committable resources associated
with that journal have been removed
from commitment control.

Commit Cycle Identifier

A commit cycle is the time from one commitment boundary to the next. The
system assigns a commit cycle identifier to associate all of the journal entries for a
particular commit cycle together. Each journal that participates in an LUW has its
own commit cycle and its own commit cycle identifier.

The commit cycle identifier is the journal sequence number of the ’C SC’ journal
entry written for the commit cycle. The commit cycle identifier is placed in each
journal entry written during the commit cycle. If more than one journal is used
during the commit cycle, the commit cycle identifier for each journal is different.

You can use the QJOSJRNE API to write journal entries for API resources. You
have the option of including the commit cycle identifier on those journal entries.

You can use the commit cycle identifier to apply or remove journaled changes to a
commitment boundary using the APYJRNCHG command or the RMVJRNCHG
command. These limitations apply:
| v Most object-level changes made under commitment control are written to the
| journal but are not applied or removed using the APYJRNCHG and
| RMVJRNCHG commands.
v The QJOSJRNE API writes user-created journal entries with a journal code of U.
These entries cannot be applied or removed using the APYJRNCHG and
RMVJRNCHG commands. They must be applied or removed with a user-written
program.

Commit and Rollback Operations

Two operations affect changes made under commitment control:
v Commit
All changes made under commitment control since the previous commit or
rollback operation are made permanent and all locks related to the LUW are
released.
v Rollback
All changes made since the previous commit or rollback operation are removed
and all locks related to the LUW are released.

The commit and rollback operations are supported in these programming

languages:
Table 78. Language Support for Commit and Rollback Operations
Language Commit Rollback
CL COMMIT command ROLLBACK command

Chapter 22. Commitment Control 547

 Table 78. Language Support for Commit and Rollback Operations (continued)
Language Commit Rollback
ILE RPG/400 COMIT operation code ROLBK operation code
ILE COBOL/400® COMMIT verb ROLLBACK verb
ILE C/400 _Rcommit function _Rrollbck function
PL/I PLICOMMIT subroutine PLIROLLBACK subroutine
SQL/400® COMMIT statement ROLLBACK statement

Roles in Commit Processing

If a commit of an LUW involves more than one resource manager, each resource
manager plays a role in the LUW. A resource manager is responsible for
committing or rolling back changes made during the LUW. The resource managers
by resource type are:
FILE Database manager
DDM Database manager
DDL Database manager
DRDA
Communications transaction program
LU62 Communications transaction program
API API exit program

Figure 53 shows the basic roles in an LUW. The structure shown in the figure is
called a transaction program network.

┌───────────────┐
│ System B │
┌────────S│ Agent │
│ └───────────────┘
│
┌───────────────┐ │ ┌───────────────┐
│ System A ├───┼────────S│ System C │
│ Initiator │ │ │ Agent │
└───────────────┘ │ └───────────────┘
│
│ ┌───────────────┐
└────────S│ System D │
│ Agent │
└───────────────┘

Figure 53. Roles in Two-Phase Commit Processing–Single-Level Tree

When an application on System A issues a commit request, the resource manager

| on System A becomes the initiator. For DRDA distributed unit of work over
| TCP/IP, the initiator is called the coordinator. The resource managers for the other
| three systems (B, C, and D) become agents for this LUW. For DRDA distributed
| unit of work over TCP/IP, the agent is also called the participant.

| If the application is using APPC communications to perform the two-phase

| commit, the relationship between systems can change from one LUW to the next.
Figure 54 on page 549 shows the same systems when an application on System B
issues the commit request:

548 OS/400 Backup and Recovery V5R1

 ┌───────────┐
┌──S│ System C │
│ │ Agent │
┌────────────┐ ┌────────────┐ │ └───────────┘
│ System B │ │ System A │ │
│ Initiator ├────S│ Agent ├───┤
│ │ │ Cascaded │ │
└────────────┘ │ Initiator │ │ ┌───────────┐
└────────────┘ │ │ System D │
└──S│ Agent │
└───────────┘

Figure 54. Roles in Two-Phase Commit Processing–Multi-Level Tree

| The roles in Figure 54 do not apply to DRDA distributed unit of work over TCP/IP
| because multi-level transactions trees are not supported.

The transaction program network has another level because System B is not
communicating directly with System C and System D. The resource manager in
System A now has the roles of agent and cascaded initiator.

| To improve performance of LU6.2 two-phase transactions, the initiator may assign

| the role of last agent to one of the agents. The last agent does not participate in the
prepare wave. In the committed wave, the last agent commits first. If the last agent
does not commit successfully, the initiator instructs the other agents to roll back.

| For DRDA distributed unit of work over TCP/IP, there is the role of the resync
| server. The resync server is a participant whose responsibility is to resynchronize
| the other participants in the event that there is a communications failure with the
| coordinator, or the coordinator has a systems failure.

How the System Performs a Commit Operation

The system performs the following steps when it receives the request to commit
(when the committed wave is started).
v The system saves the commit identification, if one is provided, for use at
recovery time.
v Records are written to the file prior to the commit operation being performed, if:
– Records were added to a local or remote database file under commitment
control, and
– SEQONLY(*YES) was specified when the file was opened so that blocked I/O
feedback is used by the system and a partial block of records exists.

Otherwise, the I/O feedback area and I/O buffers are not changed.
v A call is made to the commit and rollback exit program for each API
commitment resource that is present in the commitment definition. If a location
has more than one exit program registered, the exit programs for that location
are called in the order that they were registered.
v A ’C CM’ journal entry is written to every local journal associated with the
commitment definition. See Table 77 on page 546.
v Pending object-level changes are made permanent.
v Record and object locks that were acquired and kept for commitment control
purposes are unlocked and those resources are made available to other users.
v Information is changed in the commitment definition to show that the current
LUW has been ended.

Chapter 22. Commitment Control 549

 All of the steps in this topic must perform correctly for the commit operation to be
successful.

When remote resources are under commitment control, the initiator sends a
commit request to all remote agents. The request is propagated throughout the
transaction program network. Each agent responds with the results of the commit
operation.

If errors occur during the prepare wave, the initiator sends a rollback request to all
agents. If errors occur during the committed wave, the system attempts to bring as
many locations as possible to committed status. This may result in a heuristic
mixed state. See “States of the Logical Unit of Work (LUW)” on page 564 for more
information about the possible states.

Any errors are sent back to the initiator where they are signaled to the user. If a
default journal was specified on the STRCMTCTL command, ’C LW’ entries are
written. If errors occur, these entries are written, even if OMTJRNE(*LUWID) was
specified. You can use these entries, along with the error messages and the status
information for the commitment definition, to attempt to synchronize the
committable resources manually.

How the System Performs a Rollback Operation

The system performs the following steps when it receives the request to roll back:
v For database files under commitment control:
– Records are cleared from the I/O buffer:
- If records were added to a local or remote database file under commitment
control, and
- If SEQONLY(*YES) was specified when the file was opened so that blocked
I/O is used by the system and a partial block of records exists that has not
yet been written to the database.

Otherwise, the I/O feedback area and I/O buffers remain unchanged.
– A call is made to the commit or rollback exit program for each API
commitment resource that is present in the commitment definition. If a
location has more than one exit program registered, the exit programs for that
location are called in reverse order from the order in which they were
registered.
– If a record was deleted from a file, the record is added back to the file.
– Any changes to records that have been made during this LUW are removed,
and the original records (the before-images) are placed back into the file.
– If any records were added to the file during this LUW, they remain in the file
as deleted records.
– If any record changes were made to resources assigned to a journal during
the LUW, a journal entry of ’C RB’ is added to the journal indicating that a
rollback operation occurred. The journal also contains images of the record
changes that were rolled back. Before the rollback operation was requested,
the before-images and after-images of changed records were placed in the
journal. A ’C RB’ entry is also written to the default journal if any
committable resources are assigned to that journal.
– The system positions the open files under commitment control at the last
record accessed in the previous LUW or at the open position if no commit
operation has been performed for the file using this commitment definition.
This is an important consideration if you are doing sequential processing.

550 OS/400 Backup and Recovery V5R1

 – Non-committable changes for database files are not rolled back. For example,
opened files are not closed, and cleared files are not restored. Any files that
were closed during this LUW are not reopened or repositioned.
– Record locks that were acquired for commitment control purposes are
unlocked and those records are made available to other users.
v The commit identification currently saved by the system remains the same as the
commit identification provided with the last commit operation for the same
commitment definition.
v Object-level committable changes made during this LUW are reversed, or rolled
back.
v Object locks that were acquired for commitment control purposes are unlocked
and those objects are made available to other users.
v The previous commitment boundary is established as the current commitment
boundary.
v Information is changed in the commitment definition to show that the current
LUW has been ended.

All of the above must perform correctly for the rollback operation to be successful.

When remote resources are under commitment control, the initiator sends a
rollback request to all remote agents. The request is propagated throughout the
transaction program network. Each agent responds with the results of the rollback
operation.

Flow of Commit Processing

The actual flows that occur during a commit can be different depending on the
shape of the transaction program tree and the settings of the various commitment
options. These differences are shown below.

| The flows demonstrated in Figure 55 on page 552 and Figure 56 on page 553 are
| specific to LU6.2. For a desciption of DRDA distributed unit of work over TCP/IP
| flow, refer to the SYNCPTOV chapter of the Open Group Technical Standard,
| DRDA V2 Vol. 3: Distributed Relational Database Architecture at the following Web
| site:
| http://www.opengroup.org/

| Figure 55 on page 552 shows the flow of messages among the application programs
and the transaction managers when an application program issues a commit
instruction when last agent optimization is not used. Both the initiator application
program and the agent application programs are unaware of the two-phase commit
processing. The numbers in parentheses () in the figure correspond to the
numbered items in the description that follows.

Chapter 22. Commitment Control 551

 I TM-I TM-A A
│ │ │ Receive (1) │
│ │ │W──────────────────────┤
│Commit (2) │ │ │
├──────────S│ Prepare (3) │ │
│ ├─────────────────S│ RCVTKCMT indicator (4)│
│ │ ├──────────────────────S│
│ │ │ │
│ │ │ Commit (5) │
│ │Request commit (6)│W──────────────────────┤
│ │W─────────────────┤ │
│ │ │ │
│ │ Commit (7) │ │
│ ├─────────────────S│ │
│ │ Reset (8) │ │
│ Return (9)│W─────────────────┤ Return (9) │
│W──────────┤ ├──────────────────────S│

Legend
I = Initiator (Application that initiates commit
request)
TM-I = Transaction manager for initiator
A = Agent (Application that receives commit request)
TM-A = Transaction manager for agent

Figure 55. Flow of Commit Processing without Any Optimization

Following is a description of the events for normal processing without last agent
optimization. This describes a basic flow. The sequence of events can become much
more complex when the transaction program network has multiple levels or when
errors occur.
1. Application program A does a receive request to indicate that it is ready to
receive a request from program I.
2. The initiator application (I) issues a commit instruction.
3. The transaction manager for the initiator (TM-I) takes the role of initiator for
this LUW. It starts the prepare wave by sending a prepare message to all the
other locations that are participating in the LUW.
4. The transaction managers for every other location take the role of agent
(TM-A). The application program A is notified by TM-A that a request to
commit has been received. For ICF files, the notification is in the form of the
Receive Take Commit (RCVTKCMT) ICF indicator being set on.
5. The application program A responds by issuing a commit instruction (or a
rollback instruction). This is the application program’s vote.
6. The agent (TM-A) responds to the initiator (TM-I) with a request commit
message.
7. When the initiator (TM-I) receives all the votes, the TM-I sends a commit
message. This starts the committed wave.
8. Each agent (TM-A) commits and responds with a reset message.
9. If all agents commit successfully, the LUW is complete. A return is sent to the
application programs (I and A). This return indicates that the commit operation
was successful.

Figure 56 on page 553 shows the flow of messages among the application programs
and the transaction managers when an application program issues a commit
instruction when last agent optimization is used. Both the initiator application
program and the agent application programs are unaware of the two-phase commit

552 OS/400 Backup and Recovery V5R1

processing. The numbers in parentheses () in the figure correspond to the
numbered items in the description that follows.

I TM-I TM-A A
│ │ │ Receive (1) │
│ │ │W──────────────────────┤
│Commit (2) │ │ │
├──────────S│Request commit (3)│ │
│ ├─────────────────S│ RCVTKCMT indicator (4)│
│ │ ├──────────────────────S│
│ │ │ │
│ │ │ Commit (5) │
│ │ Committed (6) │W──────────────────────┤
│ Return (8)│W─────────────────┤ Return (7) │
│W──────────┤ ├──────────────────────S│
│ │ │ │
│any verb │Implied reset (9) │ │
├──────────S├─────────────────S│ │
│ │ │ │

Legend
I = Initiator (Application that initiates commit
request)
TM-I = Transaction manager for initiator
A = Agent (Application that receives commit request)
TM-A = Transaction manager for last agent

Figure 56. Flow of Commit Processing with Last Agent Optimization

Following is a description of the events for normal processing with last agent
optimization. This describes a basic flow. The sequence of events can become much
more complex when the transaction program network has multiple levels or when
errors occur.
1. Application program A does a receive request to indicate that it is ready to
receive a request from program I.
2. The initiator application (I) issues a commit instruction.
3. The transaction manager for the initiator (TM-I) takes the role of initiator for
this LUW. It chooses TM-A as its last agent. A request commit message is then
sent to TM-A.

Note: If there are agents other than TM-A in the LUW, they take the role of
agent, and the prepare wave as described in steps 3-6 of Figure 55 on
page 552 is completed for those agents before the request commit
message is sent to TM-A
4. When TM-A receives the request commit message, without having received a
prepare message, TM-A knows it is the last agent. The application program A is
notified by TM-A that a request to commit has been received. For ICF files, the
notification is in the form of the Receive Take Commit (RCVTKCMT) ICF
indicator being set on.
5. The application program A responds by issuing a commit instruction (or a
rollback instruction). This is the application program’s vote.
6. The agent (TM-A) responds to the initiator (TM-I) with a committed message.
7. A return is sent to the application program (A) to indicate that the LUW is
complete at agent TM-A.
8. The initiator (TM-I) receives the vote from its last agent (TM-A). A return is
sent to the application program (I) to indicate that the LUW is complete.

Chapter 22. Commitment Control 553

 Note: If there are agents other than TM-A in the LUW, the committed wave is
completed before the return is sent to the application program (I) as
described in steps 7-9 of Figure 55 on page 552.
9. The next time the initiator (TM-I) issues any message to the last agent (TM-A),
either a data flow or a commitment instruction, an implied reset indicator is
sent with the message to inform TM-A that TM-I completed the commit
successfully. The reason for this is that TM-A must retain information about the
completed LUW until it has confirmed that TM-I successfully received the
committed message in step 6.

Errors During Commit Processing

If a communications or system failure occurs during a commit operation,
resynchronization may need to be performed to ensure that the transaction
managers keep the data consistent on all the systems involved in the LUW. The
behavior of the resynchronization and how it affects the commit operation depends
on several factors :
v The Wait for outcome commitment option. Refer to “Changing the Commitment
Definition to Not Wait for Outcome” for details.
v The state of the LUW. Refer to “States of the Logical Unit of Work (LUW)” on
page 564 for details.
v The protocol of the locations that resynchronization is performed with. The
protocol is either presumed nothing or presumed abort. Beginning with
V3R7M0, the presumed abort protocol will be used with all locations that
support it.

If the failure is catastrophic such that it can never be repaired, or it cannot be

repaired in a timely manner, the system operators for other systems involved in
the LUW may have to make a heuristic decision. The heuristic decision commits or
rolls back the changes made on that system during the LUW. If the failure is
repaired after such a decision, and the resynchronization detects that the decision
caused data integrity problems, message CPD83D9 or CPD83E9 is sent to the
QSYSOPR message queue.

Changing the Commitment Definition to Not Wait for Outcome

| This section does not apply if you are using a DRDA distributed unit of work over
| TCP/IP connection. DRDA distributed unit of work over TCP/IP connections
| never wait for outcome.

When a communications or systems failure occurs during a commit operation such

that resynchronization is required, the default is to wait until the resynchronization
is finished before the commit operation completes.

You may want to change this behavior if the following conditions are true:
v The applications that participate are independent of each other.
v Your program logic does not need the results of previous transactions to ensure
that your database files remain synchronized.

After you start commitment control, you can use the Change Commitment Options
(QTNCHGCO) API to specify that the commitment definition does not wait for the
outcome of resynchronization. If you specify N (No) for the Wait for outcome
option, the system uses a database server job (QDBSRVnn) to handle
resynchronization asynchronously.

554 OS/400 Backup and Recovery V5R1

Note: These database server jobs are started during the IPL process. If you change
the options for commitment control, this has no effect on the number of jobs
that the system starts.

This section only refers to two values for the resolved Wait for outcome option, Y
(Yes) and N (No). There are actually two more values that can be specified, L (Yes
or Inherit from Initiator) and U (No or Inherit from Initiator). When these values
are used, the actual value used during each commit operation is resolved to Yes or
No by the system. Refer to the System API Reference for details.

Note: The initiator’s value can only be inherited by an agent if both the initiator
and the agent support presumed abort (available at V3R7 and later releases).

The wait for outcome (WFO) option does not affect normal, error-free commit
processing. If an error occurs, the WFO option determines whether the application
waits for resynchronization or not, with the following conditions:
1. If the resolved WFO option is Y (Yes), the application will wait for the result of
the resynchronization.
2. If the resolved WFO option is N (No) and a communication failure occurred
during the prepare wave or rollback of a location that supports presumed abort
protocols, no resynchronization is performed and the commitment definition is
rolled back.
3. If the commitment definition is in doubt (LUW state is prepared or Last Agent
Pending), the application will wait for the result of the resynchronization
regardless of the resolved WFO value. For further information on the
commitment definition being in doubt, refer to Table 79 on page 565.
4. If the resolved WFO option is N and neither one of conditions two or three are
true, the system attempts to resynchronize once. If it is not successful, the
system signals STATUS message CPF83E6 to the application to indicate that
resynchronization is in progress.
Since CPF83E6 is a STATUS message, it will only have an effect if the
application is monitoring for it. Normally, your application can treat this as an
informational message. The systems that are participating in the LUW will
attempt to resynchronize the LUW until the failure is repaired. These
subsequent resynchronization attempts are performed in the database server
jobs. If a subsequent resynchronization attempt performed in a database server
job fails, the message CPI83D0 will be sent to QSYSOPR.

For example, in Figure 57 on page 556, the commitment definition for the initiator
(I) uses the default value of Y (Yes) for the Wait for outcome option. When
communications between TM-I and TM-A is lost, both application A and
application I wait until the LUW is resynchronized.

Chapter 22. Commitment Control 555

 I TM-I TM-A A
│ │ │ Receive │
│ │ │W───────────────────┤
│ Commit │ │ │
├────────S│ Prepare │ │
│ ├─────────────────S│ RCVTKCMT indicator │
│ │ ├───────────────────S│
│ │ │ │
│ │ │ Commit │
│ │ Request commit │W───────────────────┤
│ │W─────────────────┤ │
│ │ │ │
│ │ Commit │ │
│ ├─────────────────S│ (Communications is │
│ │ Resynchronize lost) │
│ W─────────────────S │
│ │ Resynchronize │
│ W─────────────────S │
│ │ Resynchronize │ (Communications is │
│ W─────────────────S│ reestablished) │
│ │ │ │
│ Return │ │ Return │
│W────────┤ ├───────────────────S│
│ │ │ │

Figure 57. Wait for Outcome–Yes

In Figure 58 on page 557, the commitment definition for the initiator has the
resolved WFO set to N (No). TM-A meets condition 3 in the preceding list, while
TM-I meets condition 4. Control is returned to application I after one attempt to
resynchronize with TM-A. A database server job attempts to resynchronize.
Application I never receives the return indicator when the commit request has
completed successfully. Note that control is not returned to the agent application
(A) until after communications is reestablished. This depends on the timing of the
failure. In this case, the communications failure occurred before the commit
message was received from the initiator, leaving TM-A in doubt as to whether to
commit or rollback. When the transaction manager is in doubt, it retains control
until the resynchronization is completed, regardless of the resolved WFO value at
that system.

If you want the applications at all systems to continue before resynchronization

completes, you must either change the resolved WFO option to N (No) on all
systems, or set the initiator to N and the rest of the systems to U (No or Inherit
from Initiator). But remember that the resolved WFO option is ignored when the
transaction manager is in doubt as to whether to commit or rollback, and always
waits until resynchronization completes before returning control.

556 OS/400 Backup and Recovery V5R1

 I TM─I TM─A A
│ │ │ Receive │
│ │ │W───────────────────┤
│ Commit │ │ │
├────────S│ Prepare │ │
│ ├─────────────────S│ RCVTKCMT indicator │
│ │ ├───────────────────S│
│ │ │ │
│ │ │ Commit │
│ │ Request commit │W───────────────────┤
│ │W─────────────────┤ │
│ │ │ │
│ │ Commit │ │
│ ├─────────────────S│ (Communications is │
│ │ Resynchronize lost) │
│ W─────────────────S │
│ Resync. │ Passes │
│W────────┤ resynchronization │
│ in │ task to a │
│ progress│ database server │
│ │ job │
│ │ │
│ │ DB │
│ │ Server │
│ │ Job │
│ │ │ │
│ │ │ Resynchronize │
│ │ │W─────────────S │
│ │ │ Resynchronize │
│ │ │W─────────────S │
│ │ │ Resynchronize (Communications is │
│ │ │W─────────────S│ reestablished) │
│ │ │ │ │
│ │ │ │ │
│ │ │ Return │
│ │ ├───────────────────S│
│ │ │ │

Figure 58. Wait for Outcome–No

When a connection is made to a remote relational database, and no protected

conversations have already been started, the system implicitly changes the Wait
for outcome value to N. The reason for this is that the performance of commit
operations is improved when the Wait for outcome value is N and the remote
system supports presumed abort. (Presumed abort is available at V3R7 and later
releases.) This implicit change of the Wait for outcome value is only performed for
DRDA and DDM applications. APPC applications will use the default Wait for
outcome value of Y unless they invoke the QTNCHGCO API to change it.

The Information Center at the following address:

http://www.ibm.com/eserver/iseries/infocenter provides more information about
the QTNCHGCO API.

Changing the Commitment Definition to Not Select a Last

Agent
| This section does not apply if you are using a DRDA distributed unit of work over
| TCP/IP connection. By default, the transaction manager for the initiator is free to
select any agent as a last agent during a commit operation. In case of a multi-level
tree, any agent selected as a last agent by its initiator is also free to select a last
agent of its own. Performance is improved when a last agent is selected during the

Chapter 22. Commitment Control 557

 commit operation because two communications flows are eliminated between an
initiator and its last agent (the prepare phase is eliminated for these systems). See
Figure 55 on page 552 and Figure 56 on page 553 to see the differences between the
commit flows.

However, once the initiator sends the request commit to its last agent, it must wait
until it has received the last agent’s vote before it can continue. This is regardless
of the Wait for outcome value for the commitment definition. During normal,
error-free commit processing, this is not an issue. But, if an error occurs during this
window, the initiator cannot continue until resynchronization completes. If the
initiator application is handling requests from a user at a terminal, this could be a
usability consideration.

You should consider whether the improved performance during normal commit
operations is more important than the impact on usability when such an error
occurs. Note that if the error occurs before the request commit is sent to the last
agent, the LUW will immediately roll back and the initiator will not wait.
Therefore, the window when an error can cause the initiator to wait is quite small,
so such an error should be rare.

If you decide that the usability impact is not worth the improved performance, you
can change your commitment definitions to not select a last agent. After you start
commitment control, you can use the Change Commitment Options (QTNCHGCO)
API to change the Last agent permitted option to N.

For more information about the QTNCHGCO API, look in the Information Center
at the following Web site: http://www.ibm.com/eserver/iseries/infocenter.

Changing the Commitment Definition to Indicate OK to Leave

Out
| This section does not apply if you are using a DRDA distributed unit of work over
| TCP/IP connection. Normally, the transaction manager at every location in the
transaction program network participates in every commit or rollback operation. To
improve performance, you can set up some or all locations in an LUW to allow the
transaction manager to indicate OK to Leave Out. If no communications flows are
sent to the location during an LUW, the location is left out when a commit or
rollback operation is performed. This improves overall performance because the
communications flows that normally occur during the commit or rollback are
eliminated during LUWs in which no data is sent to one or more remote locations.

After you start commitment control, you can use the Change Commitment Options
(QTNCHGO) API to change the OK to Leave Out option to Y (Yes). You may want
to do this if one or more remote systems often are not involved in an LUW.

If your commitment definition is set up to indicate OK to Leave Out, the

application waits for the next message flow from another location. For example, in
Figure 55 on page 552, program A would not receive the return after issuing the
commit instruction. Program A would wait until a data flow was received from
system I.

The OK to Leave Out option is intended for applications that are client/server in
nature. If the only purpose of program A is to satisfy requests from program I and
not do any independent work, then it is appropriate to allow the OK to Leave Out
option for program A.

558 OS/400 Backup and Recovery V5R1

Figure 59 shows the flow of messages among the application programs and the
transaction managers when an application program issues a commit instruction
without last agent optimization when the agent indicates OK to leave out. Both the
initiator application program and the agent application programs are unaware of
the two-phase commit processing. The numbers in parentheses () in the figure
correspond to the numbered items in the description that follows.

I TM─I TM─A O
│ │ │ Receive (1) │
│ │ │W──────────────────────┤
│Commit (2) │ │ │
├──────────S│ Prepare (3) │ │
│ ├─────────────────S│ RCVTKCMT indicator (4)│
│ │ ├──────────────────────S│
│ │ │ │
│ │ │ Commit (5) │
│ │Request commit (6)│W──────────────────────┤
│ │(ok to leave out) │ │
│ │W─────────────────┤ │
│ │ │ │
│ │ Commit (7) │ │
│ ├─────────────────S│ │
│ │ Reset (8) │ │
│ Return (9)│W─────────────────┤ │
│W──────────┤ │ │
' ' ' '
' 0 or more ' ' (left out)(10) '
' LUWs ' ' '
' ' ' '
│ data │ new LUW ID (11) │ │
├──────────S├─────────────────S│ Return (12) │
│ │ ├──────────────────────S│

Legend
I = Initiator (Application that initiates commit
request)
TM-I = Transaction manager for initiator
A = Agent (Application that receives commit request)
TM-A = Transaction manager for agent

Figure 59. Flow of Commit Processing without Last Agent Optimization when agent votes OK
to leave out

Following is a description of the events for normal processing without last agent
optimization when the agent votes OK to leave out. This describes a basic flow.
The sequence of events can become much more complex when the transaction
program network has multiple levels or when errors occur.
1. Application program A does a receive request to indicate that it is ready to
receive a request from program I.
2. The initiator application (I) issues a commit instruction.
3. The transaction manager for the initiator (TM-I) takes the role of initiator for
this LUW. It starts the prepare wave by sending a prepare message to all the
other locations that are participating in the LUW.
4. The transaction managers for every other location take the role of agent
(TM-A). The application program A is notified by TM-A that a request to
commit has been received. For ICF files, the notification is in the form of the
Receive Take Commit (RCVTKCMT) ICF indicator being set on.

Chapter 22. Commitment Control 559

 5. The application program A responds by issuing a commit instruction (or a
rollback instruction). This is the application program’s vote.
6. If application program A has used the Change Commitment Options API
(QTNCHGCO) to set the OK to leave out commitment option to Y, an
indicator is sent when the agent (TM-A) responds to the initiator (TM-I) with
a request commit message.

Note: Any change to the OK to Leave Out commitment option does not take
effect until the next successful commit operation.
7. When the initiator (TM-I) receives all the votes, the TM-I sends a commit
message. This starts the committed wave.
8. Each agent (TM-A) commits and responds with a reset message.
9. A return is sent to the application program (I) to indicate that the LUW is
complete at the initiator.
10. Any number of LUWs may occur on TM-I, none of which require changes to
TM-A or data from TM-A. TM-A is not included in these LUWs.
11. The next time the initiator (TM-I) issues a message to the agent (A), a new
LUW ID is sent with the message. If the initiator performs any commit or
rollback operations before sending a message to the agent, no messages are
sent to the agent during those operations (the agent is ’left out’ of those
commits or rollbacks). Because one or more LUWs may have been committed
or rolled back at the initiator while the agent was left out, the initiator must
communicate its current LUW ID when the next message is sent to the agent.
12. A return is sent to the application program (A) to indicate that the original
commit is complete and that it is participating in the current LUW.

Changing the Commitment Definition to Allow Vote Read-Only

Normally, a transaction manager participates in both phases of commit processing.
To improve the performance of commit processing, you can set up some or all
locations in an LUW to allow the transaction manager to vote read-only. If the
location has no committable changes during an LUW, the transaction manager
votes read-only during the prepare wave. The location does not participate in the
committed wave. This improves overall performance because the communication
flows that normally occur during the committed wave are eliminated during
LUWs in which no updates are made at one or more remote locations.

After you start commitment control, you can use the Change Commitment Options
(QTNCHGCO) API to change the Vote read-only permitted option to Y. You may
want to do this if the following is true:
v One or more remote systems often do not have any committable changes for an
LUW.
v An LUW does not depend on where the file cursor (next record) was set by the
previous LUW. When a location votes read-only, the application is never notified
if the LUW is rolled back. The location has committed any read operations to the
database files and, thus, moved the cursor position. The position of the file
cursor is usually important only if you do sequential processing.

If your commitment definition is set up to allow vote read-only, the application

waits for the next message flow from another location. For example, in Figure 55
on page 552, program A would not receive the return after issuing the commit
instruction. Program A would wait until another message, either a data flow or a
commitment instruction, was received from system I.

560 OS/400 Backup and Recovery V5R1

The Vote read-only permitted option is intended for applications that are
client/server in nature. If the purpose of program A is only to satisfy requests from
program I, not to do any independent work, it is appropriate to allow the Vote
read-only option for program A.

Figure 60 shows the flow of messages among the application programs and the
transaction managers when an application program issues a commit instruction
without last agent optimization when the agent votes read only. Both the initiator
application program and the agent application programs are unaware of the
two-phase commit processing. The numbers in parentheses () in the figure
correspond to the numbered items in the description that follows.

I TM─I TM─A A
│ │ │ Receive (1) │
│ │ │W──────────────────────┤
│Commit (2) │ │ │
├──────────S│ Prepare (3) │ │
│ ├─────────────────S│ RCVTKCMT indicator (4)│
│ │ ├──────────────────────S│
│ │ │ │
│ │ │ Commit (5) │
│ │ Reset (6) │W──────────────────────┤
│ │W─────────────────┤ │
│ Return (7)│ │ │
│W──────────┤ │ │
│any verb │ (new LUW ID) (8) │ │
├──────────S├─────────────────S│ Return (9) │
│ │ ├──────────────────────S│

Legend
I = Initiator (Application that initiates commit
request)
TM-I = Transaction manager for initiator
A = Agent (Application that receives commit request)
TM-A = Transaction manager for agent

Figure 60. Flow of Commit Processing without Last Agent Optimization when agent votes
read only

Following is a description of the events for normal processing without last agent
optimization when the agent votes read only. This describes a basic flow. The
sequence of events can become much more complex when the transaction program
network has multiple levels or when errors occur.
1. Application program A does a receive request to indicate that it is ready to
receive a request from program I.
2. The initiator application (I) issues a commit instruction.
3. The transaction manager for the initiator (TM-I) takes the role of initiator for
this LUW. It starts the prepare wave by sending a prepare message to all the
other locations that are participating in the LUW.
4. The transaction managers for every other location take the role of agent
(TM-A). The application program A is notified by TM-A that a request to
commit has been received. For ICF files, the notification is in the form of the
Receive Take Commit (RCVTKCMT) ICF indicator being set on.
5. The application program A responds by issuing a commit instruction (or a
rollback instruction). This is the application program’s vote.
6. If application program A has used the Change Commitment Options API
(QTNCHGCO) to set the Vote read-only permitted commitment option to Y,

Chapter 22. Commitment Control 561

 and no changes have been made at the agent during the LUW, the agent
(TM-A) responds to the initiator (TM-I) with a reset message. There will be no
committed wave for the agent.
7. A return is sent to the application program (A) to indicate that the LUW is
complete at agent TM-A.
8. The next time the initiator (TM-I) issues any message to the agent (TM-A),
either a data flow or a commitment instruction, TM-I causes its current LUW
ID to be sent with the message. The reason for this is that a new LUW ID may
have been generated at TM-I if a communications failure had occurred between
TM-I and another system during the commit operation.
9. A return is sent to the application program (A) to indicate that the LUW is
complete at agent TM-A. The return is delayed until after the next message is
received because a new LUW ID must be received from TM-I before the next
LUW can be started by application A.

For more information about the QTNCHGCO API, look in the Information Center
at the following Web site: http://www.ibm.com/eserver/iseries/infocenter.

Vote Reliable Affect on Flow of Commit Processing

Vote reliable is an optimization that is part of the presumed abort level of
commitment control that is supported by OS/400 starting with release V3R7. The
optimization improves performance by returning earlier to the initiator application
after a commit operation and eliminating one message during a commit operation.
| There is no explicit vote reliable optimization for DRDA distributed unit of work
| over TCP/IP. However, OS/400 never requests a reset (forget) confirmation for
| TCP/IP connections. Therefore a reset (forget) is always implied for TCP/IP
| connections.

Vote reliable can be thought of as a promise by an agent to its initiator that no

heuristic decisions will be made at the agent if communications failure occurs
while the agent is in doubt. An agent that is using the vote reliable optimization
sends an indicator to the initiator during the prepare wave of the commit. If the
initiator is also using the vote reliable optimization, it then sends an indicator to
the agent that no reset is required in response to the commit message. This
eliminates the reset message, and allows the transaction manager to return to the
application at the initiator as soon as the commit message is sent.

You should consider using the vote reliable optimization if the following
conditions are true:
v It is unlikely that a heuristic decision will be made at an in doubt agent in the
event of a systems or communications failure unless the failure cannot be
repaired.
v Your program logic does not need the results of previous transactions to ensure
that your database files remain synchronized.

The vote reliable optimization will be used by OS/400 only if all the following
conditions are true:
v The initiator and agent locations support the presumed abort level of
commitment control.
v The initiator location accepts the vote reliable indication from the agent. On
OS/400 initiators, this depends on the value of two commitment options (refer
to the QTNCHGCO API in the System API Reference for details on changing
commitment options):

562 OS/400 Backup and Recovery V5R1

 – The value of the Wait for outcome commitment option must be No (Yes is the
default).
– The value of the Accept vote reliable commitment option must be Yes (Yes is
the default).
v The agent location votes reliable during the prepare wave. OS/400 agents
always vote reliable. This is because heuristic decisions can be made only
through a manual procedure that warns of the possible negative side effects of
making a heuristic decision.

Figure 61 shows the flow of messages among the application programs and the
transaction managers when the vote reliable optimization is used. Both the initiator
application program and the agent application programs are unaware of the
two-phase commit processing. The numbers in parentheses () in the figure
correspond to the numbered items in the description that follows.

I TM─I TM─A A
│ │ │ Receive (1) │
│ │ │W──────────────────────┤
│Commit (2) │ │ │
├──────────S│ Prepare (3) │ │
│ ├─────────────────S│ RCVTKCMT indicator (4)│
│ │ ├──────────────────────S│
│ │ │ │
│ │ │ Commit (5) │
│ │Request Commit (6)│W──────────────────────┤
│ │(vote reliable) │ │
│ │W─────────────────┤ │
│ │ │ │
│ │ Commit (7) │ │
│ │ (no reset) │ │
│ ├─────────────────S│ │
│ Return (8)│ │ Return (8) │
│W──────────┤ ├──────────────────────S│
│ │ │ │
│ │ Implied reset (9)│ any verb │
│ │W─────────────────┤W──────────────────────┤

Legend
I = Initiator (Application that initiates commit
request)
TM-I = Transaction manager for initiator
A = Agent (Application that receives commit request)
TM-A = Transaction manager for agent

Figure 61. Flow of Commit Processing with Vote Reliable optimization

Following is a description of the events for normal processing without last agent
optimization when the agent votes reliable. This describes a basic flow. The
sequence of events can become much more complex when the transaction program
network has multiple levels or when errors occur.
1. Application program A does a receive request to indicate that it is ready to
receive a request from program I.
2. The initiator application (I) issues a commit instruction.
3. The transaction manager for the initiator (TM-I) takes the role of initiator for
this LUW. It starts the prepare wave by sending a prepare message to all the
other locations that are participating in the LUW.
4. The transaction managers for every other location take the role of agent
(TM-A). The application program A is notified by TM-A that a request to

Chapter 22. Commitment Control 563

 commit has been received. For ICF files, the notification is in the form of the
Receive Take Commit (RCVTKCMT) ICF indicator being set on.
5. The application program A responds by issuing a commit instruction (or a
rollback instruction). This is the application program’s vote.
6. The agent (TM-A) responds to the initiator (TM-I) with a request commit
message. For OS/400 systems that support presumed abort, a vote reliable
indicator is sent with the request commit.
7. When the initiator (TM-I) receives all the votes, the TM-I sends a commit
message. If the Wait for outcome commitment option is N (No) and the Accept
vote reliable commitment option is Y (Yes), a no reset indicator is sent with the
commit message. This tells the agent that no reset message is required in
response to the commit.
8. The LUW is complete. A return is sent to the application programs (I and A).
This return indicates that the commit operation was successful. If a heuristic
damage occurs at system A due to a heuristic decision being made before the
committed message is received, application I would not be informed. Instead, a
message would be sent to the QSYSOPR message queue. However, application
A would receive the heuristic damage indication.
9. The next time the agent (TM-A) sends any message to the initiator (TM-I),
either a data flow or a commitment instruction, an implied reset indicator is
sent with the message to inform TM-I that TM-A completed the commit
successfully. The reason for this is that TM-I must retain information about the
completed LUW until it has confirmed that TM-A successfully received the
commit message in step 7.

Retrieving Commitment Control Information

To retrieve commitment control information, you can use the Retrieve Commitment
Information (QTNRCMTI) API. Some of the information you can retrieve is:
v Whether commitment control is active for the activation group performing the
retrieve request.
v The scope of the commitment definition.
v The default lock level and journal associated with this commitment definition.
v The settings for the commitment options.
v The logical unit of work identifier.

For information on the QTNRCMTI API, look in the Information Center at the
following Web site: http://www.ibm.com/eserver/iseries/infocenter.

States of the Logical Unit of Work (LUW)

A commitment definition is established at each location that is part of the
transaction program network. For each commitment definition, the system keeps
track of the state of its current LUW and previous LUW. The system uses the state
to decide whether to commit or roll back if an LUW is interrupted. If multiple
locations are participating in an LUW, the states of the LUWs at each location may
be compared to determine the correct action (commit or rollback). This process of
communicating between locations to determine the correct action is called
resynchronization.

Table 79 on page 565 shows:

v The basic states that may occur during an LUW.
v Additional states that may occur.

564 OS/400 Backup and Recovery V5R1

 v Whether a state requires resynchronization if the LUW is interrupted. The
possible values are:
Not needed
Each location can make the correct decision independently.
May be necessary
Each location can make the correct decision, but the initiator may need
to be informed of the decision.
Required
The state of each location must be determined before the correct decision
can be made.
v Action Taken
Table 79. Logical Unit of Work (LUW) States
Resynchronization if
State Name Description LUW Is Interrupted Action Taken
Basic states during two-phase commit processing:
Reset (RST) From the commitment boundary until a Not needed. Pending changes are rolled back.
program issues a request to commit or roll
back.
Prepare in The initiator has started the prepare wave. May be necessary. Pending changes are rolled back.
Progress All locations have not yet voted.
(PIP)
Prepared This location and all locations further down Required. In doubt. Depends on the results
(PRP) in the transaction program network have of the resynchronization process.
voted to commit. This location has not yet
received notification from the initiator to
commit.
Commit in All locations have voted to commit. The Required. Pending changes are committed.
Progress initiator has started the committed wave. Resynchronization is performed
(CIP) to ensure that all locations have
committed. If a heuristic rollback
is reported by another location,
an error is reported.
Committed All agents have committed and returned a May be necessary. None.
(CMT) reply to this node.
Additional states during two-phase commit processing:
Last Agent If a last agent is selected, this state occurs at Required In doubt. Depends on the results
Pending the initiator between the PIP state and the of the resynchronization process.
(LAP) CIP state. The initiator has instructed the last
agent to commit and has not yet received a
response.
Vote-Read- This agent responded to the prepare wave May be necessary. None.
Only (VRO) by indicating that it has no pending changes.
If the vote-read-only state is permitted, this
agent is not included in the committed wave.
Rollback One of the following occurred: May be necessary. Pending changes are rolled back.
Required v An agent issued a rollback request before
(RBR) the commit operation.
v A transaction failure has occurred.
v The QTNRBRQD API was used to place
the LUW in a rollback required state.
The transaction program is not allowed to
perform any additional changes under
commitment control.
Conditions that occur because of operator actions or errors:

Chapter 22. Commitment Control 565

Table 79. Logical Unit of Work (LUW) States (continued)
Resynchronization if
State Name Description LUW Is Interrupted Action Taken
Forced This location and all locations further down May be necessary Pending changes have already
Rollback the transaction program network, except the been rolled back.
last agent, have been rolled back through
operator intervention.
Forced This location and all locations further down May be necessary Pending changes have already
Commit the transaction program network, except the been committed.
last agent, have committed through operator
intervention.
Heuristic Some resource managers have committed. May be necessary User intervention is required.
Mixed Some have rolled back. Operator
(HRM) intervention was used or a system error
occurred. Heuristic mixed does not appear as
a status on the commitment definition
displays. Notification messages are sent to
the operator.

XA Transaction Support
DB2 UDB for iSeries can participate in X/Open global transactions. The Open
Group has defined an industry standard model for transactional work that allows
changes made against unrelated resources to be part of a single global transaction.
An example of this would be changes to databases that are provided by two
separate vendors. This model is called the Distributed Transaction Processing
(DTP) model. This model is described in detail in X/Open Publication Set T008,
ISBN 1–872630–89–8. You should be familiar with the information in these books,
particularly the XA Specification, before attempting to use the XA transaction
support provided by DB2 UDB for iSeries.

There are five components to the DTP model:

Application Program (AP)
Implements the desired function of the end-user by specifying a sequence
of operations that involves resources such as databases. It defines the start
and end of global transactions, accesses resources within transaction
boundaries, and normally makes the decision whether to commit or roll
back each transaction.
Transaction Manager (TM)
Manages global transactions and coordinates the decision to start them,
and commit them or roll them back in order to ensure atomic transaction
completion. The TM also coordinates recovery activities with the RMs after
a component fails.
Resource Manager (RM)
Manages a defined part of the computer’s shared resources, such as a
database management system. The AP uses interfaces defined by each RM
to interfaces provided by the RM to carry out transaction completion.
Communications Resource Manager (CRM)
Allows an instance of the model to access another instance either inside or
outside the current TM domain. CRMs are outside the scope of DB2 UDB
for iSeries and are not discussed here.

566 OS/400 Backup and Recovery V5R1

 Communication Protocol
The protocols used by CRMs to communicate with each other. This is
outside the scope of DB2 UDB for iSeries and is not discussed here.

The XA Specification is the part of the DTP model that describes a set of interfaces
that is used by the TM and RM components of the DTP model. DB2 UDB for
iSeries implements these interfaces as a set of UNIX style APIs and exit programs.
Look in the Information Center at the following Web site:
http://www.ibm.com/eserver/iseries/infocenter for detailed documentation of
these APIs and for more information on how to use DB2 UDB for iSeries as an RM.

The following considerations and restrictions should be understood before using

DB2 UDB for iSeries as a RM. The term ″thread″ refers to either a job that is not
thread-capable, or a single thread within a thread-capable job.
v XA transactions may only be performed in jobs that are running in SQL server
mode. One effect of this is that applications are limited to SQL interfaces when
making changes to DB2 UDB for iSeries during an XA transaction. If the
db2xa_open() API is used in a job that is not already running in SQL server
mode, SQL server mode is implicitly started. Refer to “Thread Scoped
Transactions” on page 535 for more information on SQL server mode.
v In order to use a Structured Query Language (SQL) connection for XA
transactions, you must use the db2xa_open() application program interface (API)
before the SQL connection is made. The relational database that will be
connected to must be passed to the db2xa_open() API by the Xainfo parameter.
The user profile and password to be used in the job that the connection is routed
to may be passed to the db2xa_open() API. If it is not passed, the profile will
default to the one that was specified or defaulted during the connection attempt.
v If the CLI is used to perform XA transactions, more than one connection may be
made in the same thread after the db2xa_open() API is used. The connections
may be used in other threads to perform XA transactions, as long as those other
threads first use the db2xa_open() API with the same Xainfo parameter value.
v If embedded SQL is used to perform XA transactions, the work performed for
each connection is routed to a different job, even if the connections are made in
the same thread. This is different than SQL server mode without XA, where
work performed for all connections in a single thread is routed to the same job.
This is because the XA specification requires a separate prepare, commit or
rollback invocation for each resource manager instance.
| v XA connections to a remote relational database are supported only if the
| relational database resides on a system that supports Distributed Unit of Work
| (DUW) DRDA connections. This includes systems using V3R1 or later releases
| when running DRDA over SNA LU6.2 conversations. This includes systems
| using V5R1 when running DRDA using TCP/IP connections.
v Before using the XA join function, the db2xa_open() API must be used in the
joining thread. The same relational database name and RMID must be specified
on the db2xa_open() API in both the thread that started the transaction branch
and the joining thread. If the transaction branch is active when a join is
attempted, the joining thread is blocked. The joining thread remains blocked
until the active thread suspends or ends its association with the transaction
branch.
| v If the CLI is used to perform XA transactions, the connection that is used to start
| a transaction branch must be used for all work on that transaction branch. If
| another thread is to join the transaction branch, the connection handle for the
| connection used to start the transaction branch must be passed to the joining

Chapter 22. Commitment Control 567

| thread so that it can perform work over that same connection. Likewise, if a
| thread is to resume the transaction branch, the same connection must be used.
Since CLI connection handles can not be used in a different job, the join function
is limited to threads running in the same job that started the transaction branch
when the CLI is used.
v If embedded SQL is used to perform XA transactions, only one connection per
relational database can be made per thread. Whenever the thread is not actively
associated with a transaction branch, work requested over one of the thread’s
connections will cause the RM to use the TM’s ax_reg() exit program to
determine whether the work is to start, resume or join a transaction branch.
If the work is to start a transaction branch, it is performed over that thread’s
connection to the corresponding relational database.
If the work is to join a transaction branch, it is rerouted over the connection to
the corresponding relational database that was made in the thread that started
the transaction branch. Note that the system does not enforce that the user
profile for that connection is the same as the one for the connection of the
joining thread. The TM is responsible to ensure that this is not a security
concern. Typical TMs use the same user profile for all connections. This user
profile is authorized to all data that is managed by the TM. Further security of
access to this data is managed by the TM or AP instead of using the standard
iSeries security mechanisms.
If the work is to resume a transaction branch, the connection that is used
depends on whether the suspended transaction branch association was
established by starting or joining the transaction branch.
Subsequent work is performed over the same connection until the db2xa_end()
API is used to suspend or end the thread’s association with that transaction
branch.
| v If an association between a thread and an existing transaction branch is
| suspended or ended using the db2xa_end() API, the thread may start a new
| transaction branch. If the connection used to start the new transaction branch
| was used earlier to start a different transaction branch, and the thread’s
| association with that transaction branch has been ended or suspended by the
| db2xa_end() API, a new SQL server job may be started. A new SQL server job is
| needed only if the first transaction branch has not yet been completed by the
| db2xa_commit() or db2xa_rollback() API. In this case, another completion
| message SQL7908 is sent to the joblog identifying the new SQL server job, just
| as the connection’s original SQL server job was identified when the connection
| was established. All SQL requests for the new transaction branch are routed to
| the new SQL server job. When the transaction branch is completed by the
| db2xa_commit() or db2xa_rollback() API, the new SQL server job is recycled and
| returned to the prestart job pool.
v Any errors that are detected by DB2 UDB for iSeries during the XA API
invocations are reported through return codes according to the XA specification.
Diagnostic messages are left in the job log when the meaning of the error may
not be clear from the return code alone.
v Information about XA transaction branches is shown as part of the commitment
control information displayed by the Work with Job (WRKJOB), Display Job
(DSPJOB), and Work with Commitment Definition (WRKCMTDFN) commands.
The TM name, transaction branch state, transaction identifier and branch
qualifier are all shown. The commitment definitions related to all currently
active XA transactions can be displayed by using the command WRKCMTDFN
JOB(*ALL) STATUS(*XOPEN).

568 OS/400 Backup and Recovery V5R1

 v The manual heuristic commit and rollback support that is provided for all
commitment definitions can be used if it becomes necessary to force a
transaction branch to commit or roll back while it is in a prepared state. Refer to
“Making Heuristic Decisions and Cancelling Resynchronization” on page 570 for
details.
v A transaction branch will be marked Rollback Only by the system when the
following situations occur:
– A thread ends when it is still associated with the transaction branch.
– The db2xa_close() API is used in a thread that has an active association with
the transaction branch.
v A transaction branch will be rolled back by the system if any threads are still
associated with it when any of the following situations occur:
– The connection that is related to the transaction branch is ended.
– The job that started the transaction branch is ended.
– The system fails.
v There is one situation where a transaction branch will be rolled back by the
system, regardless of whether there are still associated threads. This occurs when
the SQL server job that the connection’s work is being routed to is ended. This
can only happen when the End Job (ENDJOB) CL command is used against that
job.
v A transaction branch will not be affected if no threads have an active association
with it when any of the following situations occur. The TM may commit or roll
back the transaction branch from any thread that has used the db2xa_open() API
with the same Xainfo parameter value that was specified in the thread that
started the transaction branch.
– The connection that is related to the transaction branch is ended.
– A thread or job that performed work for the transaction branch, but not
longer has an active association with it, uses the db2xa_close() API.
– A thread or job that performed work for the transaction branch, but no longer
has an active association with it, uses the db2xa_close() API.
– The system fails. In this case, the transaction branch is not affected only if it is
in prepared state. If it is in idle state, the system rolls it back.

Commitment Control Recovery during Initial Program Load

When you perform an IPL after your system ends abnormally, the system attempts
to recover all the commitment definitions that were active when the system ended.
The recovery is performed by database server jobs that are started by the system
during IPL. Database server jobs are started by the system to handle work that
cannot or should not be performed by other jobs.

The database server jobs are named QDBSRVnn, where nn is a two-digit number.
The number of database server jobs depends on the size of your system.

Table 79 on page 565 shows the actions that the system takes, depending on the
state of the LUW when the system ended. For two states, PRP and LAP, the system
action is in doubt. The system cannot determine what to do until it performs
resynchronization with the other locations that participated in the LUW. This
resynchronization is performed after the IPL completes.

The system uses the database server jobs to perform this resynchronization. The
commitment definitions that need to be recovered are associated with the database
server jobs. During the IPL, the system acquires all record locks and other object
locks that were held by the commitment definition before the system ended. These

Chapter 22. Commitment Control 569

 locks are necessary to protect the local commitment resources until
resynchronization is complete and the resources can be committed or rolled back.

Messages are sent to the job logs of the database server jobs to indicate the status
of resynchronization with the remote locations. If the LUW is in doubt,
resynchronization must be completed with the location that owns the decision for
the LUW before local resources can be committed or rolled back.

When the decision for an LUW is made, the following messages may be sent to the
job log for the database server job.
CPI8351
&1 pending changes being rolled back
CPC8355
Post-IPL recovery of commitment definition &8 for job &19/&18/&17
completed.
CPD835F
IPL recovery of commitment definition &8 for job &19/&18/&17 failed.

Other messages related to the recovery may also be sent. These messages are sent
to the history (QHST) log. If errors occur, messages are also sent to the QSYSOPR
message queue.

You can determine the progress of the recovery by displaying the job log for the
database server job or by using the Work with Commitment Definitions
(WRKCMTDFN) command. Although the Work with Commitment Definitions
display allows you to force the system to commit or roll back, you should use this
only as a last resort. If you anticipate that all of the locations that participated in
the LUW will eventually be returned to operation, you should allow the systems to
resynchronize themselves. This ensures the integrity of your databases.

Making Heuristic Decisions and Cancelling Resynchronization

A heuristic decision is the action that you take when you force the system to
commit or to roll back an LUW. When you make a heuristic decision, the state of
the LUW becomes heuristic mixed (HRM) if your decision is inconsistent with the
results of the other locations in the LUW. You must take responsibility for
determining the action taken by all the other locations that participated in the
LUW and resynchronizing the database records.

Before you make a heuristic decision, gather as much information as you can about
the LUW. Display the jobs that are associated with the commitment definition and
make a record of what journals and files are involved. You can use this information
later if you need to display journal entries and apply or remove journaled changes
manually.

The best place to find out information about an LUW is to look at the location that
was the initiator for that LUW. However, the decision to commit or to roll back
may be owned by an API resource or by a last agent.

If an API resource was registered as a last agent resource, the final decision of
whether to commit or to roll back is owned by the API resource. You need to look
at information about the application and how it uses the API resource to determine
whether to commit or to roll back.

570 OS/400 Backup and Recovery V5R1

If the LUW has a last agent selected, the last agent owns the decision to commit or
rollback. You should look at the status of the last agent for information about the
LUW.

When you must make heuristic decisions or cancel resynchronization due to a

system or communications failure that cannot be repaired, you can find all LUWs
in doubt by using the following command:
WRKCMTDFN JOB(*ALL) STATUS(*UNDECIDED)

To find all LUWs being resynchronized, use the following command:

WRKCMTDFN JOB(*ALL) STATUS(*RESYNC)

To view a listing of the commitment definitions, use the following command:

WRKCMTDFN JOB(*ALL) STATUS(*ALL)

A list of commitment definitions is displayed showing the resynchronization status,

the current LUW ID, and the current LUW state for each commitment definition
(use F11 to display all of the fields).
Work with Commitment Definitions
System: RCXBSL22
Type options, press Enter.
5=Display status 12=Work with job 14=Forced commit
16=Forced rollback ...

Commitment Resync In
Opt Definition Job User Number Progress
__ *DFTACTGRP QPADEV0013 XPF9473 100192 NO
__ *DFTACTGRP RBMCFGS1 XPF9476 040291 YES
__ *JOB RBMCFGS1 XPF9465 040291 NO
__ RM400 RBMCFGS1 XPF9512 040291 NO
__ *DFTACTGRP RCHASR7D QUSER 097909 NO
__ *DFTACTGRP R214PAC180 $TNTESTUSR 098549 YES

You can use option 12 to work with the job that is participating in the LUW on
this system. You can choose options 14, 16, and 19 to force commit, force rollback,
or cancel resynchronization, respectively.

Before making a heuristic decision or cancelling resynchronization, you might want

to check the status of the jobs on other systems associated with the LUW. Checking
the jobs on remote systems might help you avoid decisions that cause database
inconsistencies between systems.

To determine the other systems associated with the commitment definition, use
option 5 on the Work with Commitment Definitions display to see the commitment
definition status.

Chapter 22. Commitment Control 571

 Display Commitment Definition Status RCXBSL22
mm/dd/yy hh:mm:ss
Job: RBMCFGS1 User: XPF9473 Number: 040291

Commitment definition . . . . . . : *DFTACTGRP

Activation group . . . . . . . . : 2

Logical Unit of Work ID . . . . . : APPN.RCHASL97.X'112233445566'.00001

Job active . . . . . . . . . . . : YES
Server job . . . . . . . . . . . :
Resource location . . . . . . . . : BOTH
Default lock level . . . . . . . : *ALL
Role . . . . . . . . . . . . . . : AGENT
State . . . . . . . . . . . . . . : PREPARED
Date/time stamp . . . . . . . . : mm/dd/yy hh:mm:ss
Resync in progress . . . . . . . : YES
Number of commits . . . . . . . . : 0
Number of rollbacks . . . . . . . : 0
More...
Press Enter to continue.

F3=Exit F5=Refresh F6=Display resource status F9=Command line

F12=Cancel F14=Previous F15=Next F22=Display entire field

From the Display Commitment Definition Status display, you can use F6 to display
a pop-up menu for selecting resources.
Display Commitment Definition Status

Job: RBMCFGS1 User: XZY0019 Number: 040291

Commitment definition . . . . . . : *DFTACTGRP

Activation group . . . . . . . . : 2
------------------------------------------------------------------
: Display Resource Status :
: :
: Type option, press Enter. :
: 1=Select :
: :
: Opt Resource :
: __ Record level :
: __ Object level :
: 1 Conversation :
: __ TCP/IP Connection :
: __ Remote file :
: __ RDB :
: __ API :

Type a 1 in the option column next to Conversation to display information about

the conversation resources associated with the commitment definition.

572 OS/400 Backup and Recovery V5R1

 Display Conversation Status
System: RCHASL97
Job: RBMCFGS1 User: XZY0019 Number: 040291

Commitment definition . . . . . . : *DFTACTGRP

Type options, press Enter.

5=Display details 19=Cancel resync

Remote
Remote Local Network Resync
Opt Location Location Identifier Role Required
__ RMTLOC1 LCLLOC1 APPN INITIATOR YES
__ RMTLOC2 LCLLOC2 APPN AGENT NO

Each conversation resource represents a remote system that is participating in the

LUW. On the remote systems, the Work with Commitment Definitions command
can be used to find the definitions associated with the LUW. Because the base
portion of the LUW ID is the same on all the systems, you can use the following
command to find the commitment definition on the other systems:
WRKCMTDFN JOB(*ALL) LUWID('luwid_base_portion*')

where luwid_base_portion is the LU name and instance number portion of the LUW
ID. For example, to view the commitment definitions related to the *DFTACTGRP
commitment definition for job 040291/XZY0019/RBMCFGS1 (displayed above), use the
following command:
WRKCMTDFN JOB(*ALL)
LUWID('APPN.RCHASL97.X'112233445566'.*')

Implicit Commit and Rollback Operations

Usually, a commit or rollback operation is initiated from an application program
using one of the available programming languages that supports commitment
control. These types of commit and rollback operations are known as explicit
commit and rollback requests. However, in some instances the system initiates a
commit or rollback operation for a commitment definition. Commit and rollback
operations initiated by the system are known as implicit commit and rollback
requests.

Table 80 on page 574 and Table 81 on page 576 show what the system does when
certain events occur related to a commitment definition that has pending changes.
A commitment definition has pending changes if any of the following is true:
v Any committable resource has been updated.
v A database file opened under commitment control has been read because
reading a file changes the file position.
v The commitment definition has an API resource. Because changes to API
resources are done by a user program, the system must assume that all API
resources have pending changes.

The ’C CM’ (commit operation) journal entry and ’C RB’ (rollback operation)
journal entry indicate whether the operation was explicit or implicit.

Table 80 on page 574 shows the actions the system takes when a job ends, either
normally or abnormally, based on the following:

Chapter 22. Commitment Control 573

 v The state of the LUW.
v The action-if-end job value for the commitment definition.
v Whether an API resource is the last agent
Table 80. Commit and Rollback Operations During Job End
Action If Endjob1
State Last Agent API? Option Commit Rollback Operation
RST N/A N/A If the commitment definition is not associated with an
X/Open global transaction, an implicit rollback is performed.

If the commitment definition is associated with an X/Open

global transaction, the following occurs:
| v If the transaction branch state is not Active (S1), no action
| is performed and the transaction branch is left in the same
| state.
| v If the transaction branch state is Active (S1), an implicit
| rollback is performed.
PIP N/A N/A If the commitment definition is not associated with an
X/Open global transaction, an implicit rollback is performed.

If the commitment definition is associated with an X/Open

global transaction, the transaction branch is in the Idle (S2)
state, and it is left in the Idle (S2) state.
PRP N/A WAIT If the commitment definition is not associated with an
X/Open global transaction, the following occurs:
1. Resynchronization is started to receive the decision from
the initiator of the commit operation.
2. The returned decision to commit or rollback is performed.
It is considered an explicit operation.
If the commitment definition is associated with an X/Open
global transaction, the following occurs:
v If the job that started the transaction ends, the transaction
is left in a prepared state until the XA TM either commits it
or rolls it back. The XA transaction branch state will be left
at Prepared (S3) in this case.
v If the SQL server job that the transaction’s work is being
routed to is ended, a forced rollback is implicitly
performed. The XA transaction branch state will be
changed to Heuristically Completed (S5) in this case.

574 OS/400 Backup and Recovery V5R1

Table 80. Commit and Rollback Operations During Job End (continued)
Action If Endjob1
State Last Agent API? Option Commit Rollback Operation
PRP N/A C If the commitment definition is not associated with an
X/Open global transaction, an implicit commit operation is
performed.

If the commitment definition is associated with an X/Open

global transaction, the following occurs:
v If the job that started the transaction ends, the transaction
is left in a prepared state until the XA TM either commits it
or rolls it back. The XA transaction branch state will be left
at Prepared (S3) in this case.
v If the SQL server job that the transaction’s work is being
routed to is ended, a forced rollback is implicitly
performed. The XA transaction branch state will be
changed to Heuristically Completed (S5) in this case.
R If the commitment definition is not associated with an
X/Open global transaction, an implicit rollback operation is
performed.

If the commitment definition is associated with an X/Open

global transaction, the following occurs:
v If the job that started the transaction ends, the transaction
is left in a prepared state until the XA TM either commits it
or rolls it back. The XA transaction branch state will be left
at Prepared (S3) in this case.
v If the SQL server job that the transaction’s work is being
routed to is ended, a forced rollback is implicitly
performed. The XA transaction branch state will be
changed to Heuristically Completed (S5) in this case.
CIP N/A N/A An explicit commit operation is performed.
LAP NO WAIT 1. Resynchronization to the last agent is used to retrieve the
decision to commit or to roll back.
2. The returned decision to commit or to roll back is
performed. It is considered an explicit operation.
LAP YES WAIT 1. The last agent API is called to retrieve the commit or
rollback decision.
2. The commit or rollback operation is performed. It is
considered an explicit operation.
LAP N/A C An implicit commit operation is performed.
R An implicit rollback operation is performed.
CMT N/A N/A A commit operation has already completed for this
commitment definition and any downstream locations. The
commit operation is complete.
VRO N/A N/A The local and remote agents voted to read only. All
downstream agents must also have voted to read only. No
action is required.
RBR N/A N/A A rollback operation is required. An explicit rollback
operation is performed.

1 The Action if Endjob option can be changed with the QTNCHGCO API. For more information, refer to the
Information Center at the following Web site: http://www.ibm.com/eserver/iseries/infocenter.

Table 81 on page 576 shows the actions the system takes when an activation group
ends. The system actions are based on the following:
v The state of the LUW. (It is always reset (RST) when an activation group ends).

Chapter 22. Commitment Control 575

 v How the activation group ends–normally or abnormally.
v Whether an API resource is the last agent.

Note: If an API resource is registered as the last agent, this gives control of the
commit or rollback decision to the last agent. The result of the decision is
considered an explicit operation.
Table 81. Commit and Rollback Operations during Activation Group Ending
Last Agent
State API? Type of End Commit or Rollback Operation
RST No Normal An implicit commit operation is performed. If protected
conversations exist, the commitment definition will become the
root initiator of the commit operation.
RST No Abnormal An implicit rollback is performed.
RST Yes Normal The API exit program is called. The commit or rollback operation
is determined by the API.
RST Yes Abnormal The API exit program is called. The commit or rollback operation
is determined by the API.

Commitment Control Status

You can use the Work with Commitment Definition (WRKCMTDFN) command to
display information about all the commitment definitions on the system:
Work with Commitment Definitions
System: RCHASL97
Type options, press Enter.
5=Display status 12=Work with job 14=Forced commit
16=Forced rollback ...

Commitment Resync In
Opt Definition Job User Number Progress
__ *DFTACTGRP RBMCFGS1 XZY0019 040291 NO
__ *JOB RBMCFGS1 XZY0019 040291 NO
__ RM400 RBMCFGS1 XZY0019 040291 NO
__ *JOB JCHCFGS2 JCH0418 040763 NO
__ *JOB RAJCFGS3 RAJ0692 040814 NO

From the Work with Commitment Definitions display, you can look at information
about a specific commitment definition. You can also look at information about the
job associated with a commitment definition.

Another way to look at commitment definitions is by using the Work with Job
(WRKJOB command) or Display Job (DSPJOB command). From the Display Job
menu, do the following:
1. Select the Display Commitment Control Status option. A list of commitment
definitions currently established for the job is shown.
2. Select the Display option for a commitment definition. The Display
Commitment Definition Status display is shown:

576 OS/400 Backup and Recovery V5R1

 Display Commitment Definition Status RCHASL97
mm/dd/yy hh:mm:ss
Job: RBMCFGS1 User: XZY0019 Number: 038777

Commitment definition . . . . . . : QNSCRCDFNM

Activation group . . . . . . . . : 2

Logical Unit of Work ID . . . . . : APPN.RCHASL97.X'112233445566'.00001

Job active . . . . . . . . . . . : YES
Server job . . . . . . . . . . . :
Resource location . . . . . . . . : BOTH
Default lock level . . . . . . . : *ALL
Role . . . . . . . . . . . . . . : CASCADER
State . . . . . . . . . . . . . . : PREPARED
Date/time stamp . . . . . . . . : mm/dd/yy hh:mm:ss
Resync in progress . . . . . . . : YES
Number of commits . . . . . . . . : 0
Number of rollbacks . . . . . . . : 0
More...
Press Enter to continue.

F3=Exit F5=Refresh F6=Display resource status F9=Command line

F12=Cancel F14=Previous F15=Next F22=Display entire field

Note: If this commitment definition is associated with an X/Open global

transaction, the TM name, transaction branch state, and XID will be shown
before the Logical Unit of Work ID field.

You can page forward to see additional information about the commitment
definition:
Display Commitment Definition Status

Job: QNSCRMON User: QSVSM Number: 048193

Commitment definition . . . . . . : QNSCRCDFNM

Activation group . . . . . . . . :

Heuristic operation . . . . . . . :
Default journal . . . . . . . . . :
Library . . . . . . . . . . . . :
Notify object . . . . . . . . . . : *NONE
Library . . . . . . . . . . . . :
Object type . . . . . . . . . . :
Member . . . . . . . . . . . . :

You can use F6 to display a pop-up menu for selecting resources:

Chapter 22. Commitment Control 577

 Display Commitment Definition Status

Job: QNSCRMON User: QSVSM Number: 048193

Commitment definition . . . . . . : QNSCRCDFNM

Activation group . . . . . . . . :
------------------------------------------------------------------
: Display Resource Status :
: :
: Type option, press Enter. :
: 1=Select :
: :
: Opt Resource :
: 1 Record level :
: __ Object level :
: __ Conversation :
: __ TCP/IP Connection :
: __ Remote file :
: __ RDB :
: __ API :

You can type a 1 in the option column to see more information about any
resources of that type that are associated with the commitment definition.
Following is an example of a display for resources:
Display Record Level Status
System: RCHASDP4
Job: QNSCRMON User: QSVSM Number: 048193

Commitment definition . . . . . . . . : QNSCRCDFNM

-------------Changes--------------
File Library Member Commit Rollback Pending
QANSCRAC QSMU QANSCRAC 0 0 0
QANSCRAN QSMU QANSCRAN 0 0 0
QANSCRA1 QSMU QANSCRA1 0 0 0
QANSCRA2 QSMU QANSCRA2 0 0 0
QANSCRA3 QSMU QANSCRA3 0 0 0
QANSCRCN QSMU QANSCRCN 0 0 0
QANSCRCR QSMU QANSCRCR 0 0 0
QANSCRC1 QSMU QANSCRC1 0 0 0

On-line help provides information about all the status displays and the fields on
each display.

On the Work with Active Jobs (WRKACTJOB) display, the Status column may
show a value of CMTW (commit wait) for one or more jobs. This value indicates
that the job is being prevented from moving beyond a commitment boundary due
to a save-while-active operation. During a save-while-active operation, the system
delays every job with committable resources in the save request at a commitment
boundary until those objects have reached a checkpoint. This allows the
save-while-active operation to save the objects at commitment boundaries so that
no partial LUWs are saved to the save media. The COMMIT and ROLLBACK
values are shown on the WRKACTJOB Function field during a commit or rollback.
If the Function remains COMMIT or ROLLBACK for a long time, one of the
following situations might have occurred:
v A resource failure during the commit or rollback requires resynchronization.
Control will not return to the application until the resynchronization completes
or is cancelled.
v This system voted read-only during the commit. Control will not return to the
application until the system that initiated the commit sends data to this system.

578 OS/400 Backup and Recovery V5R1

 v This system voted ok to leave out during the commit. Control will not return to
the application until the system that initiated the commit sends data to this
system.

End Commitment Control (ENDCMTCTL) Command

Commitment control may be ended for either the job-level or activation-group-
level commitment definition using the End Commitment Control (ENDCMTCTL)
command. Issuing the ENDCMTCTL command indicates to the system that the
commitment definition in use by the program making the request is to be ended.
The ENDCMTCTL command ends only one commitment definition for the job and
all other commitment definitions for the job remain unchanged.

If the activation-group-level commitment definition is ended, then programs

running within that activation group can no longer make changes under
commitment control, unless the job-level commitment definition is already started
for the job. If the job-level commitment definition is active, then it is immediately
made available for use by the programs running within the activation group that
just ended commitment control.

If the job-level commitment definition is ended, then any program running within
the job that was using the job-level commitment definition can no longer make
changes under commitment control without first starting commitment control
again with the STRCMTCTL command.

Before issuing the ENDCMTCTL command, the following must be satisfied for the
commitment definition to be ended:
v All files opened under commitment control for the commitment definition to be
ended must first be closed. When ending the job-level commitment definition,
this includes all files opened under commitment control by any program
running in any activation group that is using the job-level commitment
definition.
v All API commitment resources for the commitment definition to be ended must
first be removed using the QTNRMVCR API. When ending the job-level
commitment definition, this includes all API commitment resources added by
any program running in any activation group that is using the job-level
commitment definition.
v A remote database associated with the commitment definition to be ended must
be disconnected.
v All protected conversations associated with the commitment definition should be
ended normally using the correct synchronization level.

If commitment control is being ended in an interactive job and one or more

committable resources associated with the commitment definition have pending
changes, inquiry message CPA8350 is sent to the user asking whether to commit
the pending changes, roll back the pending changes, or cancel the ENDCMTCTL
request.

If commitment control is being ended in a batch job, and one or more closed files
associated with the commitment definition to be ended still have pending changes,
the changes are rolled back and a message is sent:
v CPF8356 if only local resources are registered
v CPF835C if only remote resources are registered
v CPF83E4 if both local and remote resources are registered

Chapter 22. Commitment Control 579

 If a notify object is defined for the commitment definition being ended it may be
updated. See ″Updates Made to the Notify Object″ on page 4-13 for more
information regarding the updating of a notify object by the system.

When an activation group that has an API registered as the last agent is ending,
the exit program for the API is called to receive the commit or rollback decision. In
this case, even though the activation group is ending normally, a rollback request
could still be returned from the API exit program. Thus, the implicit commit
operation might not be performed.

After the commitment definition has successfully ended, all the necessary recovery,
if any, has been performed. No additional recovery is performed for the
commitment resources associated with the commitment definition just ended.

After the commitment definition is ended, the job-level or activation-group-level

commitment definition can then be started again for the programs running within
the activation group. The job-level commitment definition may be started only if it
is not already started for the job.

Although commitment definitions can be started and ended many times by the
programs that run within an activation group, the amount of system resources
required for the repeated start and end operations can cause a decrease in job
performance and overall system performance. Therefore, it is recommended that a
commitment definition be left active if a program to be called later will use it.

Commitment Control during Activation Group End

The system automatically ends an activation-group-level commitment definition
when an activation group ends. If pending changes exist for an
activation-group-level commitment definition and the activation group is ending
normally, the system performs an implicit commit operation for the commitment
definition before it is ended. Otherwise, an implicit rollback operation is performed
for the activation-group-level commitment definition before being ended if the
activation group is ending abnormally, or if errors were encountered by the system
when closing any files opened under commitment control scoped to the activation
group.

NOTE
An implicit commit or rollback operation is never performed during
activation-group end processing for the *JOB or *DFTACTGRP commitment
definitions. This is because the *JOB and *DFTACTGRP commitment
definitions are never ended due to an activation group ending. Instead, these
commitment definitions are either explicitly ended with an ENDCMTCTL
command or ended by the system when the job ends.

The system automatically closes any files scoped to the activation group when the
activation group ends. This includes any database files scoped to the activation
group opened under commitment control. The close for any such file occurs before
any implicit commit operation that may be performed for the activation-group-
level commitment definition. Therefore, any records that reside in an I/O buffer
are first forced to the database before any implicit commit operation is performed.

As part of the implicit commit or rollback operation that may be performed, a call
is made to the API commit and rollback exit program for each API commitment

580 OS/400 Backup and Recovery V5R1

 resource associated with the activation-group-level commitment definition. The exit
program must complete its processing within 5 minutes. After the API commit and
rollback exit program is called, the system automatically removes the API
commitment resource.

If an implicit rollback operation is performed for a commitment definition that is

being ended due to an activation group being ended, then the notify object, if one
is defined for the commitment definition, may be updated. See “Updates Made to
the Notify Object” on page 531 for more information regarding the updating of a
notify object by the system.

Commitment Control during Normal Routing Step End

The system ends all commitment definitions for a job when a routing step is
normally ended. A routing step ends normally by one of the following:
v A normal end for a batch job.
v A normal sign-off for an interactive job.
v The Reroute Job (RRTJOB), Transfer Job (TFRJOB), or Transfer Batch Job
(TFRBCHJOB) command ends the current routing step and starts a new routing
step.

Any other end of a routing step is considered abnormal and is recognized by a

nonzero completion code in the job completion message (CPF1164) in the job log.

Prior to ending a commitment definition during routing step end, the system
performs an implicit rollback operation if the commitment definition has pending
changes. This includes calling the API commit and rollback exit program for each
API commitment resource associated with the commitment definition. The exit
program must complete its processing within 5 minutes. After the API commit and
rollback exit program is called, the system automatically removes the API
commitment resource.

If a notify object is defined for the commitment definition, it may be updated. See
“Updates Made to the Notify Object” on page 531 for more information about the
updating of a notify object by the system.

Commitment Control during Abnormal System or Job End

The system ends all commitment definitions for a job when the job ends
abnormally. These commitment definitions are ended during the end job
processing. If the system ends abnormally, the system ends all commitment
definitions that were started and being used by all active jobs at the time of the
abnormal system end. These commitment definitions are ended as part of the
database recovery processing that is performed during the next IPL after the
abnormal system end.

The system performs the following for commitment definitions being ended during
an abnormal job end or during the next IPL after an abnormal system end:
v Prior to ending a commitment definition, the system performs an implicit
rollback operation if the commitment definition has pending changes, unless
processing for the commitment definition was interrupted in the middle of a
commit operation. If ended in the middle of a commit operation, the LUW may
be rolled back, resynchronized, or committed, depending on its state. See
Table 80 on page 574 and Table 81 on page 576. The processing to perform the
implicit rollback operation or to complete the commit operation includes calling

Chapter 22. Commitment Control 581

 the API commit and rollback exit program for each API commitment resource
associated with the commitment definition. After the API commit and rollback
exit program is called, the system automatically removes the API commitment
resource.

ATTENTION!
1. Ending the job while an LUW is in doubt (LUW state is LAP or PRP)
can cause inconsistencies in the database (changes might be committed
on one or more systems and rolled back on other systems).
– If the Action if Endjob commitment option is COMMIT, changes on
this system are committed if the job is ended, without regard to
whether changes on the the other systems participating in the
transaction are committed or rolled back.
– If the Action if Endjob commitment option is ROLLBACK, changes on
this system are rolled back if the job is ended, without regard to
whether changes on the other systems participating in the transaction
are committed or rolled back.
– If the Action if Endjob commitment option is WAIT, the job will not
end until resynchronization completes to the system that owns the
commit or rollback decision. To make the job end before
resynchronization is complete, a heuristic decision must be made and
resynchronization must be cancelled.
2. Ending the job or system abnormally during a long-running rollback is
| not recommended. This will cause another rollback to occur as the job
| ends (or during the next IPL if the system is ended). The subsequent
rollback will repeat the work performed by the original rollback and
take significantly longer to run.

v If a notify object is defined for the commitment definition, it may be updated.

See “Updates Made to the Notify Object” on page 531 for more information
about the system updating a notify object.

If a process ends before commitment control is ended and protected conversations

are still active, the commitment definition may be required to commit or roll back.
The action is based on the State option and the Action if end job option for the
commitment definition.

582 OS/400 Backup and Recovery V5R1

 ATTENTION!
The recovery for commitment definitions described in this section refers to an
abnormal end for the system or a job due to a power failure, a hardware
failure, or a failure in the operating system or licensed internal code.
However, forcing a job to end abnormally with the End Job Abnormal
(ENDJOBABN) command may result in pending changes for any currently
active transactions for the job being ended to be partially committed or rolled
back. The next IPL may attempt recovery for any partial transactions for the
job ended with the ENDJOBABN command.

However, any commitment control recovery that may be performed during

an IPL for a job that is ended with the ENDJOBABN command may or may
not be desirable. All locks for commitment resources are released when the
job is ended abnormally and any pending changes due to partial transactions
are made available to other jobs. These pending changes may then cause
other application programs to make additional erroneous changes to the
database. Likewise, any ensuing IPL recovery that would be performed later
may adversely effect the changes made by any applications after the job was
ended abnormally. For example, an SQL table may be dropped during IPL
recovery as the rollback action for a pending create table. However, other
applications may have already inserted several rows into the table after the
job was ended abnormally.

Commitment Control during a Save-While-Active Operation

The save-while-active checkpoint processing waits until all committable resources
in the save request are at a commitment boundary (with respect to all jobs making
committable changes to the objects being saved) prior to actually saving the objects
to the save media. This is done so that no partial transactions are saved to the save
media by a save-while-active operation.

When commitment control is active for any job on the system, the system performs
the following during the save-while-active checkpoint processing:
v Identifies all jobs that have one or more commitment definitions with pending
committable changes related to the objects being saved by the save-while-active
operation.
v For identified jobs, allows additional committable changes to be made for any
commitment definitions already started, or to be started. Additional committable
changes are allowed for the objects so that all pending changes for the objects
saved by the save-while-active operation can be committed or rolled back.
v Delays any job that attempts to make a committable change to an object being
saved by the save-while-active operation and all commitment definitions for the
job do not have any pending changes for any objects being saved by the
save-while-active operation. The job is delayed only until the checkpoint
processing is completed for the save-while active operation.

Any job not making committable changes to the objects being saved by the
save-while-active operation is not delayed, except for the following cases:
v Object-level changes made to local resources using SQL.
Any job that attempts to make an object-level committable change within a
library that is having objects saved from it by a save-while-active operation.

Chapter 22. Commitment Control 583

 v API commitment resources.
Any job that attempts to add an API commitment resource and all other
commitment definitions for the job do not have any pending changes for any
objects being saved by the save-while-active operation.
In addition, any job with an added API commitment resource is held during a
commit or rollback request for the commitment definition with the API
commitment resource added. This is because the system does not know which
objects are being changed for an API commitment resource.
Therefore, for a save-while-active request, the system delays all jobs during the
checkpoint processing that have a commitment definition with an API
commitment resource. Because the job is held during the commit or rollback
request, and because a commit or rollback request can be performed only for a
single commitment definition at a time, jobs with more than one commitment
definition with API commitment resources added always prevent a save-while-active
operation from completing.

Note: This does not apply to API resources that were added using the
QTNADDCR API if the Allow normal save processing field has a value of Y.

For all the cases mentioned previously that delay a job due to a save-while-active
operation, the job is delayed only if all other commitment definitions for the job do
not have any pending changes for any objects being saved by the save-while-active
operation. The job is delayed only until the checkpoint processing is completed for
the save-while-active operation.

The save active wait (SAVACTWAIT) parameter value on the save commands can
be used to control the amount of time allowed for jobs to reach, and be delayed at,
a commitment boundary.

Other considerations and restrictions apply when performing a save operation

from a job that has active commitment definitions. For more information, see
“Miscellaneous Considerations and Restrictions for Commitment Control” on
page 591.

For more information about commitment control and save-while-active operations,

see the Backup and Recovery topic on the Information Center at the following Web
site: http://www.ibm.com/eserver/iseries/infocenter.

Commitment Control Considerations and Restrictions

The following sections describe programming considerations and restrictions when
using commitment control.

The Size of a Transaction

For this discussion, a transaction is interactive. (Commitment control can also be
used for batch applications, which often can be considered a series of transactions.
Many of the same considerations apply to batch applications, which are discussed
in “Commitment Control for Batch Applications” on page 588.)

| You can lock a maximum of 500 000 000 records during a transaction for each
| journal associated with the transaction. You can reduce this limit by using a Query
| Options File (QAQQINI). Use the QRYOPTLIB parameter of the Change Query
| Attributes (CHGQRYA) command to specify a Query Options File for a job to use.
| Use the COMMITMENT_CONTROL_LOCK_LEVEL value in the Query Options
| File as the lock limit for the job.

584 OS/400 Backup and Recovery V5R1

 When choosing the lock level for your records, consider the size of your
transactions. Size should determine how long records are locked before a
transaction ends. You have to decide if a commit or rollback operation for
commitment control is limited to a single use of the Enter key, or if the transaction
consists of many uses of the Enter key.

Note: The shorter the transaction, the earlier the job waiting to start
save-while-active checkpoint processing can continue and complete.

For example, for an order entry application, a customer might order several items
in a single order requiring an order detail record and an inventory master record
update for every item in the order. If the transaction is defined as the entire order
and each use of the Enter key orders an item, all records involved in the order are
locked for the duration of the entire order. Therefore, often-used records (such as
inventory master records) could be locked for long periods of time, preventing
other work from progressing. If all items are entered with a single Enter key using
a subfile, the duration of the locks for the entire order is minimized.

In general, the number and duration of locks should be minimized so several

workstation users can access the same data without long waiting periods. You can
do this by not holding locks while the user is entering data on the display. Some
applications may not require more than one workstation user accessing the same
data. For example, in a cash posting application with many open item records per
customer, the typical approach is to lock all the records and delay them until a
workstation user completes posting the cash for a given receipt.

If the workstation user presses the Enter key several times for a transaction, it is
possible to perform the transaction in a number of segments. For example:
v The first segment is an inquiry in which the workstation user requests the
information.
v The second segment is a confirmation of the workstation user’s intent to
complete the entire transaction.
v The third segment is retrieval and update of the affected records.

This approach allows record locking to be restricted to a single use of Enter.

This inquiry-first approach is normally used in applications where a decision

results from information displayed. For example, in an airline reservation
application, a customer may want to know what flight times, connecting flights,
and seating arrangements are available before making a decision on which flight to
take. Once the customer makes a decision, the transaction is entered. If the
transaction fails (the flight is now full), the rollback function can be used and a
different request entered. If the records were locked from the first inquiry until a
decision is made, another reservation clerk would be waiting until the other
transaction is complete.

Record Locking
When a job holds a record lock and another job attempts to retrieve that record for
update, the requesting job waits and is removed from active processing until one
of the following occurs:
v The record lock is released.
v The specified wait time ends.

Chapter 22. Commitment Control 585

 More than one job can request a record to be locked by another job. When the
record lock is released, the first job to request the record receives that record. When
waiting for a locked record, specify the wait time in the WAITRCD parameter on
the following create, change, or override commands:
v Create Physical File (CRTPF)
v Create Logical File (CRTLF)
v Create Source Physical File (CRTSRCPF)
v Change Physical File (CHGPF)
v Change Logical File (CHGLF)
v Change Source Physical File (CHGSRCPF)
v Override Database File (OVRDBF)

When specifying wait time, consider the following:

v If you do not specify a value, the program waits the default wait time for the
process.
v If the record cannot be allocated within the specified time, a notify message is
sent to the high-level language program.
v If the wait time for a record is exceeded, the message sent to the job log gives
the name of the job holding the locked record that caused the requesting job to
wait. If you experience record lock exceptions, you can use the job log to help
determine which programs to alter so they will not hold locks for long
durations. Programs keep record locks over long durations for one of the
following reasons:
– The record remains locked while the workstation user is considering a
change.
– The record lock is part of a long commitment transaction. Consider making
smaller transactions so a commit operation can be performed more frequently.
– An undesired lock has occurred. For example, assume a file is defined as an
update file with unique keys, and the program updates and adds additional
records to the file. If the workstation user wants to add a record to the file,
the program may attempt to access the record to determine whether the key
already exists. If it does, the program informs the workstation user that the
request made is not valid. When the record is retrieved from the file, it is
locked until it is implicitly released by another read operation to the same
file, or until it is explicitly released.

Note: For more information about how to use each high-level language
interface to release record locks, see the appropriate high-level
language reference manual.

The duration of the lock is much longer if LCKLVL(*ALL) is specified because

the record that was retrieved from the file is locked until the next commit or
rollback operation. It is not implicitly released by another read operation and
cannot be explicitly released.

Minimizing Locks
A typical way to minimize record locks is to release the record lock. (This
technique does not work if LCKLVL(*ALL) has been specified.) For example, a
single file maintenance application typically does the following:
v Displays a prompt for a record identification to be changed.
v Retrieves the requested record.

586 OS/400 Backup and Recovery V5R1

v Displays the record.
v Allows the workstation user to make the change.
v Updates the record.

In most cases, the record is locked from the access of the requested record through
the update. The record wait time may be exceeded for another job that is waiting
for the record. To avoid locking a record while the workstation user is considering
a change, release the record after it is retrieved from the database (before the
record display appears). You then need to access the record again before updating.
If the record was changed between the time it was released and the time it was
accessed again, you should inform the workstation user. The program can
determine if the record was changed by saving one or more fields of the original
record and comparing them to the fields in the same record after it is retrieved, as
follows:
v Use an update count field in the record and add 1 to the field just before an
update. The program saves the original value and compares it to the value in
the field when the record is retrieved again. If a change has occurred, the
workstation user is informed and the record appears again. The update count
field is changed only if an update occurs. The record is released while the
workstation user is considering a change. If you use this technique, you must
use it in every program that updates the file.
v Save the contents of the entire data record and compare it to the record the next
time it is retrieved.

In both cases above, the sequence of operations prevents the simple use of
externally described data in RPG where the same field names are used in the
master record and in the display file. Using the same field names (in RPG) does
not work because the workstation user’s changes are overlaid when the record is
retrieved again.

You can solve this problem by moving the record data to a data structure or
continue to use externally described data if you use the DDS keyword RTNDTA.
The RTNDTA keyword allows your program to reread data on the display without
the operating system having to move data from the display to the program. This
allows the program to do the following:
1. Prompt for the record identification.
2. Retrieve the requested record from the database.
3. Release the record.
4. Save the field or fields used to determine if the record was changed.
5. Display the record and wait for the workstation user to respond.

If the workstation user changes the record on the display, the program uses the
following sequence:
1. Retrieves the record from the database again.
2. Compares the saved fields to determine if the database record has been
changed. If it has been changed, the program releases the record and sends a
message when the record appears.
3. Retrieves the record from the display by running a read operation with the
RTNDTA keyword and updates the record in the database record.
4. Proceeds to the next logical prompt because there are no additional records to
be released if the workstation user cancels the request.

Chapter 22. Commitment Control 587

 LCKLVL(*CHG) and LCKLVL(*CS) work in this situation. If LCKLVL(*ALL) is
used, you must release the record lock by using a commit or rollback operation.

Commitment Control for Batch Applications

Batch applications may or may not need commitment control. In some cases, a
batch application can perform a single function of reading an input file and
updating a master file. However, you can use commitment control for this type of
application if it is important to start it again after an abnormal end.

The input file is an update file with a code in the records to indicate that a record
was processed. This file and any files updated are placed under commitment
control. When the code is present in the input file, it represents a completed
transaction. The program reads through the input file and bypasses any records
with the completed code. This allows the same program logic to be used for
normal and starting again conditions.

If the batch application contains input records dependent on one another and
contains switches or totals, a notify object can be used to provide information on
starting again. The values held in the notify object are used to start processing
again from the last committed transaction within the input file. See the topic
“Notify Object Parameter” on page 530.

| If input records are dependent on one another, they can be processed as a

| transaction. A batch job can lock a maximum of 500 000 000 records. You can
| reduce this limit by using a Query Options File (QAQQINI). Use the QRYOPTLIB
| parameter of the Change Query Attributes (CHGQRYA) command to specify a
| Query Options File for a job to use. Use the
| COMMITMENT_CONTROL_LOCK_LEVEL value in the Query Options File as the
| lock limit for the job.

Any commit cycle that exceeds 2000 locks will probably slow down system
performance noticeably. Otherwise, the same locking considerations exist as for
interactive applications, but the length of time records are locked in a batch
application may be less important than in interactive applications.

Performance Considerations for Commitment Control

Using commitment control requires resources that can affect system performance.
Several factors affect system performance regarding commitment control:
v Journaling
Journaling a file requires system resources. If you specify only after-images,
commitment control changes this to both before- and after-images while
commitment control is in effect. Usually this is a space, not a performance,
consideration.
v Commit operation
| If any changes were made to journaled resources during the transaction, each
| commit of a transaction adds two entries to each journal related to those
| resources. The number of entries added can increase significantly for a large
volume of small transactions. You may want to place the journal receivers in a
user auxiliary storage pool (ASP).
v Rollback operation

588 OS/400 Backup and Recovery V5R1

 Because commitment control must rollback the pending changes recorded in the
database, additional system resources are required whenever a rollback occurs.
Also, if record changes are pending, a rollback operation causes additional
entries to be added to the journal.
v Start Commitment Control (STRCMTCTL) and End Commitment Control
(ENDCMTCTL) commands
Additional overhead is incurred by the system each time a commitment
definition is started and ended using the STRCMTCTL and ENDCMTCTL
commands respectively. Avoid using the STRCMTCTL and ENDCMTCTL
commands for each LUW. Use them only when necessary. You can establish a
commitment definition at the beginning of an interactive job and use it for the
duration of the job.
v Opening a file
If you open a file without specifying the commit open option, no additional
system resource is used even if a commitment definition has been started. For
more information about specifying the commit open option, see the appropriate
high-level language reference manual.
v Using more than one journal for commitment control LUWs
With two-phase commit, files that are opened under commitment control can be
journaled to more than one journal. However, using more than one journal takes
additional system resources to manage the commitment definition. Using more
than one journal can also make recovery more complicated.
v Record locking
Record locking can affect other applications. The number of records locked
within a particular job increases the overall system resources used for the job.
Applications needing to access the same record must wait for the transaction to
end.
v SEQONLY
If you request the SEQONLY(*YES) option (by using the OVRDBF command or
the application program implicitly attempts to use SEQONLY(*YES)) and the file
is opened for input only under commitment control with LCKLVL(*ALL), the
option is changed to SEQONLY(*NO). This option can affect the performance of
input files because records will not be blocked.
v Requesting a record-level change for a database file when save-while-active
processing is active
A request to make a record-level change under commitment control for a
database file may be delayed if the commitment definition is at a commitment
boundary and a save-while-active operation is running in a different job. This
can happen when a file is journaled to the same journal as some of the objects
on the save request.

Note: The Status column on the Work with Active Jobs (WRKACTJOB
command) display shows CMTW (commit wait) when a job is being held
due to save-while-active checkpoint processing.
v Committing or rolling back changes when save-while-active processing is active
A commit or rollback operation may be delayed at a commitment boundary
when a save-while-active operation is running in a different job. This can
happen when an API commitment resource was previously added to the
commitment definition, unless the API resource was added using the
QTNADDCR API and the Allow normal save processing field has a value of Y.
Because the job is held during the commit or rollback request, and because a
commit or rollback request can be performed only for a single commitment

Chapter 22. Commitment Control 589

 definition at a time, jobs with more than one commitment definition with API
commitment resources added always prevent a save-while-active operation from
completing.
v Requesting an object-level change when save-while-active processing is active
A request to make an object-level change under commitment control may be
delayed if the commitment definition is at a commitment boundary and a
save-while-active operation is running in a different job. This can happen when
an object-level change is made while the save-while-active operation is running
against the library the object is in. For example, the create SQL table operation
under commitment control for table MYTBL in library MYSQLLIB may be
delayed when a save-while-active operation is running against library
MYSQLLIB.

Note: If the wait time exceeds 60 seconds, inquiry message CPA8351 is sent to
ask the user whether to continue waiting or cancel the operation.
v Adding an API resource using the QTNADDCR API
A request to add an API commitment resource using the QTNADDCR API may
be delayed if all commitment definitions for the job are at a commitment
boundary and a save-while-active operation is running in a different job.
Notes:
1. If the wait time exceeds 60 seconds, inquiry message CPA8351 is sent to ask
the user whether to continue waiting or cancel the operation.
2. This does not apply to API resources that were added using the
QTNADDCR API if the Allow normal save processing field has a value of Y.
v Using a default journal can help performance if you close and reopen all files
under commitment control while the commitment definition is active.
v Using a default journal with OMTJRNE(*NONE) degrades the performance of
commit and rollback operations.
v Last agent selected.
Performance is enhanced when a last agent is selected because fewer interactions
between the system and the last agent are required during a commit operation.
However, if a communications failure occurs during a commit operation, the
commit operation will not complete until resynchronization completes regardless
of the value of the wait for outcome option. Such a failure is rare but this option
allows the application writer to consider the negative impact of causing the user
to wait indefinitely for the resynchronization to complete when a failure does
occur. This should be weighed against the performance enhancement that is
provided by last agent optimization during successful commit operations. This
consideration is generally more significant for interactive jobs than for batch
jobs.
The default is that a last agent is permitted to be selected by the system but the
user can modify this value using the QTNCHGCO API.
v Wait for Outcome option
When remote resources are under commitment control, performance is improved
when the Wait for Outcome option is set to N (No) and all remote systems
support presumed abort. (Presumed abort is available at V3R7 and later
releases.) The Wait for Outcome option is set to N by the system for DRDA and
DDM application when the first connection is made to a remote system. APPC
applications must explicitly set the Wait for Outcome option, or the default
value of Y will be used.
v OK to Leave Out option

590 OS/400 Backup and Recovery V5R1

 Performance is improved when the OK to Leave Out option is selected. For
further information on this option, see “Changing the Commitment Definition to
Indicate OK to Leave Out” on page 558.
v Vote Read Only
Performance is improved when the Vote Read Only option is selected. For
further information on this option, see “Changing the Commitment Definition to
Allow Vote Read-Only” on page 560.

Miscellaneous Considerations and Restrictions for

Commitment Control
v If you specify that a shared file be opened under commitment control, all
subsequent uses of that file must be opened under commitment control.
v If SEQONLY(*YES) is specified for the file opened for read only with
LCKLVL(*ALL) (either implicitly or by the high-level language program, or
explicitly by the OVRDBF command), then SEQONLY(*YES) is ignored and
SEQONLY(*NO) is used.
v Record-level changes made under commitment control are recorded in a journal.
These changes can be applied to or removed from the database with the Apply
Journaled Changes (APYJRNCHG) command or the Remove Journaled Changes
(RMVJRNCHG) command.
Images of the files before (before-image) and after (after-image) they are changed
are journaled under commitment control. If you specified only to journal the
after-images of the files, the system also automatically journals the before-image
of the file changes that occurred under commitment control. However, because
the before-images are not captured for all changes made to the files, the
RMVJRNCHG command cannot be used for these files.
| v Most object-level changes made under commitment control are written to the
| journal but are not applied or removed using the APYJRNCHG and
| RMVJRNCHG commands. However, you can use the QJOSJRNE API to send
journal entries for other events. During recovery, you can retrieve those entries
and process them with a user-written program.
v Object-level and record-level changes made under commitment control using
SQL will use the commitment definition that is currently active for the activation
group that the requesting program is running in. If neither the job-level nor the
activation-group-level commitment definition is active, SQL will start an
activation-group-level commitment definition. For more information regarding
changes made under commitment control using SQL, look in the Information
Center at the following Web site:
http://www.ibm.com/eserver/iseries/infocenter.
| v While a one-phase remote conversation or connection is established, remote
| conversations or connections to other locations are not allowed. If a commitment
boundary is established and all resources are removed, the location can be
changed.
v If you are using two-phase commit, you do not need to use the Submit Remote
Command (SBMRMTCMD) command to start commitment control or perform
any other commitment control operations at the remote locations. The system
performs these functions for you.
v For a one-phase remote location, the COMMIT and ROLLBACK CL commands
will fail if SQL is in the call stack and the remote relational database is not on a
system. If SQL is not on the call stack, the COMMIT and ROLLBACK
commands will not fail.

Chapter 22. Commitment Control 591

 v For a one-phase remote location, commitment control must be started on the
source system before making committable changes to remote resources. The
system automatically starts commitment control for distributed database SQL on
the source system at connection time if the SQL program is running with the
commitment control option other than *NONE. When the first remote resource is
placed under commitment control, the system starts commitment control on the
target system.
v A save operation is prevented if the job performing the save has one or more active
commitment definitions with any of the following types of committable changes.
– A record change to a file that resides in the library being saved.

Note: For logical files, all the related physical files are checked.
– Any object-level changes within a library that is being saved.
– Any API resource that was added using the QTNADDCR API and with the
Allow normal save processing field set to the default value of N.

This prevents pending changes due to a partial transaction from being saved to
the save media.

Note: Object locks and record locks prevent pending changes from commitment
definitions in other jobs from being saved to the save media. This is true
only for API commitment resources if locks are acquired when changes
are made to the object or objects associated with the API commitment
resource.
v Before upgrading your system to a new release, all pending resynchronizations
should either be completed or cancelled. See Software Installation, SC41-5120-05
for more details.
v The COMMIT and ROLLBACK values are shown on the WRKACTJOB Function
field during a commit or rollback. If the Function remains COMMIT or
ROLLBACK for a long time, one of the following situations might have
occurred:
– A resource failure during the commit or rollback requires resynchronization.
Control will not return to the application until the resynchronization
completes or is cancelled.
– This system voted read-only during the commit. Control will not return to the
application until the system that initiated the commit sends data to this
system.
– This system voted ok to leave out during the commit. Control will not return to
the application until the system that initiated the commit sends data to this
system.

Commitment Control Errors

When you use commitment control, it is important to understand which conditions
cause errors and which do not. In general, errors occur when commitment control
functions are used inconsistently, such as running an End Commitment Control
(ENDCMTCTL) command when files that use the commitment definition are still
open.

Error Conditions
The following are some typical errors related to commitment control.

592 OS/400 Backup and Recovery V5R1

 If an error occurs, an escape message is sent that you can monitor for in a
program.
v Consecutive STRCMTCTL commands are run without an intervening
ENDCMTCTL command.
v Files are opened under commitment control, but no STRCMTCTL command was
run.
This is not an error condition for programs that run within an activation group
that are to use the job-level commitment definition. The job-level commitment
definition can be started only by a single program, but once started by a
program, the job-level commitment definition is used by any program running
in any activation group that is not using an activation-group-level commitment
definition. Programs that run within an activation group that are to use the
activation-group-level commitment definition must first start the
activation-group-level commitment definition with the STRCMTCTL command.
v Files that are opened for output under commitment control are not journaled.
v The first open operation of a shared file places the file under commitment
control, but subsequent open operations of the same shared file do not.
v The first open operation of a shared file does not place the file under
commitment control, but subsequent open operations of the same shared file do.
| v The record lock limit for the job is reached in a single LUW.
v The program issues a read operation, a commit operation, and a change to the
same record. The read operation must be issued again after the commit
operation because the commit operation has freed the lock on the record.
v For a one-phase location, resources placed under commitment control do not
reside at the same location as resources already under commitment control for
the commitment definition.
v Uncommitted changes exist when an ENDCMTCTL command is issued.
This is not an error condition for the ENDCMTCTL command if all files are
closed, any remote database is disconnected, and no API commitment resource is
still associated with the commitment definition to be ended.
v A commit, rollback, or ENDCMTCTL command is run, and a STRCMTCTL
command was not run.
This is not an error condition for programs that run within an activation group
and the job-level commitment definition is active. The job-level commitment
definition can be started only by a single program, but once started by a
program, the job-level commitment definition is used by any program running
in any activation group that is not using an activation-group-level commitment
definition. Programs that run within an activation group and are to use the
activation-group-level commitment definition must first start the
activation-group-level commitment definition with the STRCMTCTL command.
v An ENDCMTCTL command is run with files still open under commitment
control for the commitment definition.
v A job performing a save operation has one or more commitment definitions that
are not at a commitment boundary.
v A save-while-active operation ended because other jobs with committable
resources did not reach a commitment boundary in the time specified for the
SAVACTWAIT parameter.
v A save-while-active process was not able to continue because of API-committable
resources being added to more than one commitment definition for a single job.
v More than 1023 commitment definitions for a single job.

Chapter 22. Commitment Control 593

 v The conversation to a remote location is lost due to a resource failure. This may
cause the LUW to be rolled back.
v A one-phase resource that is opened for update is present at a node that did not
initiate the commit operation. You must remove either the resource or the node
that initiated the commit request.
v A commit operation is requested while the LUW is in rollback required (RBR)
state. A rollback operation must be done.
v An API exit program issues a commit request or a rollback request.
v A trigger program issues a commit request or a rollback request for the
commitment definition under which the trigger program was called.

Note: The trigger program can start a separate commitment definition and issue
a commit or rollback request for that definition.

Non-Error Conditions
The following are some situations for commitment control in which no errors
occur:
v A commit or rollback operation is run and no resources are under commitment
control. This allows you to include commit or rollback operations in your
program without considering whether there are resources under commitment
control. It also allows you to specify a commit identification before making any
committable changes.
v A commit or rollback operation is run and there are no uncommitted resource
changes. This allows you to include commit and rollback operations within your
program without considering whether there are uncommitted resource changes.
v A file under commitment control is closed and uncommitted records exist. This
situation allows another program to be called to perform the commit or rollback
operation. This occurs regardless of whether or not the file is shared. This
function allows a subprogram to make database changes that are part of a LUW
involving multiple programs.
v A job ends, either normally or abnormally, with uncommitted changes for one or
more commitment definitions. The changes for all commitment definitions are
rolled back.
v An activation group ends with pending changes for the activation-group-level
commitment definition. If the activation group is ending normally and there are
no errors encountered when closing any files opened under commitment control
scoped to the same activation group that is ending, an implicit commit is
performed by the system. Otherwise, an implicit rollback is performed.
v A program accesses a changed record again that has not been committed. This
allows a program to:
– Add a record and update it before specifying the commit operation.
– Update the same record twice before specifying the commit operation.
– Add a record and delete it before specifying the commit operation.
– Access an uncommitted record again by a different logical file (under
commitment control).
v You specify LCKLVL(*CHG or *CS) on the STRCMTCTL command and open a
file with a commit operation for read only. In this case, no locks occur on the
request. It is treated as if commitment control is not in effect, but the file does
appear on the WRKJOB menu option of files under commitment control.

594 OS/400 Backup and Recovery V5R1

 v You issue the STRCMTCTL command and do not open any files under
commitment control. In this situation, any record-level changes made to the files
are not made under commitment control.

Monitoring for Errors after a CALL Command

When a program that uses commitment control is called, you should monitor for
unexpected errors and perform a rollback operation if an error occurs. For
example, uncommitted records can exist when a program encounters an
unexpected error such as an RPG divide-by-zero error. Depending on the status of
the inquiry message reply (INQMSGRPY) parameter for a job, the program sends
an inquiry message or performs a default action. If the operator response or the
default action ends the program, uncommitted records still exist waiting for a
commit or rollback operation.

If another program is called and causes a commit operation, the partially

completed LUW from the previous program is committed.

To prevent partially completed LUWs from being committed, monitor for escape
messages after the CALL command. For example, if it is an RPG program, use the
following coding:
CALL RPGA
MONMSG MSGID(RPG9001)
EXEC(ROLLBACK) /*Rollback if pgm is canceled*/

If it is a COBOL program:
CALL COBOLA
MONMSG MSGID(CBE9001)
EXEC(ROLLBACK) /*Rollback if pgm is canceled*/

Error Messages to Monitor for Commitment Control

Several different error messages can be returned by the commit or rollback
operations or sent to the job log, depending on the type of message and when the
error occurred. Some of the messages related to commitment control to look for
are:
CPF8350
Commitment definition not found
CPD8351
Changes may not have been committed.
CPD8352
Changes not committed at remote location &3;
CPD8353
Changes to relational database &1 may not have been committed.
CPD8354
Changes to DDM file &1 may not have been committed.
CPD8355
Changes to DDL object &1 may not have been committed.
| CPF8355
| ENDCMTCTL not allowed. Pending changes active.
CPD8356
Rolled back changes may have been committed.

Chapter 22. Commitment Control 595

| CPF8356
| Commitment control eneded with &1 local changes not committed.
CPD8358
Changes to relational database &1 may not have been rolled back.
| CPF8358
| Notify object &1 in &2 not updated.
CPD8359
Changes to DDM file &1 may not have been rolled back.
| CPF8359
| Rollback operation failed.
CPD835A
Changes to DDL object &3 may not have been rolled back.
| CPF835A
| End of commitment definition &1 canceled.
CPF835B
Errors occurred while ending commitment control.
| CPD835C
| Notify object &1 in &2 not updated.
CPF835C
Commitment control ended with remote changes not committed.
CPD835D
DRDA resource does not allow SQL cursor hold
| CPD8360
| Members or files or both were already deallocated.
CPD8361
API exit program &1 failed during commit.
CPD8362
API exit program &1 failed during roll back.
CPD8363
API exit program &1 ended after &4 minutes during commit.
| CPF8363
| Commit operation failed.
CPD8364
API exit program &1 ended after &4 minutes during rollback.
CPF8367
Cannot perform commitment control operation.
CPF8369
Cannot place API commitment resource under commitment control; reason
code &1.
CPD836F
Protocol error occurred during commitment control operation.
CPD83D1
API resource &4; cannot be last agent.
CPD83D2
Resource not compatible with commitment control.

596 OS/400 Backup and Recovery V5R1

| CPF83D2
| Commit complete — Resynchronization in progress has been returned.
| CPF83D3
| Commit complete — Heuristic Mixed has been returned.
| CPF83D4
| Logical unit of work journal entry not sent.
CPD83D7
Commit operation changed to rollback.
CPD83D9
A heuristic mixed condition occurred.
CPF83DB
Commit operation resulted in rollback.
CPD83DC
Action If Problems Used to determine commit or rollback operation; reason
&2;
CPD83DD
Conversation ended; Reason &1;
CPD83DE
Return information not valid.
CPF83E4
Commitment control ended with resources not committed.
CPF83E6
Commitment control operation completed with resynchronization in
progress.
| CPF83E7
| Commit or rollback of X/Open global transaction not allowed.
CPD83EC
API exit program &1; voted rollback.
CPD83EF
Rollback operation started for next logical unit of work.
CPF835F
Commit or rollback operation failed.
CPF8364
Commitment control parameter value is not valid. Reason code &3;
CPF83D0
Commitment operation not allowed.
CPF83E1
Commit operation failed due to constraint violation.
CPF83E2
Rollback operation required.
CPF83E3
Requested nesting level is not active.

These messages can occur during:

v Normal commit or rollback processing
v Commit or rollback processing during job process end

Chapter 22. Commitment Control 597

 v Commit or rollback processing during activation group end

Note: None of the above messages can be monitored for during activation group
end or job process end. Any errors encountered when ending an
activation-group-level commitment definition during activation group end or
any commitment definition during job end are left in the job log as
diagnostic messages. The following sections show you when you can expect
certain types of messages.

Normal Commit or Rollback Processing

Errors may occur at any time during commit or rollback processing. The following
table divides this processing into four situations. The middle column describes the
actions taken by the system when it encounters errors during each situation. The
third column suggests what you or your application should do in response to the
messages. These suggestions are consistent with the way commitment control
processing is handled by the system.
Table 82. Commit and Rollback Processing for Failures.
Situation Commit or Rollback Processing Suggested Action
Record-level I/O commit v If the error occurs during the prepare wave, the LUW Monitor for messages;
fails is rolled back and message CPF83DB is sent. handle as desired
v If the error occurs during the committed wave,
commit processing continues to commit as many
remaining resources as possible. Message CPF8363 is
sent at the end of commit processing.
Object-level or commit and v If the error occurs during the prepare wave, the LUW Monitor for messages;
rollback exit program for is rolled back and message CPF83DB is sent. handle as desired
API commitment resource
v If the error occurs during the committed wave,
fails during commit
processing continues to commit or roll back as many
remaining resources as possible. One of the following
messages is returned, depending on the commitment
resource type:
– CPD8353
– CPD8354
– CPD8355
– CPD8361
Message CPF8363 is sent at the end of commit
processing.
Record-level I/O rollback 1. Returns CPD8356 Monitor for messages;
fails handle as desired
2. Attempt to continue processing to rollback
object-level or API commitment resources
3. Returns CPF8359 at end of processing
Object-level or commit and 1. Returns one of the following messages depending on Monitor for messages;
rollback exit program for the commitment resource type: handle as desired
API commitment resources
v CPD8358
fails during rollback
v CPD8359
v CPD835A
v CPD8362
2. Continues processing
3. Returns CPF8359 at end of processing

598 OS/400 Backup and Recovery V5R1

 Commit or Rollback Processing During Job End
All of the situations described in Table 82 on page 598 also apply when a job is
ending except that one of the following messages is sent:
v CPF8356 if only local resources are registered
v CPF835C if only remote resources are registered
v CPF83E4 if both local and remote resources are registered
In addition, one of two messages may appear specific to job completion if a
commit and rollback exit program for an API-committable resource has been
called. If the commit and rollback exit program does not complete within 5
minutes, the program is canceled, a diagnostic message CPD8363 (for commit) or
CPD8364 (for rollback) is sent, and the remainder of the commit or rollback
processing continues.

Commit or Rollback Processing During IPL

All of the situations described in Table 82 on page 598 also apply during IPL
recovery for commitment definitions except that message CPF835F is sent instead
of message CPF8359 or CPF8363. Messages that get sent for a particular
commitment definition may appear in the job log for one of the QDBSRVxx jobs or
the QHST log. In the QHST log, message CPI8356 indicates the beginning of IPL
recovery for a particular commitment definition. Message CPC8351 indicates the
end of IPL recovery for a particular commitment definition and any other
messages regarding the recovery of that commitment definition is found between
those two messages.

One of two messages may appear specific to a commitment definition if a commit

and rollback exit program for an API-committable resource has been called. If the
commit and rollback exit program does not complete within 5 minutes, the
program is canceled, a diagnostic message CPD8363 (for commit) or CPD8364 (for
rollback) is sent, and the remainder of the commit or rollback processing continues.

Example of Using Commitment Control

Note: The following example does not include the Integrated Language
Environment C/400.

The advantages of using commitment control are illustrated in the following

example. Assume that the following application program does not use commitment
control. The records read for updating are locked by the system. The following
steps describe how the application program transfers funds from a savings account
to a checking account:
v Program A locks and retrieves the savings record. (This action could require a
wait if the record is locked by another program.)
v Program A locks and retrieves the checking record. (This could also require a
wait.) Program A now has both records locked, and no other program can
change them.
v Program A updates the savings record. This causes the record to be released so it
is now available to be read for update by any other program.
v Program A updates the checking record, which causes the record to be released
so it is now available to be read for update by any other program.

Without using commitment control, a problem needs to be solved to make this

program work properly in all circumstances. For example, a problem occurs if

Chapter 22. Commitment Control 599

 program A does not update both records because of a job or system failure. In this
case, the two files are not consistent — funds are removed from the savings
account, but they are not added to the checking account. Using commitment
control allows you to ensure that all changes involved in the LUW are completed
or that the files are returned to their original state if the processing of the LUW is
interrupted.

If commitment control is used, the preceding example is changed as follows:

1. Commitment control is started.
2. Program A locks and retrieves the savings record. (This action could require a
wait if the record is locked by another program.)
3. Program A locks and retrieves the checking record. (This could also require a
wait.) Program A now has both records locked, and no other program can
change them.
4. Program A updates the savings record, and commitment control keeps the lock
on the record.
5. Program A updates the checking record, and commitment control keeps the
lock on the record.
6. Program A commits the LUW. The changes to the savings record and the
checking record are made permanent in the files. The changes are recorded in
the journal, which assumes they will appear on disk. Commitment control
releases the locks on both records. The records are now available to be read for
update by any other program.

Because the locks on both records are kept by commitment control until the LUW
is committed, a situation cannot arise in which one record is updated and the other
is not. If a routing step or system failure occurs before the LUW is committed, the
system removes (rolls back) the changes that have been made so that the files are
updated to the point where the last LUW was committed.

For each routing step in which files are to be under commitment control, the steps
shown in Figure 62 occur:

┌───────────────────────────────────────────┐
│ STRCMTCTL - establishes a commitment │
│ definition │
├───────────────────────────────────────────┤
│ CALL PGMA - calls program whose files ├─────┐
│ are under commitment control │ │
├───────────────────────────────────────────┤ │
│ ENDCMTCTL - notifies the system to end │ │
│ the commitment definition │ │
└───────────────────────────────────────────┘ │
┌───────────────────────────────┘
│
PGMA
N
┌───────────────────────────────────────────┐
│* Opens files under commitment control │
│* Processes an LUW │
│* Commits or rolls back the LUW │
│* Closes the files under commitment control│
│* Processes additional LUWs or returns │
└───────────────────────────────────────────┘

Figure 62. Routing Steps with Files under Commitment Control

600 OS/400 Backup and Recovery V5R1

The operations that are performed under commitment control are journaled to the
journal. Using the transfer of funds from a savings account to a checking account
as an example, the major entries shown in Figure 63 are placed in the journal. (For
a description of all the journal entries and their specific content, see “Appendix D.
Journal Entry Information” on page 805.)

Journal Entries
┌──────────────────────────────────┐W──────┐
│ Start commitment control │ │
├──────────────────────────────────┤ │
│ Savings record before-image │ │
├──────────────────────────────────┤ │
│ Savings record after-image │ LUW │
├──────────────────────────────────┤ │
│ Checking record before-image │ │
├──────────────────────────────────┤ │
│ Checking record after-image │ │
├──────────────────────────────────┤W──────┘
│ COMMIT -- Customer number 772389 │ LUW
└──────────────────────────────────┘ committed by
the program

Figure 63. Journal Entries for Transfer of Funds from Savings to Checking

The start commitment control journal entry appears after the first file open entry
under commitment control. This is because the first file open entry determines
what journal is used for commitment control. The journal entry from the first open
operation is then used to check subsequent open operations to ensure all files are
using the same journal.

When a job failure or system failure occurs, the resources under commitment
control are updated to a commitment boundary. If a LUW is started but is not
completed before a routing step ends, that LUW is rolled back by the system and
does not appear in the file after the routing step ends. If the system abnormally
ends before an LUW is completed, that LUW is rolled back by the system and does
not appear in the file after a subsequent successful initial program load (IPL) of
licensed internal code. Anytime a rollback occurs, reversing entries are placed in
the journal.

For example, assume you have a balance of $100.00 and a customer takes out
$20.00, for a new balance of $80.00. The database update causes both before-image
($100.00) and after-image ($80.00) journal entries. The journal entries appear as in
the example in Figure 64

Journal Entries
┌──────────────────────────┐W──── Last committed
│COMMIT -- │ LUW
│ Customer number 772389 │
├──────────────────────────┤W──── Start of new
│Savings record │ LUW
│ before-image ($100.00) │
├──────────────────────────┤
│Savings record │
│ after-image ($80.00) │
└──────────────────────────┘W──── System ends
abnormally

Figure 64. Journal Entries for a System That Ended Abnormally

Chapter 22. Commitment Control 601

 Assume the system abnormally ended after journaling the entries, but before
reaching the commitment point or rollback point. After the IPL, the system reads
the journal entry and updates the corresponding database record. This update
produces two journal entries that reverse the update: the first entry is the
before-image ($80.00) and the second entry is the after-image ($100.00) as shown in
Figure 65.

Journal Entries
┌──────────────────────────┐W──── Last committed
│COMMIT -- │ LUW
│ Customer number 772389 │
├──────────────────────────┤W──── Start of new
│Savings record │ LUW
│ before-image ($100.00) │
├──────────────────────────┤
│Savings record │
│ after-image ($80.00) │ System ends
├──────────────────────────┤W──── abnormally
│Savings record │W────┐
│ before-image ($80.00) │ │
│ updated for rollback │ │ Changes are
├──────────────────────────┤ │ rolled
│Savings record │ │ back
│ after-image ($100.00) │ │
│ updated for rollback │ │
├──────────────────────────┤ │
│ROLLBACK │ │
└──────────────────────────┘W────┘

Figure 65. Journal Entries for Rollback Changes

When the IPL is successfully completed after the abnormal end, the system
removes (or rolls back) any database changes that are not committed. In the
preceding example, the system removes the changes from the savings record
because a commit operation is not in the journal for that LUW. In this case, the
before-image of the savings record is placed in the file. The journal contains the
rolled back changes, and an indication that a rollback operation occurred, as in the
example shown in Figure 65.

Example of Transaction Logging File

A transaction logging file is used to start an application again after a system or
job failure when a notify object is not used. A transaction logging file is often used
in interactive applications to summarize the effects of an LUW.

For example, in an order entry application, a record is usually written to a

transaction logging file for each item ordered. The record contains the item
ordered, the quantity, and the price. In an accounts payable application, a record is
written to a transaction logging file for each account number that is to receive a
charge. This record normally contains such information as the account number, the
amount charged, and the vendor.

In many of the applications where a transaction logging file already exists, a

workstation user can request information about the last transaction entered. By
adding commitment control to the applications in which a transaction logging file
already exists, you can:
v Ensure that the database files are updated to a commitment boundary.
v Simplify the starting of the transaction again.

602 OS/400 Backup and Recovery V5R1

 You must be able to uniquely identify the workstation user if you use a transaction
logging file for starting applications again under commit control. If unique user
profile names are used on the system, that profile name can be placed in a field in
the transaction logging record. This field can be used as the key to the file.

The following examples ( Figure 66 through 70) assume that an order inventory file
is being used to perform LUWs and that a transaction logging file already exists.
The program does the following:
1. Prompts the workstation user for a quantity and item number.
2. Updates the quantity in the production master file (PRDMSTP).
3. Writes a record to the transaction logging file (ISSLOGL).

If the inventory quantity on hand is insufficient, the program rejects the

transaction. The workstation user can ask the program where the data entry was
interrupted, since the item number, description, quantity, user name, and date are
written to the transaction logging file.

SEQNBR *... ... 1 ... ... 2 ... ... 3 ... ... 4 ... ... 5 ... ... 6 ... ... 7

1.00 A R PRDMSTR TEXT('Master record')

2.00 A PRODCT 3 COLHDG('Product' 'Number')
3.00 A DESCRP 20 COLHDG('Description')
4.00 A ONHAND 5 0 COLHDG('On Hand' 'Amount')
5.00 A EDTCDE(Z)
6.00 A K PRODCT

Figure 66. DDS for Physical File PRDMSTP

SEQNBR *... ... 1 ... ... 2 ... ... 3 ... ... 4 ... ... 5 ... ... 6 ... ... 7

1.00 A R ISSLOGR TEXT('Product log record')

2.00 A PRODCT 3 COLHDG('Product' 'Number')
3.00 A DESCRP 20 COLHDG('Description')
4.00 A QTY 3 0 COLHDG('Quantity')
5.00 A EDTCDE(Z)
6.00 A USER 10 COLHDG('User' 'Name')
7.00 A DATE 6 0 EDTCDE(Y)
8.00 A COLHDG('Date')

Figure 67. DDS for Physical File ISSLOGP Used by ISSLOGP

SEQNBR *... ... 1 ... ... 2 ... ... 3 ... ... 4 ... ... 5 ... ... 6 ... ... 7

1.00 A LIFO
2.00 A R ISSLOGR PFILE(ISSLOGP)
3.00 A K USER

Figure 68. DDS for Logical File ISSLOGL

Chapter 22. Commitment Control 603

SEQNBR *... ... 1 ... ... 2 ... ... 3 ... ... 4 ... ... 5 ... ... 6 ... ... 7 ..

1.00 A REF(ISSLOGP)
2.00 A R PROMPT
3.00 A CA03(98 'End of program')
4.00 A CA02(97 'Where am I')
5.00 A 1 20'ISSUES PROCESSING'
6.00 A 3 2'Quantity'
7.00 A QTY R I +1
8.00 A 62 ERRMSG('Not enough +
9.00 A Qty' 62)
10.00 A +6'Product'
11.00 A PRODCT R I +1
12.00 A 61 ERRMSG('No Product +
13.00 A record found' 62)
14.00 A 55 15 2'No Previous record exists'
15.00 A 24 2'CF2 Last transaction'
16.00 A R RESTART
17.00 A 1 20'LAST TRANSACTION +
18.00 A INFORMATION'
19.00 A 5 2'Product'
20.00 A PRODCT R +1
21.00 A 7 2'Description'
22.00 A DESCRP R +1
23.00 A 9 2'Qty'
24.00 A QTY R +1REFFLD(QTY)

Figure 69. DDS for Display File PRDISSD Used in the Program

This process is outlined in Figure 70 on page 605.

604 OS/400 Backup and Recovery V5R1

 Begin
N
┌─────────────┐
│ Prompt │
└─────┬───────┘
N
┌─────────────┐
Yes │ F3? │
W─────┤ (*IN98) │
N └─────┬───────┘
End N No
Program ┌─────────────┐ ┌──────────────────┐
│ F2? │Yes │ Read last record │
│(Review last)├───S│ for this user │
└─────┬───────┘ └────────┬─────────┘
N No N
┌─────────────┐ ┌──────────────┐ Yes
│Read product │ │ Record found?├───┐
│record │ └────────┬─────┘ │
└─────┬───────┘ │ │
N N │
Yes ┌─────────────┐ ┌─────────────┐ │
┌──────┤ Found? │ │Display error│ │
│ └─────┬───────┘ │ No review │ │
│ │ No │ record │ │
│ N └────┬────────┘ │
│ ┌─────────────────┬ │ N
│ │ Display error │ │ ┌─────────────┐
│ │ No product │────┐│ │ Display │
│ │ record │ ││ │ Last │
│ └─────────────────┴ ││ │ Transaction │
│ ││ └─────┬───────┘
│ ┌─────────────┐ NN N
└─────S│ Check │ ┌──────────────────┐
│ Quantity │ │ Go to Begin │
│ On Hand │ └──────────────────┘
└─────────┬───┘
│
┌───────┐ │
│Release│ N
│lock on│ No┌─────────────┐Yes
│product│W───┤Sufficient? ├───┐ ┌─────────┐
│record │ └─────────────┘ │ │ Update │
└─────┬─┘ └────S│ product │
N │ record │
┌──────────────────┐ └─────┬───┘
│ Display error │ N
│ Not enough │ ┌───────────┐
│ quantity │ │Write │
└────┬─────────────┘ │transaction│
│ │record │
│ └───────┬───┘
│ N
│ ┌────────┐
│ │ Commit │
│ └────┬───┘
│ ┌──────────────────┐ │
└──────────S│ Go to Begin │W─────┘
└──────────────────┘

Figure 70. Program Flow

Chapter 22. Commitment Control 605

 The RPG COMMIT operation code is specified after the PRDMSTP file is updated
and the record is written to the transaction logging file. Since each prompt to the
operator represents a boundary for a new LUW, the LUW is considered a single
Enter transaction.

The user name is passed to the program when it is called. The access path for the
transaction logging file is defined in last-in-first-out (LIFO) sequence so the
program can easily access the last record entered.

The workstation user can start the program again after a system or job failure by
using the same function that identified where data entry was stopped. No
additional code needs to be added to the program. If you are currently using a
transaction logging file but are not using it to find out where you are, add the user
name to the transaction logging file (assuming user names are unique) and use this
approach in the program.

Figure 71 on page 607 shows the RPG program used. Statements required for
commitment control are marked with arrows (==>).

606 OS/400 Backup and Recovery V5R1

 SEQNBR *... ... 1 ... ... 2 ... ... 3 ... ... 4 ... ... 5 ... ... 6 ... .. 7 ..
=>1.00 FPRDMSTP UP E K DISK KCOMIT
=>2.00 FISSLOGL IF E K DISK KCOMIT
3.00 PRDISSD CP E WORKSTN
4.00 *ENTRY PLIST
5.00 PARM USER 10
6.00 C*
7.00 C* Initialize fields used in Trans Log Rcd
8.00 C*
9.00 C MOVE UDATE DATE
10.00 C*
11.00 C* Basic processing loop
12.00 C*
13.00 C LOOP TAG
14.00 C EXFMTPROMPT
15.00 C 98 GOTO END End of pgm
16.00 C 97 DO Where am I
17.00 C EXSR WHERE
18.00 C GOTO LOOP
19.00 C END
20.00 C PRODCT CHAINPRDMSTR 61 Not found
21.00 C 61 GOTO LOOP
22.00 C ONHAND SUB QTY TEST 50 62 Less than
23.00 C 62 DO Not enough
24.00 C EXCPTRLSMST Release lock
25.00 C GOTO LOOP
26.00 C END
27.00 C*
28.00 C* Update master record and output the Transaction Log Record
29.00 C*
30.00 C Z-ADDTEST ONHAND
31.00 C UPDATPRDMSTR
32.00 C WRITEISSLOGR
=>33.00 C COMIT
34.00 C GOTO LOOP
35.00 C*
36.00 C* End of program processing
37.00 C*
38.00 C END TAG
39.00 C SETON LR
40.00 C*
41.00 C* WHERE subroutine for "Where am I" requests
42.00 C*
43.00 C WHERE BEGSR
44.00 C USER CHAINISSLOGL 55 Not found
45.00 C N55 EXFMTRESTART
46.00 C ENDSR
47.00 OPRDMSTR E RLSMST

Figure 71. RPG Program

SEQNBR *... ... 1 ... ... 2 ... ... 3 ... ... 4 ... ... 5 ... ... 6 ... ... 7 ..

1.00 PGM
2.00 DCL &USER *CHAR LEN(10)
3.00 STRCMTCTL LCKLVL(*CHG)
4.00 RTVJOBA USER(&USER)
5.00 CALL PRDISS PARM(&USER)
6.00 MONMSG MSGID(RPG900l) EXEC(ROLLBACK)
7.00 ENDCMTCTL
8.00 ENDPGM

Figure 72. CL Program Used to Call RPG Program PRDISS

Chapter 22. Commitment Control 607

 To use commitment control in this program, a lock level of *CHG would normally
be specified. The record is locked by the change until a commit operation is run.
Note that if there is an insufficient quantity of inventory, the record is explicitly
released. (If the record were not explicitly released in the program, it would be
released when the next record is read for update from the file.)

In this example, there is no additional advantage to using the lock level *ALL. If
*ALL were used, a rollback or commit operation would have to be used to release
the record when an insufficient quantity existed.

Figure 72 on page 607 is a CL program that calls the RPG program PRDISS. Note
the use of STRCMTCTL/ENDCMTCTL commands. The unique user name is
retrieved (RTVJOBA command) and passed to the program. The use of the
MONMSG command to cause a rollback is described in “Standard Commit
Processing Program” on page 618.

Starting Application Programs Using a Notify Object

When a program is started after an abnormal end, it can look for an entry in the
notify object. If one exists, the program can start an LUW again. After the LUW
has been started again, the notify object is cleared by the program to prevent it
from starting the same LUW yet another time.

Following are ways you can use a notify object:

v If the commit identification is placed in a database file, query this file to
determine where to start each application or workstation job again.
v If the commit identification is placed in a message queue for a particular
workstation, a message can be sent to the work station users when they sign on
to inform them of the last transaction committed.
v If the commit identification is placed in a database file that has a key or user
name, the program can read this file when it is started. If a record exists in the
file, start the program again. The program can send a message to the
workstation user identifying the last transaction committed. Any recovery is
performed by the program. If a record existed in the database file, the program
deletes that record at the end of the program.
v For a batch application, the commit identification can be placed in a data area
that contains totals, switch settings, and other status information necessary to
start the application again. When the application is started, it accesses the data
area and verifies the values stored there. If the application ends normally, the
data area is set up for the next run.
v For a batch application, the commit identification can be sent to a message
queue. A program that is run when the application is started can retrieve the
messages from the queue and start the programs again.

There are several techniques for starting your applications again depending on
your application needs. In choosing the technique, consider the following:
v When there are multiple users of a program at the same time, a single data area
cannot be used as the notify object because after an abnormal system end, the
commit identification for each user would overlay each other in the data area.
v Your design for deleting information in the notify object should handle the
situation when a failure occurs immediately following use of the information:
– If information is deleted immediately, it would not exist if another failure
occurs before processing the interrupted LUW.

608 OS/400 Backup and Recovery V5R1

 – The information in the notify object should not be deleted until the successful
processing of the interrupted LUW. In this case, more than one entry will
exist in the notify object if it is a database file or message queue.
– The program should access the last record if there is more than one entry.
v A notify object cannot be used to provide the work station user with the last
transaction committed because the notify object is updated only if a system or
job failure occurs or if uncommitted changes exist at the normal end of a job.
v If information is displayed to the workstation user, it must be meaningful. To
accomplish this may require that the program translate codes kept in the notify
object into information that will help the user start again.
v Information for starting again should be displayed if the work station user needs
it. Additional logic in the program is required to prevent information from being
displayed again when it is no longer meaningful.
v A single notify object and a standard processing program can provide a starting
again function if the notify object is a database file. This standard processing
program is called by the programs that require the ability to start again to
minimize the changes to each individual program.

Using a Unique Notify Object for Each Program

Using a single, unique notify object for each job allows use of an externally
described commit identification even though there may be multiple users of the
same program. In the following example ( Figure 73 through Figure 76 on
page 612), a database file is used as a notify object and it is used only by this
program.

The program has two database files (PRDMSTP and PRDLOCP) that must be
updated for receipts to inventory. The display file used by the program is named
PRDRCTD. A database file, PRDRCTP, is used as the notify object. This notify
object is defined to the program as a file and is also used as the definition of a
data structure for the notify function.

The DDS for the physical file PRDMSTP is shown in Figure 66 on page 603.

SEQNBR *... ... 1 ... ... 2 ... ... 3 ... ... 4 ... ... 5 ... ... 6 ... ... 7

1.00 A R PRDLOCR TEXT('Location record')

2.00 A PRODCT 3 COLHDG('Product' 'Number')
3.00 A LOCATN 6 COLHDG('Location')
4.00 A LOCAMT 5 0 COLHDG('Location' 'Amount')
5.00 A EDTCDE(Z)
6.00 A K PRODCT
7.00 A K LOCATN

Figure 73. DDS for Physical File PRDLOCP

Chapter 22. Commitment Control 609

SEQNBR *... ... 1 ... ... 2 ... ... 3 ... ... 4 ... ... 5 ... ... 6 ... ... 7 ..

1.00 A REF(PRDMSTP)
2.00 A R PROMPT
3.00 A CA03(98 'End of program')
4.00 A SETOFF(71 'RESTART')
5.00 A 1 20'PRODUCT RECEIPTS'
6.00 A 3 2'Quantity'
7.00 A QTY 3 OI +1
8.00 A +6'Product'
9.00 A PRODCT R I +1
10.00 A 61 ERRMSG('No record +
11.00 A found in the +
12.00 A master file' 62)
13.00 A +6'Location'
14.00 A LOCATN R I +1REFFLD(LOCATN PRDLOCP)
15.00 A 62 ERRMSG('No record +
16.00 A found in the +
17.00 A location file' 62)
18.00 A 9 2'Last Transaction'
19.00 A 71 +6'This is restart +
20.00 A information'
21.00 A DSPATR(HI BL)
22.00 A 12 2'Quantity'
23.00 A 12 12'Product'
24.00 A 12 23'Location'
25.00 A 12 35'Description'
26.00 A LSTPRD R 14 15REFFLD(PRODCT)
27.00 A LSTLOC R 14 26REFFLD(LOCATN *SRC)
28.00 A LSTQTY R 14 5REFFLD(QTY *SRC)
29.00 A EDTCDE(Z)
30.00 A LSTDSC R 14 35REFFLD(DESCRP)

Figure 74. DDS for Display File PRDRCTD

SEQNBR *... ... 1 ... ... 2 ... ... 3 ... ... 4 ... ... 5 ... ... 6 ... ... 7 ..

1.00 A LIFO
2.00 A REF(PRDMSTP)
3.00 A R PRDRCTR
4.00 A USER 10
5.00 A PRODCT R
6.00 A DESCRP R
7.00 A QTY 3 0
8.00 A LOCATN R REFFLD(LOCATN PRDLOCP)
9.00 A K USER

Figure 75. DDS for Notify Object and Externally Described Data Structure (PRDRCTP)

The program processes the notify object as follows:

v At the beginning, the program randomly processes the notify object and displays
a record if it exists for the specific key:
– If multiple records exist, the last record for this key is used because the
PRDRCTP file is in LIFO sequence.
– If no record exists, an LUW was not interrupted so it is not necessary to start
again.
– If the program fails before the first successful commit operation, it does not
consider that starting again is required.
v The routine to clear the notify object occurs at the end of the program:

610 OS/400 Backup and Recovery V5R1

 – If there were multiple failures, the routine can handle deletion of multiple
records in the notify object.
– Although the system places the commit identification in a database file, the
commit identification must be specified as a variable in the RPG program.
– Because RPG allows a data structure to be externally described, a data
structure is a convenient way of specifying the commit identification. In this
example, the data structure uses the same external description that the
database file used as the notify object.

The processing for this program prompts the user for a product number, a location,
and a quantity:
v Two files must be updated:
– Product master file (PRDMSTP)
– Product location file (PRDLOCP)
v A record in each file must exist before either is updated.
v The program moves the input fields to corresponding last fields after each
transaction is successfully entered. These last fields are displayed to the operator
on each prompt as feedback for what was last entered.
v If information for starting again exists, it is moved to these last fields and a
special message appears on the display.

This process is outlined in Figure 76 on page 612. The user name is passed to the
program to provide a unique record in the notify object.

Chapter 22. Commitment Control 611

 B
Begin

Read
Read Last Product
Notify Object Location
Record Record

No Release
Lock on No
Found Found
Product
Master

Yes Yes

Move Restart
Set for Add
Information to
Error Quantity
’Last’ Fields
62

Set for A Update

Restart Product
Message Master
A
71

Update
Basic Product
Prompt Prompt Location

98 = F3
End of No Read Product
Program Master Record Commit

Yes

No Move Restart
Read Notify Found Set for Information to
Object Record Error ’Last’ Fields
61
Yes

No B A A
Found End

Yes

Delete

RV2W415-0

Figure 76. Program Flow

The following is the RPG source code for this example. The notify object (file
PRDRCTP) is used as a normal file at the beginning and end of the program, and
is also specified as the notify object in the CL (STRCMTCTL command) before
calling the program.

612 OS/400 Backup and Recovery V5R1

SEQNBR *... ... 1 ... ... 2 ... ... 3 ... ... 4 ... ... 5 ... ... 6 ... ... 7 ..

1.00 FPRDMSTP UF E K DISK KCOMIT

2.00 FPRDLOCP UF E K DISK KCOMIT
3.00 FPRDRCTD CF E WORKSTN
4.00 F*
5.00 F* The following file is the specific notify object for this pgm.
6.00 F* It is accessed only in a restart situation and at the
7.00 F* end of the program to delete any records. The records
8.00 F* are written to the notify object by Commitment Control.
9.00 F*
10.00 FPRDRCTP UF E K DISK
11.00 ICMTID E DSPRDRCTP

12.00 C *ENTRY PLIST

13.00 C PARM USER10 10
14.00 C MOVE USER10 USER
15.00 C*
16.00 C* Check for restart information - get last rcd per user
17.00 C* PRDRCTP file access path is in LIFO sequence
18.00 C*
19.00 C USER CHAINPRDRCTR 20 Not found
20.00 C N20 DO Restart
21.00 C EXSR MOVLST Move to last
22.00 C SETON 71 Restart
23.00 C END
24.00 C*
25.00 C* Basic processing loop
26.00 C*
27.00 C L00P TAG
28.00 C EXFMTPROMPT
29.00 C 98 GOTO END End of pgm
30.00 C PRODCT CHAINPRDMSTR 61 Not found
31.00 C 61 GOTO L00P
32.00 C KEY KLIST
33.00 C KFLD PRODCT
34.00 C KFLD LOCATN
35.00 C KEY CHAINPRDLOCR 62 Not found
36.00 C 62 DO
37.00 C EXCPTRLSMST Release lck
38.00 C GOTO L00P
39.00 C END
40.00 C ADD QTY ONHAND Add
41.00 C ADD QTY LOCAMT
42.00 C UPDATPRDMSTR Update
43.00 C UPDATPRDLOCR Update
44.00 C*

45.00 C* Commit and move to previous fields

46.00 C*
47.00 C CMTID COMIT
48.00 C EXSR MOVLST Move to last
49.00 C GOTO L00P
50.00 C*
51.00 C* End of program processing

Figure 77. RPG Source (Part 1 of 2)

Chapter 22. Commitment Control 613

 52.00 C*
53.00 C END TAG
54.00 C SETON LR
55.00 C*56.00 C* Delete any records in the notify object
57.00 C*
58.00 C DLTLP TAG
59.00 C USER CHAINPRDRCTR 20 Not found
60.00 C N20 DO
61.00 C DELETPRDRCTR Delete
62.00 C GOTO DLTLP
63.00 C END
64.00 C*
65.00 C* Move to -Last Used- fields for operator feedback
66.00 C*
67.00 C MOVLST BEGSR
68.00 C MOVE PRODCT LSTPRD
69.00 C MOVE LOCATN LSTLOC
70.00 C MOVE QTY LSTQTY
71.00 C MOVE DESCRP LSTDSC
72.00 C ENDSR
73.00 OPRDMSTR E RLSMST

Figure 77. RPG Source (Part 2 of 2)

Using a Single Notify Object for All Programs

Using a single notify object for all programs is advantageous since all information
required to start again is in the same object and a standard approach to the notify
object can be used in all programs. In this situation, use a unique combination of
user and program identifications to make sure that the program accesses the
correct information when it starts again.

Because the information required to start again may vary from program to
program, an externally described data structure for the commit identification
should not be used. If a single notify object is used, the preceding program could
describe the data structure within the program rather than externally. For example:
1 10 USER
11 20 PGMNAM
21 23 PRODCT
24 29 LOCATN
30 49 DESC
50 51 0 QTY
52 220 DUMMY

In each program that uses this notify object, the information specified for the
commit identification would be unique to the program (the user and program
names are not unique). The notify object must be large enough to contain the
maximum information that any program would place in the commit identification.

Using a Standard Processing Program

A standard processing program is one way to start your application again using
one database file as the notify object for all applications. This approach assumes
that user profile names are unique by user for all applications using the standard
program.

For this approach, the physical file NFYOBJP is used as the notify object and
defined as:
Unique user profile name 10 characters
Program identification 10 characters
Information for

614 OS/400 Backup and Recovery V5R1

 starting again Character field
(This should be large
enough to contain the maximum
amount of information for starting
programs again that require
information for starting again.
This field is required by
the application programs.
In the example, it is
assumed to be a length of 200.)

The file is created with SHARE(*YES). The first two fields in the file are the key to
the file. (This file can also be defined as a data structure in RPG programs.)

Processing Flow
The standard program is called from applications that must start again. The
application programs pass this parameter list to the standard program:
v Request code
v Return code
v Data structure name (the contents of the notify object)

Request codes do the following:

v R (Read)
Retrieves the last record added to the notify object with the same key. The return
code is set as:
0 No record is available (no start again required).
1 Record returned in the information field for starting again (start again
required).
v W (Write)
Writes a record to the file. This code could be used if you use a notify object for
your own purposes. For example, if the program determines that the LUW
needs to be started again, the program could write a record to the notify object
to simulate what the system will do if a job or the system fails.
v D (Delete)
Deletes all records in the notify object with the same key. The return code is set
as:
0 No records exist to be deleted.
1 One or more records were deleted.
v O (Open)
The O request code is optional and is used to avoid having to start the
processing program each time it is called.
v C (Close)
After the open request code is used, using the close request code ensures the file
is closed.
v S (Search)
Returns the last record for this user. The program name is not used. This code
can be used in an initial program to determine whether starting again is
required.

Application Program Example

Following is an example of using a standard processing program. The application
shown in Figure 78 on page 617 performs as follows:

Chapter 22. Commitment Control 615

 1. The application program receives the user name in a parameter and uses it
with the program name as a unique identifier in the notify object.
2. The application program passes a request code of R to the standard commit
processing program, which determines if a record exists in the notify object.
3. If the standard commit processing program returns a code of 1, a record was
found and the application program presents the information needed to start
again to the user.
4. The application program proceeds with normal processing.
5. When an LUW is completed, values are saved for reference so the workstation
user can see what was done for the previous transaction.
The information saved is not provided by the notify object because the notify
object is updated only if a job or system failure occurs.

616 OS/400 Backup and Recovery V5R1

SEQNBR *... ... 1 ... ... 2 ... ... 3 ... ... 4 ... ... 5 ... ... 6 ... ... 7 ..

1.00 FPRDMSTP UF E K DISK KCOMIT

2.00 FPRDLOCP UF E K DISK KCOMIT
3.00 FPRDRCTD CF E WORKSTN
4.00 F*
5.00 F* The following is a compile time array which contains the
6.00 F* restart information used in the next example
7.00 F*
8.00 E RTXT 50 50 1 Restart text
9.00 I*
10.00 I* Data structure used for info passed to notify object
11.00 I*
12.00 ICMTID DS
13.00 I 1 10 USER
14.00 I 11 20 PGMNAM
15.00 I 21 23 PRODCT
16.00 I 24 29 LOCATN
17.00 I 30 49 DESCRP
18.00 I P 50 510QTY
19.00 I 52 170 DUMMY
20.00 I 171 220 RSTART
21.00 C *ENTRY PLIST
22.00 C PARM USER10 10
23.00 C*
24.00 C* Initialize fields used to communicate with std program
25.00 C*
26.00 C MOVE USER10 USER
27.00 C MOVEL'PRDRC2' PGMNAM
28.00 C MOVE 'R' RQSCOD Read Rqs
29.00 C CALL 'STDCMT'
30.00 C PARM RQSCOD 1
31.00 C PARM RTNCOD 1
32.00 C PARM CMTID 220 Data struct
33.00 C RTNCOD IFEQ '1' Restart
34.00 C EXSR MOVLST Move to last
35.00 C SETON 71 Restart
36.00 C END
37.00 C*
38.00 C* Initialize fields used in notify object
39.00 C*
40.00 C MOVEARTXT,1 RSTART Move text
41.00 C*
42.00 C* Basic processing loop
43.00 C*
44.00 C LOOP TAG
45.00 C EXFMTPROMPT
46.00 C 98 GOTO END
47.00 C PRODCT CHAINPRDMSTR 61 Not found
48.00 C 61 GOTO LOOP
49.00 C KEY KLIST
50.00 C KFLD PRODCT
51.00 C KFLD LOCATN

Figure 78. Application Program Example (Part 1 of 2)

Chapter 22. Commitment Control 617

SEQNBR *... ... 1 ... ... 2 ... ... 3 ... ... 4 ... ... 5 ... ... 6 ... ... 7 ..

52.00 C KEY CHAINPRDLOCR 62 Not found

53.00 C 62 DO
54.00 C EXCPTRLSMST Release lck
55.00 C GOTO LOOP
56.00 C END
57.00 C ADD QTY ONHAND Add
58.00 C ADD QTY LOCAMT
59.00 C UPDATPRDMSTR Update
60.00 C UPDATPRDLOCR Update
61.00 C*
62.00 C* Commit and move to previous fields
63.00 C*
64.00 C CMTID COMIT
65.00 C EXSR MOVLST Move to last
66.00 C GOTO LOOP
67.00 C* End of program processing
68.00 C*
69.00 C END TAG
70.00 C MOVE 'D' RQSCOD Dlt Rqs
71.00 C CALL 'STDCMT'
72.00 C PARM RQSCOD
73.00 C PARM RTNCOD
74.00 C PARM CMTID
75.00 C SETON LR
76.00 C*
77.00 C* Move to -Last Used- fields for operator feedback
78.00 C*
79.00 C MOVLST BEGSR
80.00 C MOVE PRODCT LSTPRD
81.00 C MOVE LOCATN LSTLOC
82.00 C MOVE DESCRP LSTDSC
83.00 C MOVE QTY LSTQTY
84.00 C ENDSR
85.00 OPRDMSTR E RLSMST
86.00 ** RTXT Restart Text
87.00 Inventory Menu - Receipts Option

Figure 78. Application Program Example (Part 2 of 2)

Standard Commit Processing Program

The standard commit (STDCMT) processing program performs the functions
required to communicate with a single notify object used by all applications. While
the commitment control function automatically writes an entry to the notify object,
a user-written standard program must process the notify object. The standard
program simplifies and standardizes the approach.

The program is written to verify the parameters that were passed and perform the
appropriate action as follows:
O=Open
The calling program requests the notify object file be kept open on return.
Because the notify object is opened implicitly by the RPG program, the
program should not close it. Indicator 98 is set so the program returns with
LR off to keep the program’s work areas and leaves the notify object open
so it can be called again without excess overhead.

618 OS/400 Backup and Recovery V5R1

C=Close
The calling program has determined it no longer needs the notify object
and requests a close. Indicator 98 is set off to allow a full close of the
notify object.
R=Read
The calling program requests that a record with matching key fields be
read and passed back. The program uses the passed key fields to attempt
to retrieve a record from NFYOBJP. If duplicate records exist for the same
key, the last record is returned. The return code is set accordingly and, if
the record existed, it is passed back in the data structure CMTID.
W=Write
The calling program requests a record to be written to the notify object to
allow the calling program to start again the next time it is called. The
program writes the contents of the passed data as a record in NFYOBJP.
D=Delete
The calling program requests that records for this matching key be deleted.
This function is usually performed at the successful completion of the
calling program to remove any information on starting again. The program
attempts to delete any records for passed key fields. If no records exist, a
different return code is passed back.
S=Search
The calling program requests a search for a record for a particular user
regardless of which program wrote it. This function is used in the program
for sign-on to indicate that starting again is required. The program uses
only the user name as the key to see if records exist. The return code is set
appropriately, and the contents of the last record for this key (if it exists)
are read and passed back.

Figure 79 on page 620 shows the standard commit processing program, STDCMT.

Chapter 22. Commitment Control 619

SEQNBR *... ... 1 ... ... 2 ... ... 3 ... ... 4 ... ... 5 ... ... 6 ... ... 7 ..

1.00 FNFYOBJP UF E K DISK A

2.00 ICMTID DS
3.00 I 1 10 UNQUSR
4.00 I 11 20 UNQPGM
5.00 I 21 220 BIGFLD
6.00 C *ENTRY PLIST
7.00 C PARM RQSCOD 1
8.00 C PARM RTNCOD 1
9.00 C PARM CMTID 220
10.00 C UNQUSR CABEQ*BLANKS BADEND H1 Invalid
11.00 C UNQPGM CABEQ*BLANKS BADEND H2 Invalid
12.00 C*
13.00 C* 'O' for Open
14.00 C*
15.00 C RQSCOD IFEQ 'O' Open
16.00 C SETON 98 End LR
17.00 C GOTO END
18.00 C END
19.00 C*
20.00 C* 'C' for Close
21.00 C*
22.00 C RQSCOD IFEQ 'C' Close
23.00 C SETOF 98
24.00 C GOTO END
25.00 C END
26.00 C*
27.00 C* 'R' for Read - Get last record for the key
28.00 C*
29.00 C RQSCOD IFEQ 'R' Read
30.00 C KEY KLIST
31.00 C KFLD UNQUSR
32.00 C KFLD UNQPGM
33.00 C KEY CHAINNFYOBJR 51 Not found
34.00 C 51 MOVE '0' RTNCOD
35.00 C 51 GOTO END
36.00 C MOVE '1' RTNCOD Found
37.00 C LOOPl TAG
38.00 C KEY READENFYOBJR 20 EOF
39.00 C 20 GOTO END
40.00 C GOTO LOOP1
41.00 C END
42.00 C*
43.00 C* 'W' FOR Write
44.00 C*
45.00 C RQSCOD IFEQ 'W' Write
46.00 C WRITENFYOBJR
47.00 C GOTO END
48.00 C END
49.00 C*
50.00 C* 'D' for Delete - Delete all records for the key
51.00 C*

Figure 79. Standard Commit Processing Program (Part 1 of 2)

620 OS/400 Backup and Recovery V5R1

SEQNBR *... ... 1 ... ... 2 ... ... 3 ... ... 4 ... ... 5 ... ... 6 ... ... 7 ..

52.00 C RQSCOD IFEQ 'D' Delete

53.00 C KEY CHAINNFYOBJR 51 Not found
54.00 C 51 MOVE '0' RTNCOD
55.00 C 51 GOTO END
56.00 C MOVE '1' RTNCOD Found
57.00 C LOOP2 TAG
58.00 C DELETNFYOBJR
59.00 C KEY READENFYOBJR 20 EOF
60.00 C N20 GOTO LOOP2
61.00 C GOTO END
62.00 C END
63.00 C*
64.00 C* 'S' for Search for the last record for this user
65.00 C* (Ignore the -Program- portion of the key)
66.00 C*
67.00 C RQSCOD IFEQ 'S' Search
68.00 C UNQUSR SETLLNFYOBJR 20 If equal
69.00 C N20 MOVE '0' RTNCOD
70.00 C N20 GOTO END
71.00 C MOVE '1' RTNCOD Found
72.00 C LOOP3 TAG
73.00 C UNQUSR READENFYOBJR 20 EOF
74.00 C N20 GOTO LOOP3
75.00 C GOTO END
76.00 C END
77.00 C*
78.00 C* Invalid request code processing
79.00 C*
80.00 C SETON H2 Bad RQS code
81.00 C GOTO BADEND
82.00 C*
83.00 C* End of program processing
84.00 C*
85.00 C END TAG
86.00 C N98 SETON LR
87.00 C RETRN
88.00 C* BADEND tag is used then fall thru to RPG cycle error return
89.00 C BADEND TAG

Figure 79. Standard Commit Processing Program (Part 2 of 2)

Deciding If It Is Necessary to Start Again

The initial program can call the standard commit processing program to determine
if it is necessary to start again. The workstation user can than decide whether or
not to start again.

The initial program passes a request code of S (search) to the standard program,
which searches for any record for the user. If a record exists, the information for
starting again is passed to the initial program and the information is displayed to
the workstation user.

The commit identification in the notify object should contain information that the
initial program can display identifying what program needs to be started again.
For example, the last 50 characters of the commit identification can be reserved to
contain this information. In the application program, this information could be in a
compile-time array and moved to the data structure in an initialization step.
Figure 78 on page 617 shows how to include this in the application program.

Chapter 22. Commitment Control 621

 Figure 80 is an example of an initial program that determines if a record exists in
the notify object.

SEQNBR *... ... 1 ... ... 2 ... ... 3 ... ... 4 ... ... 5 ... ... 6 ... ... 7

1.00 PGM
2.00 DCLF CMTINLD
3.00 DCL &RQSCOD *CHAR LEN(1) VALUE(S) /* Search */
4.00 DCL &RTNCOD *CHAR LEN(1)
5.00 DCL &CMTID *CHAR LEN(220)
6.00 DCL &USER *CHAR LEN(10)
7.00 DCL &INFO *CHAR LEN(50)
8.00 RTVJOBA USER(&USER)
9.00 CHGVAR &CMTID (&USER *CAT XX)
10.00 /* The XX is reqrd to prevent a blank Pgm nam */
11.00 CALL STDCMT PARM(&RQSCOD &RTNCOD &CMTID)
12.00 IF (&RTNCOD *EQ '1') DO /* RESTART REQD */
13.00 CHGVAR &INFO %SST(&CMTID 171 50)
14.00 SNDRCVF RCDFMT(RESTART)
15.00 ENDDO
16.00 /* */
17.00 /* Enter normal initial program statements */
18.00 /* or -TFRCTL- to first menu program */
19.00 /* */
20.00 ENDPGM

Figure 80. Initial Program Example

Commitment Control Practice Problem

This practice problem will assist you in understanding commitment control and its
requirements. The following steps assume you are familiar with the OS/400
licensed program and the data file utility (DFU), and have read this chapter. Before
beginning this problem, do the following:
v Create a special library for this practice problem. In the instructions, the library
is called CMTLIB. Substitute the name of your library where you see CMTLIB.
v Create source files and a job description.

Perform the following steps:

1. Create a physical file named ITMP (item master file). The data description
specification (DDS) for this file is:
10 A R ITMR
20 A ITEM 2
30 A ONHAND 5 0
40 A K ITEM
2. Create a physical file named TRNP (transaction file). This file is used as a
transaction log file. The DDS for this file is:
10 A R TRNR
20 A QTY 5 0
30 A ITEM 2
40 A USER 10
3. Create a logical file named TRNL (transaction logical). This file is used to
assist in starting the application again. The USER field is the type LIFO
sequence. The DDS for this file is:
10 LIFO
20 A R TRNR PFILE (TRNP)
30 A K USER

622 OS/400 Backup and Recovery V5R1

 4. Enter the STRDFU command, and create a DFU application named ITMU for
the ITMP file. Accept the defaults offered by DFU during the application
definition.
5. Type the command CHGDTA ITMU and enter the following records for the
ITMP file:
Item On Hand
AA 450
BB 375
CC 4000
6. End the program using F3. This entry provides some data against which the
program will operate.
7. Create the CL program Item Process (ITMPCSC) as follows:
PGM
DCL &USER *CHAR LEN(10)
RTVJOBA USER(&USER)
CALL ITMPCS PARM(&USER)
ENDPGM

This is the control program that calls the ITMPCS program. It retrieves the
user name and passes it to the processing program. This application assumes
that unique user names are used.
8. Create a display file named ITMPCSD from the DDS shown in Figure 82 on
page 629.
There are two formats, the first for the basic prompt display and the second to
allow the operator to review the last transaction entered. This display file is
used by the ITMPCS program.
9. Study the logic flow provided in Figure 83 on page 630.
10. Enter the STRSEU command and type the source shown in Figure 81 on
page 624.

Chapter 22. Commitment Control 623

SEQNBR *... ... 1 ... ... 2 ... ... 3 ... ... 4 ... ... 5 ... ... 6 ... ... 7 ..

1.00 FITMP UF E K DISK

2.00 F* KCOMIT
3.00 FTRNP O E DISK
4.00 F* KCOMIT
5.00 FTRNL IF E K DISK
6.00 F TRNR KRENAMETRNR1
7.00 FITMPCSD CF E WORKSTN
8.00 C* Enter parameter with User name for -TRNP- file
9.00 C *ENTRY PLIST
10.00 C PARM USER 10
11.00 C LOOP TAG
12.00 C EXFMTPROMPT
13.00 C* Check for CF3 for end of program
14.00 C 93 DO End of Pgm
15.00 C SETON LR
16.00 C RETRN
17.00 C END
18.00 C* Check for CF4 for review last transaction
19.00 C 94 DO Review last
20.00 C* Check for existence of a record for this user in -TRNL- file
21.00 C USER CHAINTRNR1 64 Not found
22.00 C 64 GOTO LOOP
23.00 C EXFMTREVW
24.00 C GOTO LOOP
25.00 C END
26.00 C* Access Item record
27.00 C ITEM CHAINITMR 62 Not found
28.00 C* Handle -not found- Condition
29.00 C 62 GOTO LOOP
30.00 C* Does sufficient quantity exist
31.00 C ONHAND SUB QTY TEST 50 61 Minus
32.00 C* Handle insufficient quantity
33.00 C 61 DO
34.00 C* Release Item record which was locked by the CHAIN for update
35.00 C EXCPTRLSITM
36.00 C GOTO LOOP
37.00 C END
38.00 C* Change ONHAND and update the Item record
39.00 C Z-ADDTEST ONHAND
40.00 C UPDATITMR
41.00 C* Test for Special Simulation Conditions
42.00 C ITEM IFEQ 'CC'
43.00 C* Simulate program need for rollback
44.00 C QTY IFEQ 100
45.00 C SETON 63 Simult Rlbck
46.00 C* ROLBK
47.00 C GOTO LOOP
48.00 C END

Figure 81. RPG Source Program for ITMPTCS (Part 1 of 2)

624 OS/400 Backup and Recovery V5R1

| 49.00 C* Simulate an abnormal program cancellation by Div by zero
| 50.00 C* Operator Should respond -C- to inquiry message
| 51.00 C QTY IFEQ 101
| 52.00 C Z-ADD0 ZERO 30
| 53.00 C TESTZ DIV ZERO TESTZ 30 Msg occurs
| 54.00 C END
| 55.00 C* Simulate an abnormal job cancellation by DSPLY.
| 56.00 C* Operator Should System Request to another job
| 57.00 C* and cancel this one with OPTION(*IMMED)
| 58.00 C QTY IFEQ 102
| 59.00 C 'CC=102' DSPLY Msg occurs
| 60.00 C END
| 61 00 C END ITEM=CC
| 62.00 C* Write the -TRNP- file
| 63 00 C WRITETRNR
| 64.00 C* Commit the update to -ITMP- and write to -TRNP-
| 65.00 C* COMIT
| 66.00 C GOTO LOOP
| 67.00 OITMR E RLSITM

|
| Figure 81. RPG Source Program for ITMPTCS (Part 2 of 2)
|
11. Enter the CRTRPGPGM command to create program ITMPCS from the source
entered in the previous step.
12. Type the command CALL ITMPCSC, press Enter, and press F4. A message
should appear stating that there are no entries for this operator.
13. Enter the following data to see if the program operates correctly:
Quantity
Item
3 AA
4 BB
14. Press F4. The review display should appear with the BB item last entered.
Enter the following data:
Quantity
Item
5 FF (Invalid item number message should occur.)
9000 BB (Insufficient quantity error message should occur.)
100 CC (Rollback message should occur.)
102 CC (RPG DSPLY operation should occur. Press the Enter key.)
101 CC (The program should display an inquiry message stating that a
divide-by-zero condition has occurred or end, depending on the
setting of job attribute INQMSGRPY. If the inquiry message appears,
enter C to cancel the RPG program and then C to cancel the CL
program on the subsequent inquiry. This simulates an unexpected
error condition.)
15. Type the Display Data command DSPDTA ITMP.
See if the records AA and BB have been updated correctly. The values should
be AA = 447, BB = 371, and CC = 3697. Note that the quantities subtracted
from CC occurred, but the transaction records were not written.
16. Create a journal receiver for commitment control. Use the Create Journal
Receiver (CRTJRNRCV) command to create a journal receiver called RCVR1 in
| the CMRLIB library. Specify a threshold of at least 5000KB.A larger threshold
| is recommended if your system has sufficient space in order to maximize the

Chapter 22. Commitment Control 625

| time between generation of new journal receivers to minimize the
| performance impacts of too frequent change journals
17. Create a journal for commitment control. Use the Create Journal (CRTJRN)
command to create a journal called JRNTEST in the CMTLIB library. Because
this journal is used only for commitment control, specify MNGRCV(*SYSTEM)
DLTRCV(*YES). For the JRNRCV parameter, specify the journal receiver that
you created in step 16 on page 625.
18. Use the Start Journal Physical File (STRJRNPF) command with the parameters
FILE(CMTLIB/ITMP CMTLIB/TRNP) JRN(CMTLIB/JRNTEST) to journal the
files to be used for commitment control.
The IMAGES parameter uses a default of *AFTER, meaning only after-image
changes of the records appear in the journal. The files ITMP and TRNP have
now started journaling.
Normally, you would save the files after starting journaling. You cannot apply
journaled changes to a restored file that does not have the same JID as the
journal entries. Because this practice problem does not require you to apply
journaled changes, you can skip saving the journaled files.
19. Type the command CALL ITMPCSC and enter the following transactions:
Quantity
Item
5 AA
6 BB

End the program by pressing F3.

20. Type the Display Journal command: DSPJRN CMTLIB/JRNTEST.
Note the entries appearing in the journal. The same sequence of entries (’R
UP’ = update of ITMP followed by ’R PT’ = record added to TRNP) occurs in
the journal as was performed by the program. This is because a logical file is
defined over the physical file TRNP and the system overrides the RPG
default. If no logical file existed, the RPG assumption of SEQONLY(*YES)
would be used, and a block of PT entries would appear because the records
would be kept in the RPG buffer until the block is full.
21. Change the CL program ITMPCSC as follows (the new statements are shown
with an asterisk):
PGM
DCL &USER *CHAR LEN(10)
RTVJOBA USER(&USER)
* STRCMTCTL LCKLVL(*CHG)
CALL ITMPCS PARM(&USER)
* MONMSG MSGID(RPG9001) EXEC(ROLLBACK)
* ENDCMTCTL
ENDPGM

The STRCMTCTL command sets up the commitment control environment.

The LCKLVL word specifies that records read for update but not updated can
be released during the transaction. The MONMSG command handles any RPG
escape messages and performs a ROLLBACK in case the RPG program
abnormally ends. The ENDCMTCTL command ends the commitment control
environment.
22. Delete the existing ITMPCSC program and create it again.
23. Change the RPG program to remove the comment symbols at statements 2.00,
4.00, 46.00, and 65.00. The source is now ready for use with commitment
control.

626 OS/400 Backup and Recovery V5R1

24. Delete the existing ITMPCS program and create it again. The program is now
ready to operate under commitment control.
25. Type the command CALL ITMPCSC and the following transactions:
Quantity
Item
7 AA
8 BB
26. Use System Request and request the option to display the current job. When
the Display Job display appears, select option 16 to request the display of the
commitment control status.
Note the values on the display. There should be two commits because two
commit statements were run in the program.
27. Press F9 to see a list of the files under commitment control and the amount of
activity for each file.
28. Return to the program and end it by pressing F3.
29. Type DSPJRN CMTLIB/JRNTEST and note the entries for the files and the special
journal entries for commitment control:
’C BC’ STRCMTCTL command occurred.
’C SC’ Start commit cycle. This occurs whenever the first database operation
in the transaction causes a record to be inserted, updated, or deleted
as part of commitment control.
’C CM’
Commit operation has occurred.
’C EC’ ENDCMTCTL command occurred.

Note how the commitment control before- and after-images (’R UB’ and ’R
UP’ types) automatically occur even though you had originally requested
IMAGES(*AFTER) for the journal.
30. Type the command CALL ITMPCSC and the following transactions:
Quantity
Item
12 AA
100 CC (This is the condition to simulate the need for an application use
of rollback. The CC record in the ITMP file, which was updated by
RPG statement 40.00 is rolled back.)
31. Press F4 to determine the last transaction entered.
The last committed transaction is the entry for item AA.
32. Use System Request and request the Display Current Job option. When the
Display Job display appears, request the display of the commitment control
status.
Note the values on the display and how they have been changed by the
rollback.
33. Return to the program.
34. Return to the basic prompt display and end the program by pressing F3.
35. Type the command DSPJRN CMTLIB/JRNTEST.
Note the additional entries that appear in the journal for the use of the
rollback entry (’C RB’ entry). When the ITMP record is rolled back, three

Chapter 22. Commitment Control 627

 entries are placed in the journal. This is because any change to the database
file under commitment control produces a before (’R BR’) and after (’R UR’)
entry.
36. Display the entries with journal code R and these entry types: UB, UP, BR,
and UR. Use option 5 to display the full entries. Because the Quantity field is
in packed decimal, use F11 to request a hex display. Note the following:
v The on-hand value of the ITMP record in the UB record
v How the on-hand value is reduced by the UP record
v How the BR record is the same as the UP record
v How the UR record returns the value as originally displayed for the UB
record

The last entry is the RB entry for the end of the rollback.
37. Type the command CALL ITMPCSC, press Enter, and press F4. Note the last
transaction entered.
38. Type the following transactions:
Quantity
Item
13 AA
101 CC (This is the condition to simulate an unexpected error condition,
which causes the program to end. The simulation occurs by dividing a
field by 0. The program will display an inquiry message or end,
depending on the setting of the job attribute INQMSGRPY. If the
inquiry message appears, enter C to end the program. Because the CL
program was changed to monitor for RPG program errors, the second
inquiry which occurred does not occur.)
39. Type the command DSPJRN CMTLIB/JRNTEST.
The same type of rollback handling has occurred, but this time the rollback
was caused by the EXEC parameter of the MONMSG command in the CL
program instead of the RPG program. Display the two RB entries to see which
program caused them.
40. Type the command WRKJOB and write down the fully qualified job name to
be used later.
41. Type the command CALL ITMPCSC and enter the following transaction:
Quantity
Item
14 AA
102 CC (The RPG DSPLY operation should occur to the external message
queue. Use the System Request key and select option 1 on the system
request menu to transfer to a secondary job.)
42. Sign on to the second job and reestablish your environment.
43. Type the command ENDJOB and specify the fully qualified job name
identified earlier and OPTION(*IMMED). This simulates an abnormal job or
system end.
44. Wait about 30 seconds, type the command CALL ITMPCSC and press F4.
Note the last committed transaction. It should be the AA item entered earlier.
45. Return to the basic prompt display and end the program by pressing F3.
46. Type the command DSPJRN CMTLIB/JRNTEST.

628 OS/400 Backup and Recovery V5R1

 The same type of rollback handling has occurred, but this time the rollback
was caused by the system instead of one of the programs. The RB entry was
written by the program QWTPITPP, which is the work management abnormal
end program.

You have now used the basic functions of commitment control. You can proceed
with commitment control on your applications or try some of the other functions
such as:
v Using a notify object
v Locking records that are only read with LCKLVL(*ALL)
v Locking multiple records in the same file with LCKLVL(*ALL)
Figure 82 shows the DDS for the display file.

SEQNBR *... ... 1 ... ... 2 ... ... 3 ... ... 4 ... ... 5 ... ... 6 ... ... 7 ..

1.00 A R PROMPT
2.00 A CA03(93 'End of program')
3.00 A CA04(94 'Review last')
4.00 A SETOFF(64 'No rcd to rvw')
5.00 A 1 2'INVENTORY TRANSACTIONS'
6.00 A 3 2'Quantity'
7.00 A QTY 5 0I +1
8.00 A 61 ERRMSG('Invalid +
9.00 A quantity' 61)
10.00 A +5'ITEM'
11.00 A ITEM 2 I +1
12.00 A 62 ERRMSG('Invalid +
13.00 A Item number' 62)
14.00 A 63 ERRMSG('Rollback +
15.00 A occurred' 63)
16.00 A 64 24 2'CF4 was pressed and +
17.00 A there are no +
18.00 A transactions for +
19.00 A this user'
20.00 A DSPATR(HI)
21.00 A 23 2'CF4 Review last +
22.00 A transaction'
23.00 A R REVW
24.00 A 1 2'INVENTORY TRANSACTIONS'
25.00 A +5'REVIEW LAST TRANSACTION'
26.00 A 3 2'Quantity'
27.00 A QTY 5 0 +1EDTCDE(Z)
28.00 A +5'Item'
29.00 A ITEM 2 +1

Figure 82. DDS for the Display File

Figure 83 on page 630 illustrates the logic flow for the practice problem.

Chapter 22. Commitment Control 629

 B Begin

17 1
Write
Transaction Retrieve
Record User Name
A

18 2
Basic
Commit Prompt Prompt

23 3
End of Yes End of
Program Program
Functions F3

No

5 4 19
Read No Yes Read Last
Item Review
for This
Record Last
F4 User

6 7 20 21
No No Item No No Review
Found Record Found Record
Error Error
62 64

Yes Yes

8
A A
Check
on Hand

9 10 22
Release Last
Sufficient No Lock on
Display Transaction
Quantity Item Information
Record

Yes

12 11
Modify/ A
Insufficient
Update
Quantity
Item
Error
Record
61

13
No A
Check
B
Simulation

Yes

RV2W416-0

Figure 83. Logic Flow of the Practice Problem (Part 1 of 2)

630 OS/400 Backup and Recovery V5R1

 C

14
No
Quantity
= 100

Yes

15

Quantity No
Rollback
= 101

Yes

16
A
Divide by No
Quantity B
Zero = 102

Yes
Response to
inquiry should
be c = cancel.
Display

Operator should
system request
to a separate
job and end
this one.

RSLS817-3

Figure 83. Logic Flow of the Practice Problem (Part 2 of 2)

Steps Associated with Logic Flow

The following steps are associated with the logic flow of the practice problem.
1. Retrieve the user name that is passed in as a parameter. This is used to write
to the TRNP file and also used to retrieve the last transaction entered by each
operator. This application assumes unique user names for operators.
2. Prompt for the basic display using the format name PROMPT.
3. If F3 is pressed, start an end of program function.
4. If F4 is pressed, start a routine to access the last transaction entered by the
operator.
5. Read the item record using the field ITEM. Since the file is an update file, this
request locks the record.
6. Check for a not found condition in the file ITMP.
7. If no ITMP record exists, set on indicator 62 to cause the error message and
return to step 2.
8. Subtract the quantity requested (QTY) from the on hand balance (ONHAND)
into a work area.
9. Check to see if sufficient quantity exists to meet the request.
10. If insufficient quantity exists, release the lock on the record in the ITMP file.
This step is needed because of insufficient quantity.

Chapter 22. Commitment Control 631

 11. Set on indicator 61 to signal an insufficient quantity display error message and
return to step 2 on page 631.
12. Change the ONHAND field for the new balance and update the ITMR record.
13. Check for special entry in the ITEM field that can be used to simulate
conditions where ROLLBACK is required.
14. Check for QTY=100. Issue a ROLLBACK operation. This simulates a condition
where the program senses a need for rollback.
15. Check for QTY=101. Cause an exception in the program that will produce an
inquiry message. Use divide by zero for this function. The operator should
enter C to cancel the program unless the job description INQMSGRPH option
provides an automatic reply. This simulates a condition where an unexpected
error has occurred and the operator cancels the program.
16. Check for QTY=102. Issue a display with inquiry operation. This stops the
program at this step and allows the use of the System Request key to get to a
different job. Cancel the updating job. This simulates a condition where an
abnormal job or system end has occurred in the middle of a commit
boundary.
17. Write the transaction record to TRNP.
18. Commit the records for the transaction and return to step 2 on page 631.
19. Read the first record on the access path for file TRNL, using USER as the key.
Since this file is in LIFO sequence, this will be the last transaction record
entered by this user.
20. Check for a record not found condition in the TRNL file that would be caused
if the file does not contain entries for this user.
21. If there is no record for this user, set on indicator 64 to cause an error message
and return to step 2 on page 631.
22. Display the last transaction entered for this user. This information can be used
if the operator forgets what was previously entered or when the transaction is
restarted. When the operator responds, return to step 2 on page 631.
23. Perform any end of program functions.

632 OS/400 Backup and Recovery V5R1

Part 7. Alternate Installation Device
Chapter 23. Using an Alternate Installation
Device . . . . . . . . . . . . . . . 635
Alternate Installation Device-Overview . . . . . 635
Install and Recovery Implications for Models
600, 620, 720, and S20 . . . . . . . . . 635
How to Set Up an Alternate Installation Device 635
How to Disable an Alternate Installation Device 638
Recovering Your System Using an Alternate
Installation Device. . . . . . . . . . . 639

© Copyright IBM Corp. 1997, 2000, 2001 633

634 OS/400 Backup and Recovery V5R1
Chapter 23. Using an Alternate Installation Device
Alternate Installation Device-Overview
New function was added to V4R1M0 of OS/400 allowing you to perform
installation and recovery procedures using a combination of devices. Previously,
these types of activities could only be performed using devices attached to the first
system bus. Now, you can use a combination of devices that are attached on the
first system bus and on additional buses. You can attach the alternate installation
to the first system bus.

If you use this function, the system uses existing support (a device on the first
system bus) to install or recover a portion of the Licensed Internal Code. This
allows you to perform an IPL-type D. With the new alternate installation device
support; the system continues the operation by using media in the alternate
installation device. This new function supports installation and recovery from
storage media. This media is a SAVSYS media volume or distribution media
volume which you created, that contains Licensed Internal Code and may contain
the operating system, licensed programs, and data.

Some models, typically with 3590 tape devices attached, may see a performance
improvement when using an alternate installation device for save operations.

Install and Recovery Implications for Models 600, 620, 720,

and S20
You can use the alternate installation device function for any installation or
recovery that requires replacing Licensed Internal Code. Some models may require
that you set up an alternate installation device. After which, you use the alternate
installation device to install the distribution media created by a central site or for
recovering using a SAVSYS media volume.

Attention!
In Models 600, 620, 720, and S20, the IOPs to which certain older tape devices
attach require an expansion unit in order to use the devices. The following
tape devices are affected: 2440, 3422, 3430, 9347, 3480, some models of 3490,
and 7208-002. Other 7208 models and 3490 models Exx, C11 and C22 are
supported without an expansion unit. 3490 models C1A and C2A can be
converted to SCSI format which is supported without an expansion unit. If
you want to use these older tape devices as alternate installation devices on
models 600 and 620, you need the expansion unit and you need to set up the
devices as alternate installation devices.

How to Set Up an Alternate Installation Device

Before you use an alternate installation device, you need to ensure that you define
it on a bus, and you must enable the device. You need to record and keep the
logical address of the system bus and system card at which the alternate
installation device is attached. If you have the device set up on different bus and
you do not have this information available, you will not be able to complete
installations or recoveries.

© Copyright IBM Corp. 1997, 2000, 2001 635

 If you change the configuration of your system, you need to ensure that this
information is correct before you start to use the alternate installation device.

Do the following to set the addresses and enable the alternate installation device:

Note: You need to know the password for Dedicated Service Tools to perform this
procedure.
1. Use the control panel to set the mode to Manual. Then perform an IPL using
the command: PWRDWNSYS OPTION(*IMMED) RESTART(*YES) IPLSRC(B).

Attention logical partitioning users!

If you are going to use this command on the primary partition, be sure
to power off all secondary partitions before running the command.

2. When the IPL or Install the System display appears, select option 3 (Use
Dedicated Service Tools (DST)) and press the Enter key.
3. The Dedicated Service Tools (DST) Sign On display appears.

Dedicated Service Tools (DST) Sign On

System: SYSTEMA
Type choices, press Enter.
DST user . . . . . . . . . . . . QSECOFR
DST password . . . . . . . . . . _______

Sign on using the QSECOFR user profile.

Note: Do not sign on with a profile other than QSECOFR.

4. The Use Dedicated Service Tools (DST) menu appears. From the Use
Dedicated Service Tools (DST) menu, do the following:

┌──────────────────────┐ Select option 5

│ DST menu │ (Work with DST environment)
│ │
│ │
│ ┌──────────────────┴─────┐ Select option 2
│ 5 │ Work with DST │ (System devices)
└───┤ Environment │
│ │
│ ┌────────────────────┴───┐ Select option 5
│ 2 │ Work with System │ (Alternate installation device)
└───┤ Devices │
│ │
│ ┌────────────────────┴───┐
│ 5 │ Select Alternate │
└───┤ Installation Device │
│ │
│ │
│ │
└────────────────────────┘

5. From the Select Alternate Installation Device display, type a 5 (Display details)
next to the resource you want and press the Enter key.

636 OS/400 Backup and Recovery V5R1

 Select Alternate Installation Device
System: SYSTEMA
Type option, press Enter.
1=Select 5=Display details

Resource Serial
Option Name Type Model Number Selected
_ TAP01 6380 001 00-00000
_ TAP08 3287 030 00-00000
_ TAP02 6380 001 00-00000
_ TAP05 3287 030 00-00000 *
_ TAP09 6380 001 00-00000
_ TAP16 3287 030 00-00000

F2=Deselect device F3=Exit F5=Refresh F12=Cancel

6. The Display Device Details display appears.

Display Device Details

System: SYSTEMA
Resource Serial
Name Type Model Number
TAP02 3490 C11 00-0000000

Physical location:
Location text . . . . . . . :
Frame ID . . . . . . . . . :
Card slot . . . . . . . . . :

Logical address:
SPD bus:
System bus . . . . . . . . : 0002
System board . . . . . . . : 0000
System card . . . . . . . . : 0002

Storage:
I/O bus number . . . . . . : 0000
Controller . . . . . . . . : 0007
Device address . . . . . . : 0000

F3=Exit F12=Cancel

You need to have a record of the addresses assigned to the alternate

installation device selected for installing and recovering your system. Record
the following information:
v Type/Model: _________
v System bus: _________
v System card: _________
Notes:
a. You may want to repeat this process to record the addresses for all
alternate installation devices that appear in step 5 on page 636. You should
store this information in a safe place, such as the location where your
recovery information and recovery media are stored.
b. If there more than one alternate installation device is defined, only one can
be enabled.
c. The alternate installation device is actually the first device that is found on
an SPD bus that contains installation media. You should ensure that only

Chapter 23. Using an Alternate Installation Device 637

 one device contains valid installation media. This prevents you from
loading the wrong version of the Licensed Internal Code.

Press the Enter key.

7. The Select Alternate Installation Device display appears. Type a 1 (Select) next
to the resource you want and press the Enter key.
8. You should see the following message at the bottom of the display:
Alternate installation device updated
9. Press F3 (Exit) to return to the Use Dedicated Service Tools (DST) display.
10. Press F3 (Exit) again. The Exit Dedicated Service Tools (DST) display appears.

Exit Dedicated Service Tools

System: SYSTEMA
Select one of the following:

1. Exit Dedicated Service Tools (DST)

2. Resume Dedicated Service Tools

Type a 1 (Exit Dedicated Service Tools (DST)) and press the Enter key.
11. The next display display you see is the IPL or Install the System display. Type
a 1 (Perform an IPL) and press the Enter key to complete the procedure.

How to Disable an Alternate Installation Device

You may need to disable an alternate installation device for one of the following
reasons:
v To continue with an installation using CD-ROM media.
v To enable a different device as an alternate installation device.
v To correct the logical address if hardware has been moved or changed.
1. Use the control panel to set the mode to Manual. Then perform an attended
IPL using the command: PWRDWNSYS OPTION(*IMMED) RESTART(*YES) IPLSRC(B).

Attention logical partitioning users!

If you are going to use this command on the primary partition, be sure to
power off all secondary partitions before running the command.

Note: An alternative to this step is to use the control panel to select function
21. (Dedicated Service Tools). If you use this alternative, the next step is
step 3.
2. When the IPL or Install the System display appears, select option 3 (Use
Dedicated Service Tools (DST)) and press the Enter key.
3. The Dedicated Service Tools (DST) Sign On display appears. Sign on using the
QSECOFR user profile.
4. The Use Dedicated Service Tools (DST) menu appears. From the Use Dedicated
Service Tools (DST) menu, do the following:

638 OS/400 Backup and Recovery V5R1

 ┌──────────────────────┐ Select option 5
│ DST menu │ (Work with DST environment)
│ │
│ │
│ ┌──────────────────┴─────┐ Select option 2
│ 5 │ Work with DST │ (System devices)
└───┤ Environment │
│ │
│ ┌────────────────────┴───┐ Select option 5
│ 2 │ Work with System │ (Alternate installation device)
└───┤ Devices │
│ │
│ ┌────────────────────┴───┐
│ 5 │ Select Alternate │
└───┤ Installation Device │
│ │
│ │
│ │
└────────────────────────┘

5. At the Select Alternate Installation Device display, type a 1 next to the device
you want to disable, and press F2 (Deselect device).
6. You should see the following message at the bottom of the display:
Alternate installation device disabled
7. Press F3 (Exit) to return to the Use Dedicated Service Tools (DST) display.
8. Press F3 (Exit) again. The Exit Dedicated Service Tools (DST) display appears.
Type a 1 (Exit Dedicated Service Tools (DST)) and press the Enter key.
9. The next display you see is the IPL or Install the System display. Type a 1
(Perform an IPL) and press the Enter key to complete the procedure.

Recovering Your System Using an Alternate Installation

Device
If you use an alternate installation device, you need to ensure that you set up the
device and that you enable the device. You also need to have the CD-ROM for
Licensed Internal Code as well as your save media.

Chapter 23. Using an Alternate Installation Device 639

640 OS/400 Backup and Recovery V5R1
Part 8. Disk Configuration and Protection — Procedures
Chapter 24. Procedures for Configuring Disks How to Change the Storage Threshold for an
and Disk Protection. . . . . . . . . . . 643 Auxiliary Storage Pool . . . . . . . . . . 672
Choosing the Right Procedure for Configuring How to Change the Storage Threshold for the
Disks . . . . . . . . . . . . . . . . 643 System Auxiliary Storage Pool. . . . . . . . 673
Configuring Disks on a New System–Checklist 1 644 How to Move a Disk Unit to a Different Auxiliary
Adding Disk Units without Device Parity Storage Pool . . . . . . . . . . . . . . 676
Protection–Checklist 2 . . . . . . . . . 645 How to Remove a Disk Unit from an Auxiliary
Adding Disk Units to a 9337 Disk Array Storage Pool . . . . . . . . . . . . . . 678
Subsystem–Checklist 3 . . . . . . . . . 646 How to Delete an Auxiliary Storage Pool . . . . 680
Adding a New 9337 Disk Array Subsystem and How to Calculate Space Requirements for an
Starting Device Parity Protection on It–Checklist Auxiliary Storage Pool . . . . . . . . . . 681
4 . . . . . . . . . . . . . . . . 647 Task 1–Calculating the Current Storage Used 682
Adding a New 9337 Disk Array Subsystem That Task 2–Calculating Storage Needed . . . . . 683
Has Device Parity Protection Already How to Display the Objects in a User ASP . . . 684
Started–Checklist 5 . . . . . . . . . . 648 Balancing an Auxiliary Storage Pool . . . . . . 685
Adding Disk Units to an Existing Input/Output Capacity Balancing . . . . . . . . . . 685
Processor–Checklist 6. . . . . . . . . . 649 Usage Balancing . . . . . . . . . . . 685
Adding a New Input/Output Hierarchical Storage Management (HSM)
Processor–Checklist 7. . . . . . . . . . 650 Balancing. . . . . . . . . . . . . . 685
Moving Disk Units Between Non-Mirrored Transferring Objects between Auxiliary Storage
Auxiliary Storage Pools–Checklist 8 . . . . . 652 Pools . . . . . . . . . . . . . . . . 686
Moving Disk Units Between Mirrored Auxiliary How to Move Authorities to a Different ASP 686
Storage Pools–Checklist 9 . . . . . . . . 653 How to Transfer a Library to a Different ASP 687
Deleting an Auxiliary Storage Pool–Checklist 10 654 How to Transfer a Folder to a Different ASP . . 687
Removing Disk Units Without Device Parity How to Transfer Journals and Objects to a
Protection–Checklist 11 . . . . . . . . . 655 Different ASP . . . . . . . . . . . . 687
Removing Disk Units with Device Parity How to Create Objects in a Library User ASP 688
Protection (9337) from an ASP Without Mirrored Placing a Document in a User ASP–Example 689
Protection–Checklist 12 . . . . . . . . . 656 Placing an Object in a User ASP–Example 689
Removing Disk Units That Have Device Parity | Creating a UDFS in User ASP–Example . . 689
Protection(9337) from an ASP with Mirrored How to Place Journal Receivers in a User ASP 689
Protection–Checklist 13 . . . . . . . . . 657 Placing Journal Receivers in a Library User
Removing Disk Units That Have Device Parity ASP . . . . . . . . . . . . . . 690
Protection from an ASP Without Mirrored How to Move Journal Receivers From an
Protection–Checklist 14 . . . . . . . . . 658 Overflowed User ASP . . . . . . . . . 690
Removing Disk Units That Have Device Parity How to Reset a Journal with a Status of
Protection from an ASP With Mirrored Overflowed . . . . . . . . . . . . . 691
Protection–Checklist 15 . . . . . . . . . 659 How to Work with Nonlibrary User ASPs . . . . 692
Using System Service Tools and Dedicated Creating Objects in a Nonlibrary User ASP . . 692
Service Tools . . . . . . . . . . . . 660 Transferring an Object to a Nonlibrary User ASP 693
How to Start Dedicated Service Tools (DST) 660 Transferring a Journal to a Nonlibrary User ASP 693
How to Stop Dedicated Service Tools (DST) 662 Placing Journal Receivers in a Nonlibrary
Starting System Service Tools (SST) . . . . 662 User ASP . . . . . . . . . . . . . 694
Stopping System Service Tools (SST). . . . 662
How to Display Your Disk Configuration . . . 663 Chapter 26. Working with Device Parity
Displaying Your Disk Protection . . . . . . . . . . . . . . 695
Configuration–Hardware View . . . . . 663 Starting Device Parity Protection . . . . . . . 695
Displaying Your Disk How to Start Device Parity Protection for a 9337
Configuration–Software View . . . . . . 665 Disk Array Subsystem . . . . . . . . . 695
Interpreting Disk Configuration and Status How to Start Device Parity Protection for an
Displays . . . . . . . . . . . . . 667 Input/Output Processor . . . . . . . . . 699
Stopping Device Parity Protection . . . . . . 701
Chapter 25. Working with Auxiliary Storage How to Stop Device Parity Protection for a 9337
Pools . . . . . . . . . . . . . . . . 669 Disk Array Subsystem . . . . . . . . . 701
How to Add Disk Units to an Auxiliary Storage How to Stop Device Parity Protection on an
Pool . . . . . . . . . . . . . . . . 669 Input/Output Processor . . . . . . . . . 704

© Copyright IBM Corp. 1997, 2000, 2001 641

How to Include a Disk Unit in Device Parity
Protection . . . . . . . . . . . . . . 705
How to Exclude a Disk Unit from Device Parity
Protection . . . . . . . . . . . . . . 707
How to Display Device Parity Status . . . . . 708
How to Enable Disk Units Attached to the MFIOP
to Use Device Parity Protection . . . . . . . 710

Chapter 27. Working with Mirrored Protection 719

Mirrored Protection–Configuration Rules . . . . 719
How to Start Mirrored Protection. . . . . . . 719
What the System Does When You Start Mirrored
Protection . . . . . . . . . . . . . 722
Mirrored Protection Configuration Errors . . . . 722
How to Stop Mirrored Protection . . . . . . . 723

Chapter 28. Working with Disk Compression 725

Introduction to Disk Compression . . . . . . 725
Restrictions and Considerations . . . . . . 725
Disk Compression and Capacity . . . . . . 726
Disk Unit Full Considerations . . . . . . . 728
How The System Responds to Disk Unit Full 728
SRC Code A6xx 0277 . . . . . . . . . . . 729
User Action 1 . . . . . . . . . . . . 730
User Action 2 . . . . . . . . . . . . 731
User Action 3 . . . . . . . . . . . . 731
User Action 4 . . . . . . . . . . . . 731
Examples of A6xx 0277 . . . . . . . . . 732
How to Start Disk Compression . . . . . . . 732
How to Stop Disk Compression . . . . . . . 735
Procedural Sequences for Configuring Disks and
Protection . . . . . . . . . . . . . . 737
Adding a New I/O Compression-Capable
Storage Controller . . . . . . . . . . . 737
Adding Disk Units to an Existing
Compression-Capable Storage Controller . . . 738
Moving Disk Units from the System ASP to a
User ASP . . . . . . . . . . . . . . 739
Recovering from Error Codes . . . . . . . . 740
Recovering from SRC 6xxx 7051 . . . . . . 740
Recovering from SRC 6xxx 7052 . . . . . . 741

Chapter 29. Managing Auxiliary Storage Pools 743

Working with ASP Trace and ASP Balance. . . . 743
Capacity Balance . . . . . . . . . . . 744
Hierarchical Storage Management (HSM)
Balance . . . . . . . . . . . . . . 745
Usage Balance . . . . . . . . . . . . 745
ASP Trace . . . . . . . . . . . . . 746
Determining adequate disk storage . . . . . . 746

Chapter 30. Creating Logical Partitions . . . . 747

Preparing for Logical Partitions . . . . . . . 747
Hardware and Software Requirements for
Logical Partitions . . . . . . . . . . . 747
Before You Partition Your System. . . . . . . 748
Creating a Logical Partition. . . . . . . . . 749

642 OS/400 Backup and Recovery V5R1

Chapter 24. Procedures for Configuring Disks and Disk
Protection
This chapter describes the procedures for configuring and protecting disks on your
system. It includes checklists for performing the disk configuration tasks in the
correct sequence.

You can use System Service Tools (SST) to do some disk configuration procedures
while your system is active.For other procedures, you must stop your system and
use Dedicated Service Tools (DST). This chapter provides information about both
SST and DST.

Choosing the Right Procedure for Configuring Disks

This chapter contains several checklists for performing configuration procedures.
Use Table 83 to determine which checklist to use for your situation.
Table 83. Choosing the Right Disk Procedure
Procedure to
Task Description Follow Requires DST?
Configure your system for the first time Checklist 1 on Yes
page 644.
Add one or more disk units that will not have device parity protection. This Checklist 2 on No
checklist applies to disk units that are capable of device parity protection if page 645.
you do not plan to start device parity protection for the disks.
Add one or more disks to an existing 9337 Disk Array Subsystem if you plan Checklist 3 on No
to protect the disks with device parity protection. Use this checklist if device page 646.
parity protection is already started for the 9337 Disk Array Subsystem.
Add a new 9337 Disk Array Subsystem. Use this checklist if you plan to Checklist 4 on Yes
protect some or all of the new disks with device parity protection and the page 647.
9337 Disk Array Subsystem did not arrive with device parity protection
already started.
Add a new 9337 Disk Array Subsystem. Use this checklist if you plan to Checklist 5 on No
protect some or all of the new disks with device parity protection and the page 648.
9337 Disk Array Subsystem arrived with device parity protection already
started.
Add one or more disks to an existing 6502, 6512, 6751, or 6532 Input/Output Checklist 6 on No
Processor. Use this checklist if you plan to protect some or all of the new page 649.
disks with device parity protection.
Add a new 6502, 6512, 6751, or 6532 IOP. Use this checklist if you plan to Checklist 7 on Yes
protect some or all of the new disks with device parity protection. page 650.
Move disk units between existing ASPs without mirrored protection. Checklist 8. on Yes
page 652.
Move disk units between existing ASPs with mirrored protection. Checklist 9 on Yes
page 653.
Delete a user ASP. Checklist 10 on Yes
page 654.
Remove one or more disk units without device parity protection. Checklist 11 on Yes
page 655.

© Copyright IBM Corp. 1997, 2000, 2001 643

Table 83. Choosing the Right Disk Procedure (continued)
Procedure to
Task Description Follow Requires DST?
Remove one or more disk units from a 9337 Disk Array Subsystem. Use this Checklist 12 on Yes
checklist if device parity protection is started for some or all of the disks in page 656.
the 9337 Disk Array Subsystem and they are in ASPs without mirrored
protection.
Remove one or more disk units from a 9337 Disk Array Subsystem. Use this Checklist 13 on Yes
checklist if device parity protection is started for some or all of the disks in page 657.
the 9337 Disk Array Subsystem and they are in ASPs with mirrored
protection.
Remove one or more disk units from a 6502, 6512, 6751, or 6532 IOP. Use this Checklist 14 on Yes
checklist if device parity protection is started for some or all of the disk units page 658.
that are attached to the IOP and they are in ASPs without mirrored
protection.
Remove one or more disk units from a 6502, 6512, 6751, or 6532 IOP. Use this Checklist 15 on Yes
checklist if device parity protection is started for some or all of the disks page 659.
units that are attached to the IOP and they are in ASPs with mirrored
protection.

Configuring Disks on a New System–Checklist 1

This checklist shows the sequence of tasks that you use to configure disks on a
new iSeries server. Whether you need to perform all the tasks depends on the disk
protection that you want on your system. The topic Setting up disk protection for
your data in the iSeries Information Center provides more information about the
disk protection that is available. You can access the Information Center from the
iSeries Information Center CD-ROM or from the following Web site:
http://www.ibm.com/eseries/iserver/infocenter

Before you begin, make a copy of this checklist. Fill in the appropriate areas as
you or the service representative perform the configuration tasks. This checklist
provides an important record of your actions. It may help you diagnose any
problems that occur.

Attention: When you perform the tasks in this checklist, the system moves large
amounts of data. Make sure that you have saved your system completely in the
event that you need to recover from an error situation.

Most tasks in the checklist include references to other topics in this book. Refer to
these topics if you need more information about how to perform a particular task.
Table 84. Configuring Disks on a New System–Tasks
Task What To Do Where To Read More About It
Task 1 Start DST. “How to Start Dedicated Service Tools
(DST)” on page 660.
Task 2 Display your disk configuration. Currently, “How to Display Your Disk Configuration”
all of your disk units except the load source on page 663.
unit appear as nonconfigured.
Task 3 If you plan to have device parity protection “How to Start Device Parity Protection for a
on any of your disk units, start it using the 9337 Disk Array Subsystem” on page 695 or
procedure for the types of disk units that “How to Start Device Parity Protection for an
you have. Input/Output Processor” on page 699.

644 OS/400 Backup and Recovery V5R1

Table 84. Configuring Disks on a New System–Tasks (continued)
Task What To Do Where To Read More About It
Task 4 Add nonconfigured disk units to the correct “How to Add Disk Units to an Auxiliary
ASPs. Storage Pool” on page 669.
Task 5 The default storage threshold for each ASP is “How to Change the Storage Threshold for
90%. If you want a different storage an Auxiliary Storage Pool” on page 672.
threshold for any ASP, change it.
Task 6 Specify the storage threshold for the system “How to Change the Storage Threshold for
ASP. If you use the QSTGLOWLMT and the System Auxiliary Storage Pool” on
QSTGLOWACN system values, you can page 673.
prevent the system ASP from filling to
capacity and causing an abnormal shutdown.
Task 7 If you plan to have mirrored protection for “How to Start Mirrored Protection” on
any ASPs, start it. page 719.
Task 8 If you started mirrored protection, wait until “Starting System Service Tools (SST)” on
the system completes the initial program page 662.
load. Then sign on and start SST.
Task 9 Verify that your disk configuration is correct “How to Display Your Disk Configuration”
and print a copy for your records. on page 663.
Task 10 End DST or SST. “How to Stop Dedicated Service Tools
(DST)” on page 662 or “Stopping System
Service Tools (SST)” on page 662.

Adding Disk Units without Device Parity Protection–Checklist

2
This checklist shows the sequence of tasks that you use to add one or more disks
to your system when you do not plan to protect the new disks with device parity
protection. You can use either DST or SST to perform the tasks in this checklist. If
you use SST, you can perform the tasks while your system is active. If you use
DST, you must stop your system to perform the tasks in this checklist.

Adding to an ASP with Mirrored Protection?

You can add disk units to an ASP that has mirrored protection without
stopping and starting mirrored protection. You must add disk units in pairs
with equal capacities. The added units will always be paired with each other.
You may want to choose a later time, when your system can be unavailable
for several hours, to stop and start mirrored protection. When you start
mirrored protection again, the system evaluates the pairing for all disk units
on your system. This may provide a higher level of availability for failures
that affect a controller, an IOP, or a bus.

Before you begin, make a copy of this checklist. Fill in the appropriate areas as
you or the service representative perform the configuration tasks. This checklist
provides an important record of your actions. It may help you diagnose any
problems that occur.

Attention: When you perform the tasks in this checklist, the system moves large
amounts of data. Make sure that you have saved your system completely in the
event that you need to recover from an error situation.

Chapter 24. Procedures for Configuring Disks and Disk Protection 645
 Most tasks in the checklist include references to other topics in this book. Refer to
these topics if you need more information about how to perform a particular task.
Table 85. Adding Disk Units without Device Parity Protection–Tasks
Task What To Do Where To Read More About It
Task 1 Physically attach disk units. This is normally
done by a service representative.
Task 2 Start DST or SST “How to Start Dedicated Service Tools
(DST)” on page 660 or “Starting System
Service Tools (SST)” on page 662.
Task 3 Print your current disk configuration. “How to Display Your Disk Configuration”
on page 663.
Task 4 Add nonconfigured disk units to the correct “How to Add Disk Units to an Auxiliary
ASPs. See note 1 and note 2. Storage Pool” on page 669.
Task 5 If you created a new ASP on your system “How to Change the Storage Threshold for
when you added disk units, the system set an Auxiliary Storage Pool” on page 672.
the storage threshold of the ASP to 90%. If
you want a different threshold, change it.
Task 6 Specify the storage threshold for the system “How to Change the Storage Threshold for
ASP. If you use the QSTGLOWLMT and the System Auxiliary Storage Pool” on
QSTGLOWACN system values, you can page 673.
prevent the system ASP from filling to
capacity and causing an abnormal shutdown.
Task 7 Verify that your disk configuration is correct “How to Display Your Disk Configuration”
and print a copy for your records. on page 663.
Task 8 End DST or SST. “How to Stop Dedicated Service Tools
(DST)” on page 662 or “Stopping System
Service Tools (SST)” on page 662.

1
You can add the disk units to an existing ASP or you can add them to a new ASP.
2
If you are adding disk units to an ASP that has mirrored protectionand the new disk units do not have
device parity protection, you must add pairs of disk units that have identical capacities.

Adding Disk Units to a 9337 Disk Array Subsystem–Checklist

3
This checklist shows the sequence of tasks that you use to add one or more disks
to an existing 9337 Disk Array Subsystem. Use this checklist if you plan to protect
the new disk units with device parity protection. You can use either DST or SST to
perform the tasks in this checklist. If you use SST, you can perform the tasks while
your system is active. If you use DST, you must stop your system to perform the
tasks in this checklist.

Before you begin, make a copy of this checklist. Fill in the appropriate areas as
you or the service representative perform the configuration tasks. This checklist
provides an important record of your actions. It may help you diagnose any
problems that occur.

Attention: When you perform the tasks in this checklist, the system moves large
amounts of data. Make sure that you have saved your system completely in the
event that you need to recover from an error situation.

646 OS/400 Backup and Recovery V5R1

 Most tasks in the checklist include references to other topics in this book. Refer to
these topics if you need more information about how to perform a particular task.
Table 86. Adding Disk Units to an Existing 9337 Disk Array Subsystem–Tasks
Task What To Do Where To Read More About It
Task 1 Physically attach disk units. This is normally
done by a service representative. When the
new disk units are attached, device parity
protection is started automatically if device
parity protection is active on the 9337 Disk
Array Subsystem.
Task 2 Start DST or SST. “How to Start Dedicated Service Tools
(DST)” on page 660 or “Starting System
Service Tools (SST)” on page 662.
Task 3 Print your current disk configuration. “How to Display Your Disk Configuration”
on page 663.
Task 4 Add nonconfigured disk units to the correct “How to Add Disk Units to an Auxiliary
ASPs. See note 1. Storage Pool” on page 669.
Task 5 If you created a new ASP on your system “How to Change the Storage Threshold for
when you added disk units, the system set an Auxiliary Storage Pool” on page 672.
the storage threshold of the ASP to 90%. If
you want a different threshold, change it.
Task 6 Specify the storage threshold for the system “How to Change the Storage Threshold for
ASP. If you use the QSTGLOWLMT and the System Auxiliary Storage Pool” on
QSTGLOWACN system values, you can page 673.
prevent the system ASP from filling to
capacity and causing an abnormal shutdown.
Task 7 Verify that your disk configuration is correct “How to Display Your Disk Configuration”
and print a copy for your records. on page 663.
Task 8 End DST or SST. “How to Stop Dedicated Service Tools
(DST)” on page 662 or “Stopping System
Service Tools (SST)” on page 662.

1
You can add the disk units to an existing ASP or you can add them to a new ASP.

Adding a New 9337 Disk Array Subsystem and Starting Device

Parity Protection on It–Checklist 4
This checklist shows the sequence of tasks that you use to add a new 9337 Disk
Array Subsystemto your system. Use this checklist if the 9337 Disk Array
Subsystem arrived without device parity protection started. You can use this
procedure whether or not you have mirrored protection on your system because
you start device parity protection before you add the disk units to an ASP. You can
use either DST or SST to perform the tasks in this checklist. If you use SST, you
can perform the tasks while your system is active. If you use DST, you must stop
your system to perform the tasks in this checklist.

Note: If you do not plan to start device parity protection for any of the new disks,
use the procedure in “Adding Disk Units without Device Parity
Protection–Checklist 2” on page 645 to add them.

Chapter 24. Procedures for Configuring Disks and Disk Protection 647
 Before you begin, make a copy of this checklist. Fill in the appropriate areas as
you or the service representative perform the configuration tasks. This checklist
provides an important record of your actions. It may help you diagnose any
problems that occur.

Attention: When you perform the tasks in this checklist, the system moves large
amounts of data. Make sure that you have saved your system completely in the
event that you need to recover from an error situation.

Most tasks in the checklist include references to other topics in this book. Refer to
these topics if you need more information about how to perform a particular task.
Table 87. Adding a New 9337 Disk Array Subsystem That Does Not Have Device Parity Protection Started–Tasks
Task What To Do Where To Read More About It
Task 1 Physically attach disk units. This is normally
done by a service representative.
Task 2 Start DST. “How to Start Dedicated Service Tools
(DST)” on page 660.
Task 3 Print your current disk configuration. “How to Display Your Disk Configuration”
on page 663.
Task 4 Start device parity protection for the 9337 “How to Start Device Parity Protection for a
Disk Array Subsystem. 9337 Disk Array Subsystem” on page 695.
Task 5 Add nonconfigured disk units to the correct “How to Add Disk Units to an Auxiliary
ASPs. Storage Pool” on page 669.
Task 6 If you created a new ASP on your system “How to Change the Storage Threshold for
when you addeddisk units, the system set the an Auxiliary Storage Pool” on page 672.
storage threshold of the ASP to 90%. If you
want a different threshold, change it.
Task 7 Specify the storage threshold for the system “How to Change the Storage Threshold for
ASP. If you use the QSTGLOWLMT and the System Auxiliary Storage Pool” on
QSTGLOWACN system values, you can page 673.
prevent the system ASP from filling to
capacity and causing an abnormal shutdown.
Task 8 Verify that your disk configuration is correct “How to Display Your Disk Configuration”
and print a copy for your records. on page 663.
Task 9 End DST. “How to Stop Dedicated Service Tools
(DST)” on page 662.

Adding a New 9337 Disk Array Subsystem That Has Device

Parity Protection Already Started–Checklist 5
This checklist shows the sequence of tasks that you use to add a new 9337 Disk
Array Subsystem to your system. Use this checklist if the 9337 Disk Array
Subsystem arrived with device parity protection already started. You can use this
procedure whether or not you have mirrored protection on your system because
you start device parity protection before you add the disk units to an ASP. You can
use either DST or SST to perform the tasks in this checklist. If you use SST, you
can perform the tasks while your system is active. If you use DST, you must stop
your system to perform the tasks in this checklist.

648 OS/400 Backup and Recovery V5R1

 Before you begin, make a copy of this checklist. Fill in the appropriate areas as
you or the service representative perform the configuration tasks. This checklist
provides an important record of your actions. It may help you diagnose any
problems that occur.

Attention: When you perform the tasks in this checklist, the system moves large
amounts of data. Make sure that you have saved your system completely in the
event that you need to recover from an error situation.

Most tasks in the checklist include references to other topics in this book. Refer to
these topics if you need more information about how to perform a particular task.
Table 88. Adding a New 9337 Disk Array Subsystem That Has Device Parity Protection Already Started–Tasks
Task What To Do Where To Read More About It
Task 1 Physically attach disk units. This is normally
done by a service representative.
Task 2 Start DST or SST. “How to Start Dedicated Service Tools
(DST)” on page 660 or “Starting System
Service Tools (SST)” on page 662.
Task 3 Print your current disk configuration. “How to Display Your Disk Configuration”
on page 663.
Task 4 Add nonconfigured disk units to the correct
ASPs. See note 1.
Task 5 If you created a new ASP on your system “How to Change the Storage Threshold for
when you added disk units, the system set an Auxiliary Storage Pool” on page 672.
the storage threshold of the ASP to 90%. If
you want a different threshold, change it.
Task 6 Specify the storage threshold for the system “How to Change the Storage Threshold for
ASP. If you use the QSTGLOWLMT and the System Auxiliary Storage Pool” on
QSTGLOWACN system values, you can page 673.
prevent the system ASP from filling to
capacity and causing an abnormal shutdown.
Task 7 Verify that your disk configuration is correct “How to Display Your Disk Configuration”
and print a copy for your records. on page 663.
Task 8 End DST or SST. “How to Stop Dedicated Service Tools
(DST)” on page 662 or “Stopping System
Service Tools (SST)” on page 662.

1
You can add the disk units to an existing ASP or you can add them to a new ASP.

Adding Disk Units to an Existing Input/Output

Processor–Checklist 6
This checklist shows the sequence of tasks that you use to add one or more disks
to an existing 6502, 6512, 6532, 6533, 6751, 6754, or 2741 Input/Output Processor.
Use this checklist if you plan to protect some or all of the new disk units with
device parity protection. If you do not plan to protect any of the new disk units,
use checklist 2.

You can use this procedure whether or not you have mirrored protection on your
system because you start device parity protection protection before you add the
disk units to an ASP. You can use either DST or SST to perform the tasks in this

Chapter 24. Procedures for Configuring Disks and Disk Protection 649
 checklist. If you use SST, you can perform the tasks while your system is active. If
you use DST, you must stop your system to perform the tasks in this checklist.

Before you begin, make a copy of this checklist. Fill in the appropriate areas as
you or the service representative perform the configuration tasks. This checklist
provides an important record of your actions. It may help you diagnose any
problems that occur.

Attention: When you perform the tasks in this checklist, the system moves large
amounts of data. Make sure that you have saved your system completely in the
event that you need to recover from an error situation.

Most tasks in the checklist include references to other topics in this book. Refer to
these topics if you need more information about how to perform a particular task.
Table 89. Adding Disk Units to an Existing Input/Output Processor–Tasks
Task What To Do Where To Read More About It
Task 1 Physically attach disk units. This is normally
done by a service representative.
Task 2 Start DST or SST “How to Start Dedicated Service Tools
(DST)” on page 660 or “Starting System
Service Tools (SST)” on page 662.
Task 3 Print your current disk configuration. “How to Display Your Disk Configuration”
on page 663.
Task 4 Include the disk units that you want to “How to Include a Disk Unit in Device
protect in device parity protection. Parity Protection” on page 705.
Task 5 Add nonconfigured disk units to the correct “How to Add Disk Units to an Auxiliary
ASPs. See note 1 and note 2. Storage Pool” on page 669.
Task 6 If you created a new ASP on your system “How to Change the Storage Threshold for
when you added disk units, the system set an Auxiliary Storage Pool” on page 672.
the storage threshold of the ASP to 90%. If
you want a different threshold, change it.
Task 7 Specify the storage threshold for the system “How to Change the Storage Threshold for
ASP. If you use the QSTGLOWLMT and the System Auxiliary Storage Pool” on
QSTGLOWACN system values, you can page 673.
prevent the system ASP from filling to
capacity and causing an abnormal shutdown.
Task 8 Verify that your disk configuration is correct “How to Display Your Disk Configuration”
and print a copy for your records. on page 663.
Task 9 End DST or SST. “How to Stop Dedicated Service Tools
(DST)” on page 662 or “Stopping System
Service Tools (SST)” on page 662.

1
You can add the disk units to an existing ASP or you can add them to a new ASP.
2
If you are adding disk units to an ASP that has mirrored protection and the new disk units do not have
device parity protection, you must add pairs of disk units that have identical capacities.

Adding a New Input/Output Processor–Checklist 7

This checklist shows the sequence of tasks that you use to add a new 6502, 6512,
6532, 6533, 6751, 6754, or 2741 Input/Output Processor (IOP) and disk units to
your system. Use this checklist if you plan to protect some or all of the new disks
with device parity protection.

650 OS/400 Backup and Recovery V5R1

 You can use this procedure whether or not you have mirrored protection on your
system because you start device parity protection before you add the disk units to
an ASP. If you do have mirrored protection and you are adding disks that do not
have device parity protection, you must add them in pairs that have equal
capacities. You can use either DST or SST to perform the tasks in this checklist. If
you use SST, you can perform the tasks while your system is active. If you use
DST, you must stop your system to perform the tasks in this checklist.

Note: If you do not plan to start device parity protection for any of the new disks,
use the procedure in checklist 2 to add them.

Before you begin, make a copy of this checklist. Fill in the appropriate areas as
you or the service representative perform the configuration tasks. This checklist
provides an important record of your actions. It may help you diagnose any
problems that occur.

Attention: When you perform the tasks in this checklist, the system moves large
amounts of data. Make sure that you have saved your system completely in the
event that you need to recover from an error situation.

Most tasks in the checklist include references to other topics in this book. Refer to
these topics if you need more information about how to perform a particular task.
Table 90. Adding a New Input/Output Processor–Tasks
Task What To Do Where To Read More About It
Task 1 Install the new Input/Output Processor in
the system. This is normally done by a
service representative.
Task 2 Physically attach disk units to the new IOP.
This is normally done by a service
representative.
Task 3 Start DST. “How to Start Dedicated Service Tools
(DST)” on page 660.
Task 4 Print your current disk configuration. “How to Display Your Disk Configuration”
on page 663.
Task 5 Start device parity protection for the IOP. “How to Start Device Parity Protection for an
Input/Output Processor” on page 699.
Task 6 Add nonconfigured disk units to the correct “How to Add Disk Units to an Auxiliary
ASPs. Storage Pool” on page 669.
Task 7 If you created a new ASP on your system “How to Change the Storage Threshold for
when you added disk units, the system set an Auxiliary Storage Pool” on page 672.
the storage threshold of the ASP to 90%. If
you want a different threshold, change it.
Task 8 Specify the storage threshold for the system “How to Display Your Disk Configuration”
ASP. If you use the QSTGLOWLMT and on page 663.
QSTGLOWACN system values, you can
prevent the system ASP from filling to
capacity and causing an abnormal shutdown.
Task 9 Verify that your disk configuration is correct “How to Display Your Disk Configuration”
and print a copy for your records. on page 663.
Task 10 End DST. “How to Stop Dedicated Service Tools
(DST)” on page 662.

Chapter 24. Procedures for Configuring Disks and Disk Protection 651
Table 90. Adding a New Input/Output Processor–Tasks (continued)
Task What To Do Where To Read More About It

Notes:
1. You can add the disk units to an existing ASP or you can add them to a new ASP.
2. If you are adding disk units to an ASP that has mirrored protection and the new disk units do not have device
parity protection, you must add pairs of disk units that have identical capacities.

Moving Disk Units Between Non-Mirrored Auxiliary Storage

Pools–Checklist 8
This checklist shows the sequence of tasks that you use to move one or more disk
units from one ASP to another ASP. Use these tasks when you do not have
mirrored protection active for the ASPs. You must stop your system and use DST
to perform the tasks in this checklist.

Before you begin, make a copy of this checklist. Fill in the appropriate areas as
you or the service representative perform the configuration tasks. This checklist
provides an important record of your actions. It may help you diagnose any
problems that occur.

Attention: When you perform the tasks in this checklist, the system moves large
amounts of data. Make sure that you have saved your system completely in the
event that you need to recover from an error situation.

Most tasks in the checklist include references to other topics in this book. Refer to
these topics if you need more information about how to perform a particular task.
Table 91. Moving Disk Units Between ASPs–Tasks
Task What To Do Where To Read More About It
Task 1 Print your current disk configuration. “How to Display Your Disk Configuration”
on page 663.
Task 2 Calculate the space requirements for both the “How to Calculate Space Requirements for
source and target ASPs for the disk units. an Auxiliary Storage Pool” on page 681.
Task 3 Use option 21 from the Save menu to save “Saving your server with the GO SAVE
your entire system. command” on page 15.
Task 4 Start DST. “How to Start Dedicated Service Tools
(DST)” on page 660.
Task 5 Move the disk units. “How to Move a Disk Unit to a Different
Auxiliary Storage Pool” on page 676.
Task 6 If you created a new ASP on your system “How to Change the Storage Threshold for
when you moved disk units, the system set an Auxiliary Storage Pool” on page 672.
the storage threshold for the ASP to 90%. If
you want a different threshold, change it.
Task 7 Specify the storage threshold for the system “How to Change the Storage Threshold for
ASP. If you use the QSTGLOWLMT and the System Auxiliary Storage Pool” on
QSTGLOWACN system values, you can page 673.
prevent the system ASP from filling to
capacity and causing an abnormal shutdown.
Task 8 Verify that your disk configuration is correct “How to Display Your Disk Configuration”
and print a copy for your records. on page 663.
Task 9 End DST. “How to Stop Dedicated Service Tools
(DST)” on page 662.

652 OS/400 Backup and Recovery V5R1

Table 91. Moving Disk Units Between ASPs–Tasks (continued)
Task What To Do Where To Read More About It
Task 10 If necessary, move objects between ASPs. “Transferring Objects between Auxiliary
Storage Pools” on page 686.

Moving Disk Units Between Mirrored Auxiliary Storage

Pools–Checklist 9
This checklist shows the sequence of tasks that you use to move one or more disk
units from one ASP to another ASP. Use these tasks when one or more or the ASPs
involved in the move has mirrored protection. You cannot use the move unit
procedure when mirrored protection is active. Instead, you remove mirrored pairs
from the source ASP and add them to the target ASP. You must stop your system
and use DST to perform the tasks in this checklist.

Before you begin, make a copy of this checklist. Fill in the appropriate areas as
you or the service representative perform the configuration tasks. This checklist
provides an important record of your actions. It may help you diagnose any
problems that occur.

Attention: When you perform the tasks in this checklist, the system moves large
amounts of data. Make sure that you have saved your system completely in the
event that you need to recover from an error situation.

Most tasks in the checklist include references to other topics in this book. Refer to
these topics if you need more information about how to perform a particular task.
Table 92. Moving Disk Units Between ASPs with mirrored protection–Tasks
Task What To Do Where To Read More About It
Task 1 Print your current disk configuration. “How to Display Your Disk Configuration”
on page 663.
Task 2 Calculate the space requirements for the “How to Calculate Space Requirements for
ASPs that are involved in moving disk units. an Auxiliary Storage Pool” on page 681.
Task 3 Use option 21 from the Save menu to save “Saving your server with the GO SAVE
your entire system. command” on page 15
Task 4 Start DST. “How to Start Dedicated Service Tools
(DST)” on page 660.
Task 5 Remove disk units that you plan to add to a “How to Remove a Disk Unit from an
different ASP. Auxiliary Storage Pool” on page 678.
Task 6 Add nonconfigured disk units to the correct
ASPs. See note 1.
Task 7 If you created a new ASP on your system “How to Change the Storage Threshold for
when you added disk units, the system set an Auxiliary Storage Pool” on page 672.
the storage threshold of the ASP to 90%. If
you want a different threshold, change it.
Task 8 Specify the storage threshold for the system “How to Change the Storage Threshold for
ASP. If you use the QSTGLOWLMT and the System Auxiliary Storage Pool” on
QSTGLOWACN system values, you can page 673.
prevent the system ASP from filling to
capacity and causing an abnormal shutdown.
Task 9 If you created any new ASPs and you want “How to Start Mirrored Protection” on
those ASPs to have mirrored protection, start page 719.
mirrored protection now.

Chapter 24. Procedures for Configuring Disks and Disk Protection 653
Table 92. Moving Disk Units Between ASPs with mirrored protection–Tasks (continued)
Task What To Do Where To Read More About It
Task 10 Verify that your disk configuration is correct “How to Display Your Disk Configuration”
and print a copy for your records. on page 663.
Task 11 End DST. “How to Stop Dedicated Service Tools
(DST)” on page 662.
Task 12 If necessary, move objects between ASPs. “Transferring Objects between Auxiliary
Storage Pools” on page 686.

1
If you are adding disk units to an ASP that has mirrored protectionand the new disk units do not have
device parity protection, you must add pairs of disk units that have identical capacities.

Deleting an Auxiliary Storage Pool–Checklist 10

This checklist shows the sequence of tasks that you use to delete a user ASP. You
must stop your system and use DST to perform the tasks in this checklist.

Before you begin, make a copy of this checklist. Fill in the appropriate areas as
you or the service representative perform the configuration tasks. This checklist
provides an important record of your actions. It may help you diagnose any
problems that occur.

Attention: When you perform the tasks in this checklist, the system moves large
amounts of data. Make sure that you have saved your system completely in the
event that you need to recover from an error situation. Also note that when an
ASP is deleted, all data remaining in that ASP is lost.

Most tasks in the checklist include references to other topics in this book. Refer to
these topics if you need more information about how to perform a particular task.
Table 93. Deleting an Auxiliary Storage Pool–Tasks
Task What To Do Where To Read More About It
Task 1 Print your current disk configuration. “How to Display Your Disk Configuration”
on page 663.
Task 2 Calculate the space requirements for the “How to Calculate Space Requirements for
remaining ASPs. an Auxiliary Storage Pool” on page 681.
Task 3 Use option 21 from the Save menu to save “Saving your server with the GO SAVE
your entire system. command” on page 15.
Task 4 Remove objects from the ASP that you are “Transferring Objects between Auxiliary
deleting or move the objects to a different Storage Pools” on page 686.
ASP.
Task 5 Start DST. “How to Start Dedicated Service Tools
(DST)” on page 660.
Task 6 Delete the ASP. This procedure places all of “How to Delete an Auxiliary Storage Pool”
the disks that were assigned to the deleted on page 680.
ASP in nonconfigured status.
Task 7 Add nonconfigured disk units to the correct “How to Add Disk Units to an Auxiliary
ASPs. See note 1. Storage Pool” on page 669.
Task 8 If you created a new ASP on your system “How to Change the Storage Threshold for
when you added disk units, the system set an Auxiliary Storage Pool” on page 672.
the storage threshold of the ASP to 90%. If
you want a different threshold, change it.

654 OS/400 Backup and Recovery V5R1

Table 93. Deleting an Auxiliary Storage Pool–Tasks (continued)
Task What To Do Where To Read More About It
Task 9 Specify the storage threshold for the system “How to Change the Storage Threshold for
ASP. If you use the QSTGLOWLMT and the System Auxiliary Storage Pool” on
QSTGLOWACN system values, you can page 673.
prevent the system ASP from filling to
capacity and causing an abnormal shutdown.
Task 10 Verify that your disk configuration is correct “How to Display Your Disk Configuration”
and print a copy for your records. on page 663.
Task 11 End DST. “How to Stop Dedicated Service Tools
(DST)” on page 662.
Task 12 If necessary, move objects between ASPs. “Transferring Objects between Auxiliary
Storage Pools” on page 686.

1
If you are adding disk units to an ASP that has mirrored protectionand the new disk units do not have
device parity protection, you must add pairs of disk units that have identical capacities.

Removing Disk Units Without Device Parity

Protection–Checklist 11
This checklist shows the sequence of tasks that you use to remove one or more
disk units from your system when the disk units do not have device parity
protection. Use these tasks when you are permanently removing disk units from
your system.Do not use these tasks when you are repairing or replacing a failed
disk unit. You must stop your system and use DST to perform the tasks in this
checklist.

Before you begin, make a copy of this checklist. Fill in the appropriate areas as
you or the service representative perform the configuration tasks. This checklist
provides an important record of your actions. It may help you diagnose any
problems that occur.

Attention: When you perform the tasks in this checklist, the system moves large
amounts of data. Make sure that you have saved your system completely in the
event that you need to recover from an error situation.

Most tasks in the checklist include references to other topics in this book. Refer to
these topics if you need more information about how to perform a particular task.
Table 94. Removing Disk Units That Do Not Have Device Parity Protection–Tasks
Task What To Do Where To Read More About It
Task 1 Print your current disk configuration. “How to Display Your Disk Configuration”
on page 663.
Task 2 Calculate the space requirements for the “How to Calculate Space Requirements for
ASPs that are involved in removing disk an Auxiliary Storage Pool” on page 681.
units.
Task 3 Use option 21 from the Save menu to save “Saving your server with the GO SAVE
your entire system. command” on page 15
Task 4 Start DST. “How to Start Dedicated Service Tools
(DST)” on page 660.
Task 5 Remove disk units that you plan to remove “How to Remove a Disk Unit from an
from the system. Auxiliary Storage Pool” on page 678.

Chapter 24. Procedures for Configuring Disks and Disk Protection 655
Table 94. Removing Disk Units That Do Not Have Device Parity Protection–Tasks (continued)
Task What To Do Where To Read More About It
Task 6 Verify that your disk configuration is correct “How to Display Your Disk Configuration”
and print a copy for your records. on page 663.
Task 7 End DST. “How to Stop Dedicated Service Tools
(DST)” on page 662.

Note: This checklist works only if at least one unit remains in the ASP and there is
enough capacity remaining.

Removing Disk Units with Device Parity Protection (9337) from

an ASP Without Mirrored Protection–Checklist 12
This checklist shows the sequence of tasks that you use to remove one or more
disk units from a 9337 Disk Array Subsystem. Use these steps when device parity
protection is active for the 9337 Disk Array Subsystem and when the ASPs
containing the disk units do not have mirrored protection. Use these tasks when
you are permanently removing disk units from your system. Do not use these
tasks when you are repairing or replacing a failed disk unit. You must stop your
system and use DST to perform the tasks in this checklist.

Before you begin, make a copy of this checklist. Fill in the appropriate areas as
you or the service representative perform the configuration tasks. This checklist
provides an important record of your actions. It may help you diagnose any
problems that occur.

Attention: When you perform the tasks in this checklist, the system moves large
amounts of data. Make sure that you have saved your system completely in the
event that you need to recover from an error situation.

Most tasks in the checklist include references to other topics in this book. Refer to
these topics if you need more information about how to perform a particular task.
Table 95. Removing Disk Units with Device Parity Protection from an ASP without Mirrored Protection–Tasks
Task What To Do Where To Read More About It
Task 1 Print your current disk configuration. “How to Display Your Disk Configuration”
on page 663.
Task 2 Calculate the space requirements for the “How to Calculate Space Requirements for
ASPs that are involved in removing disk an Auxiliary Storage Pool” on page 681.
units.
Task 3 Use option 21 from the Save menu to save “Saving your server with the GO SAVE
your entire system. command” on page 15.
Task 4 Start DST. “How to Start Dedicated Service Tools
(DST)” on page 660.
Task 5 Remove disk units that you plan to remove “How to Remove a Disk Unit from an
from the system. Auxiliary Storage Pool” on page 678.
Task 6 Exclude the disk units from the 9337. The “How to Exclude a Disk Unit from Device
service representative can perform the Parity Protection” on page 707.
exclude function on the 9337.
Task 7 Verify that your disk configuration is correct “How to Display Your Disk Configuration”
and print a copy for your records. on page 663.

656 OS/400 Backup and Recovery V5R1

Table 95. Removing Disk Units with Device Parity Protection from an ASP without Mirrored
Protection–Tasks (continued)
Task What To Do Where To Read More About It
Task 8 End DST. “How to Stop Dedicated Service Tools
(DST)” on page 662.

Removing Disk Units That Have Device Parity Protection(9337)

from an ASP with Mirrored Protection–Checklist 13
This checklist shows the sequence of tasks that you use to remove one or more
disk units from a 9337 Disk Array Subsystem. These tasks apply when the ASPs
containing the disk units have mirrored protection and when device parity
protection is active for the 9337 Disk Array Subsystem. Use these tasks when you
are permanently removing disk units from your system. Do not use these tasks
when you are repairing or replacing a failed disk unit. You must stop your system
and use DST to perform the tasks in this checklist.

Before you begin, make a copy of this checklist. Fill in the appropriate areas as
you or the service representative perform the configuration tasks. This checklist
provides an important record of your actions. It may help you diagnose any
problems that occur.

Attention: When you perform the tasks in this checklist, the system moves large
amounts of data. Make sure that you have saved your system completely in the
event that you need to recover from an error situation.

Most tasks in the checklist include references to other topics in this book. Refer to
these topics if you need more information about how to perform a particular task.
Table 96. Removing Disk Units from a 9337 Disk Array Subsystem and a Mirrored ASP–Tasks
Task What To Do Where To Read More About It
Task 1 Print your current disk configuration. “How to Display Your Disk Configuration”
on page 663.
Task 2 Calculate the space requirements for the “How to Calculate Space Requirements for
ASPs that are involved in removing disk an Auxiliary Storage Pool” on page 681.
units.
Task 3 Use option 21 from the Save menu to save “Saving your server with the GO SAVE
your entire system. command” on page 15.
Task 4 Start DST. “How to Start Dedicated Service Tools
(DST)” on page 660.
Task 5 Remove disk units that you plan to remove “How to Remove a Disk Unit from an
from the system. Auxiliary Storage Pool” on page 678.
Task 6 Stop mirrored protection for the ASPs that “How to Stop Mirrored Protection” on
will have disk units removed. When you page 723.
stop mirrored protection, one disk unit from
each mirrored pair becomes nonconfigured.
See note 1.
Task 7 Exclude the disk units from the 9337. The “How to Exclude a Disk Unit from Device
service representative can perform the Parity Protection” on page 707.
exclude function on the 9337.

Chapter 24. Procedures for Configuring Disks and Disk Protection 657
Table 96. Removing Disk Units from a 9337 Disk Array Subsystem and a Mirrored ASP–Tasks (continued)
Task What To Do Where To Read More About It
Task 8 Add nonconfigured disk units to the correct “How to Add Disk Units to an Auxiliary
ASPs. These disks became nonconfigured Storage Pool” on page 669.
when you stopped mirrored protection in
task 6.
Task 9 Start mirrored protection for the ASPs that “How to Start Mirrored Protection” on
had mirrored protection stopped in task 6. page 719.
Task 10 Verify that your disk configuration is correct “How to Display Your Disk Configuration”
and print a copy for your records. on page 663.
Task 11 End DST. “How to Stop Dedicated Service Tools
(DST)” on page 662.

1
You need to stop mirrored protection only if the ASP contains other disk units that are attached to the 9337
Disk Array Subsystem and have device parity protection.

Removing Disk Units That Have Device Parity Protection from

an ASP Without Mirrored Protection–Checklist 14
This checklist shows the sequence of tasks that you use to remove one or more
disk units from a 6502, 6512, 6532, 6533, 6751, 6754 or 2741 Input/Output
Processor. These tasks apply when the ASPs containing the disk units do not have
mirrored protection and when device parity protection is started for the IOP. Use
these tasks when you are permanently removing disk units from your system. Do
not use these tasks when your are repairing or replacing a failed hard disk. You
must stop your system and use DST to perform the tasks in this checklist.

Before you begin, make a copy of this checklist. Fill in the appropriate areas as
you or the service representative perform the configuration tasks. This checklist
provides an important record of your actions. It may help you diagnose any
problems that occur.

Attention: When you perform the tasks in this checklist, the system moves large
amounts of data. Make sure that you have saved your system completely in the
event that you need to recover from an error situation.

Most tasks in the checklist include references to other topics in this book. Refer to
these topics if you need more information about how to perform a particular task.
Table 97. Removing Disk Units from an IOP and a Non-Mirrored ASP–Tasks
Task What To Do Where To Read More About It
Task 1 Print your current disk configuration. “How to Display Your Disk Configuration”
on page 663.
Task 2 Calculate the space requirements for the “How to Calculate Space Requirements for
ASPs that are involved in removing disk an Auxiliary Storage Pool” on page 681.
units.
Task 3 Use option 21 from the Save menu to save “Saving your server with the GO SAVE
your entire system. command” on page 15.
Task 4 Start DST. “How to Start Dedicated Service Tools
(DST)” on page 660.
Task 5 Remove disk units that you plan to remove “How to Remove a Disk Unit from an
from the system. Auxiliary Storage Pool” on page 678.

658 OS/400 Backup and Recovery V5R1

Table 97. Removing Disk Units from an IOP and a Non-Mirrored ASP–Tasks (continued)
Task What To Do Where To Read More About It
Task 6 Exclude the disk units from device parity “How to Exclude a Disk Unit from Device
protection. If you were successful in Parity Protection” on page 707.
excluding the disk units, skip to task 8.
Otherwise, continue with task 7.
Task 7 Stop device parity protection for the IOP. “How to Stop Device Parity Protection on an
Input/Output Processor” on page 704.
Task 8 Physically remove disk units. This is
normally done by a service representative. If
you stopped device parity protection in task
7, then continue with task 9. If you did not
stop device parity protection, then skip to
task 10.
Task 9 Start device parity protection for the IOP. “How to Start Device Parity Protection for an
Input/Output Processor” on page 699.
Task 10 Verify that your disk configuration is correct “How to Display Your Disk Configuration”
and print a copy for your records. on page 663.
Task 11 End DST. “How to Stop Dedicated Service Tools
(DST)” on page 662.

Removing Disk Units That Have Device Parity Protection from

an ASP With Mirrored Protection–Checklist 15
This checklist shows the sequence of tasks that you use to remove one or more
disk units from a 6502, 6512, 6532, 6533, 6751, 6754, or 2741 Input/Output
Processor. These tasks apply when the ASPs that contain the disk units have
mirrored protection and when the disk units have device parity protection. Use
these tasks when your are permanently removing disk units from your system. Do
not use these tasks when you are repairing or replacing a failed disk unit. You
must stop your system and use DST to perform the tasks in this checklist.

Before you begin, make a copy of this checklist. Fill in the appropriate areas as
you or the service representative perform the configuration tasks. This checklist
provides an important record of your actions. It may help you diagnose any
problems that occur.

Attention: When you perform the tasks in this checklist, the system moves large
amounts of data. Make sure that you have saved your system completely in the
event that you need to recover from an error situation.

Most tasks in the checklist include references to other topics in this book. Refer to
these topics if you need more information about how to perform a particular task.
Table 98. Removing Disk Units from an IOP and a Mirrored ASP–Tasks
Task What To Do Where To Read More About It
Task 1 Print your current disk configuration. “How to Display Your Disk Configuration”
on page 663.
Task 2 Calculate the space requirements for the “How to Calculate Space Requirements for
ASPs that are involved in removing disk an Auxiliary Storage Pool” on page 681.
units.
Task 3 Use option 21 from the Save menu to save “Saving your server with the GO SAVE
your entire system. command” on page 15.

Chapter 24. Procedures for Configuring Disks and Disk Protection 659
Table 98. Removing Disk Units from an IOP and a Mirrored ASP–Tasks (continued)
Task What To Do Where To Read More About It
Task 4 Start DST. “How to Start Dedicated Service Tools
(DST)”.
Task 5 Remove disk units that you plan to remove “How to Remove a Disk Unit from an
from the system. Auxiliary Storage Pool” on page 678.
Task 6 Exclude the disk units from device parity “How to Exclude a Disk Unit from Device
protection. If you were successful in Parity Protection” on page 707.
excluding the disk units, skip to task 9.
Otherwise, continue with task 7.
Task 7 Stop mirrored protection for the ASPs that “How to Stop Mirrored Protection” on
will have disk units removed. When you page 723.
stop mirrored protection, one disk unit from
each mirrored pair becomes nonconfigured.
See note 1.
Task 8 Stop device parity protection for the IOP. “How to Stop Device Parity Protection on an
Input/Output Processor” on page 704.
Task 9 Physically remove disk units. This is
normally done by a service representative. If
you stopped device parity protection in task
8, then continue with task 10. If you did not
stop device parity protection, then skip to
task 14.
Task 10 Start device parity protection for the IOP. “How to Start Device Parity Protection for an
Input/Output Processor” on page 699.
Task 11 Add nonconfigured disk units to the correct “How to Add Disk Units to an Auxiliary
ASPs. These disks became nonconfigured Storage Pool” on page 669.
when you stopped mirrored protectionin
task 7.
Task 12 If you created a new ASP on your system “How to Change the Storage Threshold for
when you added disk units, the system set an Auxiliary Storage Pool” on page 672.
the storage threshold of the ASP to 90%. If
you want a different threshold, change it.
Task 13 Start mirrored protection for the ASPs that “How to Start Mirrored Protection” on
had mirrored protection stopped in task 7. page 719.
Task 14 Verify that your disk configuration is correct “How to Display Your Disk Configuration”
and print a copy for your records. on page 663.
Task 15 End DST. “How to Stop Dedicated Service Tools
(DST)” on page 662.

1
You need to stop mirrored protection only if the ASP contains other disk units that are attached to the IOP
and have device parity protection.

Using System Service Tools and Dedicated Service Tools

Several backup and recovery procedures, including disk storage management,
require the use of Dedicated Service Tools (DST) or System Service Tools (SST).
This topic describes how to start and end SST and DST. It also provides a list of
the options available through these tools.

How to Start Dedicated Service Tools (DST)

Use this procedure to start DST. If the IPL or Install the System menu is already
displayed, start with step 5 on page 661.

660 OS/400 Backup and Recovery V5R1

 1. Ensure that the keystick is in the system unit control panel.
2. Place the system in manual mode.
3. Power down the system:
PWRDWNSYS OPTION(*CNTRLD) DELAY(600)
RESTART(*YES) IPLSRC(B)

Attention logical partitioning users!

If you are going to use this command on the primary partition, be sure to
power off all secondary partitions before running the command.

Note: If you are sure that no jobs are running on your system, you can specify
OPTION(*IMMED) when you power down the system. Otherwise, specify a
delay time that is sufficient to allow jobs to end normally.
4. When the IPL completes, the IPL or Install the System menu appears.

IPL or Install the System

Select one of the following:
1. Perform an IPL
2. Install the operating system
3. Use Dedicated Service Tools (DST)
4. Perform automatic installation of the operating system
5. Save Licensed Internal Code

5. Select option 3 (Use Dedicated Service Tools (DST)) and press the Enter key.
The Dedicated Service Tools (DST) Sign On display is shown.

Dedicated Service Tools (DST) Sign On

Type choice, press Enter.

DST user . . . . . . . . . . _______

DST password . . . . . . . . _______

| 6. In the DST user field, type QSECOFR. In the DST password field, type your DST
| security-level or full-level password. On a new system, the password is
| QSECOFR. The password is case sensitive; use all capital letters. You should
| change this password after you complete your installation procedures. The
| iSeries Security Reference book has more information about DST passwords.
| The Use Dedicated Service Tools (DST) menu is shown.
|

Chapter 24. Procedures for Configuring Disks and Disk Protection 661
 Use Dedicated Service Tools (DST)
Select one of the following:
1. Perform an IPL
2. Install the operating system
3. Work with licensed internal code
4. Work with disk units
5. Work with DST environment
6. Select DST console mode
7. Start a service tool
8. Perform automatic installation of the operating system
9. Work with save storage and restore storage
10. Work with remote DST support

How to Stop Dedicated Service Tools (DST)

Use this procedure to end DST:
1. If you do not want to see the displays for a manual initial program load (IPL),
return the system to automatic mode. If you want to see the displays, leave the
system in manual mode.
2. Press F3 until you return to the Use Dedicated Service Tools (DST) menu.
3. From the Use Dedicated Service Tools (DST) menu, select option 1 (Perform an
IPL).

Note: Do not perform an IPL if you are performing a complete system

recovery.
The system may take significantly longer than normal to complete the IPL.
Some functions that you perform by using DST, such as starting mirrored
protection, require the system to do additional work during the IPL before the
system is available for your use.

Starting System Service Tools (SST)

You can access system service tools doing the following:
1. Use the Start System Service Tools (STRSST) command or by selecting the
option for problem handling from the iSeries Main menu. From the Problem
Handling menu, select the option for system service tools.
2. At the Start Service Tools (STRSST) Sign On display, enter your service tools
user profile and password. For more information about about Service tools user
profiles, see Tips and Tools for Securing Your iSeries.
3. Press Enter.
4. The System Service Tools (SST) menu appears:
System Service Tools (SST)
Select one of the following:
1. Start a service tool
2. Work with active service tools
3. Work with disk units
4. Work with diskette data recovery
5. Work with system partitions

Stopping System Service Tools (SST)

To end system service tools, do the following:
1. Press F3 (Exit) until you return to the System Service Tools (SST) menu.
2. Press F3 (Exit) again. You are shown the Exit System Service Tools display.
3. Press the Enter key to end SST.

662 OS/400 Backup and Recovery V5R1

How to Display Your Disk Configuration
This topic describes how to display or print your current disk configuration. It also
explains some of the fields that appear on the display or listing. For some
purposes, such as planning a mirrored configuration, you need to view your disk
configuration both from a hardware perspective and from a software perspective.
The hardware perspective shows how the disk units are attached by bus, IOP, and
controller. The software perspective shows how disk units are assigned to ASPs
and how they are protected.

You can use DST, SST, or commands to display your disk configuration. When you
are planning changes to your disk configuration, use SST and commands to print
your current configuration before you begin to make changes. After you have
made changes, you can use DST to verify the new configuration before you end
DST.

Displaying Your Disk Configuration–Hardware View

When you display your hardware disk configuration, you see all disk-related
components that are physically attached to your system. This includes disk units
whose software status is nonconfigured because they are not assigned to an ASP
yet.

This topic describes both the DST method and the command method for
displaying your disk hardware configuration.

Displaying Disk Hardware Configuration–Command Method: To display the

hardware configuration of disk units on your system, do the following:
1. On a command line, type WRKHDWRSC TYPE(*STG) and press the Enter key. The
Work with Storage Resources display appears. The display shows buses, IOPs,
and controllers.

Work with Storage Resources

System: RCHASDP4
Type options, press Enter.
9=Work with resource

Opt Resource Status Text

CMB01 Operational Combined function IOP
DC01 Operational Disk Storage Controller
DC02 Operational Disk Storage Controller
DC05 Operational Tape Controller

2. If you want to see the detail about disk units that are attached to a controller,
type 9 (Work with resource) in the Option column for the controller.

To print the hardware configuration of disk units on your system, do the

following:
1. On a command line, type DSPHDWRSC TYPE(*STG) OUTPUT(*PRINT) and press the
Enter key. Figure 84 on page 664 shows part of the listing that you receive:

Chapter 24. Procedures for Configuring Disks and Disk Protection 663
 Display Spooled File
File . . . . . : QSYSPRT Page/Line 1/1
Control . . . . . +15 Columns 1 - 78
Find . . . . . .
*...+....1....+....2....+....3....+....4....+....5....+....6....+....7....+...
Display Hardware Resources
5716SS1 V3R6M0 950602
Storage Resources List
--------
Serial Part Frame
Resource Type-Model Number Number ID
CMB01 9162-001 10-00000 0000086G7917 1
DC01 6602-030 00-0193825 1
DD001 6602-030 00-0193825 1
DC02 6602-030 00-17900 1
DD002 6602-030 00-17900 1

Figure 84. Display Hardware Resource Listing

Displaying Disk Hardware Configuration–DST Method: When you are using

DST, you can use the following method to display your disk hardware
configuration:
1. If you are not already using DST, perform a manual IPL to start DST. See “How
to Start Dedicated Service Tools (DST)” on page 660.
2. From the Use Dedicated Service Tools (DST) menu, do the following:

┌────────────────────────┐ Select option 7

│ DST Menu │ (Start a service tool)
│ │
│ ┌───────────────────┴─────┐Select option 4
│ 7 │ Start a Service Tool │(Hardware service manager)
└────┤ │
│ ┌─────────────────────┴──┐
│ │ Hardware Service │
│ 4 │ Manager │
└───┤ │
│ │
│ │
└────────────────────────┘

3. To print the hardware disk configuration, press F6 (Print configuration). If your

system already has a printer defined for DST, the output is sent to that printer.
If you do not have a DST printer defined, you are prompted with a list of
attached printers. When you are using DST, output goes directly to the printer
because spooling is not active.
4. To display the configuration, select option 2 (Logical hardware resources) from
the Hardware Service Manager menu. From this display, you can select to show
system bus, processor, or main storage resources.
5. To see additional detail, type 5 (Display detail) in the Option column next to
each controller and press the Enter key.
6. If you have no other tasks to perform, end DST. (See “How to Stop Dedicated
Service Tools (DST)” on page 662.)

664 OS/400 Backup and Recovery V5R1

Displaying Your Disk Configuration–Software View
When you display your software disk configuration, you see how disk units are
assigned to ASPs and how they are protected. A separate display shows disk units
that are attached to the system but have not been assigned to an ASP
(nonconfigured status).

To display the software configuration of disk units on your system, do the

following:
1. If you are not already using DST, perform a manual IPL to start DST. See “How
to Start Dedicated Service Tools (DST)” on page 660.
2. From the Use Dedicated Service Tools (DST) menu, do the following:
or from the System Service Tools (SST) menu, do the following:
┌────────────────────────┐ Select option 4
│ DST Menu │ (Work with disk units)
│ │
│ ┌───────────────────┴────┐ Select option 1
│ 4 │ Work with Disk Units │ (Work with disk configuration)
└────┤ │
│ ┌───────────────────┴────┐ Select option 1
│ 1 │Work with Disk │ (Display disk configuration)
└────┤ Configuration │
│ ┌───────────────────┴──────┐
│ 1 │Display Disk Configuration│ Select option 1
└────┤ │ (Display disk configuration status
│ │
│ 1 │
└──────────────────────────┘

The Display Disk Configuration menu appears.

┌────────────────────────┐ Select option 3
│ SST Menu │ (Work with disk units)
│ │
│ ┌───────────────────┴────┐ Select option 1
│ 3 │ Work with Disk Units │ (Display disk configuration)
└────┤ │
│ 1 ┌───────────────────┴──────┐
│ │Display Disk Configuration│
└────┤ │
│ │
│ │
└──────────────────────────┘

Display Disk Configuration

Select one of the following:
1. Display disk configuration status
2. Display disk configuration capacity
3. Display disk configuration protection
4. Display non-configured units
5. Display device parity status
6. Display disk hardware status

3. Select option 1 to see the Display Disk Configuration Status display:

Chapter 24. Procedures for Configuring Disks and Disk Protection 665
 Display Disk Configuration Status
Serial Resource
ASP Unit Number Type Model Name Status
1 Unprotected
1 00-0193825 6602 030 DD001 Configured
2 00-0163477 6602 074 DD019 DPY/Active
3 00-0190494 6602 070 DD036 DPY/Active
6 00-17900 6602 030 DD002 Configured
3 Unprotected
4 00-0330477 6602 074 DD005 DPY/Active
5 00-0323200 6602 074 DD033 DPY/Active

Press Enter to continue.

F3=Exit F5=Refresh F9=Display disk unit details

F11=Disk configuration capacity F12=Cancel

Note: If you are performing a complete system restore, all the disk units on the
system may not report in right away. Verify that the number of disk
units displayed matches the number of disk units physically attached to
the system. If they do not match, wait a few minutes and press F5
(Refresh) until all of the disk units report in.
4. If the lower right of the display says More..., you can page forward to see
additional units.
5. To display the capacity of your disk units and how much capacity is used,
press F11 from the Display Disk Configuration Status display or select option 2
from the Use Dedicated Service Tools (DST) menu:

Display Disk Configuration Capacity

--Protected-- --Unprotected--
ASP Unit Type Model Threshold Overflow Size %Used Size %Used
1 90% No 1805 * 2063 *
1 6602 030 0 0.00% 1031 *
2 6602 074 773 * 0 0.00%
3 6602 070 1031 * 0 0.00%
6 6602 030 0 0.00% 1031 *
3 90% No 1547 * 0 0.00%
4 6602 074 773 * 0 0.00%
5 6602 074 773 * 0 0.00%

6. To display the disk protection that is configured for each disk unit, press F11
again:
7. To display nonconfigured disk units, press F11 from the Display Disk
Configuration Protection display or select option 4 from the Display Disk
Configuration menu:

Display Non-Configured Units

Serial Resource
Number Type Model Name Capacity Status
00-0313374 6602 074 DD003 773 DPY/Active

8. To print the software disk configuration, use the print key from the displays. If
your system already has a printer defined for DST, the output is sent to that

666 OS/400 Backup and Recovery V5R1

 printer. If you do not have a DST printer defined, you are prompted with a list
of attached printers. When you are using DST, output goes directly to the
printer because spooling is not active.
9. If you have no other tasks to perform, end DST or SST.(See “How to Stop
Dedicated Service Tools (DST)” on page 662 or “Stopping System Service Tools
(SST)” on page 662.)

Interpreting Disk Configuration and Status Displays

This topic explains some of the fields that appear on the displays that you use to
look at your disk configuration and status. You can view online information for all
the fields and their possible values.

Unit Field: A unit number is assigned by the system to identify a specific disk unit.
The unit number is a software function and does not appear when you display the
hardware configuration. When disk units are protected by mirrored protection,
both disk units in a mirrored pair are assigned the same unit number.

Resource Name Field: The system resource manager assigns a resource name to
every hardware device that is physically attached to the system. This resource
name is the link between the hardware and the software definition of the
hardware. When you add a disk unit to an ASP, you use the resource name to
identify which disk unit to add.

Status Field for the Auxiliary Storage Pool: The display shows the status of an
entire ASP. This status indicates the software disk protection that is in effect for the
ASP. The possible values are:

Unprotected Mirrored protection is not active for the ASP. However, device parity
protection may be active for some or all of the disk units in the ASP. You
need to look at the individual disk units to determine the level of
protection for the ASP.
Mirrored The ASP is fully protected. Mirrored protection has been started for the
ASP. All of the disk units in the ASP are protected either by mirrored
protectionor by device parity protection.

Status–Disk Unit: The display also shows the status of individual disk units. The
possible values are:

Operational The disk unit is operational and ready to accept input or output
operations.
Not operational The device cannot communicate with the IOP. You should verify
that the unit is powered on.
Not ready The device cannot perform media-related functions, but it can still
communicate with the IOP.
Busy The device is not available for processing any commands on this
connection.
Read/write protected The device cannot process either a read or a write operation. A
device may be in this state due to a cache problem, a device
configuration problem, or other types of problems that could cause
a data integrity exposure.
Write protected The device cannot accept write operations. Read operations are
allowed.
Performance degraded The device is functional, but performance may be impacted due to
other hardware problems (such as a problem with the IOP cache).

Chapter 24. Procedures for Configuring Disks and Disk Protection 667
 Redundant failure The device is functional, but availability may be impacted due to
other problems (such as a redundant power supply problem).
Service is required to prevent additional failures that will stop input
and output operations to the device.
DPY/Failed This unit is part of a disk unit subsystem that has device parity
protection. The disk unit failed within its device parity set, causing
the loss of data protection for the device parity set.
DPY/Unprotected This unit is part of a disk unit subsystem that has device parity
protection. Data protection is no longer in effect due to a failure in
another resource.
DPY/Rebuilding This unit is part of a disk unit subsystem that has device parity
protection. Data protection is being rebuilt.
DPY/Active This unit is part of a disk unit subsystem that has device parity
protection. The unit is operational and ready to accept input or
output operations.
DPY/Resyncing This unit is part of a disk unit subsystem that has device parity
protection. The subsystem is in the process of recreating the
redundancy data for the device parity set. All units in the set that
are being synchronized will have this status.
DPY/Unknown This unit is part of a disk unit subsystem that has device parity
protection. The status of this unit is not known to the system.
Active This unit is one of a mirrored pair. It is capable of having data
written to it or read from it.
Suspended This unit is one of a mirrored pair. It is not capable of havingdata
written to it or read from it. The data on this unit is not current. For
example, if the disk needs repair action or has been manually
suspended, it would be in a Suspended state.
Resuming This unit is one of a mirrored pair. The current data is being copied
(or will be copied) to this unit from the other active unit of the
mirrored pair.
Unprotected The device is in a state that cannot be determined.

668 OS/400 Backup and Recovery V5R1

 Chapter 25. Working with Auxiliary Storage Pools
This chapter contains procedures for working with auxiliary storage pools (ASPs).
When you are making changes to the disk configuration on your system, refer to
Chapter 24. Procedures for Configuring Disks and Disk Protection for the correct
sequence of steps for your situation.

| Support for independent ASPs is provided through Operations Navigator. If you

| want to work with independent ASPs, you can find information in the Information
| Center at the following Web site:
| http://www.ibm.com/eserver/iseries/infocenter

| Both the Information Center and Operations Navigator refer to ASPs as Disk Pools.

How to Add Disk Units to an Auxiliary Storage Pool

Do Things in the Right Sequence

v If you want to have device parity protection for the disks that you are
adding, you should start device parity protection before you add the disk
units to an ASP.
v If you have more than one ASP on your system, you should plan how you
want to add the new disk units before you begin this procedure. The topic,
Information Center contains more information. You can access the
Information Center from the from the following Web site:
http://www.ibm.com/eserver/iseries/infocenter

When you (or your service representative)physically attach a new disk unit to your
system, the status of the new disk unit is nonconfigured. Nonconfigured status
means that a disk unit has not yet been assigned to an ASP on the system. You can
assign disk units to an existing ASP or to a new ASP. You create a new ASP simply
by assigning disk units to it.

To assign nonconfigured disks to an ASP, do the following:

1. If you are not already using DST, perform a manual IPL to start DST. See “How
to Start Dedicated Service Tools (DST)” on page 660.
2. From the Use Dedicated Service Tools (DST) menu, do the following:

© Copyright IBM Corp. 1997, 2000, 2001 669

 ┌────────────────────────┐ Select option 4
│ DST Menu │ (Work with disk units)
│ │
│ ┌───────────────────┴─────┐ Select option 1
│ 4 │ Work with Disk Units │ (Work with disk configuration)
└────┤ │
│ ┌────────────────────┴────┐ Select option 3
│ │ Work with Disk │ (Work with ASP configuration)
│ 1 │ Configuration │
└────┤ ┌────────────────────┴─────┐
│ 3 │ Work with ASP │
└────┤ Configuration │
│ │
│ │
└──────────────────────────┘

or from the System Service Tools (SST) menu,

┌────────────────────────┐ Select option 3

│ SST Menu │ (Work with disk units)
│ │
│ ┌───────────────────┴───┐ Select option 2
│ 3 │ Work with Disk Units │ (Work with disk configuration)
└────┤ │
│ ┌──────────────────┴──────┐
│ 2 │ Work with Disk │
└────┤ Configuration │
│ │
│ │
└─────────────────────────┘

3. Select the option to add units to ASPs and balance data. The Specify ASPs to
Add Units to display appears. It lists all the disk units that have a status of
nonconfigured.

Specify ASPs to Add Units to

Specify the ASP to add each unit to.

Specify Serial Resource

ASP Number Type Model Capacity Name
___ 00-0103706 6602 030 1031 DD031
___ 00-1000341 9337 211 542 DD012
___ 00-5000341 9337 211 542 DD015
___ 00-7000341 9337 211 542 DD011
___ 00-3000341 9337 211 542 DD014
___ 00-2000341 9337 211 542 DD013
1 00-61300 6603 074 1475 DD006
1 00-52262 6606 074 1475 DD008
1 00-86978 6606 050 1967 DD009
2 00-95744 6603 074 1475 DD005
2 00-47657 6606 074 1475 DD007
___ 00-0238703 6602 074 773 DD052
___ 00-0128330 6602 074 773 DD051

More...
F3=Exit F5=Refresh F11=Display disk configuration capacity
F12=Cancel

Note: If you are performing a complete system restore, all the disk units on the
system may not report in right away. Verify that the number of disk
units displayed matches the number of disk units physically attached to
the system. If they do not match, wait a few minutes and press F5
(Refresh) until all of the disk units report in.

670 OS/400 Backup and Recovery V5R1

 4. Type an ASP number next to each disk unit that you want to configure. If you
type an ASP number that does not currently exist on your system, the system
will create that ASP.
| Number 1 is reserved for the system ASP. You can enter a number from 2 to 32.
| Numbers 33 to 99 are reserved for independent ASPs.
The Confirm Add Units display appears:

Confirm Add Units

Add will take several minutes for each unit. The system will
have the displayed protection after the unit(s) are added.

Press Enter to confirm your choice for 1=Add units.

Press F9=Capacity Information to display the resulting capacity.
Press F12=Cancel to return and change your choice.

Serial Resource
ASP Unit Number Type Model Name Protection
1 Unprotected
1 00-48519 6606 030 DD010 Unprotected
2 00-86978 6606 050 DD009 Unprotected
3 00-52262 6606 074 DD008 Device Parity
4 00-61300 6603 074 DD006 Device Parity
2 Unprotected
5 00-95744 6603 074 DD005 Device Parity
6 00-47657 6606 074 DD007 Device Parity

F9=Resulting Capacity F12=Cancel

The Confirm Add Units display shows what the entire system configuration
will be when you add the units. If you have more than one ASP on your
system, verify this configuration against your planned configuration.
5. You can press F9 (Resulting capacity) to see how the change will affect your
disk utilization. You are shown the following display:

Resulting Capacity
The configuration change that you requested would result in the
following ASP capacities.

Press Enter to continue.

-----------Current---------- ----------Proposed----------
--Protected-- -Unprotected- --Protected-- -Unprotected-
ASP Threshold Size %Used Size %Used Size %Used Size %used
1 90% 0 0.00% 1967 23.98% 2950 0.07% 3934 12.02%
2 90% 2950 0.07% 0 0.00%

6. Press F12 (Cancel) to return to the Confirm Add Units display.

7. If you are satisfied with the configuration, press the Enter key to add the disk
units to the ASP. If you want to make changes, press F12 to return to step 4
Adding units can take from several minutes to several hours. See “Appendix C.
Example of Configuration Planning for Mirrored Protection” on page 795 for
information on how to estimate the amount of time required. During that time,
you are shown the Function Status display:

Chapter 25. Working with Auxiliary Storage Pools 671

 Function Status

You selected to add units.

5 % Complete

The system updates the display periodically.

Note: You can press F16 to return to the Use Dedicated Service Tools (DST)
menu if you have other tasks to perform. However, you cannot perform
any disk configuration tasks or end DST until the system has finished
adding disk units.

The time it takes the system to add units depends on the type, model, and size
of each unit being added and the ability of the system to do multiple adds at
the same time.
8. If you have no other tasks to perform, end DST or SST. (See “How to Stop
Dedicated Service Tools (DST)” on page 662 or “Stopping System Service Tools
(SST)” on page 662.)

How to Change the Storage Threshold for an Auxiliary Storage Pool

The storage threshold for an ASP determines when the system warns you that the
space allocated for the ASP is almost full. The default for an ASP is 90%. To change
the storage threshold for an ASP, do the following:
1. From the System Service Tools (SST) menu, do the following:
or from the Use Dedicated Service Tools (DST) menu, do the following:
┌────────────────────────┐ Select option 3
│ SST Menu │ (Work with disk units)
│ │
│ ┌───────────────────┴───┐ Select option 2
│ 3 │ Work with Disk Units │ (Work with disk configuration)
└────┤ │
│ ┌──────────────────┴──────┐
│ 2 │ Work with Disk │
└────┤ Configuration │
│ │
│ │
└─────────────────────────┘

Note: If you are not already using DST, see “How to Start Dedicated Service
Tools (DST)” on page 660.

672 OS/400 Backup and Recovery V5R1

 ┌────────────────────────┐ Select option 4
│ DST Menu │ (Work with disk units)
│ │
│ ┌───────────────────┴─────┐ Select option 1
│ 4 │ Work with Disk Units │ (Work with disk configuration)
└────┤ │
│ ┌────────────────────┴────┐ Select option 3
│ │ Work with Disk │ (Work with ASP configuration)
│ 1 │ Configuration │
└────┤ ┌────────────────────┴─────┐
│ 3 │ Work with ASP │
└────┤ Configuration │
│ │
│ │
└──────────────────────────┘

If you are not already using DST, perform a manual IPL to start DST. See “How
to Start Dedicated Service Tools (DST)” on page 660.
2. Select the option to work with the ASP threshold. The Select ASP to Change
Threshold display is shown.

Select ASP to Change Threshold

Type option, press Enter.
1=Select
--Protected-- --Unprotected--
Option ASP Threshold Overflow Size %Used Size %Used
1 90% No 6046 0.31% 7676 6.36%
1 2 90% No 2950 * 0 0.00%

3. On the Select the ASP to Change Threshold display, select the ASP that you
want to have a different threshold. Press the Enter key. The following display is
shown.

Change Storage Threshold

--Protected-- --Unprotected--
ASP Threshold Overflow Size %Used Size %Used
2 90% No 2950 * 0 0.00%

This is an unprotected ASP. The threshold represents the amount of

unprotected storage used before a warning message is sent to the system
operator. Type choice, press Enter.
New threshold . . . . . . . . 88% 1-100

4. Type your choice for the New threshold prompt and press the Enter key.
5. If you have no other tasks to perform, end DST or SST. (See “How to Stop
Dedicated Service Tools (DST)” on page 662 or “Stopping System Service Tools
(SST)” on page 662.)

How to Change the Storage Threshold for the System Auxiliary

Storage Pool
It is important to prevent the system ASP from filling to capacity. If this occurs, the
system will end abnormally. You can try to avoid this by specifying a storage
threshold that warns you of a potential space shortage.

One way to establish this threshold is through Dedicated Service Tools (DST) or
System Service Tools (SST). Use the same procedures as you would when setting

Chapter 25. Working with Auxiliary Storage Pools 673

 the storage threshold for any other ASP. See “How to Change the Storage
Threshold for an Auxiliary Storage Pool” on page 672 for information on this
procedure.

Note: Establishing the threshold through DST will not prevent the system from
ending abnormally. It will only notify you when the system ASP reaches the
capacity threshold.

You can also protect the system ASP from filling to capacity by using the
QSTGLOWLMT and QSTGLOWACN system values. The QSTGLOWLMT system
value specifies the percentage of unallocated auxiliary storage that is remaining
when the critical storage lower limit is reached. If the system reaches that limit, the
QSTGLOWACN system value specifies what action the system should take. Using
this method allows the system to actively prevent an abnormal shutdown instead
of simply sending a warning of the condition.

Note: Using these system values does not affect any existing storage threshold that
you may have set through DST.

You can use the QSTGLOWLMT and QSTGLOWACN system values on the
following commands:

CHGSYSVAL RTVSYSVAL
DSPSYSVAL WRKSYSVAL

The following procedure demonstrates how to use these sytem values. (The
WRKSYSVAL command is used as an example.)
1. At a command line, type WRKSYSVAL and press Enter. You are shown the Work
with System Values display.

Work with System Values

System: YOURSYS
Position to . . . . . _______ Starting characters of system value
Subset by Type . . . . _______ F4 for list

Type options, press Enter.

2=Change 5=Display

System
Option Value Type Description
_ QSTGLOWACN *STG Auxiliary storage lower limit action
_ QSTGLOWLMT *STG Auxiliary storage lower limit

2. Type a 2 in the option field to change QSTGLOWACN and press Enter. You
must have *ALLOBJ and *SECADM authority to change QSTGLOWACN. You
are shown the Change System Value display.

674 OS/400 Backup and Recovery V5R1

 Change System Value

System value . . . . . : QSTGLOWACN

Description . . . . . : Auxiliary storage lower limit action

Type choice, press Enter.

Action . . . . . . . . *MSG_______ *MSG

*CRITMSG
*REGFAC
*ENDSYS
*PWRDWNSYS

3. On the Change System Value display, type the name of the action that you
want the system to perform after reaching the critical storage lower limit. Press
the Enter key. The actual actions that are performed by the action names are as
follows:
*MSG
The system sends the CPI099C message to the QSYSMSG and QSYSOPR
message queues. (The system also sends this message when you select any
one of the other actions.)
*CRITMSG
The system sends the CPI099B critical message to the user who is specified
in the service attribute to receive critical messages.
*REGFAC
The system submits a job to call exit programs that are registered for the
QIBM_QWC_QSTGLOWACN exit point.
*ENDSYS
The system ends to the restricted state.
*PWRDWNSYS
The system powers down immediately and restarts
4. From a command line, type DSPSYSVAL and press the Enter key. The Display
System Value screen is shown.
The lower limit value is the lowest amount of unused storage that can exist in
Display System Value

System value . . . . . : QSTGLOWLMT

Description . . . . . : Auxiliary storage lower limit

Lower limit . . . . . : 1.0000 0-100 percent

the system ASP before the system performs the QSTGLOWACN action. (You
can use the WRKSYSSTS command to view the amount of storage that is
currently being used in the system ASP.) The system is shipped with the
QSTGLOWLMT system value set to 5.0. Any change you make to this system
value takes effect immediately.

Note: If the DST threshold is above 95%, the lower limit value will be set to
the difference between 100% and the threshold setting. For example, if
the DST threshold is set to 98%, QSTGLOWLMT will be set to 2.0. (100
— 98 = 2.) This only occurs at the time you install V4R2.

Chapter 25. Working with Auxiliary Storage Pools 675

How to Move a Disk Unit to a Different Auxiliary Storage Pool
You may want to move a disk unit from one ASP to another. For example, you
want to create a user ASP for journal receivers and move some of the disk units on
your system to the new user ASP. You can accomplish this in one process. When
you move a disk unit to an ASP that does not exist, the system creates the ASP.

You may also decide to move disk units because you no longer need to have user
ASPs on your system and you want to move all the disk units back to the system
ASP.

Restrictions When Changing Your ASP Configuration: Consider these things when
you are planning to move disk units from an ASP:
v The system may take a long time to move the unit because it must copy the data
from that unit to other units in the ASP.
v You cannot move unit 1 (the load source unit) from the system ASP.
v You cannot move disk units from a user ASP that is overflowed.
v You cannot move units in and out of the same ASP in the same operation.
v When mirrored protection is active for an ASP, you cannot move disk units in
and out of the ASP. You must remove disk units in pairs from a mirrored ASP.
You can then add them to a different ASP.
v When mirrored protection is active for the ASP that contains the disk units, you
must remove both units of a mirrored pair.
v When you remove a disk unit, it becomes nonconfigured.
To move disk units between ASPs, do the following:
1. If you are not already using DST, perform a manual IPL to start DST. See
“How to Start Dedicated Service Tools (DST)” on page 660.
2. From the Use Dedicated Service Tools (DST) menu, do the following:

┌────────────────────────┐ Select option 4

│ DST Menu │ (Work with disk units)
│ │
│ ┌───────────────────┴─────┐ Select option 1
│ 4 │ Work with Disk Units │ (Work with disk configuration)
└────┤ │
│ ┌────────────────────┴────┐ Select option 3
│ │ Work with Disk │ (Work with ASP configuration)
│ 1 │ Configuration │
└────┤ ┌────────────────────┴─────┐
│ 3 │ Work with ASP │
└────┤ Configuration │
│ │
│ │
└──────────────────────────┘

3. Select option 6 (Move units from one ASP to another) from the Work with
ASP Configuration display. The Specify ASP to Move Disk Units display is
shown.

676 OS/400 Backup and Recovery V5R1

 Specify ASP to Move Disk Units
To move units to different ASPs, specify the ASP that you want to move each
one to in the 'New ASP' field. Specify the units to be moved, press Enter.

New Current Serial --Protected-- --Unprotected--

ASP ASP Unit Number Type Model Size %Used Size %Used
1 0 0.00% 4124 41.50%
1 00-0193825 6602 030 0 0.00% 1031 82.00%
2 00-0163477 6602 030 0 0.00% 1031 29.00%
2 3 00-0190494 6602 030 0 0.00% 1031 27.00%
2 4 00-17900 6602 030 0 0.00% 1031 28.00%

4. Type the number of the ASP you want to move the units to in the New ASP
column and press the Enter key. If you specify an ASP that does not
currently exist on your system, the system creates a new ASP. If the move
operation would leave the source ASP with insufficient storage, you receive
an error message.
If you see the Confirm Move of Unit display, skip to step 6.
The Confirm Continuation display is shown if the storage management
directories are not usable:

Confirm Continuation
In order to proceed the system must perform internal processing that may
take several minutes during which the system may appear inactive. Press
Enter to continue. Press F12=Cancel to return and change your choice.

5. Determine whether you want to cancel the procedure or continue. If you

want to continue, press the Enter key.
6. The Confirm Move of Unit display is shown:

Confirm Move of Unit

Moving units will take several minutes. Press Enter to confirm your
choice to move the units. Press F9=Capacity information to display the capacity.
Press F12=Cancel to return to change your choice.
Serial --Protected-- --Unprotected--
ASP Unit Number Type Model Size %Used Size %Used
1 0 0.00% 2062 83.00%
1 00-0193825 6602 030
2 00-0163477 6602 030
2 0 0.00% 2062 0.01%
3 00-0190494 6602 030
4 00-17900 6602 030

Press F9 (Capacity information) to display the resulting capacity.

Resulting Capacity
The configuration change that you requested would result in the following
ASP capacities. Press Enter to continue.
-----------Current---------- ----------Propose-----------
--Protected-- -Unprotected- --Protected-- -Unprotected-
ASP Threshold Size %Used Size %Used Size %Used Size %Used
1 90% 0 0.00% 4124 41.50% 0 0.00% 2062 83.00%
2 90% 0 0.00% 2062 0.01%

7. Press the Enter key to return to the Confirm Move of Unit display.

Chapter 25. Working with Auxiliary Storage Pools 677

 8. Press the Enter key on the Confirm Move of Units display to move the
selected units. The system will move the data off the selected units to the
remaining units in the source ASP. The move can take several minutes
during which the system appears inactive.
9. When the move operation is complete, you return to the Work with ASP
Configuration display.
10. If you have no other tasks to perform, end DST. (See “How to Stop
Dedicated Service Tools (DST)” on page 662.)

How to Remove a Disk Unit from an Auxiliary Storage Pool

Use the procedure for removing disk units from your system for the following
reasons:
v You want to physically remove a disk unit from your system.
v You want to move a disk unit from an ASP that has mirrored protection to
another ASP. You can remove pairs of disk units from a mirrored ASP without
stopping mirrored protection. You then add the disk units to the target ASP.

Considerations When Changing Your ASP Configuration: Consider these things

when you are planning to remove disk units from an ASP:
v The system may take a long time to remove the unit because it must copy the
data from that unit to other units in the ASP.
v To proceed, the system must perform internal processing that may take several
minutes during which the system may appear inactive.
v When you remove a disk unit, it becomes nonconfigured.

Restrictions When Changing Your ASP Configuration: Consider these restrictions

when you are planning to remove disk units from an ASP:
v You cannot remove unit 1 (the load source unit) from the system ASP.
v You cannot remove disk units from a user ASP that is overflowed.
v When mirrored protection is active for the ASP that contains the disk units, you
must remove both units of a mirrored pair.
v If you are going to physically remove a disk unit from a 6502, 6512, 6751, or
6532 IOP, you must either exclude the disk unit or stop device parity protection
first.
v If you are going to physically remove a disk unit from a 9337 Disk Array
Subsystem, you must either exclude the disk unit or stop device parity
protection.

To remove a disk unit, do the following:

1. If you are not already using DST, perform a manual IPL to start DST. See “How
to Start Dedicated Service Tools (DST)” on page 660.
2. From the Use Dedicated Service Tools (DST) menu, do the following:

678 OS/400 Backup and Recovery V5R1

┌────────────────────────┐ Select option 4
│ DST Menu │ (Work with disk units)
│ │
│ ┌───────────────────┴─────┐ Select option 1
│ 4 │ Work with Disk Units │ (Work with disk configuration)
└────┤ │
│ ┌────────────────────┴────┐ Select option 3
│ │ Work with Disk │ (Work with ASP configuration)
│ 1 │ Configuration │
└────┤ ┌────────────────────┴─────┐
│ 3 │ Work with ASP │
└────┤ Configuration │
│ │
│ │
└──────────────────────────┘

3. You are shown the Remove Units from Configuration display.

Remove Units from Configuration

Type options, press Enter.

4=Remove unit from configuration

Serial Resource
OPT Unit ASP Number Type Model Name Status
2 1 10-00A7529 9332 400 DD010 Configured
3 1 10-00A4936 9332 400 DD012 Configured
4 1 10-00A4936 9332 400 DD019 Configured
4 5 1 10-00A7498 9332 400 DD025 Configured
4 6 1 10-00A7498 9332 400 DD036 Configured
7 1 10-00A7530 9332 400 DD042 Configured
8 1 10-00A7530 9332 400 DD052 Configured

4. Type a 4 (Remove unit from configuration) in the OPT column for each unit that
you want to remove and press the Enter key. If the remove operation would
leave the ASP with insufficient storage, you receive an error message.
If you see the Confirm Remove Disk Units display, skip to 6.
The Confirm Continuation display may be shown before the Confirm Remove
Disk Units display if the storage management directories are not usable.

Confirm Continuation

To proceed, the system must perform internal processing

that may take several minutes during which the system may
appear inactive.

Press Enter to continue.

Press F12=Cancel to return and change your choice.

5. Determine whether you want to cancel the procedure or continue. If you want
to continue, press the Enter key.
6. The Confirm Remove Disk Units display is shown:

Chapter 25. Working with Auxiliary Storage Pools 679

 Confirm Remove Disk Units

Removing disk units will take several minutes.

Press Enter to confirm remove of disk units.

Press F9=Capacity information to display the capacity information.
Press F12=Cancel to return to change your choice.

Serial Resource
OPT Unit ASP Number Type Model Name Status
4 5 1 10-00A7498 9332 400 DD010 Configured
4 6 1 10-00A7498 9332 400 DD012 Configured

Press F9 (Capacity information to display the resulting capacity.

Resulting Capacity

The configuration change that you requested would result in the

following ASP capacities.

Press Enter to continue.

-----------Current---------- ----------Modified--------
--Protected-- -Unprotected- --Protected-- -Unprotected-
ASP Threshold Size %Used Size %Used Size %Used Size %Used
1 90% 0 0.00% 1600 52.70% 0 0.00% 1200 70.26%

7. Press the Enter key to return to the Confirm Remove Disk Units display.
8. Press the Enter key on the Confirm Remove Disk Units display to remove the
selected units. The system moves the data off the units selected to be removed
to the remaining units in the source ASP. The remove can take several minutes
or several hours during which the system appears inactive.
Notes:
a. The time it takes to remove a unit depends on the disk unit type and
model.
b. If the data on the unit being removed is severely fragmented and the
amount of storage used is high, the remove operation could take several
hours.
9. When the remove operation is complete, you return to the Work with ASP
Configuration display.
If you have no other tasks to perform, end DST. (See “How to Stop Dedicated
Service Tools (DST)” on page 662.)

How to Delete an Auxiliary Storage Pool

When you delete a user ASP, the status of all the units that are assigned to the ASP
becomes nonconfigured. Any data that is in the ASP is deleted. The normal
procedure is to remove all objects from the ASP before you delete the ASP. You can
do this either by moving the objects to a different ASP or by deleting the objects. If
you delete an ASP that has objects in it, those objects are marked by the system as
damaged or destroyed.

You cannot delete ASP 1, which is the system ASP and holds the operating system.

To delete a user ASP, do the following:

680 OS/400 Backup and Recovery V5R1

 1. If you are not already using DST, perform a manual IPL to start DST. See “How
to Start Dedicated Service Tools (DST)” on page 660.
2. From the Use Dedicated Service Tools (DST) menu, do the following:

┌────────────────────────┐ Select option 4

│ DST Menu │ (Work with disk units)
│ │
│ ┌───────────────────┴─────┐ Select option 1
│ 4 │ Work with Disk Units │ (Work with disk configuration)
└────┤ │
│ ┌────────────────────┴────┐ Select option 3
│ │ Work with Disk │ (Work with ASP configuration)
│ 1 │ Configuration │
└────┤ ┌────────────────────┴─────┐
│ 3 │ Work with ASP │
└────┤ Configuration │
│ │
│ │
└──────────────────────────┘

3. Select option 2 (Delete user ASP) on the Work with ASP Configuration display
and press the Enter key.

Delete User ASP

Type option, press Enter 4=Delete
--Protected-- --Unprotected--
Option ASP Threshold Overflow Size %Used Size %Used
1 90% No 600 77.84% 0 0.00%
2 90% No 0 0.00% 200 0.53%
3 90% No 0 0.00% 200 0.53%

4. Type a 4 in the Option field of the ASP you want to delete and press the Enter
key. The Confirm Delete of User ASP display is shown.

Confirm Delete Of User ASP

Warning: Deleting a user ASP will remove all units of that ASP from the
configuration. The units will become nonconfigured. Press F10 to confirm
your choice for 4=delete Press F12=Cancel to return to change your choice
--Protected-- --Unprotected--
Option ASP Threshold Overflow Size %Used Size %Used
4 2 90% No 0 0.00% 200 0.53%

5. Press F10 (Confirm) to confirm that delete of the ASP. The delete operation may
take several minutes.
6. If you have no other tasks to perform, end DST. (See “How to Stop Dedicated
Service Tools (DST)” on page 662.)

How to Calculate Space Requirements for an Auxiliary Storage Pool

When you are planning to make changes to the disk configuration or disk
protection on your system, you need to calculate the space requirements for the
change before you begin. You want to ensure that your system has sufficient disk
storage for the changes.

Table 99 on page 682 shows an example of a worksheet that you can use to
calculate your requirements. The following topics describe the steps for completing
the worksheet.

Chapter 25. Working with Auxiliary Storage Pools 681

 Table 99. Worksheet for Calculating ASP Capacity
Amount of Storage Used or Estimated
Type of Storage Current Interim Final
ASP Number: 1
Protected
Unprotected
Total
Total Capacity
Percent Used
ASP Number: 2
Protected
Unprotected
Total
Total Capacity
Percent Used
ASP Number: 3
Protected
Unprotected
Total Used
Total Capacity
Percent Used

Task 1–Calculating the Current Storage Used

Begin by determining how your current system configuration uses the disk
capacity. The following steps use the disk configuration shown in Figure 85 as an
example.

Display Disk Configuration Capacity

--Protected-- --Unprotected--
ASP Unit Type Model Threshold Overflow Size %Used Size %Used
1 90% No 1804 38.86% 2062 54.00%
1 6602 030 0 0.00% 1031 82.00%
2 6602 074 773 48.00% 0 0.00%
3 6602 070 1031 32.00% 0 0.00%
6 6602 030 0 0.00% 1031 26.00%
3 90% No 1546 19.50% 0 0.00%
4 6602 074 773 18.00% 0 0.00%
5 6602 074 773 21.00% 0 0.00%

More... Press Enter to continue. F3=Exit F5=Refresh

F11=Display disk configuration protection F12=Cancel

Figure 85. Sample Disk Configuration

If you have not already done so, print a copy of your current disk configuration.
See “Displaying Your Disk Configuration–Software View” on page 665.

682 OS/400 Backup and Recovery V5R1

 For each ASP on your system, calculate the number of megabytes that are being
used for protected and unprotected storage. Do the following:
1. Multiply the value in the Size column under protected for the ASP by the value
in the % Used column. The result is the megabytes of protected storage for the
ASP.
In the example, ASP 1 has 1804 of protected storage. The protected storage in
ASP 1 is 38.86 percent used:
1804 x .3886 = 701MB used
2. Multiply the value in the Size column under Unprotected for the ASP by the
value in the % Used column. The result is the megabytes of unprotected storage
for the ASP.
In the example, ASP 1 has 2062 of unprotected storage. The unprotected storage
in ASP 1 is 54.00 percent used:
2062 X .54 = 1114MB used
3. Determine the total storage used by adding the protected storage and the
unprotected storage:
701 + 1114 = 1815MB used
4. Calculate the total storage capacity for the ASP by adding the total protected
storage and the total unprotected storage:
1804 + 2062 = 3866MB total storage
5. Calculate the total percent used. Divide the total storage (from step 3) by the
total storage used (from step 4):
1815 / 3866 = 47%

Table 100 shows these amounts filled in:

Table 100. Calculating ASP Capacity–Example 2
Amount of Storage Used or Estimated
Type of Storage Current Interim Final
ASP Number: 1
Protected 701
Unprotected 1114
Total used 1815
Total capacity 3866
Percent used 47%

Task 2–Calculating Storage Needed

Some changes that you make to your disk configuration require work space on the
system. For example, if you move a disk unit from one ASP to another, the system
must copy the information from the disk unit to the other disk units in the source
ASP. This topic has several examples of changes to disk configuration. It explains
how to calculate the amount of space that is required during the procedure and
when the procedure has completed. Find the example that is closest to your
situation and use it as a guide in calculating your disk requirements.

Moving Disk Units to a New Auxiliary Storage Pool–Example: Look at Figure 85

on page 682. For this example, assume that you want to move disk units 2 and 3
from ASP 1 to a new user ASP, ASP 2. Do the following to determine the amount
of interim space required by the system for this procedure:

Chapter 25. Working with Auxiliary Storage Pools 683

 1. Calculate the number of megabytes that you plan to remove from the ASP by
adding the capacities of the units to be moved:
773 + 1031 = 1804MB
2. Calculate the amount of storage that the ASP will have after the disk units are
removed. Subtract the amount in step 1 from the total capacity of the ASP:
3866 - 1804 = 20622MB
3. Determine what the interim-percent-used amount will be. Divide the amount
from step 3 on page 683 by the amount from step 2 on page 684:
1815 / 2062 = 88.02 % used
4. Estimate the amount of data that you will move from the source ASP to the
target ASP. For example, if you plan to move a library you would use one of
the following:
v The DSPLIB LIB(library-name) OUTPUT(*PRINT) command
v The QLIRLIBD API
5. Subtract the amount in step 4 from the total storage used for the source ASP.
Add it to the target ASP. For purposes of this example, assume you are moving
several large libraries whose size totals 196MB.

Table 101 shows these amounts filled in:

Table 101. Calculating ASP Capacity–Example 3
Amount of Storage Used or Estimated
Type of Storage Current Interim Final
ASP Number: 1
Protected 701
Unprotected 1114
Total used 1815 1815 1619
Total capacity 3866 2062 2062
Percent used 47% 88.02% 78.52%
ASP Number: 2
Protected
Unprotected
Total used 196
Total capacity 1804
Percent used 10.86%

How to Display the Objects in a User ASP

To print a list of all the objects in a nonlibrary user ASP, use the DSPOBJD
command. Specify object types *SAVF, *JRN, and *JRNRCV. The object description
information includes the ASP where the object is located.

To list all the documents in a user ASP, use the Query Document Library
(QRYDOCLIB) command:
QRYDOCLIB ... QRYDFN(*IF(*ASP *EQ 4))

684 OS/400 Backup and Recovery V5R1

 To determine which ASP an object is in, use the DSPOBJD command and look at
the number shown on the Auxiliary storage pool field. To determine which ASP a
DLO is in, use the DSPDLONAM command. Look at the number shown on the
Auxiliary storage pool field.

| If the object is an IFS object, use the Display Object Links (DSPLNK) command.
| Select option 8 (Display attributes) to determine which ASP the object is in.

Balancing an Auxiliary Storage Pool

Beginning with V4R4M0, you can use the ASP balancing function. This function
improves system performance by balancing disk utilization across all of the disk
arms in an ASP. You can use the Start ASP Balance (STRASPBAL) command to
start the function. You will need to select the method of balancing that you wish to
use:
v Capacity balancing
v Usage balancing
v Hierarchical Storage Management (HSM) balancing

| Note: You cannot balance journal receivers among the disk units of an ASP
| because journal receivers can only be spread across 10 disk arms (100 if you
| specify RCVSIZOPT(*MAXOPT1) or RCVSIZOPT(*MAXOPT2)).

Before using usage balancing or HSM balancing, you must run the Trace ASP
Balance (TRCASPBAL) command. This command starts a trace function that
collects statistics on the data in the ASPs that you wish to balance. Data that is
used often is referred to as high use or hot data. Data that is not used often is
referred to as low use or cold data.

To end the ASP balancing function, use the End ASP Balance (ENDASPBAL)
command.

Capacity Balancing
When you use capacity balancing, the data on the disk units within an ASP is
distributed evenly across all the units. Instead of some units containing the
majority of the data, each unit has an equal percentage of used and unused space.
This type of balancing is useful when you add new disk units to an ASP.

Usage Balancing
Usage balancing is useful when the ASP contains some disk units which are
utilized to a higher degree than other disk units in the ASP. The TRCASPBAL
command must finish collecting statistics before usage balancing can begin. When
you use usage balancing, the high use and low use data on each unit in the ASP is
redistributed to balance the arm utilization of each unit within the specified ASP.

Hierarchical Storage Management (HSM) Balancing

Hierarchical Storage Management balancing can only be used ASPs that contain a
mixture of compressed and non-compressed disk units. The TRCASPBAL
command must finish collecting statistics before hierarchical storage management
balancing can begin. When you use hierarchical storage management balancing,
the high use and low use data on each unit in the ASP is redistributed. The high
use data is moved to high performance units and the low use data is moved to
low performance units. After the balance activity has completed, the system clears
the trace information.

Chapter 25. Working with Auxiliary Storage Pools 685

 Transferring Objects between Auxiliary Storage Pools
This topic explains how to move entire libraries or folders from one ASP to
| another ASP. It also describes special procedures when you are moving a library
| that contains journals because a journal and the journaled objects must be in the
| same ASP. “How to Work with Nonlibrary User ASPs” on page 692 discusses the
procedures for working with nonlibrary user ASPs.

You cannot directly move objects between ASPs because the MOVOBJ command
and the MOVDOC command move only the pointer to the object. They do not
physically copy data from one location to another. In general, do the following to
move an object to a different ASP:
1. Save the object.
2. Delete the object from the system.
3. Restore the object to the target ASP by using the RSTASP parameter on the
RSTxxx command.

These restrictions apply when specifying the RSTASP parameter:

v When you attempt to restore an object to a different ASP from its library, the
ASP must be a nonlibrary user ASP and the object must be a journal, a journal
receiver, or a save file.
– For journals, journal receivers, and save files, if a library exists on the ASP,
you receive an error message and the object is not restored.
– For other object types, the object is restored to the ASP that contains the
library.
v If you try to restore an object to a user ASP by explicitly specifying the desired
user ASP for the RSTASP parameter and the designated user ASP does not exist,
you receive a message. The object is not restored.
v If you restore an object and specify RSTASP(*SAVASP), and if the ASP from
which the object was saved no longer exists, the object is restored to the system
ASP. You receive an informational message.

How to Move Authorities to a Different ASP

Because you must delete an object in order to move it to another ASP, private
authorities to the object are lost. To move authorities for an object, do the
following:
1. Sign on as QSECOFR.
2. Copy the object’s authorities to a temporary object.
a. Create a temporary object:
CRTDTAARA QTEMP/X *CHAR AUT(*EXCLUDE)
b. Copy the authorities:
RVKOBJAUT QTEMP/X *DTAARA QSECOFR *ALL
GRTOBJAUT OBJ(QTEMP/X) OBJTYPE(*DTAARA)
REFOBJ(object)
REFOBJTYPE(object-type)
3. Save the object to your save media.
4. Delete the object from the system.
5. Restore the object to the target ASP.
6. Copy the authorities to the restored object.
GRTOBJAUT OBJ(object) OBJTYPE(object-type)
REFOBJ(QTEMP/X)
REFOBJTYPE(*DTAARA)

686 OS/400 Backup and Recovery V5R1

 7. Delete the temporary object:
DLTDTAARA QTEMP/X

How to Transfer a Library to a Different ASP

Use the following procedure to move a library to a different ASP. This example
moves the CUSTLIB library from ASP 1 to ASP 2.
1. Save the private authorities for the library: SAVSECDTA DEV(TAP01).
2. Save the library: SAVLIB LIB(CUSTLIB) DEV(TAP01) ACCPTH(*YES). Consider
saving the object twice to 2 different media volumes.
3. Delete the library: DLTLIB LIB(CUSTLIB).
4. Place the system in a restricted state: ENDSBS *ALL *IMMED.
5. Restore the private authorities you saved in step 1: RSTUSRPRF USRPRF(*ALL)
DEV(TAP01)
6. Restore the library to the new user ASP: RSTLIB SAVLIB(CUSTLIB) RSTASP(2)
7. Restore authority to the library and its objects: RSTAUT

How to Transfer a Folder to a Different ASP

Use the following procedure to move a folder to a different ASP. This example
moves the HRFLR folder from ASP 1 to ASP 2. Do not move IBM-supplied folders
(those starting with Q) to a user ASP. These folders must be in the system ASP.
1. Save the private authorities for the folder: SAVSECDTA DEV(TAP01).
2. Save the folder: SAVDLO DLO(*ALL) FLR(HRFLR) DEV(TAP01). Consider saving the
object twice to 2 different media volumes.
3. Delete the folder: DLTDLO DLO(*ALL) FLR(HRFLR). Do not skip this step. If you
restore a DLO to an ASP and it already exists in another ASP, you receive an
error message. The restore operation continues with the next DLO. If you issue
a restore command for a large number of DLOs, you will receive a large
number of error messages.
4. Place the system in a restricted state: ENDSBS *ALL *IMMED.
5. Restore the private authorities you saved in step 1: RSTUSRPRF USRPRF(*ALL)
DEV(TAP01)
6. Restore the folder to the new user ASP: RSTDLO DLO(*ALL) SAVFLR(HRFLR)
RSTASP(2)
7. Restore authority to the folder and its objects: RSTAUT

You can move more than one folder at a time by specifying multiple folders on the
SAVDLO and RSTDLO commands. If you save DLOs from more than one ASP,
you must specify sequence numbers on the RSTDLO command.

How to Transfer Journals and Objects to a Different ASP

| If you use a library user ASP, both the objects you are journaling and the journal
| must be in the same ASP. For the purpose of recovery as well as performance, it is
| recommended that the journal receiver be placed in a different user ASP. If a
| failure occurs in the ASP that contains the objects and the journal, you do not lose
| both the objects and the journaled changes that are in the receiver. Placing your
| objects and journal receiver in the same user ASP also causes contention between
| access to the object and access to the journal receiver.

Use the following procedure to move a journal and the associated journaled objects
to a different ASP. This procedure applies to library user ASPs (where the journal

Chapter 25. Working with Auxiliary Storage Pools 687

 and its library are in the same ASP). If the journal is in a nonlibrary user ASP, refer
to “How to Work with Nonlibrary User ASPs” on page 692.

| Transferring journaled objects

| 1. Save private authorities for the journal and journaled objects: SAVSECDTA
| DEV(TAP01)
| 2. Save the journal using the SAV, SAVOBJ or the SAVLIB command.
| 3. Because the journal and journaled objects must be in the same ASP, the objects
| must also be moved to the same user ASP before you can resume journaling
| the objects after the move.
| 4. Save any objects that you journal and save any logical files that have their
| access paths journaled. You can use the Work with Journal Attributes
| command to determine which objects that you journal. Consider saving the
| journal and journaled objects twice to two different media volumes.
| 5. Delete the journaled objects using the appropriate delete command.
| 6. Delete the journal using the Delete Journal (DLTJRN) command.
| 7. Delete the library that contained the journal.
| 8. Create the library for the journal in the user ASP using the Create Library
| (CRTLIB) command:
| CRTLIB LIB(library-name) ASP(asp-number)

| Note: The new library must have the same name as the library in which the
| journal was originally located.
| 9. Place the system in a restricted state: ENDSBS *ALL *IMMED
| 10. Restore the user profiles you saved in step 1:
| RSTUSRPRF USRPRF(*ALL) DEV(TAP01)
| 11. Restore the journal to the library in the user ASP using the Restore Object
| (RSTOBJ) command.
| 12. Restore the previously journaled objects to the library or directory in the user
| ASP. If you want to restore the previously journaled objects to their original
| libraries or directories, you must first move those libraries or directories to the
| user ASP. You move libraries and directories to a different ASP by saving
| them, deleting them, and restoring them to the new ASP.
| Restoring the previously journaled objects automatically resumes journaling
| for the objects if the journal already exists.
| 13. Restore the private authorities you saved in step 1:
| RSTAUT
| 14. Save the journaled objects so that the journaled changes can be applied, if
| necessary. When journaling starts, the system assigns a journal identifier (JID)
| to the object. Usually the JID that is assigned is the same JID that the object
| had when it was saved. The object must be saved after the JID is assigned.

| How to Create Objects in a Library User ASP

| You create an object in a specific ASP by placing it in a library or folder that is in
| the ASP. You place a library or a folder in an ASP in one of the following ways:
| v Specify a value for the ASP parameter on either the CRTLIB command or the
| CRTFLR command. The default for the ASP parameter is 1.
| v Restore the library or the folder to a specific ASP by using the RSTASP
| parameter on the restore command.

688 OS/400 Backup and Recovery V5R1

| IFS objects can reside in a user ASP through the use of User-defined file systems
| (UDFS) by doing the following:
| 1. Create a user defined file system in the ASP.
| 2. Mount the UDFS over another directory and use it through the mounted-over
| path.
| For more information about User-defined files sytems, see the Information Center
| at the following Web site:
| http://www.ibm.com/eserver/iseries/infocenter

| Placing a Document in a User ASP–Example

The following is an example of placing a document in a specific user ASP:
1. To create a folder in a user ASP, use the ASP parameter on the CRTFLR
command:
CRTFLR FLR(ASP3FLR) ASP(3)
2. To create a document in that folder, use the CRTDOC command or a program
that creates documents.

When you create a document or another folder in ASP3FLR, the new document or
folder is automatically placed in ASP 3.

When you create the first folder in a user ASP, the system creates the
corresponding library. For example, when you create the ASP3FLR folder, the
system creates the QDOC0003 library if it does not already exist. You should never
create a QDOCnnnn library yourself. This can cause unpredictable results.

Placing an Object in a User ASP–Example

The following is an example of placing a journal receiver in a specific user ASP:
1. Create a library for the journal receiver:
CRTLIB LIB(ASP2LIB) ASP(2)
2. Create the journal receiver in the library that you created in the user ASP:
CRTJRNRCV JRNRCV(ASP2LIB/RCVINASP2)

| Creating a UDFS in User ASP–Example

| The following is an example of creating a UDFS in a user ASP using the Create
| User-Defined File System (CRTUDFS) command and the Add Mounted File System
| (MOUNT) command.
| 1. Create a user-defined file system in ASP 2. All objects created in this new file
| system will also reside in ASP 2:
| CRTUDFS UDFS('/dev/qasp02/asp2dir.udfs')
| 2. Mount the user-defined file system over directory ’/myLocalPath’:
| MOUNT TYPE(*UDFS) MFS('/dev/qasp02/asp2dir.udfs') MNTOVRDIR('/myLocalPath')
| 3. Create a directory in the user-defined file system:
| CRTDIR DIR('/myLocalPath/newDir')

| How to Place Journal Receivers in a User ASP

Placing journal receivers in a user ASP may improve your system performance.
Dedicating a user ASP to the journal receivers for a single journal is the
recommended approach. Placing journal receivers in a library user ASP simplifies
recovery. Following are the procedures for placing journal receivers in both a
library user ASP and a nonlibrary user ASP.

Chapter 25. Working with Auxiliary Storage Pools 689

 Placing Journal Receivers in a Library User ASP
The following procedure uses the example of the CUSTJRN journal and journal
receivers that use the naming convention CUSTRnnnn.
1. Create a library for the journal receiver in the desired user ASP:
CRTLIB LIB(CUSTJRNR) ASP(4)
2. Use the WRKJRNA command to find the name of the currently attached journal
receiver: WRKJRNA JRN(CUSTJRNR/CUSTJRN). For the example, assume the
currently attached journal receiver is CUSTR0005.
3. Use option 8 (Display attributes) to determine the attributes for the current
receiver.
4. Create a new journal receiver in the library that is in the user ASP. Use a name
that continues your naming convention. Specify the attributes that you
displayed. For example, if the threshold of the current receiver is 5000, you
would specify the following:
CRTJRNRCV JRNRCV(CUSTJRNR/CUSTR0006) THRESHOLD(5000)
5. Change journal receivers so the new journal receiver is attached and actively
receiving journal entries:
CHGJRN JRN(CUSTJRN) JRNRCV(CUSTJRNR/CUSTR0006)
6. You can save the detached journal receiver and delete it from the system.
7. In the future, when you change journal receivers and specify JRNRCV(*GEN), the
system will create the new journal receiver in the CUSTJRNR library in ASP 4.

How to Move Journal Receivers From an Overflowed User

ASP
| To maintain journaling for the objects, do the following steps:
1. Use the WRKJRNA command to determine the names of the journal receivers
associated with the journal: WRKJRNA JRN(library-name/journal-name)
2. Use option 8 (Display attributes) to display the attributes of the receiver that is
attached.
3. If the journal receiver to be moved is attached to a journal, create a new journal
receiver on a different ASP by using the CRTJRNRCV command. Assume the
currently attached receiver is CUSTR0005. Use a name for the journal receiver
that continues your naming conventions.
If the journal receiver is in a library user ASP, do the following:
a. Create a new library in a different ASP, such as LIBJNEW: CRTLIB
LIB(LIBJNEW) ASP(4)
b. Create a new journal receiver in the library. Specify the attributes that you
displayed. For example, if the threshold of the current receiver is 5000, you
would specify the following:
CRTJRNRCV JRNRCV(LIBJNEW/CUSTR0006) THRESHOLD(5000)

If the journal receiver is in a nonlibrary user ASP, create a new journal receiver
in a different nonlibrary user ASP or in the system ASP: CRTJRNRCV
JRNRCV(CUSTJRNR/CUSTR0006) ASP(5)
4. Change the journal by using the Change Journal (CHGJRN) command. Specify
the newly created journal receiver on the JRNRCV parameter: CHGJRN
JRN(CUSTJRNR/CUSTJRN) JRNRCV(library-name/CUSTR0006)
5. Save the journal receivers from the overflowed user ASP. If the journal receivers
are the only objects in the library, use the Save Library (SAVLIB) command. If
other objects are in the library, use the Save Object (SAVOBJ) command.

690 OS/400 Backup and Recovery V5R1

 6. If you used the SAVLIB command in step 5 on page 690, delete the library from
the overflowed user ASP by using the DLTLIB command. If you used the
SAVOBJ command, delete the journal receivers using the DLTJRNRCV
command.
7. Journal receivers can be restored only to the library from which they were
saved. The steps required to restore these journal receivers if you need them for
a recovery depend on whether they were in a library user ASP or a nonlibrary
user ASP. For journal receivers in a nonlibrary user ASP, you can restore them
to any ASP, as long as their library is in the system ASP. For libraries that were
in a library user ASP, you must ensure that the ASP has adequate space before
restoring the journal receivers to the library.

How to Reset a Journal with a Status of Overflowed

| If a journaled object has a status of overflowed, you must delete it and restore it to
| reset its status.

| Note: Use the DSPOBJD command to determine whether a specific object in a

| library has a status of overflowed. Use the DSPLNK command and select
| option 8 to determine whether a specific object in a directory has a status of
| overflowed.

| Because journals and journaled objects must be in the same ASP, the best method
| for dealing with an overflowed journal is to restore it to the same ASP. If you
| restore the journal to a different ASP, you must also move all the journaled objects
| to that ASP.

This topic describes the procedure for restoring a journal to the same ASP to reset
| its overflowed status. If you want to move the journal and journaled objects to a
| different ASP, follow the procedure in “How to Transfer Journals and Objects to a
Different ASP” on page 687.

Before beginning this procedure, ensure that you have freed enough space in the
overflowed ASP to prevent the journal from overflowing when it is restored.
1. Use the WRKJRNA command to print information about journaled objects and
the receiver directory: WRKJRNA JRN(library-name/journal-name)
OUTPUT(*PRINT).
2. Use the SAVOBJ command to save the journal that must be reset.
3. Save the journal receivers that are associated with the journal by using the Save
Object (SAVOBJ) command.
4. End journaling for any objects being journaled as follows
| a. Access paths:
| ENDJRNAP JRN(library-name/journal-name) FILE(*ALL)
| b. Physical database files:
| ENDJRNPF JRN(library-name/journal-name) FILE(*ALL)
| c. IFS objects:
| ENDJRN OBJ(*ALL) JRN('QSYS.LIB/library-name.LIB/journal-name.JRN')
| d. All other object types:
| ENDJRNOBJ OBJ(*ALL) OBJTYPE(*ALL) JRN(library-name/journal-name)
| 5. Delete the journal: DLTJRN JRN(library-name/journal-name).

Chapter 25. Working with Auxiliary Storage Pools 691

 6. Restore the journal to the same library and the same ASP. If the journal was in
library user ASP, you do not need to specify the ASP parameter on the RSTOBJ
command. If the journal was in a nonlibrary user ASP, specify RSTASP(*SAVASP)
on the RSTOBJ parameter.
| 7. Start journaling again for each object that was journaled as follows:
| a. Database physical files:
| STRJRNPF FILE(library-name/file-name) JRN(library-name/journal-name)
| b. Access paths:
| STRJRNAP FILE(library-name/file-name) JRN(library-name/journal-name)
| c. IFS objects:
| STRJRN OBJ('object-path-name')
| JRN('/QSYS.LIB/library-name.LIB/journal-name.JRN')
| d. All other object types:
| STRJRNOBJ OBJ(library-name/file-name) OBJTYPE(object-type)
| JRN(library-name/journal-name)
| You printed a list of objects in step 1 on page 691.
8. Reestablish the journal receiver chain. Do the following:
a. Type WRKJRN and press the Enter key.
b. On the prompt display, type the name of the journal and the library.
c. On the Work with Journals display, select option 9 (Associate receivers with
journal).
d. When you receive a message that the receivers have been associated, press
F12 to return.
e. Type the following:
WRKJRNA JRN(library-name/journal-name)
f. On the Work with Journal Attributes display, press F15 to display the
receiver directory. If the receiver chain is not correct, do the following. This
is most likely to be required if any journal receivers were created before
V3R1.
1) Delete all the journal receivers from the system, using the DLTOBJ
command.
2) Restore the journal receivers from the newest to oldest. This associates
them with the journal and builds the receiver chain in the correct
sequence.
9. Use the DSPOBJD command to display the object description for the journal.
Ensure that the journal is no longer in overflowed status.

How to Work with Nonlibrary User ASPs

This topic describes the procedures for working with objects in a nonlibrary user
ASP. The object types that are allowed in a nonlibrary user ASP are:
v Journal
v Journal receiver
v Save file

Creating Objects in a Nonlibrary User ASP

You can create journals, journal receivers, and save files in nonlibrary user ASPs.
When you create an object in a nonlibrary user ASP, the library for the object must
exist in the system ASP. For example, to create a save file in a nonlibrary user ASP,
do the following:

692 OS/400 Backup and Recovery V5R1

 CRTSAVF FILE(SAVFLIB/DSTSAV) ASP(4)

where 4 is the number of the user ASP where you are placing the save file. The
library for the save file is in the system ASP and ASP 4 does not contain any
libraries.

After the object is created, all storage for the object resides in the designated user
ASP. Changes and additions to that object are also made in the user ASP. If the
ASP becomes full, it overflows into the system ASP. “Chapter 25. Working with
Auxiliary Storage Pools” on page 669 describes how to reset an overflowed
auxiliary storage pool.

It is recommended that all journals and journal receivers on the system have
unique names. RCLSTG renames them if duplicate names are found when objects
are placed in library QRCL and the user cannot rename them to their original
name.

Monitor the size of objects to prevent them from overflowing into the system ASP
with the MAXRCDS parameter on the CRTSAVF command, and the THRESHOLD
parameter on the CRTJRNRCV command.

Transferring an Object to a Nonlibrary User ASP

The following procedure shows an example of moving an object to a nonlibrary
user ASP. In the example, the DSTSAVF save file is moved to ASP 4. The library
(SAVFLIB) for the DSTSAVF is in the system ASP. You can use this procedure only
for save files, journals, and journal receivers.
1. Display private authorities for the save file: DSPOBJAUT OBJ(SAVFLIB/DSTSAVF)
OBJTYPE(*SAVF) OUTPUT(*PRINT)
2. Save the save file: SAVOBJ OBJ(SAVFLIB/DSTSAVF) OBJTYPE(*SAVF) DEV(TAP01)

Note: If you want to save the data in the save file, specify SAVFDTA(*YES).
3. Delete the save file: DLTSAVF SAVF(SAVFLIB/DSTSAVF)
4. Restore the save file to ASP 4: RSTOBJ OBJ(SAVFLIB/DSTSAVF) RSTASP(4)
5. Use the Edit Object Authority (EDTOBJAUT) command to reestablish the
private authorities you printed in step 1.

Transferring a Journal to a Nonlibrary User ASP

Use the following procedure to move a journal to a different nonlibrary user ASP
and to reassociate any previously journaled objects with that journal.
1. Sign on as QSECOFR.
| 2. Save the journal with either the SAV, SAVOBJ, or SAVLIB command.
| 3. List the objects being journaled: WRKJRNA JRN(library-name/journal-name)
| OUTPUT(*PRINT)
4. Copy private authorities for the journal to a temporary object.
a. Create a temporary object:
CRTDTAARA DTAARA(QTEMP/X)
TYPE(*CHAR)
AUT(*EXCLUDE)
b. Make sure that there are no authorities to the temporary object:
RVKOBJAUT OBJ(QTEMP/X) OBJTYPE(*DTAARA)
USER(QSECOFR) AUT(*ALL)
c. Copy the authorities from the journal to the temporary object:

Chapter 25. Working with Auxiliary Storage Pools 693

 GRTOBJAUT OBJ(QTEMP/X) OBJTYPE(*DTAARA)
REFOBJ(library-name/journal-name)
REFOBJTYPE(*JRN)
5. Stop journaling access paths for the journal: ENDJRNAP FILE(*ALL)
JRN(library-name/journal-name)
6. Stop journaling physical files for the journal: ENDJRNPF FILE(*ALL)
JRN(library-name/journal-name)
| 7. Stop journaling IFS objects: ENDJRN OBJ(*ALL) JRN('/QSYS.LIB/library-
| name.LIB/journal-name.JRN'/)
| 8. Stop journaling all other object types for the journal: ENDJRNOBJ OBJ(*ALL)
| OBJTYPE(*ALL) JRN(library-name/journal-name)
9. Delete the journal with the DLTJRN command.
10. Restore the journal to the user ASP: RSTOBJ OBJ(journal-name) OBJTYPE(*JRN)
RSTASP(asp-number)
| 11. Use the STRJRNPF, STRJRNAP, STRJRNOBJ, and STRJRN commands to
| resume journaling for any previously journaled objects. Refer to the list you
| printed in step 3 on page 693.
12. Reestablish private authorities to the journal.
a. Copy authorities from the temporary object to the journal.
GRTOBJ OBJ(library-name/journal-name)
OBJTYPE(*JRN) REFOBJ(QTEMP/X)
REFOBJTYPE(*DTAARA)
b. Delete the temporary object. DLTDTAARA QTEMP/X

Placing Journal Receivers in a Nonlibrary User ASP

The following procedure also uses the example of the CUSTJRN journal and
journal receivers that use the naming convention CUSTRnnnn. This example
assumes that the library for the journal receiver (CUSTJRNR) already exists in the
system ASP.
1. Use the WRKJRNA command to find the name of the currently attached journal
receiver: WRKJRNA JRN(CUSTJRNR/CUSTJRN). For the example, assume that the
currently attached journal receiver is CUSTR0005.
2. Use the Display Journal Receivers Attributes (DSPJRNRCVA) command to
determine the attributes for the current receiver: DSPJRNRCVA
JRNRCV(CUSTJRNR/CUSTR0005)
3. Create a new journal receiver in a user ASP. Use a name that continues your
naming convention. Specify the attributes that you displayed, such as the
threshold:
CRTJRNRCV JRNRCV(CUSTJRNR/CUSTR0006) ASP(4)
THRESHOLD(5000)
4. Change journal receivers so the new journal receiver is attached and actively
receiving journal entries:
CHGJRN JRN(CUSTJRN) JRNRCV(CUSTJRNR/CUSTR0006)
5. You can save the detached journal receiver and delete it from the system.
6. In the future, when you change journal receivers and specify JRNRCV(*GEN), the
system will create the new journal receiver in the CUSTJRNR library in ASP 4.

694 OS/400 Backup and Recovery V5R1

Chapter 26. Working with Device Parity Protection
This chapter describes procedures for working with device parity protection on
your system. When you are making changes to the disk configuration or disk
protection on your system, you need to perform tasks in the correct sequence.
Refer to “Choosing the Right Procedure for Configuring Disks” on page 643 to
determine the correct tasks for your situation.

Starting Device Parity Protection

Usually, you start device parity protection when you attach a new disk subsystem
to your system. Ideally, you should start device parity protection before you add
the disk units to the software disk configuration. When you start device parity
protection for disk units that are already being used, the system must move data
off of the disk units to other disk units in the ASP in order to make space available
for the parity data. The performance for starting device parity protection is much
better if the system does not have to move data.

When you start device parity protection, the system does validity checking and
moves data from the required units, if necessary. For some types of disk units, you
or your service representative must perform tasks with the disk subsystem when
you start device parity protection.

Note: If you plan to start device parity protection for disk units that are already
part of your disk configuration, check the following before you start device
parity protection.
v The configuration must be complete and no disk units can be missing in
any ASPs that contain disk units that are to have device parity protection.
This is because the system must move data off the disks that are to be
protected to make room for parity information.
v The disk units that will become device-parity protected cannot be in an
ASP that has mirrored protection active. If the disk units are in an ASP
that has mirrored protection, you must stop mirrored protection before
starting device parity protection.
v When you start device parity protection, you reduce the capacity of some
of the disk units in the subsystem. The system must have sufficient
storage in each affected ASP to make room for redundant parity data.

How to Start Device Parity Protection for a 9337 Disk Array

Subsystem

Displays May Differ

When you work with device parity protection, the displays you see may
differ slightly from the displays in this book because of differences in disk
subsystem requirements. The best way to ensure that you are running the
correct step in each procedure is to refer to the titles of the displays.

1. If you are not already using DST, perform a manual IPL to start DST. See
“How to Start Dedicated Service Tools (DST)” on page 660.

© Copyright IBM Corp. 1997, 2000, 2001 695

 2. From the Use Dedicated Service Tools (DST) menu, do the following:

┌────────────────────────┐ Select option 4

│ DST Menu │ (Work with disk units)
│ │
│ ┌───────────────────┴─────┐ Select option 1
│ 4 │ Work with Disk Units │ (Work with disk configuration)
└────┤ │
│ ┌────────────────────┴───┐ Select option 5
│ 1 │ Work with Disk │ (Work with device parity protection)
└────┤ Configuration │
│ ┌───────────────────┴──────┐
│ │Work with Device Parity │
│ 5 │ Protection │
└────┤ │
│ │
└──────────────────────────┘

3. Select option 2 (Start device parity protection) on the Work with Device Parity
Protection display and press the Enter key. You are shown the Start Device
Parity Protection display. It lists all the disk unit subsystems for which you
can start device parity protection.

Start Device Parity Protection

Select the subsystems to start device parity protection. Type choices,
press Enter.
1=Start device parity protection
Parity Serial
Resource Option Set Number Type Model Name
2 3314025 6502 001 SI01
1 1 0000318 9337 225 DD011
_ ____ _______ ____ ___ _________

4. Type a 1 in the Option column for the disk unit subsystems that you want to
start device parity protection. Press the Enter key.

Note: You can select disks that are attached to a 9337 Disk Array Subsystem
and disks that are attached to a 6502 or 6512 IOP at the same time.

The following display may be shown, if so, press Enter to continue.

Press the Enter key to continue.
Confirm Continuation
To proceed, the system must perform directory recovery, which may take a s
ignificant amount of time. The system may appear inactive during this time.
Press Enter to continue. Press F12=Cancel to return and change your
choices.

5. If you select disk unit subsystems that require you to use the operating panel
to start device parity protection or if the system detects a configuration
problem, the following display is shown:

696 OS/400 Backup and Recovery V5R1

 Warning Report
Note: Some action for the warnings listed below may need to be taken.
Please select a warning to display more detailed information about the warning
and to see what possible action may be taken to correct the warning. Type
option, press Enter.
5=Display Detailed Report OPT Warning
5 Manual intervention will be required

6. Type a 5 next to the warning to see more information. You are shown the
detail display for the specific warning.
You may choose to cancel (F12) and investigate the situation, or you may
Manual Intervention Will be Required
You selected storage subsystems that will require manual intervention to
complete the procedure. The subsystems with that requirement are listed here.
The system will notify you when preparation work is complete and the manual
process can be started. The manual process should be described in the
appropriate device documentation.
Press Enter to continue. Press F12=Cancel to return to change your choices.

Parity Serial
Resource Set Number Type Model Name
2 00-00341 9337 210 DC109

press the Enter key to return to the Warning Report display. From there, press
F10 to continue.
7. If you choose to continue, you are shown the Confirm Starting Device Parity
Protection display. The display shows all the disk unit subsystems that you
have selected and the individual disk units that are eligible to be started. Disk
units that have an asterisk (*) in the ASP and Unit columns are not yet
configured.

Confirm Starting Device Parity Protection

During the preparation for starting device parity protection, data will be
moved from parts of some disk units. This may take several minutes for each
subsystem selected. Press Enter to continue. Press F12=Cancel to return and
change your choice.
Parity Serial
Resource Option Set ASP Unit Number Type Model Name
1 1 0000318 9337 225 DD010
1 1 * * 00-1000318 9337 227 DD011
1 1 * * 00-3000318 9337 227 DD013
1 1 * * 00-5000318 9337 227 DD015
1 1 * * 00-7000318 9337 227 DD017

8. Notice: At this point, pressing the Enter key initiates the procedure for
starting device parity protection. Once begun, this procedure continues to run
until it is complete. If the subsystems you have selected are correct, press the
Enter key to continue. The following display is shown.

Start Device Parity Protection Status

The operation to start device parity protection will be done in several
phases. The phases are listed here and the status will be indicated when known.
Operation Status
Prepare to start . . . . . . . . . . . . : ______________________
Start device parity protection . . . . . : ___ %

Chapter 26. Working with Device Parity Protection 697

 Note: If you have not yet received ″Completed″ status you can press F16 to
return to the Use Dedicated Service Tools (DST) menu if you have other
tasks to perform. However, you cannot perform any disk configuration
tasks or end DST until the system has finished starting device parity
protection.
After the system has completed its preparation for starting device parity
protection on the selected subsystems, the following display is shown.
At this time, you or your service representative should consult the 9337 Disk
Ready to Start Device Parity Protection
The system has completed its preparation for starting device parity
protection on these subsystems. Consult the subsystem documentation for the
procedures to complete starting device parity protection. Configured disk
units in the subsystems may become missing during this procedure. Press
Enter to continue after those procedures are completed.
Serial Resource Parity
Option Type Model Number Name Set
1 9337 225 0000318 DD010 1
1 9337 227 00-1000318 DD011 1
1 9337 227 00-3000318 DD013 1
1 9337 227 00-5000318 DD015 1
1 9337 227 00-7000318 DD017 1

Array Model 2xx/4xx Customer Information book, SA41-0014, for the procedures
to complete starting device parity protection.
9. The system has completed its preparation for starting device parity protection
on the selected subsystems. The following screen is displayed.

Start Device Parity Protection Status

The operation to start device parity protection will be done in several
phases. The phases are listed here and the status will be indicated when known.
Operation Status
Prepare to start . . . . . . . . . . . . : Completed
Start device parity protection . . . . . : Completed

WARNING: There are still unprotected disk units remaining on this system.
When a system has unprotected, exposed, or suspended disk units attached to it,
disk related failures may affect the availability of the system and can cause
loss of data.

10. If you press the Enter key too soon and the system has not yet finished
working with the hardware, you see the Preparation for Starting Device Parity
has not Completed display. This may happen for one of the following reasons:
v The 9337 was not brought online after the start device parity protection
procedure. This must be done at the operator panel of the disk unit.
v The Licensed Internal Code is still processing the newly protected disk
units.

698 OS/400 Backup and Recovery V5R1

 Preparation for Starting Device Parity has not Completed
Your request to start device parity protection required changes to the
hardware. Those changes have not been completed. You may continue without
making those changes or return to the previous screen to wait for them to be
done. Press F10=Confirm to bypass the hardware changes. Press F12=Cancel to
return to wait for the hardware changes to be made.

WARNING: If you select to bypass the hardware changes, you may not be able to
perform some operations until you IPL the system. You will be notified that
there are 'missing' units. The units are not really missing, but have been
placed into a state that prevents the system from communicating with them. An
IPL will reset the units back to their normal state.

Press F12 to return to the Ready to Start Device Parity Protection display. Wait
a few minutes and then press the Enter key again.

How to Start Device Parity Protection for an Input/Output

Processor
This topic lists the rules and describes the procedure for starting protection for
IOPs that support device parity protection. The following are the basic rules for
these types of IOPs:
v Maximum number of parity sets allowed: 2
v Maximum number of devices per parity set: 10
v Minimum number of devices per parity set: 4
v All devices in a parity set must be the same capacity
v The protection supports 66xx devices, not 61xx devices in parity sets or to use
the write cache support.
v You must remove 66xx model 30 devices from the ASP configuration before you
start device parity protection. This allows the system to reformat the device for
the advanced functions of the IOP. After you start device parity protection, add
the device to the ASP configuration.

The IOP starts the fewest number of parity sets needed to protect all the devices of
the same capacity. For example, to protect 10 devices, it starts one parity set of ten
devices. To protect 11 devices, it starts two parity sets: one parity set of seven
devices and one parity set of four devices.

Displays May Differ

When you work with device parity protection, the displays you see may
differ slightly from the displays in this book because of differences in disk
subsystem requirements. The best way to ensure that you are running the
correct step in each procedure is to refer to the titles of the displays.

1. From the Use Dedicated Service Tools (DST) menu, do the following:

Chapter 26. Working with Device Parity Protection 699

 ┌────────────────────────┐ Select option 4
│ DST Menu │ (Work with disk units)
│ │
│ ┌───────────────────┴─────┐ Select option 1
│ 4 │ Work with Disk Units │ (Work with disk configuration)
└────┤ │
│ ┌────────────────────┴───┐ Select option 5
│ 1 │ Work with Disk │ (Work with device parity protection)
└────┤ Configuration │
│ ┌───────────────────┴──────┐
│ │Work with Device Parity │
│ 5 │ Protection │
└────┤ │
│ │
└──────────────────────────┘

2. Select option 2 (Start device parity protection) on the Work with Device Parity
Protection display and press the Enter key. You are shown the Start Device
Parity Protection display. It lists all the disk unit subsystems for which you can
start device parity protection.

Start Device Parity Protection

Select the subsystems to start device parity protection.
Type choices, press Enter.
1=Start device parity protection
Parity Serial
Resource Option Set Number Type Model Name
1 2 3314025 6502 001 SI01
1 0000318 9337 225 DD013
_ ______ __________ _____ _____ ________

3. Type a 1 in the Option column for the disk unit subsystems that you want to
prepare to start device parity protection. Press the Enter key.
If you are shown the following display, press Enter to continue.

Confirm Continuation
In order to proceed the system must perform internal processing that may
take several minutes during which the system may appear inactive. Once you
confirm to continue, the system must perform an IPL when you leave Work with
Disk Configuration functions.

Press Enter to continue.

Press F12=Cancel to return to change your choice.

4. Press the Enter key to continue. You are shown the Confirm Starting Device
Parity Protection display. The display shows all the disk unit subsystems that
you have selected and the individual disk units that are eligible to be started.
Disk units that have an asterisk (*) in the ASP and Unit columns are not yet
configured.

700 OS/400 Backup and Recovery V5R1

 Confirm Starting Device Parity Protection During the
preparation for starting device parity protection, data will be moved from
parts of some disk units. This may take several minutes for each subsystem
selected. Press Enter to continue. Press F12=Cancel to return and change
your choice.
Parity Serial
Resource Option Set ASP Unit Number Type Model Name
1 2 3314025 6502 001 SI01
1 2 * * 00-0051556 6603 074 DD056
1 2 * * 00-0020525 6603 074 DD025
1 2 * * 00-0024030 6603 074 DD030
1 2 * * 00-0013026 6603 074 DD026
1 2 * * 00-0024519 6603 074 DD019
1 2 * * 00-0046338 6603 074 DD038

5. Notice: At this point, pressing the Enter key initiates the procedure for starting
device parity protection. Once begun, this procedure continues to run until it is
complete. If the subsystems that you have selected are correct, press the Enter
key to continue. The status display shows how the operation is proceeding.
When the system has completed its preparation for starting device parity
protection on the selected subsystems, the following display is shown.

Start Device Parity Protection Status

The operation to start device parity protection will be done in several
phases. The phases are listed here and the status will be indicated when
known.
Operation Status
Initialize disk units . . . . . . . . . : Running
Prepare to start . . . . . . . . . . . . : ______________________
Start device parity protection . . . . . : ___ %

6. Press the Enter key to return to the Work with Device Parity Protection menu.

Stopping Device Parity Protection

When preparing to stop device parity protection, the system does validity checking
to make sure that stopping device parity protection does not leave the system in a
configuration that is not supported. The following restrictions apply when you
stop device parity protection:
v You cannot stop device parity protectionon a subsystem when a unit in that
subsystem is in a mirrored ASP. In order to stop device parity protection,
mirrored protection must be stopped first.
v You cannot stop device parity protection on a 9337 Disk Array Subsystem that
contains data when one of the units in the subsystem has failed. Stopping device
parity protection would destroy the redundant data required to reconstruct data
on the failed unit.

How to Stop Device Parity Protection for a 9337 Disk Array

Subsystem

Displays May Differ

When you work with device parity protection, the displays you see may
differ slightly from the displays in this book because of differences in disk
subsystem requirements. The best way to ensure that you are running the
correct step in each procedure is to refer to the titles of the displays.

Chapter 26. Working with Device Parity Protection 701

 1. If you are not already using DST, perform a manual IPL to start DST. See “How
to Start Dedicated Service Tools (DST)” on page 660.
2. From the Use Dedicated Service Tools (DST) menu, do the following:

┌────────────────────────┐ Select option 4

│ DST Menu │ (Work with disk units)
│ │
│ ┌───────────────────┴─────┐ Select option 1
│ 4 │ Work with Disk Units │ (Work with disk configuration)
└────┤ │
│ ┌────────────────────┴───┐ Select option 5
│ 1 │ Work with Disk │ (Work with device parity protection)
└────┤ Configuration │
│ ┌───────────────────┴──────┐
│ │Work with Device Parity │
│ 5 │ Protection │
└────┤ │
│ │
└──────────────────────────┘

3. Select option 3 (Stop device parity protection) on the Work with Device Parity
Protection display and press the Enter key. The following display is shown.

Stop Device Parity Protection

Select the subsystems to stop device parity protection.
Type choices, press Enter.
1=Stop device parity protection
Parity Serial
Resource Option Set Number Type Model Name
__ 1 3314025 6502 001 SI01
1 2 0000318 9337 227 DD018
_ __ _________ ____ ___ ________

4. Type a 1 in the Option column for the disk unit subsystems that you want to
stop device parity protection. Press the Enter key.
5. If you select disk unit subsystems that require user action to stop device parity,
the following display is shown.
Type a 5 in the Option column next to the error to see more information. The
Warning Report
Note: Some action for the warnings listed below may need to be taken.
Please select a warning to display more detailed information about the
warning and to see what possible action may be taken to correct the warning.
Type option, press Enter.
5=Display Detailed Report OPT Warning
5 Manual intervention will be required

following display is shown:

702 OS/400 Backup and Recovery V5R1

 Manual Intervention Will be Required
You selected storage subsystems that will require manual intervention to
complete the procedure. The subsystems with that requirement are listed here.
The system will notify you when preparation work is complete and the manual
process can be started. The manual process should be described in the
appropriate device documentation.
Press Enter to continue.
Press F12=Cancel to return to change your choices.
Parity Serial
Resource Set Number Type Model Name
2 00-00341 9937 212 DC09

Press the Enter key to return to the Warning Report display.

6. Press the F10 key to continue. The following display is shown. Press the Enter
key to return to the Warning Report display. From the Warning Report display
press the F10 key to continue to the Confirm Stop Device Parity Protection
display.

Confirm Stop Device Parity Protection

Warning: Disk units connected to these subsystems will not be protected
after you confirm your choices.
Press Enter to continue.
Press F12=Cancel to return and change your choices.
Parity Serial
Resource Option Set ASP Unit Number Type Model Name
1 1 0000318 9337 227 DD010
1 1 2 5 00-1000318 9337 226 DD011
1 1 2 6 00-3000318 9337 226 DD013
1 1 2 7 00-5000318 9337 226 DD015
1 1 3 8 00-7000318 9337 226 DD017

7. Press the Enter key to continue. The system has completed its preparation for
stopping device parity protection on the selected subsystems. At this time, you
should consult the 9337 Disk Array Model 2xx/4xx Customer Information book,
SA41-0014, for the procedures to complete stopping device parity protection.

Ready to Stop Device Parity Protection

The system has completed its preparation for stopping device parity
protection on these subsystems. Consult the subsystem documentation for the
procedures to complete stopping device parity protection. Press Enter to
continue after those procedures are completed.
Parity Serial
Resource Option Set Number Type Model Name
1 2 00-00341 9337 212 DC09

8. Press the Enter key to continue. The following display is shown.

Chapter 26. Working with Device Parity Protection 703

 Ready to Stop Device Parity Protection
The system has completed its preparation for stopping device parity
protection on these subsystems. Consult the subsystem documentation for the
procedures to complete stopping device parity protection. Press Enter to
continue after those procedures are completed.
Serial
Resource Parity Option Type Model Number Name
Set
1 1 9337 227 0000318 DD010
1 1 9337 226 00-1000318 DD011
1 1 9337 226 00-3000318 DD013
1 1 9337 226 00-5000318 DD015
1 1 9337 226 00-7000318 DD017

How to Stop Device Parity Protection on an Input/Output

Processor
The following instructions only apply to IOPs that support device parity
protection.
1. If you are not already using DST, perform a manual IPL to start DST. See “How
to Start Dedicated Service Tools (DST)” on page 660.
2. From the Use Dedicated Service Tools (DST) menu, do the following:

┌────────────────────────┐ Select option 4

│ DST Menu │ (Work with disk units)
│ │
│ ┌───────────────────┴─────┐ Select option 1
│ 4 │ Work with Disk Units │ (Work with disk configuration)
└────┤ │
│ ┌────────────────────┴───┐ Select option 5
│ 1 │ Work with Disk │ (Work with device parity protection)
└────┤ Configuration │
│ ┌───────────────────┴──────┐
│ │Work with Device Parity │
│ 5 │ Protection │
└────┤ │
│ │
└──────────────────────────┘

3. Select option 3 (Stop device parity protection) on the Work with Device Parity
Protection display and press the Enter key. The following display is shown.

Stop Device Parity Protection

Select the subsystems to stop device parity protection.
Type choices, press Enter.
1=Stop device parity protection
Parity Serial
Resource Option Set Number Type Model Name
1 1 10-4453031 6502 001 SI01
2 00-00341 9337 212 DC09

4. Type a 1 in the Option column for the disk unit subsystems that you want to
stop device parity protection. Press the Enter key. The following display is
shown.

704 OS/400 Backup and Recovery V5R1

 Confirm Stop Device Parity Protection
Warning: Disk units connected to these subsystems will not be protected
after you confirm your choices.
Press Enter to continue.
Press F12=Cancel to return and change your choices.
Parity Serial
Resource Option Set ASP Unit Number Type Model Name
1 1 10-4453031 6502 001 SI01
1 1 8 3 00-0334602 6602 050 DD012
1 1 8 4 00-0334673 6602 050 DD011
1 1 8 5 00-0231915 6602 050 DD010
1 1 8 6 00-0334610 6602 050 DD009
1 1 8 7 00-0274937 6602 050 DD008

5. Notice: At this point, pressing the Enter key initiates the procedure for
stopping device parity protection. Once this procedure has begun, you may not
cancel it. If the subsystems you have selected are correct, press the Enter key to
continue. You will get status screens.

Stop Device Parity Protection Status

The operation to stop device parity protection will be done in several
phases. The phases are listed here and the status will be indicated when
known.
Operation Status
Prepare to stop . . . . . . . . . . . . : Completed
Stop device parity protection . . . . . : Completed
WARNING: There are now unprotected disk units on this system. When a system
has unprotected, exposed, or suspended disk units attached to it, disk related
failures may affect the availability of the system and can cause loss of data.

Note: If you have not yet received ″Completed″ status you can press F16 to
return to the Use Dedicated Service Tools (DST) menu if you have other
tasks to perform. However, you cannot perform any disk configuration
tasks or end DST until the system has finished starting device parity
protection.
6. When the status shows Completed, press the Enter key to return to the Work
with Device Parity Protection menu.

How to Include a Disk Unit in Device Parity Protection

When you attach a new disk unit to an existing IOP that has device parity
protection, you can include the disk unit in the device parity set. You can include a
disk unit by using either DST or SST.

This topic lists the rules and describes the procedure for starting device parity
protection for an IOP. The following are the basic rules for this type of IOP:
v Maximum number of parity sets that are allowed: 2 per IOP
v Maximum number of devices per parity set: 10
v Minimum number of devices per parity set: 4
v All devices in a parity set must be the same capacity
v The protection supports 66xx devices, not 61xx devices in parity sets or to use
the write cache support
v You must remove 66xx model 30 devices from the configuration before you start
device parity protection. This allows the system to reformat the device for the
advanced functions of the IOP. After you start device parity protection, add the
device to the ASP configuration.

Chapter 26. Working with Device Parity Protection 705

 The devices will be included into an existing array. However, if enough devices
exist to create a new array, the devices will be eligible for start but not include.

Note: You cannot include a disk unit if that disk unit has already been added to
an ASP that has mirrored protection. You must stop mirrored protection
before including the disk unit. Stopping mirrored protection must be done
from the DST menu. Adding mixed protection on the same IOP requires
mirroring to be stopped and restarted.
To include disk units in a device parity set, perform the following steps:
1. From the System Service Tools (SST) menu, do the following:
or from the Use Dedicated Service Tools (DST) menu, do the following:
┌────────────────────────┐ Select option 3
│ SST Menu │ (Work with disk units)
│ │
│ ┌───────────────────┴─────┐ Select option 2
│ 3 │ Work with Disk Units │ (Work with disk configuration)
└────┤ │
│ ┌────────────────────┴─────┐ Select option 4
│ │ Work with Disk │ (Include unit in device
│ 2 │ Configuration │ parity protection)
└────┤ │
│ 4 │
└──────────────────────────┘

Note: If you are not already using DST, see “How to Start Dedicated Service
Tools (DST)” on page 660.
The Include Disk Units in Device Parity Protection display appears:
┌────────────────────────┐ Select option 4
│ DST Menu │ (Work with disk units)
│ │
│ ┌───────────────────┴─────┐ Select option 1
│ 4 │ Work with Disk Units │ (Work with disk configuration)
└────┤ │
│ ┌────────────────────┴─────┐ Select option 5
│ 1 │ Work with Disk │ (Work with device
└────┤ Configuration │ parity protection)
│ ┌─────────────────────┴────┐
│ 5 │ Work with Device │ Select option 4
└────┤ Parity Protection │ (Include unit in
│ │ device parity
│ 4 │ configuration)
└──────────────────────────┘

Include Disk Units in Device Parity Protection

Select the units to be included in Device Parity Protection.
Type choices, press Enter.
1=Include unit in device parity protection
Parity Serial
Resource Option Set ASP Unit Number Type Model Name
1 1 8 7 00-0274937 6602 050 DD008

2. Type a 1 in the Option column for the disk units that you want to include in
device parity protection and press the Enter key. The following display is
shown.

706 OS/400 Backup and Recovery V5R1

 Confirm Disk Units to be Included
Including disk units in device parity protection may take a considerable
amount of time. Press Enter to confirm your choice to have the system include
the selected units in device parity protection
Press F12=Cancel to return to change your choice
Parity Serial
Resource Option Set ASP Unit Number Type Model Name
1 1 8 7 00-0274937 6602 070 DD008

3. If the disk units that you selected are to be included in device parity protection,
confirm this by pressing the Enter key. After the include operation has
completed, the following display is shown.

Include Disk Units in Device Parity Protection Status

The operation to include units in the device parity protection will be done
in several phases. The phases are listed here and the status will be indicated
when known.
Operation Status
Prepare to include units . . . . . . . . : Completed
Include units . . . . . . . . . . . . . : Completed

Note: You can press F16 to return to the Use Dedicated Service Tools (DST)
menu if you have other tasks to perform. However, you cannot perform
any disk configuration tasks or end DST until the system has finished
including disk units in device parity protection.
4. Press the Enter key to return to the Work with Device Parity Protection menu.

How to Exclude a Disk Unit from Device Parity Protection

You can exclude a disk unit that is attached to an IOP from device parity
protection. You cannot exclude a disk unit that has been assigned to an ASP that
has mirrored protection. To exclude a disk unit from device parity protection, do
the following.
1. If you are not already using DST, perform a manual IPL to start DST. See “How
to Start Dedicated Service Tools (DST)” on page 660.
2. From the Use Dedicated Service Tools (DST) menu, do the following:

┌────────────────────────┐ Select option 4

│ DST Menu │ (Work with disk units)
│ │
│ ┌───────────────────┴─────┐ Select option 1
│ 4 │ Work with Disk Units │ (Work with disk configuration)
└────┤ │
│ ┌────────────────────┴───┐ Select option 5
│ 1 │ Work with Disk │ (Work with device parity protection)
└────┤ Configuration │
│ ┌───────────────────┴──────┐
│ │Work with Device Parity │
│ 5 │ Protection │
└────┤ │
│ │
└──────────────────────────┘

3. Select option 5 (Exclude unit from device parity protection) on the Work with
Device Parity Protection display and press the Enter key. The following display
is shown.

Chapter 26. Working with Device Parity Protection 707

 Exclude Disk Units from Device Parity Protection
Select the units to be excluded from Device Parity Protection.
Type choices, press Enter.
1=Exclude unit from device parity protection
Parity Serial
Resource Option Set ASP Unit Number Type Model Name
1 1 8 7 00-0274937 6602 070 DD008

This display shows only disk units that are eligible to be excluded. A disk unit
is eligible to be excluded if it does not contain parity information. If the disk
units that you want to remove are not eligible to be excluded, you must stop
device parity protection instead. Then physically remove the disk units and
restart device parity protection.
4. Type a 1 in the Option column for the disk units that you want to exclude from
device parity protection and press the Enter key. The following display is
shown.

Confirm Disk Units to be Excluded

Press Enter to confirm your choice to have the system exclude the selected
units from device parity protection
Press F12=Cancel to return to change your choice
Parity Serial
Resource Option Set ASP Unit Number Type Model Name
1 1 8 7 00-0274937 6602 050 DD008

5. If the disk units you selected are to be excluded from device parity protection,
confirm this by pressing the Enter key. After the exclude operation has
completed, the following display is shown.

Exclude Disk Units in Device Parity Protection Status

The operation to exclude units from the device parity protection will be done
in several phases. The phases are listed here and the status will be indicated
when known.
Operation Status
Prepare to exclude units . . . . . . . . : Completed
Exclude units . . . . . . . . . . . . . : Completed

Note: You can press F16 to return to the Use Dedicated Service Tools (DST)
menu if you have other tasks to perform. However, you cannot perform
any disk configuration tasks or end DST until the system has finished
excluding disk units in device parity protection.
6. Press the Enter key to return to the Work with Device Parity Protection menu.

How to Display Device Parity Status

To display device parity status, do the following:
1. From the System Service Tools (SST) menu, do the following:

708 OS/400 Backup and Recovery V5R1

 ┌────────────────────────┐ Select option 3
│ SST Menu │ (Work with disk units)
│ │
│ ┌───────────────────┴────┐ Select option 1
│ 3 │ Work with Disk Units │ (Display disk configuration)
└────┤ │
│ 1 ┌───────────────────┴──────┐
│ │Display Disk Configuration│
└────┤ │
│ │
│ │
└──────────────────────────┘

or from the Use Dedicated Service Tools (DST) menu, do the following:

┌────────────────────────┐ Select option 4

│ DST Menu │ (Work with disk units)
│ │
│ ┌───────────────────┴────┐ Select option 1
│ 4 │ Work with Disk Units │ (Work with disk configuration)
└────┤ │
│ ┌───────────────────┴────┐ Select option 1
│ 1 │Work with Disk │ (Display disk configuration)
└────┤ Configuration │
│ ┌───────────────────┴──────┐
│ 1 │Display Disk Configuration│ Select option 1
└────┤ │ (Display disk configuration status
│ │
│ 1 │
└──────────────────────────┘

The Display Disk Configuration menu appears.

2. Select option 5 (Display device parity status) on the Display Disk Configuration
display and press the Enter key. The Display Device Parity Status display
appears:

Display Device Parity Status

Parity Serial
Resource Set ASP Unit Number Type Model Name Status
1 10-4251006 6502 001 SI01
* * 00-0128330 6602 074 DD055 Active
* * 00-0238703 6602 070 DD053 Active
* * 00-0162516 6602 074 DD052 Active
* * 00-0186325 6602 074 DD051 Active
* * 00-0103706 6602 074 DD050 Active
2 00-00341 9337 212 DC12
* * 00-7000341 9337 212 DD015 Active
* * 00-5000341 9337 212 DD034 Active
* * 00-3000341 9337 212 DD033 Active
* * 00-2000341 9337 213 DD012 Active
* * 00-1000341 9337 212 DD011 Active

* - See help for more information

Press Enter to continue. F3=Exit F5=Refresh
F11=Display disk hardware status F12=Cancel

The display is organized by device parity set. It includes controllers that can
support device parity protection and all of the disk units that have the
hardware capability for device parity protection. The possible values for the
Status column are the following:
Active This unit is part of a disk unit subsystem that has device parity
protection. This unit is fully operational.

Chapter 26. Working with Device Parity Protection 709

 Failed This unit is part of a disk unit subsystem that has device parity
protection. This unit has failed. If another unit in the disk unit
subsystem fails, data could be lost.
% Rebuilt
This unit is part of a disk unit subsystem that has device parity
protection. The data on this unit is being rebuilt from other units in the
disk unit subsystem.
Unprotected
This unit is part of a disk unit subsystem that has device parity
protection. This unit is operational. However, another unit in the disk
unit subsystem has failed or is being rebuilt. If another unit in the disk
unit subsystem fails, data could be lost.
% Resyncing
The parity for the parity set is being built from the data within the disk
unit subsystem.

How to Enable Disk Units Attached to the MFIOP to Use Device Parity
Protection
As illustrated in the preceding sections, some multi-function input/output
processors (MFIOPs) can support device parity protection. However, disk units that
were migrated from other RISC-based systems might not be in the correct format
to allow device parity protection to be started.

This section describes a conversion procedure for the disk units attached to the
MFIOP so that device parity protection can be started. Ensure that the disk units
and the MFIOP meet all of the following conditions before starting this procedure:
v The disk units that are currently attached to the MFIOP have mirrored
protection
v All disk units with mirrored protection have a state of ’Active’
v The MFIOP on the system supports device parity protection
v All of the disk units that are attached to the MFIOP are the same capacity

Mirrored protection cannot be running on a disk unit that is using device parity
protection. In order to use the MFIOP capability to support device parity
protection, you will have to stop mirrored protection on the loadsource disk unit.
You should be aware that if you stop mirrored protection on the loadsource disk
unit and replace mirrored protection with device parity protection, you could be
reducing the system availability.
Notes:
1. With both device parity protection and mirrored protection, the system
continues to run after a single disk failure. With mirrored protection, the
system may continue to run after the failure of a disk-related component, such
as a controller or an IOP.
2. When a second disk failure occurs such that the system has two failed disks,
the system is more likely to continue to run with mirrored protection than with
device parity protection.
1. If you are not already using DST, end all active jobs and power down the
system. Perform a manual IPL to start DST. See “How to Start Dedicated
Service Tools (DST)” on page 660, for information on starting DST.

710 OS/400 Backup and Recovery V5R1

2. From the Use Dedicated Service Tools (DST) menu, do the following:
The disk units and their status are displayed.
┌────────────────────────┐ Select option 4
│ DST Menu │ (Work with disk units)
│ │
│ ┌───────────────────┴────┐ Select option 1
│ 4 │ Work with Disk Units │ (Work with disk configuration)
└────┤ │
│ ┌───────────────────┴────┐ Select option 1
│ 1 │Work with Disk │ (Display disk configuration)
└────┤ Configuration │
│ ┌───────────────────┴──────┐
│ 1 │Display Disk Configuration│ Select option 1
└────┤ │ (Display disk configuration status
│ │
│ 1 │
└──────────────────────────┘

Display Disk Configuration Status

Serial Resource
ASP Unit Number Type Model Name Status
1 Mirrored
1 68-0C47591 6602 030 DD001 Active
1 68-0119804 6602 030 DD002 Active
2 68-0C60040 6602 030 DD003 Active
2 68-54531 6602 050 DD004 Active
3 68-0C99140 6602 030 DD012 Active
3 68-5544453 6602 050 DD011 Active
5 10-1000128 9337 221 DD005 Active
5 10-2000128 9337 221 DD006 Active
7 10-3000128 9337 221 DD007 Active
7 10-5000128 9337 221 DD008 Active

3. Press F9 to view the Display Disk Unit Details screen.

Display Disk Unit Details

Type option, press Enter.

5=Display hardware resource information details

Serial Sys Sys I/O I/O

OPT ASP Unit Number Bus Card Adapter Bus Ctl Dev
1 1 68-0C47591 1 1 0 0 1 0
1 1 68-0119804 1 1 0 0 5 0
1 2 68-0C60040 1 1 0 0 2 0
1 2 68-54531 3 1 0 0 2 0
1 3 68-0C99140 1 1 0 0 3 0
1 3 68-5544453 3 1 0 0 2 0
1 5 10-1000128 2 5 0 1 0
1 5 10-2000128 2 5 0 1 1
1 7 10-3000128 2 5 0 1 2
1 7 10-5000128 2 5 0 1 4

4. On the Display Disk Unit Details screen, locate the disk units which are on
System Bus 1 and System Card 1. Those are the units that are attached to the
Multi-function Input/Output Processor (MFIOP). Write down the unit
numbers and serial numbers of those disk units. You will need that
information in later steps. In the example above, the disk units with serial
numbers 68–0C47591, 68–0119804, 68–0C60040, and 68–0C99140 are attached to
the MFIOP.

Chapter 26. Working with Device Parity Protection 711

 5. If none of the disk units that are attached to the MFIOP have a model of 030,
you can exit this procedure at this time.
6. Check to see if the following conditions exist:
v There are only two disk units that are attached to the MFIOP
v Both of those disk units are the load source disk units (unit 1)
If these conditions exist, then you cannot use this procedure to convert the
mirror-protected disk units to device parity protection.
7. Ensure that all of the disk units that are attached to the MFIOP have identical
capacities. If they are not identical, exit this procedure.
8. The non-load source disk units which are attached to the MFIOP must be
removed from the disk configuration. The load source disk unit will be
processed in later steps.
The Remove Units from Configuration screen is displayed.
┌────────────────────────┐ Select option 4
│ DST Menu │ (Work with disk units)
│ │
│ ┌───────────────────┴─────┐ Select option 1
│ 4 │ Work with Disk Units │ (Work with disk configuration
└────┤ │
│ ┌────────────────────┴─────┐ Select option 3
│ 1 │ Work with Disk │ (Work with ASP configuration)
└────┤ Configuration │
│ ┌─────────────────────┴─────┐ Select option 7
│ 3 │Work with ASP Configuration│ (Remove units from configuration)
└────┤ │
│ ┌──────────────────────┴─────┐
│ 7 │ Remove Units from │
└────┤ Configuration │
│ │
│ │
└────────────────────────────┘

Remove Units from Configuration

Type options, press Enter

4=Remove unit from configuration

Serial Resource
OPT Unit ASP Number Type Model Name Status
4 2 1 68-0C60040 6602 030 DD003 Active
4 2 1 68-54531 6602 050 DD004 Active
4 3 1 68-0C99140 6602 030 DD012 Active
4 3 1 68-5544453 6602 050 DD011 Active
5 1 10-1000128 9337 221 DD005 Active
5 1 10-2000128 9337 221 DD006 Active
7 1 10-3000128 9337 221 DD007 Active
7 1 10-5000128 9337 221 DD008 Active

9. Type a 4 (Remove unit from configuration) in the OPT column for each unit
on the MFIOP that you want to remove and press the Enter key. In an earlier
step, you wrote down the serial numbers and units of the disk units attached
to the MFIOP. If the disk units that are attached to the MFIOP have mirrored
protection, select both units of the mirrored pair. In the example above, the
disk units with serial numbers 68–0C60040 and 68–0C99140 are attached to the
MFIOP. Those correspond to units 2 and 3, so unit 2 and unit 3 have to be
removed from the configuration. The mirrored pairs that contain those units
were selected.
10. The Confirm Continuation display may be shown before the Confirm Remove
Disk Units display if the storage management directories are not usable.

712 OS/400 Backup and Recovery V5R1

 Confirm Continuation

To proceed, the system must perform internal processing

that may take several minutes during which the system may
appear inactive.

Press Enter to continue.

Press F12=Cancel to return and change your choice.

11. Press the Enter key. The Confirm Remove Disk Units display is shown:

Confirm Remove Disk Units

Removing disk units will take several minutes.

Press Enter to confirm remove of disk units.

Press F9=Capacity information to display the capacity information
Press F12=Cancel to return to change your choice

Serial Resource
OPT Unit ASP Number Type Model Name Status
4 2 1 68-0C60040 6602 030 DD003 Active
4 2 1 68-54531 6602 050 DD004 Active
4 3 1 68-0C99140 6602 030 DD012 Active
4 3 1 68-5544453 6602 050 DD011 Active

12. Press the Enter key on the Confirm Remove Disk Units display to remove the
selected units. The system moves the data off the units that are selected to be
removed to the remaining units in the source ASP.
Notes:
a. The time it takes to remove a unit depends on the disk unit type and
model.
b. If the data on the unit being removed is severely fragmented and the
amount of storage used is high, the remove operation could take several
hours.
When the remove operation is complete, you are returned to the Work with
ASP Configuration display.
13. Exit the Work with Disk Units function and return to the Use Dedicated
Service Tools menu.
14. Power off the system.
15. Put the keylock in Normal mode.
16. Power on the system.
17. The system starts the IPL, and eventually the Sign On display appears. You
will see the message Enter your user ID and password.
18. When the IPL is complete, start the System Service Tools (SST). See “Starting
System Service Tools (SST)” on page 662 for more information.
19. The following steps will change the mirror-protected loadsource disk units
from a model 030 so that device parity protection can be enabled on the disk
units. The MFIOP cannot start device parity protection until all the disk units
that are attached to the MFIOP have been correctly formatted.
20. From the System Service Tools (SST) menu, do the following:

Chapter 26. Working with Device Parity Protection 713

 ┌────────────────────────┐ Select option 3
│ SST Menu │ (Work with disk units)
│ │
│ ┌───────────────────┴─────┐ Select option 3
│ 3 │ Work with Disk Units │ (Work with disk unit recovery)
└────┤ │
│ ┌────────────────────┴─────┐ Select option 3
│ 3 │ Work with Disk Unit │ (Suspend mirrored protection)
└────┤ Recovery │
│ │
│ │
└──────────────────────────┘

The Suspend Mirrored Protection display is shown.

Suspend Mirrored Protection

Type options, press Enter

1=Suspend Mirrored Protection

Serial Resource
OPT Unit ASP Number Type Model Name Status
_ 1 1 68-0C47591 6602 030 DD001 Active
_ 1 1 68-0119804 6602 030 DD002 Active
_ 5 1 10-1000128 9337 221 DD005 Active
_ 5 1 10-2000128 9337 221 DD006 Active
_ 7 1 10-3000128 9337 221 DD007 Active
_ 7 1 10-5000128 9337 221 DD008 Active

21. Type a 1 (Suspend Mirrored Protection) in the Option column. Select a

loadsource disk unit.
22. Replace the suspended loadsource unit. The replace function initializes the
disk units to the correct format so that device parity protection can be started
on that disk unit. The model of the disk unit will not be 030 after the replace
completes.
The Select Configured Unit to Replace display is shown.
┌────────────────────────┐ Select option 3
│ SST Menu │ (Work with disk units)
│ │
│ ┌───────────────────┴─────┐ Select option 3
│ 3 │ Work with Disk Units │ (Work with disk unit recovery)
└────┤ │
│ ┌────────────────────┴─────┐ Select option 1
│ 3 │ Work with Disk Unit │ (Replace configured unit)
└────┤ Recovery │
│ │
│ 1 │
└──────────────────────────┘

Select Configured Unit to Replace

Type option, press Enter.

1=Select

Serial Resource
OPT Unit ASP Number Type Model Name Status
1 1 1 68-0119804 6602 030 DD002 Suspended

714 OS/400 Backup and Recovery V5R1

23. Type a 1 in the Option column on the Select Configured Unit to Replace
display and press the Enter key. The only unit which should be displayed is
the loadsource disk unit that was just suspended.

Select Replacement Unit

Serial Resource
Unit ASP Number Type Model Name Status
1 1 68-0119804 6602 030 DD002 Suspended
Type option, press Enter
1=Select

Serial Resource
Option Number Type Model Name Status
1 68-0C60040 6602 030 DD003 Non-configured
_ 68-54531 6602 050 DD004 Non-configured
_ 68-0C99140 6602 030 DD012 Non-configured
_ 68-5544453 6602 050 DD011 Non-configured

24. Type a 1 in the Option column on the Select Replacement Unit display and
press the Enter key. Select a non-configured disk unit that is attached to the
MFIOP. You recorded the serial numbers of the disk units that are attached to
the MFIOP in an earlier step.

Confirm Replace of Configured Unit

This screen allows the confirmation of the configured unit to

be replaced with the selected replacement unit.

Press Enter to confirm your choices for Replace.

Press F12 to return to change your choices.

The configured unit being replaced is:

Serial Resource
Unit ASP Number Type Model Name Status
1 1 68-0119804 6602 030 DD002 Suspended

The replacement unit will be:

Serial Resource
Unit ASP Number Type Model Name Status
1 1 68-0C60040 6602 050 DD003 Resuming

25. Press Enter to confirm replacement.

26. The replacement function runs for several minutes. Wait until the replacement
function completes.
27. From the System Service Tools (SST) menu, do the following:

Chapter 26. Working with Device Parity Protection 715

 ┌────────────────────────┐ Select option 3
│ SST Menu │ (Work with disk units)
│ │
│ ┌───────────────────┴─────┐ Select option 3
│ 3 │ Work with Disk Units │ (Work with disk unit recovery)
└────┤ │
│ ┌────────────────────┴─────┐ Select option 3
│ 3 │ Work with Disk Unit │ (Suspend mirrored protection)
└────┤ Recovery │
│ │
│ │
└──────────────────────────┘

The Suspend Mirrored Protection display is shown.

Suspend Mirrored Protection

Type option, press Enter.

1=Suspend Mirrored Protection

Serial Resource
OPT Unit ASP Number Type Model Name Status
1 1 1 68-0C47591 6602 030 DD001 Active
_ 1 1 68-0C60040 6602 050 DD003 Active
_ 5 1 10-1000128 9337 221 DD005 Active
_ 5 1 10-2000128 9337 221 DD006 Active
_ 7 1 10-3000128 9337 221 DD007 Active
_ 7 1 10-5000128 9337 221 DD008 Active

28. Type a 1 (Suspend Mirrored Protection) in the Option column.

29. Replace the suspended loadsource unit. The replace function initializes the
disk units to the correct format so that device parity protection can be utilized
on that disk unit. The model of the disk unit will not be 030 when the replace
function completes.
The Select Configured Unit to Replace display is shown.
┌────────────────────────┐ Select option 3
│ SST Menu │ (Work with disk units)
│ │
│ ┌───────────────────┴─────┐ Select option 3
│ 3 │ Work with Disk Units │ (Work with disk unit recovery)
└────┤ │
│ ┌────────────────────┴─────┐ Select option 1
│ 3 │ Work with Disk Unit │ (Replace configured unit)
└────┤ Recovery │
│ │
│ 1 │
└──────────────────────────┘

Select Configured Unit to Replace

Type option, press Enter.

1=Select

Serial Resource
OPT Unit ASP Number Type Model Name Status
1 1 1 68-0C47591 6602 030 DD001 Suspended

716 OS/400 Backup and Recovery V5R1

30. Type a 1 in the Option column on the Select Configured Unit to Replace
display and press the Enter key. The only unit that should be displayed is the
loadsource disk unit that was just suspended.

Select Replacement Unit

Serial Resource
Unit ASP Number Type Model Name Status
1 1 68-0C47591 6602 030 DD001 Suspended
Type option, press Enter
1=Select

Serial Resource
Option Number Type Model Name Status
1 68-0119804 6602 030 DD002 Non-configured
_ 68-54531 6602 050 DD004 Non-configured
_ 68-0C99140 6602 030 DD012 Non-configured
_ 68-5544453 6602 050 DD011 Non-configured

31. Type a 1 in the Option column on the Select Replacement Unit display and
press the Enter key. Select a non-configured disk unit that is attached to the
MFIOP. You recorded the serial numbers of the disk units that are attached to
the MFIOP in an earlier step.

Confirm Replace of Configured Unit

This screen allows the confirmation of the configured unit to

be replaced with the selected replacement unit.

Press Enter to confirm your choices for Replace.

Press F12 to return to change your choices.

The configured unit being replaced is:

Serial Resource
Unit ASP Number Type Model Name Status
1 1 68-0C47591 6602 030 DD001 Suspended

The replacement unit will be:

Serial Resource
Unit ASP Number Type Model Name Status
1 1 68-0119804 6602 050 DD002 Resuming

32. Press Enter to confirm.

33. The replacement function runs for several minutes. Wait until the replacement
function completes.
34. Add the non-configured disk units to the system ASP. The Add function
initializes the disk units so that device parity protection can be started on the
disk units.
35. Display the disk configuration again and ensure that the disk units which are
attached to the MFIOP do not have a model 030.

Chapter 26. Working with Device Parity Protection 717

 ┌────────────────────────┐ Select option 3
│ SST Menu │ (Work with disk units)
│ │
│ ┌───────────────────┴─────┐ Select option 1
│ 3 │ Work with Disk Units │ (Work with disk configuration)
└────┤ │
│ ┌────────────────────┴─────┐ Select option 1
│ 1 │ Work with Disk │ (Display disk configuration)
└────┤ Configuration │
│ │
│ 1 │
└──────────────────────────┘

The disk units and their status are displayed. Ensure that the disk units
attached to the MFIOP are not model 030.

Display Disk Configuration Status

Serial Resource
ASP Unit Number Type Model Name Status
1 Mirrored
1 68-0119804 6602 050 DD002 Active
1 68-OC60040 6602 050 DD003 Active
5 10-1000128 9337 221 DD005 Active
5 10-2000128 9337 221 DD006 Active
7 10-3000128 9337 221 DD007 Active
7 10-5000128 9337 221 DD008 Active
8 68-54531 6602 050 DD004 Active
8 68-0C99140 6602 050 DD012 Active
9 68-5544453 6602 050 DD011 Active
9 68-OC47591 6602 050 DD001 Active

36. Stop mirrored protection on the system ASP. See “How to Stop Mirrored
Protection” on page 723 for more information.
37. Start device parity protection on the disk units that are attached to the MFIOP.
See “Starting Device Parity Protection” on page 695 for the complete
instructions on starting device parity protection

718 OS/400 Backup and Recovery V5R1

Chapter 27. Working with Mirrored Protection
This chapter describes how to start and stop mirrored protection. It also describes
the rules that apply when you set up a mirrored ASP.

Mirrored Protection–Configuration Rules

The following rules apply to mirrored configurations:
v Mirrored protection is configured by ASP number. The system configures the
two units of a mirrored pair within an ASP.
v Mirrored protection requires an even number of storage units for each size of
disk unit in the ASP that is being mirrored. An uneven number of storage units
for any size of disk unit prevents mirrored protection from starting. This
requirement does not apply to disk units that have device parity protection.
v The system attempts to assign the two storage units of a mirrored pair so that a
failed unit can be repaired while the system continues to use the other mirrored
unit. For a hardware configuration where this is not possible, the repair of the
failing unit must be delayed until the system can be powered down. This may
be true for a failing mirrored unit which is sharing the same controller or I/O
processor as its mirrored unit.
v Standard DASD mirroring support requires that the mirrored units are at
specific input and output addresses on the system. (For Version 3 Release 6, you
can override this restriction with a patch.) Both units in the pair must be
attached to the multi-function IOP because the system must be able to perform
an IPL from either unit. Therefore, the system attempts to assign the mirrored
units for unit 1 of the system ASP first. If you are mirroring the system ASP,
mirrored protection does not start if valid mirrored units for unit 1 cannot be
found.

Note: Remote mirroring support removes this requirement and provides

IOP-level or bus-level protection.

How to Start Mirrored Protection

You start mirrored protection for a specific ASP on your system. Before you
attempt to start mirrored protection, you should ensure that the ASP meets the
requirements that are described in “Mirrored Protection–Configuration Rules”. If
you are performing several disk configuration and disk protection tasks, refer to
Chapter 24 to determine the correct sequence of steps for your situation.

To start mirrored protection, do the following:

Logical partitioning users: If you perform an IPL on the primary partition, the
secondary partitions will power down. If there is any
activity on the secondary partitions when this occurs,
the next IPL may be abnormal. You should quiesce all
secondary partition activity before starting mirroring
on the primary partition.
1. If you are not already using DST, perform a manual IPL to start DST. See “How
to Start Dedicated Service Tools (DST)” on page 660.

© Copyright IBM Corp. 1997, 2000, 2001 719

 2. From the Use Dedicated Service Tools (DST) menu, do the following:

┌────────────────────────┐ Select option 4

│ DST Menu │ (Work with disk units)
│ │
│ ┌───────────────────┴─────┐ Select option 1
│ 4 │ Work with Disk Units │ (Work with disk configuration)
└────┤ │
│ ┌────────────────────┴───┐ Select option 4
│ 1 │ Work with Disk │ (Work with mirrored protection)
└────┤ Configuration │
│ ┌───────────────────┴──────┐
│ 4 │ Work with Mirrored │
└────┤ Protection │
│ │
│ │
└──────────────────────────┘

3. Select option 2 (Start mirrored protection) on the Work with Mirror Protection
display.

Select ASP to Start Mirrored Protection

Select the ASPs to start mirrored protection on.

Type options, press Enter

1=Select
Option ASP Protection
_ 1 Unprotected
1 2 Unprotected

4. Select the ASP or ASPs to be mirrored on the Select ASP to Start Mirrored
Protection display and press the Enter key. Do not select ASPs that are
protected with device parity protection.
You may see the following display:
Press the Enter key to continue.
Confirm Continuation
To proceed, the system must perform directory recovery, which may take a
significant amount of time. The system may appear inactive during this time.

Press Enter to continue. Press F12=Cancel to return and change your choices.

5. The system shows a confirmation display of the new mirrored protection

configuration, including the levels of protection. Notice that half of the
previous unit numbers for units being mirrored in the ASP no longer exist. The
storage units for those unit numbers have been paired with the storage units
for the remaining unit numbers to make mirrored pairs. Also notice that device
parity units in the ASP have not been affected.

720 OS/400 Backup and Recovery V5R1

 Confirm Start Mirrored Protection

Press Enter to confirm your choice to start mirrored

protection. During this process the system will be IPLed.
You will return to the DST main menu after the IPL is
complete. The system will have the displayed protection.

Press F12 to return to change your choice.

Serial Resource
ASP Unit Number Type Model Name Protection
1 Unprotected
1 00-48519 6606 030 DD010 Unprotected
2 Mirrored
2 00-1000341 9337 211 DD012 Disk Unit
2 00-5000341 9337 211 DD015 Disk Unit
3 00-0186325 6602 074 DD019 Device Parity
4 00-0162516 6602 074 DD025 Device Parity
5 00-0238703 6602 074 DD052 Device Parity

6. If the configuration is what you had planned and you do not have other
configuration changes to make, skip to step 7
If the configuration is not what you had planned, for example, the level of
protection is less, you have the following options:
v Verify that the correct ASP was selected. Verify that any new storage units
have been added to the correct ASP.
v Determine if additional hardware is required to achieve the planned level of
protection.
v Determine if the existing hardware needs to be connected differently to
achieve the planned level of protection. Contact your technical support
organization for assistance.
v Consider continuing the start mirrored protection process which will provide
better availability than non-mirrored protection, rather than waiting until
additional hardware arrives so that you can achieve your planned level of
protection. After you receive and install the additional hardware, use Table 83
on page 643 to determine the procedure for configuring your disk storage
correctly. Even on very large systems, the tasks to stop mirroring, add units,
and start mirrored protection can be done in several hours.
7. Place the system in Normal mode and press the Enter key to accept the
configuration. The system performs the first part of the process for starting
mirrored protection. During that time, you are shown the Function Status
display:
The system updates the display periodically.
Function Status
You selected to start mirrored protection. 5 % Complete

Note: You can press F16 to return to the Use Dedicated Service Tools (DST)
menu if you have other tasks to perform. However, you cannot perform
any disk configuration tasks or end DST until the system has finished
starting mirrored protection.

The system continues the start mirrored protection process in What the System
Does When You Start Mirrored Protection without further operator
intervention.

Chapter 27. Working with Mirrored Protection 721

 8. After the system reaches the Command Entry display, you may want to create
the QSYSMSG message queue to receive messages. If you have this message
queue, the system sends certain critical messages to it. You can monitor the
message queue with a program to make sure the messages are not missed.

What the System Does When You Start Mirrored Protection

The following steps are performed by the system when mirrored protection is
started.
1. Data is moved off half of the storage units in the selected ASPs. This can take
from a few minutes to a few hours, depending on the amount of data that must
be moved. See “Appendix C. Example of Configuration Planning for Mirrored
Protection” on page 795 for information in estimating the time required.
Objects that are created on a preferred unit may be moved to another unit. The
preferred unit number may no longer exist when mirrored protection starts.
2. New control information is written to disk, describing the new mirrored system
configuration.
| 3. After the data is moved and the control information is written, the system
| performs an IPL. However, when you start mirroring only on independent
| ASPs, the system does not perform an IPL.
4. When the system reaches DST, the previously selected ASPs are mirrored,
although the two storage units in the mirrored pairs are not yet synchronized.
If the keylock switch is in the Manual position, you have the option to perform
other configuration changes or perform an IPL. If you do not have
configuration changes to make, select the option to perform an IPL and press
the Enter key.
If the keylock switch is in the Normal position, the system automatically
continues the IPL.
5. When the system continues the IPL past DST, the mirrored pairs are
synchronized during storage management recovery. This can take a few hours,
although this long recovery time only occurs when mirrored protection is first
started, and not during every IPL on a mirrored system. The progress of the
synchronization process is displayed and periodically updated on the control
panel. The system will display the code SRC C6xx 4205 where xx indicates the
completed percentage of the synchronization process. On very large systems,
the entire start mirrored protection process can take up to approximately 8 to
10 hours. (See “Appendix C. Example of Configuration Planning for Mirrored
Protection” on page 795 for an example.)
6. After storage management recovery is completed, the selected ASPs have
mirrored protection.

Mirrored Protection Configuration Errors

There can be no missing, active disk units anywhere in the configuration when
mirroring is being started. Units with a status of missing must be powered,
repaired, or replaced.

Starting mirrored protection can fail if there is insufficient storage available in the
ASP to contain the current data in the ASP. The percentage that is used in the ASP
must normally be less than half of the ASP threshold. The exception to this occurs
when the ASP contains device parity protected disk units that can allow starting
mirrored protection with a greater percent used.

722 OS/400 Backup and Recovery V5R1

 There must be sufficient storage units in the ASP for the system to create mirrored
pairs. If you receive a message that indicates that the system cannot pair unit 1 or
other units, review “Mirrored Protection–Configuration Rules” on page 719.

How to Stop Mirrored Protection

When you stop mirrored protection, one unit from each mirrored pair becomes
nonconfigured. Before you can stop mirrored protection for an ASP, at least one
unit in each mirrored pair in that ASP must be present and not suspended. To
control which mirrored unit of each pair becomes nonconfigured, you may
suspend the storage units that you want to become nonconfigured. For units that
are not suspended, the selection is automatic.

To stop mirrored protection, do the following:

1. If you are not already using DST, perform a manual IPL to start DST. See “How
to Start Dedicated Service Tools (DST)” on page 660.
2. From the Use Dedicated Service Tools (DST) menu, do the following:

┌────────────────────────┐ Select option 4

│ DST Menu │ (Work with disk units)
│ │
│ ┌───────────────────┴─────┐ Select option 1
│ 4 │ Work with Disk Units │ (Work with disk configuration)
└────┤ │
│ ┌────────────────────┴───┐ Select option 4
│ 1 │ Work with Disk │ (Work with mirrored protection)
└────┤ Configuration │
│ ┌───────────────────┴──────┐
│ 4 │ Work with Mirrored │
└────┤ Protection │
│ │
│ │
└──────────────────────────┘

3. Select option 3 (Stop mirrored protection) on the Work with Mirror Protection
display. You are shown the Select ASP to Stop Mirrored Protection display:

Select ASP to Stop Mirrored Protection

Select the ASPs to stop mirrored protection on.

Type options, press Enter

1=Select

Option ASP Protection

1 2 Mirrored

4. Select the ASP or ASPs for which mirrored protection is to be stopped on the
Select ASP to Stop Mirrored Protection display and press the Enter key. The
Confirm Stop Mirrored Protection display is shown:

Chapter 27. Working with Mirrored Protection 723

 Confirm Stop Mirrored Protection

Press Enter to confirm your choice to stop mirrored

protection. During this process the system will be IPLed.
You will return to the DST main menu after the IPL is
complete. The system will have the displayed protection.

Press F12 to return to change your choice.

Serial Resource
ASP Unit Number Type Model Name Protection
1 Unprotected
1 00-48519 6606 030 DD010 Unprotected
2 Unprotected
2 00-1000341 9337 211 DD012 Unprotected
3 00-0186325 6602 074 DD019 Device Parity
4 00-0162516 6602 074 DD025 Device Parity
5 00-0238703 6602 074 DD052 Device Parity

5. Press the Enter key to confirm your choice. The system stops mirrored
| protection for the ASPs you requested and performs an IPL. However, when
| you stop mirroring on only independent ASPs, the system does not perform an
| IPL.

724 OS/400 Backup and Recovery V5R1

 Chapter 28. Working with Disk Compression
This chapter describes how to start and stop disk compression. It also discusses
various considerations for using disk compression.

Introduction to Disk Compression

Disk compression is a technology that increases the apparent capacity of disk
storage devices by encoding the data to take up less physical storage space on
disk. Disk compression is performed in the disk subsystem controller and does not
affect the iSeries system processor. The compression and decompression of data is
performed automatically on each write command and read command, respectively.
With the exception of a performance impact, disk compression is transparent to
applications. The performance of compressed disk drives is slower than the
performance of non-compressed disk drives. This is due to the overhead of
compression and decompression, and the variations in the length of the data that is
written to disk.

Typically, data that is found on disk units has a wide range of access requirements.
You may choose to move data that is accessed infrequently, or data that does not
require high performance I/O rates, to compressed disk units. Disk compression is
intended to make infrequently accessed data available on-line at a lower cost. This
storage alternative is positioned between non-compressed disk unit storage and
optical or tape storage.

Compressed disks have the same disk subsystem availability options of device
parity protection and mirrored protection as non-compressed disks. Disk
compression is only supported in user ASPs.

Restrictions and Considerations

The following restrictions and considerations apply to disk compression:
v A compression-capable storage controller is required for the compressed disk
units.
| v If you configured a 2748 or 2778 storage I/O controller for extended adaptive
| cache, you cannot use this controller for disk compression at the same time.
| However, you may re-configure the 2748 or the 2778 storage I/O controller for
| disk compression.
| The iSeries Information Center contains information on how to reconfigure your
| 2748 or 2778 storage I/O controller. Search for ″jumper″ in the Information
| Center. Select the page about setting or changing the mode of an I/O card from
| the search results and follow the procedure. You can access the Information
| Center form the following Web site:
| http://www.ibm.com/eserver/iseries/infocenter
| v Disk compression is only allowed on certain types of disk units. Contact your
| service provider for the most current list of disk units which are capable of disk
| compression.
v If you are using V4R3M0 on your system, you can start or stop disk
compression only on nonconfigured disk units. If you are using V4R4M0 or later
on your system, you can start or stop disk compression on configured and
nonconfigured disk units.

© Copyright IBM Corp. 1997, 2000, 2001 725

 Note: If the disk unit is configured, you can only start disk compression if the
disk unit is less than or equal to 92% full. You can only stop disk
compression if there is room in the ASP for moving data off the disk unit
so that the amount of data left on the disk unit is 92% of its
uncompressed capacity.
v Disk units that are compressed may not be added to the System ASP. You can
only add compressed units to a User ASP.
v A User ASP may contain compressed and non-compressed disk units.
v A device parity protection set may contain compressed and non-compressed disk
units.
v If you are using mirrored protection, both disk units in the mirrored pair must
have the same compression status (either both compressed or both
non-compressed).
v You may experience a zero to 30% performance degradation on I/O operations
to compressed disk units. Your results may vary based on your actual work load
characteristics.

Disk Compression and Capacity

Capacity gains may vary and are determined by the amount of compression that
can be performed on the data. An overall 2:1 compression ratio should be expected
for a disk unit. A maximum 4:1 compression ratio for a disk unit may be
achievable in certain circumstances. Additionally, a compression ratio for a set of
data on that disk unit may reach a maximum of 8:1. The iSeries software enforces
the overall disk unit maximum compression ratio of 4:1. Because capacity is
dependent on how well the data compresses, the capacity that is displayed will
change as data is written to or deleted from the disk. The capacity that is
displayed for a compressed disk unit is the amount of data on the disk plus an
estimate of how much additional data can fit on the disk.

The following example shows the calculation and display of capacity by the system
for compressed disk units. The disk unit capacities are available on the Display
Disk Configuration Capacity display under the DST menus and the SST menus.
The capacities are also available on the Work with Disk Status (WRKDSKSTS)
display.

Note: If you have the licensed program 5769PT1 Performance Tools for iSeries
installed on the system, you may use the system report to display the
compression ratio. (You can find the ratio in the″Disk Compression
Statistics″ section in the system report.)
1. Prior to starting compression, a nonconfigured 6602 Model 050 has a capacity
of 1031 megabytes.
Display Non-Configured Units

Serial Resource
Number Type Model Name Capacity Status
83-0135199 6602 050 DD005 1031 Non-configured
83-0306044 6602 050 DD006 1031 Non-configured

2. After starting compression, the 6602 model number changes to 060, and the
capacity doubles.

726 OS/400 Backup and Recovery V5R1

 Display Non-Configured Units

Serial Resource
Number Type Model Name Capacity Status
83-0135199 6602 060 DD005 2062 Non-configured
83-0306044 6602 060 DD006 2062 Non-configured

3. Two compressed disk units are added into user ASP 2.

Display Disk Configuration Capacity

----Protected---- ---Unprotected---
ASP Unit Type Model Threshold Overflow Size % Used Size %Used
1 90% No 0 0.00% 21372 17.26%
1 6607 050 0 0.00% 4194 29.25%
2 6713 050 0 0.00% 8589 14.33%
3 6713 050 0 0.00% 8589 14.34%

2 90% No 0 0.00% 4124 0.10%

5 6602 060 0 0.00% 2062 0.10%
6 6602 060 0 0.00% 2062 0.10%

4. After writing data to the user ASP, capacities and percentages that are used are
displayed.
Display Disk Configuration Capacity

----Protected---- ---Unprotected---
ASP Unit Type Model Threshold Overflow Size % Used Size %Used
1 90% No 0 0.00% 21372 17.26%
1 6607 050 0 0.00% 4194 29.25%
2 6713 050 0 0.00% 8589 14.33%
3 6713 050 0 0.00% 8589 14.34%

2 90% No 0 0.00% 5812 61.06%

5 6602 060 0 0.00% 2905 61.04%
6 6602 060 0 0.00% 2907 61.08%

5. The following calculations may be performed to determine how well the data is
being compressed, and the estimated disk unit capacity. These calculations may
be performed on a user ASP basis, as well as on individual disk units.

Amount Calculation

Logical data written Size * (% Used / 100)

Physical free space (Size * (1-(% Used / 100)) / 2
Physical space used Non-compressed size - Physical free space
Compression ratio of data Logical data written / Physical space used
written
Estimated disk capacity Logical data written + (2 * Physical free space)

Using the values listed for Unit 5 in the preceding step with these formulas
produces the following:

Amount Calculation

Logical data written 2905 * .6104 = 1773

Physical free space (2905 * (1-(.6104)) / 2 = 566
Physical space used 1031 - 566 = 465
Compression ratio 1773 / 465 = 3.8
Estimated disk capacity 1773 + (2 * 566) = 2905

Chapter 28. Working with Disk Compression 727

 Disk Unit Full Considerations
As space is being reserved or data is being written to compressed disk units, it is
possible for a unit to become temporarily full. The storage subsystem controller
will detect this situation and attempt to reposition the data on the unit in order to
maximize the amount of data that can be stored on the unit. This repositioning of
the data increases the effective size of the disk unit. If the storage subsystem
controller can not immediately service a system request, a failure will be returned
to the system. The system bases its response to this failure on the type of I/O
operation that is requested. One of the following scenarios may occur:
v The system responds to this failure by overflowing the auxiliary storage pool
(ASP). Once the overflow occurs, this I/O request will be performed in the
system ASP and will succeed.
v The system responds to this failure by displaying an A6xx 0277 system reference
code (SRC) on the control panel in the system unit. The system displays this
attention SRC until storage space becomes available on the disk unit that is
denoted in the attention SRC.
For more information, refer to “How The System Responds to Disk Unit Full”.

When the system displays an A6xx 0277 attention SRC on the control panel, it also
logs a corresponding A6xx 0277 record in the Product Activity Log. This occurs
each time this disk unit full condition is detected. The system also sends message
CPI116C ″Compressed disk unit &1 is full″ to the QSYSOPR message queue. The
system will reissue the failed I/O operation and will continue to display the
attention SRC on the control panel until the condition is corrected. When the
storage subsystem controller creates sufficient space on the compressed unit to
contain the system request, the I/O operation completes successfully and the
system resumes normal processing.

While this attention SRC is being displayed, some I/O operations to the affected
compressed disk unit may be suspended. As a result, you may observe that jobs
which issue I/O operations to the affected unit appear to hang.

To reduce the likelihood of system operations hanging while the storage subsystem
recovers from a disk unit full condition, it is recommended that ASPs with
compressed units operate with a storage threshold of less than or equal to 90%.

As space on the disk unit continues to be used, eventually the storage subsystem
controller can no longer store more data on the unit. At this point, the storage
subsystem controller will return a failure on any system requests requiring storage
space. Refer to the following section, How The System Responds to Disk Unit Full,
for more information.

How The System Responds to Disk Unit Full

The system bases its response to the disk unit full condition on the type of I/O
operation that was issued that caused the condition.

If the system request is reserving additional storage space in the ASP, the
compression recovery policy for the ASP determines the system response. You set
this policy by using the Change ASP Attribute (CHGASPA) command.

728 OS/400 Backup and Recovery V5R1

 The system may respond to the disk unit full condition in one of the following
ways:
v If the compression recovery policy is *OVERFLOW, the system responds to this
condition by overflowing the ASP. Once the overflow occurs, this I/O request
will be performed in the system ASP and will succeed. This is the system default
compression recovery policy for all user ASPs.
For more information, refer to “How to Recover An Overflowed User Auxiliary
Storage Pool” on page 191.
v If the compression recovery policy is *RETRY, the system responds to this
condition by displaying an A6xx 0277 SRC on the control panel and repeatedly
retrying the failed I/O operation. The system displays this attention SRC until
storage space becomes available on the disk unit and the I/O operation
succeeds. If storage space does not become available on the disk unit, the ASP
overflows. The system removes the attention SRC from the control panel, and
the I/O operation is successfully performed in the system ASP.
For more information, refer to “How to Recover An Overflowed User Auxiliary
Storage Pool” on page 191.
v If the compression recovery policy is *WAIT, the system responds to this
condition by displaying an A6xx 0277 SRC on the control panel and repeatedly
retrying the failed I/O operation. If storage space does not become available on
the disk unit, the ASP does not overflow. The user must take one of the
corrective actions that are discussed in “SRC Code A6xx 0277”.

SRC Code A6xx 0277

When you see the A6xx 0277 SRC code in the control panel, select the appropriate
compression reference code word for additional information. The compression
reference code word is either word 15 or word 17. In the V4R5 release, the
compression reference code word format changed.
Table 102. Word formats for SRC codes in V4R5.
Word for the Word format Description
SRC code
5 (for models CCEE 0000 This word is the compression
270 and 8xx) reference code that indicates two
CC indicates the operation in things. First, it indicates which
15 for all other progress with the following values: operation was in progress. Second,
models v 84 is an allocate operation it indicates whether or not the
v 2x is a write operation where x is storage subsystem controller can
1, 2, or 4 obtain additional storage space on
the system.
EE indicates the error code with the
following values: This reference code word was
formerly word 17 in the V4R4
v 00 is the storage subsystem release and in previous releases.
controller can not obtain any
additional storage space on the
unit
v 02 is the storage subsystem
controller will be able to obtain
additional storage space on the
unit
6 (for models uuuu uuuu This word describes the unit
270 and 8xx) address of the disk unit.

16 for all other

models

Chapter 28. Working with Disk Compression 729

 Table 102. Word formats for SRC codes in V4R5. (continued)
7 (for models BBBB ccbb This word defines the bus, card,
270 and 8xx) and board address of the disk unit.

17 for all other

models

Note: The Information Center contains information that correlates the function and
word for SRC codes on models 270, and 8xx. Look under System
Administration, Availability, and Maintenance-> Logical
Partitions->Troubleshooting logical partitions->Learning about system
reference codes (SRC) for logical partitions.
Table 103. SRC code word formats in V4R4 and previous releases.
Word for the Word format Description
SRC code
15 0000 0000 Unassigned in V4R4 and previous
releases.
16 uuuu uuuu This word describes the unit
address of the disk unit.
17 CCEE BBcb Word 17 defines: the operation in
progress, the error code; and the
CC indicates the operation in bus, card, and board address of the
progress with the following values: disk unit.
v 84 is an allocate operation
v 2x is a write operation where x is
1, 2, or 4

EE indicates the error code with the

following values:
v 00 is the storage subsystem
controller can not obtain any
additional storage space on the
unit
v 02 is the storage subsystem
controller will be able to obtain
additional storage space on the
unit

BBcb indicates the bus, card, and

board address of the disk unit.

Perform one of the three following actions in response to SRC A6xx 0277.

User Action 1
Wait for the storage subsystem controller to reposition the data on the disk unit.

If the error code for EE of the attention SRC is 02, the storage subsystem controller
will eventually obtain additional storage space on the unit, such that the I/O
operation will succeed. If the system does not resume normal processing within 20
minutes, contact your next level of support.

730 OS/400 Backup and Recovery V5R1

User Action 2
Make storage space available in the ASP that contains the disk unit that is denoted
in the SRC.

Word 16 contains the unit address of the disk unit. Word 17 (V4R4 and previous)
contains the right most characters as BBcb. Word 17 or 7 (V4R5) is BBBB ccbb.
Refer to ″Hardware Service Manager″ in the iSeries Service Functions to correlate
the unit address (logical address) with a resource name or serial number. The ASP
that contains the disk unit can be determined by using the Display Disk
Configuration Status display under the DST and SST menus.

If the error code for EE of the attention SRC is 00, the storage subsystem controller
determined that the disk unit is full.

Perform one or more of the following actions:

v Delete objects that are not needed from the ASP.
v Save objects that are not needed from the ASP by specifying STG(*FREE) on the
Save Object (SAVOBJ) command.
v Move one or more libraries to a different ASP.

Note: You cannot use the MOVOBJ command to do this. You must save the
library, delete it, and then restore it to a different ASP.
v Move one or more folders to a different ASP by saving the folder, deleting it,
and restoring it to a different ASP.
v Increase the storage capacity by adding disk units to the ASP.

User Action 3
Change the compression recovery policy to the desired system behavior. For more
information on the CHGASPA command, refer to the iSeries server online help.

User Action 4
Re-IPL the system so that additional storage space can be made available in the
ASP that contains the disk unit that was denoted in the attention SRC on the
subsequent IPL.

Word 16 contains the unit address of the disk unit. Word 17 (V4R4 and previous)
contains the right most characters as BBcb. Word 17 or 7 (V4R5) is BBBB ccbb.
Refer to ″Hardware Service Manager″ in the iSeries Service Functions to correlate
the unit address (logical address) with a resource name or serial number. The ASP
that contains the disk unit can be determined by using the Display Disk
Configuration Status display under the DST and SST menus.

If the error code for EE of the attention SRC is 00 and the system is holding critical
resources, the system will eventually hang. The recommended recovery procedure
is to re-IPL the system. The system must be in Manual mode. Perform the
following steps:
1. Force the system to write changed data in main storage to disk storage by
pressing the power push-button twice to stop the system. Wait for system
activity to stop.
There will be changed data that is in main storage which can not be written to
the disk unit. Therefore, the above system power off will eventually hang.
2. Start an IPL.

Chapter 28. Working with Disk Compression 731

 a. Ensure that the keystick is in the system unit control panel.
b. Place the system in manual mode.
c. Press the Function Select switch (or buttons) to display 03 in the Function
display.
d. Press the Enter button on the control panel.
3. On the following Manual mode IPL, perform one of the following:
v Increase the storage capacity by adding disk units to the ASP at DST
v Start the system to restricted state. Make storage space available in the ASP
that contains the disk unit denoted in the attention SRC. Make the space
available by performing one or more of the following steps:
– Delete objects that are not needed from the ASP.
– Save objects that are not needed from the ASP by specifying STG(*FREE)
on the Save Object (SAVOBJ) command.
– Move one or more libraries to a different ASP.

Note: You cannot use the MOVOBJ command to do this. You must save
the library, delete it, and restore it to a different ASP.
– Move one or more folders to a different ASP by saving the folder, deleting
it, and restoring it to a different ASP.

Examples of A6xx 0277

The following examples illustrate two situations where an A6xx 0277 SRC record
will be generated, and any associated actions you may need to take.
17 8402 0110 (V4R4 and previous) OR 15 or 5 8402 0000 (V4R5)
In this example, an allocate operation (84) is being attempted and the error
code (02) denotes the operation is being retried and will eventually
succeed. You do not need to take any additional action. If the system does
not resume normal processing within 20 minutes, contact your next level of
support.
If you want the system to overflow the user ASP into the system ASP, take
User Action 3 that specifies the compression recovery policy *OVERFLOW,
described above.
17 2000 0110 (V4R4 and previous) OR 15 or 5 2000 0000 (V4R5)
In this example, a write operation (20) is being attempted and the error
code (00) denotes that the operation is being retried indefinitely, because
the storage subsystem controller has determined that there is no storage
space available on the disk unit. Take either User Action 2 or User Action
4, described above.

How to Start Disk Compression

You can start disk compression from the Dedicated Service Tools (DST) menu.

Note:

You can use a 2748 storage I/O controller for extended adaptive cache or
disk compression, but not both at the same time. The Information Center
contains information on how to configure your 2748 storage I/O controller.
Look under the Storage I/O card modes and jumpers information in the
Information Center Website:
http://www.ibm.com/eserver/iseries/infocenter.

732 OS/400 Backup and Recovery V5R1

 The V4R5 navigation path to this information is System
Management–>System hardware–>Storage I/O card modes and
jumpers–>Setting or changing the mode of an I/O card.

To start disk compression from DST, perform the following steps.

1. If your use your 2748 storage I/O controller for compression, make sure that
you set the jumper for compression mode before you continue with the
following steps. See 732.
2. If you are not already using DST, perform a manual IPL to start DST. See “How
to Start Dedicated Service Tools (DST)” on page 660.
3. From the Use Dedicated Services Tools (DST) menu, do the following:

┌────────────────────────┐ Select option 4

│ DST Menu │ (Work with disk units)
│ │
│ ┌───────────────────┴─────┐ Select option 1
│ 4 │ Work with Disk Units │ (Work with disk configuration)
└────┤ │
│ ┌────────────────────┴───┐ Select option 6
│ 1 │ Work with Disk │ (Work with disk compression)
└────┤ Configuration │
│ ┌───────────────────┴──────┐
│ 6 │ Work with Disk │
└────┤ Compression │
│ │
│ │
└──────────────────────────┘

4. Select option 2 on the Work with Disk Compression screen.

Work with Disk Compression

Select one of the following:

1. Display disk compression status

2. Start compression on disk units
3. Stop compression on disk units

Selection

F3=Exit F12=Cancel

5. Select the disk units that you want to start compression on from the Select Disk
Units for Start Compression screen.

Note: You can only start compression on a configured unit if the disk unit is
less than or equal to 92% full.

Chapter 28. Working with Disk Compression 733

 Select Disk Units for Start Compression

Select the units for start compression.

Type choice, press Enter.

1=Start compression
Serial Resource
OPT Unit ASP Number Type Model Name Status
1 68-7F0DB 6607 050 DD005 Non-configured
68-5FB0B 6713 050 DD001 Non-configured

F3=Exit F12=Cancel

6. You are shown the Confirm Disk Units for Start Compression screen. This
display shows the approximate amount of time that is required to start disk
compression, and the current and proposed sizes of the disk unit.

Confirm Disk Units for Start Compression

Estimated time for this operation to complete : 14-16 Minutes

Press Enter to confirm your choice to compress the disk units.

Press F9=Capacity Information to display the resulting capacity.
Press F12=Cancel to return to change your choice.

Serial Current Proposed

OPT ASP Unit Number Type Model Size Size
1 68-7F0DB 6607 050 4194 8388

F9=Resulting Capacity F12=Cancel

7. At the Confirm Disk Units for Start Compression screen, press Enter to confirm
your choice of disk units that you want to start compression on. You are shown
the Start Compression on Disk Unit Status screen.

Start Compression on Disk Unit Status

Estimated time for this operation to complete : 14-16 Minutes

The operation to start compression on the selected disk units

will be done in multiple phases. The phases are listed here
and the status will be updated as the phase proceeds.

Phase Status

Prepare to start compression . . . . . . : 0 % Complete

Start compression . . . . . . . . . . . :
Prepare to compress data . . . . . . . . :
Compress data . . . . . . . . . . . . . :

Wait for next display or press F16 for DST menu

734 OS/400 Backup and Recovery V5R1

 8. After the start compression operation completes, you are returned to the Work
with Disk Compression screen, and shown a completion message.

Work with Disk Compression

Select one of the following:

1. Display disk compression status

2. Start compression on disk units
3. Stop compression on disk units

Selection

F3=Exit F12=Cancel
The requested compression operation completed successfully.

How to Stop Disk Compression

To stop disk compression from DST, perform the following steps.
1. If you are not already using DST, perform a manual IPL to start DST. See “How
to Start Dedicated Service Tools (DST)” on page 660.
2. From the Use Dedicated Services Tools (DST) menu, do the following:

┌────────────────────────┐ Select option 4

│ DST Menu │ (Work with disk units)
│ │
│ ┌───────────────────┴─────┐ Select option 1
│ 4 │ Work with Disk Units │ (Work with disk configuration)
└────┤ │
│ ┌────────────────────┴───┐ Select option 6
│ 1 │ Work with Disk │ (Work with disk compression)
└────┤ Configuration │
│ ┌───────────────────┴──────┐
│ 6 │ Work with Disk │
└────┤ Compression │
│ │
│ │
└──────────────────────────┘

3. Select option 3 on the Work with Disk Compression screen.

Work with Disk Compression

Select one of the following:

1. Display disk compression status

2. Start compression on disk units
3. Stop compression on disk units

Selection

F3=Exit F12=Cancel

4. Select the disk units that you want to stop compression on from the Select Disk
Units for Start Compression screen.

Chapter 28. Working with Disk Compression 735

 Note: Compression can only be stopped if there is room in the ASP for moving
data off of the disk unit. Once the data has been moved off, the amount
of data that is left on the disk is less than or equal to 92% of its
uncompressed capacity.

Select Disk Units for Stop Compression

Select the units for stop compression.

Type choice, press Enter.

1=Stop compression
Serial Resource
OPT Unit ASP Number Type Model Name Status
1 68-7F0DB 6607 060 DD005 Non-configured

F3=Exit F12=Cancel

5. You are shown the Confirm Disk Units for Stop Compression screen. This
display shows the approximate amount of time that is required to stop disk
compression, and the current and proposed sizes of the disk unit.

Confirm Disk Units for Stop Compression

Estimated time for this operation to complete : 1-2 Minutes

Press Enter to confirm stop compression.

Press F9=Capacity Information to display the resulting capacity.
Press F12=Cancel to return to change your choice.

Serial Current Proposed

OPT ASP Unit Number Type Model Size Size
1 68-7F0DB 6607 060 8388 4194

F9=Resulting Capacity F12=Cancel

6. At the Confirm Disk Units for Stop Compression screen, press Enter to confirm
your choice of disk units that you want to stop compression on. You are shown
the Stop Compression on Disk Unit Status screen.

Stop Compression on Disk Unit Status

Estimated time for this operation to complete : 1-2 Minutes

The operation to stop compression on the selected disk units

will be done in multiple phases. The phases are listed here
and the status will be updated as the phase proceeds.

Phase Status

Prepare to stop compression . . . . . . : 0 % Complete

Stop compression . . . . . . . . . . . . :

736 OS/400 Backup and Recovery V5R1

 7. After the stop compression operation completes, you are returned to the Work
with Disk Compression screen, and shown a completion message.

Work with Disk Compression

Select one of the following:

1. Display disk compression status

2. Start compression on disk units
3. Stop compression on disk units

Selection

F3=Exit F12=Cancel
The requested compression operation completed successfully.

Procedural Sequences for Configuring Disks and Protection

Each respective configuration change requires you to perform procedures in a
specific order. The following lists show the order that you should use when
performing the procedures.

Adding a New I/O Compression-Capable Storage Controller

This checklist shows the sequence of tasks that you use to add a new
compression-capable I/O Storage Controller and disk units to your system.

Before you begin, make a copy of this checklist. Fill in the appropriate areas as you
or the service representative perform the configuration tasks. This checklist
provides an important record of your actions. It may help you to diagnose any
problems that occur.

Most tasks in the checklist include references to other topics in this book. Refer to
these topics if you need more information about how to perform a particular task.
Table 104. Adding a New I/O Storage Controller and Disk Units
Task What to Do Where to Read More About It
Task 1 Install the new storage controller in the
system. This is normally done by a service
representative.
Task 2 Physically attach disk units to the new
storage controller. This is normally done by a
service representative.
Task 3 Start DST. “How to Start Dedicated Service Tools
(DST)” on page 660.
Task 4 Print your current disk configuration. “How to Display Your Disk Configuration”
on page 663.
Task 5 If you want to have device parity protection “How to Start Device Parity Protection for
for the storage controller, start device parity an Input/Output Processor” on page 699.
protection now.
Task 6 Start disk compression on nonconfigured disk “How to Start Disk Compression” on
units. page 732.

Chapter 28. Working with Disk Compression 737

Table 104. Adding a New I/O Storage Controller and Disk Units (continued)
Task What to Do Where to Read More About It
Task 7 Add nonconfigured disk units to the correct “How to Add Disk Units to an Auxiliary
ASPs. You can add the disk units to an Storage Pool” on page 669.
existing user ASP or to a new ASP. If you are
adding disk units to an ASP that has
mirrored protection and the new disk units
do not have device parity protection, you
must add pairs of disk units that have equal
capacities.
Task 8 If you created a new ASP on your system “How to Change the Storage Threshold for
when you added disk units, the system set an Auxiliary Storage Pool” on page 672.
the storage threshold of the ASP to 90%. If
you want a different threshold, change it at
this time.
Task 9 If you created any new ASPs and you want “How to Start Mirrored Protection” on
those ASPs to have mirrored protection, start page 719.
mirrored protection now.
Task 10 Verify that your disk configuration is correct. “How to Display Your Disk Configuration”
on page 663.
Task 11 End DST. “How to Stop Dedicated Service Tools
(DST)” on page 662.

Adding Disk Units to an Existing Compression-Capable

Storage Controller
Before you begin, make a copy of this checklist. Fill in the appropriate areas as you
or the service representative perform the configuration tasks. This checklist
provides an important record of your actions. It may help you to diagnose any
problems that occur.

Most tasks in the checklist include references to other topics in this book. Refer to
these topics if you need more information about how to perform a particular task.
Table 105. Adding Compressed Disk Units to an Existing Storage Controller
Task What to Do Where to Read More About It
Task 1 Physically attach disk units to an existing
storage controller. This is normally done by a
service representative.
Task 2 Start DST or SST. “How to Start Dedicated Service Tools
(DST)” on page 660 or “Starting System
Service Tools (SST)” on page 662.
Task 3 Print your current disk configuration. “How to Display Your Disk Configuration”
on page 663.
Task 4 Include the disk units that you want to “How to Include a Disk Unit in Device
protect in device parity protection. Parity Protection” on page 705.
Task 5 Start disk compression on nonconfigured disk “How to Start Disk Compression” on
units. page 732.

738 OS/400 Backup and Recovery V5R1

Table 105. Adding Compressed Disk Units to an Existing Storage Controller (continued)
Task What to Do Where to Read More About It
Task 6 Add nonconfigured disk units to the correct “How to Add Disk Units to an Auxiliary
ASPs. You can add the disk units to an Storage Pool” on page 669.
existing user ASP or to a new ASP. If you are
adding disk units to an ASP that has
mirrored protection and the new disk units
do not have device parity protection, you
must add pairs of disk units that have equal
capacities.
Task 7 If you created a new ASP on your system “How to Change the Storage Threshold for
when you added disk units, the system set an Auxiliary Storage Pool” on page 672.
the storage threshold of the ASP to 90%. If
you want a different threshold, change it at
this time.
Task 8 If you created any new ASPs and you want “How to Start Mirrored Protection” on
those ASPs to have mirrored protection, start page 719.
mirrored protection now.
Task 9 Verify that your disk configuration is correct. “How to Display Your Disk Configuration”
on page 663.
Task 10 End DST or SST. “How to Stop Dedicated Service Tools
(DST)” on page 662 or “Stopping System
Service Tools (SST)” on page 662.

Moving Disk Units from the System ASP to a User ASP

This checklist shows the sequence of tasks that you use to move one or more disk
units from the System ASP to a User ASP. This checklist assumes that the disk
units are attached to a compression-capable subsystem storage controller.

Before you begin, make a copy of this checklist. Fill in the appropriate areas as you
or the service representative perform the configuration tasks. This checklist
provides an important record of your actions. It may help you diagnose any
problems that occur.

Attention: When you perform the tasks in this checklist, the system moves large
amounts of data. Make sure that you have saved your system
completely in the event that you need to recover from an error
situation.

Most tasks in the checklist include references to other topics in this book. Refer to
these topics if you need more information about how to perform a particular task.
Table 106. Moving Disk Units from the System ASP to a User ASP
Task What to Do Where to Read More About It
Task 1 Print your current disk configuration. “How to Display Your Disk Configuration”
on page 663.
Task 2 Calculate the space requirements for both the “How to Calculate Space Requirements for
source and target ASPs for the disk units. an Auxiliary Storage Pool” on page 681.
Task 3 Use option 21 from the Save menu to save “Saving your server with the GO SAVE
your entire system. command” on page 15.
Task 4 Start DST. “How to Start Dedicated Service Tools
(DST)” on page 660.

Chapter 28. Working with Disk Compression 739

Table 106. Moving Disk Units from the System ASP to a User ASP (continued)
Task What to Do Where to Read More About It
Task 5 Remove disk units that you plan to add to a “How to Remove a Disk Unit from an
different ASP. Auxiliary Storage Pool” on page 678.
Task 6 If you wish to use device parity protection, “How to Start Device Parity Protection for
start device parity protection (if necessary), an Input/Output Processor” on page 699 (if
and include the disk units that you want to necessary), and “How to Include a Disk Unit
protect. If you do not wish to use device in Device Parity Protection” on page 705.
parity protection, proceed to the next step.
Task 7 Start disk compression on nonconfigured disk “How to Start Disk Compression” on
units. page 732.
Task 8 Add nonconfigured disk units to the correct “How to Add Disk Units to an Auxiliary
ASPs. You can add the disk units to an Storage Pool” on page 669.
existing user ASP or to a new ASP. If you are
adding disk units to an ASP that has
mirrored protection and the new disk units
do not have device parity protection, you
must add pairs of disk units that have equal
capacities.
Task 9 If you created a new ASP on your system “How to Change the Storage Threshold for
when you added disk units, the system set an Auxiliary Storage Pool” on page 672.
the storage threshold of the ASP to 90%. If
you want a different threshold, change it at
this time.
Task 10 If you created any new ASPs and you want “How to Start Mirrored Protection” on
those ASPs to have mirrored protection, start page 719.
mirrored protection now.
Task 11 Verify that your disk configuration is correct. “How to Display Your Disk Configuration”
on page 663.
Task 12 End DST. “How to Stop Dedicated Service Tools
(DST)” on page 662.
Task 13 If necessary, move objects between ASPs. “Transferring Objects between Auxiliary
Storage Pools” on page 686.

Recovering from Error Codes

You may encounter SRC codes when working with disk compression. This section
discusses some causes of SRC codes, and how to respond to them. You may want
to refer to “Chapter 25. Working with Auxiliary Storage Pools” on page 669 for
information on moving disk units into and out of auxiliary storage pools (ASPs).

Recovering from SRC 6xxx 7051

You are receiving this message because the compressed device and the
compression input/output adapter (IOA) are not compatible.
1. Did you remove the disk unit from another system?
Yes No
↓ Go to step 6 on page 741.
2. Was it removed from the ASP of the other system?
Yes No
↓ Go to step 4 on page 741.

740 OS/400 Backup and Recovery V5R1

 3. Stop compression on the disk unit.
This ends the procedure.
4. Do you want to save the data that is on the disk unit?
Yes No
↓ Stop compression on the disk unit.
This ends the procedure.
5. Return the disk unit to its original system and IOA and perform the following:
a. Remove the disk unit from the ASP.
b. Stop compression on the disk unit.

Reinstall the disk unit to this system.

This ends the procedure.

6. The IOP or IOA that you are using is not compatible with the disk unit.
Do you want to save the data that is on the disk unit?
Yes No
↓ Stop compression on the disk unit.
This ends the procedure.
7. If you came here from another procedure, return there and follow the
procedure, otherwise contact your next level of support.
This ends the procedure.

Recovering from SRC 6xxx 7052

You are receiving this message because the counter that tracks the number of write
operations to this drive has reached 75% of its maximum value. This counter is
used to determine if data on the drive is current with data within IOA memory.
Since this counter is large, it is not necessary to stop compression for this drive
immediately, but you should do so during your next scheduled maintenance. If
this counter wraps, data can be lost.

To stop compression and restart compression on the disk drive to reset the write
count, perform the following:
1. Perform a Manual Mode IPL to DST. (Refer to “Dedicated Service Tools (DST)”
in the iSeries Service Functions for more information.)
2. To find the resource name of the disk drive with the problem, perform the
following:
a. Select the Use Dedicated Service Tools option.
b. Select the Start a service tool option.
c. Select the Hardware service manager option.
d. Select the Work with service action log option.
e. Select the time frame of the problem.
f. Record the resource name that is associated with the 6xxx 7052 entry in the
SRC column.
3. Remove the disk unit from the ASP.
4. Stop compression on the disk unit.
5. Start compression on the disk unit.
6. Add the disk drive back to the ASP from which it was removed.

Chapter 28. Working with Disk Compression 741

 This ends the procedure.

742 OS/400 Backup and Recovery V5R1

Chapter 29. Managing Auxiliary Storage Pools
Auxiliary Storage Pools (ASPs) separate disk units into logical subsets, which can
give you a number of advantages. Using ASPs helps protect your data. By
separating libraries, documents, or other objects in an ASP, you protect them from
data loss should a disk unit in a different ASP fail. You can protect each ASP
individually, indicating the value you put on that data.

Using ASPs also increases performance. You can place libraries or objects in an
ASP, allowing you to dedicate the disk units in the ASP exclusively for the use of
those objects. If you do extensive journaling, a dedicated disk unit for the journal
receiver can also improve journaling performance.

Note: Placing many active journal receivers in the same user ASP is not
productive. The resulting contention between writing to more than one
receiver in the ASP can slow system performance. For maximum
performance, place each active journal receiver in a separate user ASP.

You can improve system performance by using the ASP Trace and ASP Balance
features.

Changing ASP size may involve determining adequate disk storage, “How to Add
Disk Units to an Auxiliary Storage Pool” on page 669 and “How to Delete an
Auxiliary Storage Pool” on page 680.

If the amount of data in a storage pool increases, you may need to increase the size
of the storage pool. Conversely, if data in a storage pool decreases, you may want
to decrease the size of that storage pool and use the disk space elsewhere.

Changing the size of an ASP can mean adding a disk unit, removing a disk unit,
moving a disk unit, or deleting an ASP from the system. You must typically have
QSECOFR authority to access these tasks.

Working with ASP Trace and ASP Balance

There are three types of balancing action a user can choose:
v Capacity Balance
v Hierarchical Storage Management (HSM)
v Usage Balance

The balancing actions use the results of previous ASP traces to determine disk unit
usage. Therefore, an ASP balance will be more effective if you perform an ASP
trace first.

The following restrictions and considerations:

v The Usage Balance uses the results of previous ASP traces to determine the disk
unit usage. You must perform an ASP trace before you run a Usage Balance.
v The Heirarchical Storage Management (HSM) Balance uses the results of
previous ASP traces to determine the disk unit usage. You must perform an ASP
trace before you can run a HSM Balance.

© Copyright IBM Corp. 1997, 2000, 2001 743

 v The Heirarchical Storage Management (HSM) Balance requires a mixture of
compressed and non-compressed disk units in the ASP.
v The system will experience degraded performance during the Trace activity and
Balance activity.
v You can only use the Trace and Balance functions in Auxiliary Storage Pools
which contain more than one disk unit.
v You can add a non-configured disk unit to the configuration during the time a
Trace is running. In this case the system will automatically include the newly
configured disk unit in the trace activity.
v You can only run a single Trace activity or Balance activity at a given time to an
Auxiliary Storage Pool.
v The amount of improvement in system throughput that is achieved by running
the balance is dependent on many factors. This includes some of the following
items.
– The amount of mainstore.
– The number of processors.
– The level of activity running on the system.
– The amount of cache in the storage subsystem.
– The quantity of disk arms under each I/O Processor in each storage
subsystem.

Capacity Balance
The Capacity Balance function rearranges the data on all of the disk units within
an Auxiliary Storage Pool. It moves data so that each disk unit has an equal
percentage of used and unused space. This is useful when you add new units to
an Auxiliary Storage Pool. We want to avoid the situation where several disk units
contain the majority of the data and the newly added disk units contain very little
data. This situation leads to poor system performace. The balance function spreads
the data in the ASP evenly across all of the disk units.

Below is a display which shows the effects of using Capacity Balance. Before using
Capacity Balance the recently added Unit 4 contained very little data. The system
storage management allocates newly created data to the disk unit that has the
lowest percentage of capacity used. Thus, the system routes all new storage
allocations to the unit 4. If the system uses that newly created data frequently, a
potential bottleneck is created. The system directs all of the I/O operations to a
single disk unit instead of spreading it across all units in the ASP. The Capacity
Balance performed on the ASP allows the data to be evenly distributed across all
of the disk units in the ASP. This means that the distribution of future space
allocation on the disk units in the ASP is even across all of the disk units in the
ASP. That ensures that the I/O to those allocates are also evenly spread out
amongst the disk units instead of being concentrated on the newly added disk
unit.

Unit Before Capacity Balancing After Capacity Balancing

Disk Size % Used Disk Size % Used
1 1967 54.59% 1967 55.69%
2 1031 68.45% 1031 55.80%
3 1031 68.41% 1031 55.76%
4 1031 0.30% 1031 55.77%

744 OS/400 Backup and Recovery V5R1

 To start Capacity Balance via a CL command use the Start ASP Balance
(STRASPBAL) command. For example, if you want to start a capacity balance on
ASP 4 and have it run for 25 minutes, enter the following command: STRASPBAL
ASP(4) TYPE(*CAPACITY) TIMLMT(25).

If you want to end a Capacity Balance before the time limit requested is reached,
use the End ASP Balance (ENDASPBAL) command. For example, if you want to
end a running capacity balance on ASP 4, enter the following command:
ENDASPBAL ASP(4).

Hierarchical Storage Management (HSM) Balance

The high-use and low-use data on each disk unit in the ASP is redistributed. This
is done so that the high-use data resides on ’fast’ disk units and the low-use data
resides on compression disk units, which are typically slower than standard disk
units. The ASP selected for a HSM Balance must have a combination of
compression disk units and non-compression disk units. You can only run a HSM
Balance following a Trace ASP Balance. The Trace ASP Balance function monitors
the I/O activity on each of the disk units in the ASP to determine where the
high-use and low-use data resides.

Compression disk units have larger capacity, but are somewhat slower than
non-compression disk units. This is due to the overhead of compression and
decompression, and the variations in the length of the data that is written to disk.
Typically, data that is found on disk units has a wide range of access requirements.
The HSM balance function moves data that is accessed infrequently to compression
disk units. Disk compression makes infrequently accessed data available on-line at
a lower cost. System throughput improves when you move high-use data off of
compression disk units. Moving the low-use data to the large compression disk
units makes additional capacity available on the standard disk units so high-use
data can be allocated.

The Start ASP Balance (STRASPBAL) command is used to perform the HSM
Balance function. For example, if you want to run a HSM balance on ASP 4 for 25
minutes, enter the following command: STRASPBAL ASP(4) TYPE(*HSM)
TIMLMT(25).

If you want to end a HSM Balance before the time limit requested is reached, use
the End ASP Balance (ENDASPBAL) command. For example, if you want to end a
running HSM balance on ASP 4, enter the following command: ENDASPBAL
ASP(4).

Usage Balance
The Usage Balance attempts to balance the usage of the disk units in an Auxiliary
Storage Pool. The Usage Balance can only be done following a Trace ASP Balance.
The Trace ASP Balance function monitors the I/O activity on each of the disk units
in the ASP. It then determines where the frequently used and infrequently used
data resides. The Usage Balance function utilizes that trace information. It adjusts
the data on the disk units so future system activity will be more equally balanced
amongst the disk units in the ASP.

If the system determines that all of the disk units are of approximately equal
utilization, the balance will end very quickly. The Balance Usage function uses the
Trace information in its calculations. If the trace data is old, or if your applications
have changed to reference different data since the trace was run, the Usage Balance

Chapter 29. Managing Auxiliary Storage Pools 745

 could result in very little improvement in your system throughput. It might even
cause the throughput to decrease somewhat.

The Start ASP Balance (STRASPBAL) command is used to perform the Archive
Balance function. For example, if you want to start a usage balance on ASP 4 to
run for 25 minutes, enter the following command: STRASPBAL ASP(4)
TYPE(*USAGE) TIMLMT(25).

If you want to end a Usage Balance before the time limit requested is reached, use
the End ASP Balance (ENDASPBAL) command. For example, if you want to end a
running usage balance on ASP 4, enter the following command: ENDASPBAL
ASP(4).

ASP Trace
The Trace ASP Balance command monitors the frequency that data is accessed on
the disk units in the Auxiliary Storage Pool. Each I/O to the disk units is
monitored, and the results are recorded for use by the Balance commands. The
statistics which are collected are cumulative. For example, suppose you start one
Trace and it runs for 35 minutes. Then you start another trace on that ASP and it
runs for 15 minutes. The second group of statistics is added to the first collection
and the cumulative result is used to balance the ASP.

Select an Auxiliary Storage Pool that you want the system to monitor. The system
will record all of the I/O activity on the disk units in that ASP. For example to
start a trace on ASP 4 which will run for 35 minutes, enter the following
command: TRCASPBAL ASP(4) SET(*ON) TIMLMT(35).

If you want to end a trace before the time limit requested on the Start Trace is
reached, use the Trace ASP Balance (TRCASPBAL) command. For example, if you
want to end the trace on ASP 4, enter the following command: TRCASPBAL
ASP(4) SET(*OFF).

The collected statistics on each disk units I/O activity can be cleared by use of the
TRCASPBAL command. You can clear old trace data if you do not want that data
to be used in determining the locations of the high-use and low-use data on the
disk units in the ASP. Use the Trace ASP Balance (TRCASPBAL) command to clear
the trace data. For example, if you want to clear the trace data which has been
collected from ASP 4, enter the following command: TRCASPBAL ASP(4)
SET(*CLEAR).

Determining adequate disk storage

If you need to know how much disk storage is available on your system, use the
Work with System Status (WRKSYSSTS) command.
1. Type WRKSYSSTS at the command line. The Work with System Status display
appears.
2. Record the values that are shown for System ASP and % system ASP used.
3. Use those values in the formula that is shown below to calculate the available
free space on your system:
System ASP x (100 - % system ASP used) x 0.01 = Available disk storage

If the amount of storage is less than you need to complete your task, you must
create more disk space. You can create more space by adding additional disk units
or by cleaning your system of files and programs you no longer use.

746 OS/400 Backup and Recovery V5R1

 Chapter 30. Creating Logical Partitions
The following information tells you how to prepare for and create logical partitions
on your system.

When you have made all the planning decisions, you can start to create logical
partitions on your iSeries server. As with any major change in system
configuration, there are a number of prerequisites that you should do before you
make create any new logical partitions. Once you have done the prerequisites, you
can start the six tasks involved in partitioning your server.

| Starting with V5R1M0 you can use Operations Navigator to create logical
| partitions. For information on using Operations Navigator to create logical
| partitions, go to the Information Center at the following Web site:
| www.ibm.com/eserver/iseries/infocenter
|
Preparing for Logical Partitions
Your preparation needs are based on the system that you want to create logical
partitions on:

Type of System Action Needed

New System You do not need to do anything special before creating logical
partitions as long as you have not restored any user data to the system.
If there is user data on the system, back up all data before creating
logical partitions.
Existing System Back up all your data before creating any partitions.

You should back up all your system data and user data before creating any logical
partitions. You should also make two copies of the backup, or perform one full
backup, and duplicate the save media. Use option 21 on the save menu. You
should have two copies because you should always store one off site in case of a
disaster.

Use the following steps when you create logical partitions on a system that already
contains user data and system data:
1. Use option 21 on the Save menu to perform a backup. You can find the steps
for using option 21 in “Chapter 2. Saving your Server” on page 15.
2. After you complete your backups, continue with the steps for creating logical
partitions.

Hardware and Software Requirements for Logical Partitions

Minimum Software Requirements

Only Version 4 Release 4 (V4R4) and newer versions of the OS/400 operating
system support logical partitions. V4R4 is the earliest release that is supported in
any logical partition.

However, the 8xx servers can only support OS/400 V4R5 or later software releases
on all logical partitions.

© Copyright IBM Corp. 1997, 2000, 2001 747

 More information about release support is in Learning about logical partitions on
the Information Center.

Minimum Hardware Requirements

The 6xx, 7xx, 8xx, and Sxx n-way base servers support logical partitions. Servers
1xx, 2xx, 4xx, and 5xx do not support logical partitions.

For upgrade customers, the current physical placement of hardware may restrict
your configuration choices. For server-specific information, consult the Technical
information section on the Logical Partition Web site, and contact your business
partner, marketing representative, or service specialist.

Be sure that you have the right hardware and software for your server. The
following minimum hardware requirements are necessary for creating logical
partitions.
Table 107. Hardware requirements for logical partitions
Hardware Primary partition Each secondary partition
Requirements
Processor One processor One processor
Main Storage 256 MB 64 MB
Load source disk One disk unit in supported load One disk unit in supported
unit source location with dedicated load source location with
load source capable IOP dedicated load source capable
IOP
Console One system console and One system console and
appropriate dedicated controlling appropriate dedicated
IOP controlling IOP
Alternate load One optical or tape drive (both are One optical or tape drive that
source recommended) and the controlling is switchable between all
IOP secondary partitions and the
controlling IOP
Recommendations
Input/Output (I/O) A minimum of one bus is recommended for each partition to enable
Bus bus level I/O partitioning
Electronic customer One for the system (electronic customer support line is currently
support line shipped with the primary)

You can create a maximum of one logical partitions for each installed processor.
The server limit is 24 processors, so you can have a potential of 24 partitions on a
server.

To see examples of logical partitions, see the partitioning examples in the

Information Center.

Before You Partition Your System

It is absolutely essential that you complete the following steps first.
v Call your service provider to Perform a logical partition product assurance
check.
v Print a copy of the Logical Partition Planning Guide Worksheet. You can
download a copy from the Logical Partitioning Web site at the following URL:
http://www.ibm.com/eserver/iseries/lpar

748 OS/400 Backup and Recovery V5R1

 v Perform a full system backup before beginning to create partitions. See
“Chapter 2. Saving your Server” on page 15 for instructions.
v You should remove disk units from the current partition configuration if you
intend to use existing disk units in any secondary partition. Go to “Chapter 25.
Working with Auxiliary Storage Pools” on page 669.
v Print the PDF file for the Creating logical partitions in the Information Center.
Use the system resources worksheet (Table 1) to keep track of how you will
divide your system’s resources. Use the system resources worksheet together
with the resource printout that is generated later.

During the process of creating partitions you are frequently prompted to IPL the
whole system or a single partition. The following situations each require an IPL.
Alternatively, a single IPL can be initiated after you have made all adjustments.
v Primary partition (system) IPL.
– Changing minimum or maximum limits for any partition.
– Changing bus ownership for any partition.
– Starting or stopping virtual OptiConnect.
– Creating or deleting a partition.
v Secondary partition system IPL.
– Increasing or decreasing the number of processors.
– Increasing or decreasing the amount of allocated memory.
– Increasing or decreasing the percentage of interactive performance.

If you plan to use Operations Console on a secondary partition, you must have an
electronic customer support resource (asynchronous communication resource)
available on the same IOP.

Creating a Logical Partition

You will begin creating logical partitions on a system that is without partitions,
and powered down. If you are ready to begin creating logical partitions, print this
entire page. If you need additional system resource worksheets, photocopy the one
provided. Use the following checklist to track your progress through the
procedure.

If questions arise during the creation of logical partitions, you have the following
options:
v Press F1 for on-line help.
v Consult the Understanding logical partition error messages and reports page in
Troubleshooting logical partitions in the Information Center.
v Check the World Wide Web under www.ibm.com/eserver/iseries/lpar for
important information.
Note: If you will be using Operations Console on a secondary partition, you must
have an electronic customer support resource (asynchronous communication
resource) available on the same IOP.

Chapter 30. Creating Logical Partitions 749

 Table 108. System Resource Worksheet
Partitions Total Primary Secondary Secondary Secondary
system partition partition (1) partition (2) partition (3)

Number of
processors

min/max

Memory

min/max

Interactive
Workload

min/max

Interpartition
communication
(OptiConnect -
yes/no)

Interpartition
communication
(High Speed
Link-yes/no)

Release Level

Bus number
ownership

Load Source IOP

logical address
feature code
(bus/board/card)

Console IOP
(bus/board/card)

Alternate IPL IOP

(bus/board/card)

ECS line IOP

750 OS/400 Backup and Recovery V5R1

Logical partition checklist

1. Generate a system resource printout

Generate a resource printout using the Start System Service Tools (STRSST)
command. You should be signed on with service tools authority such as QSECOFR.
__ Step 1. At the OS/400 command line type STRSST and press enter.
__ Step 2. Select option 1 (Start a service tool)
__ Step 3. Select option 7 (Hardware service manager)
__ Step 4. At the Hardware Service Manager display, press F6 (Print
configuration).
__ Step 5. At the Print Format options display, select either Format - 1 (132
characters wide) and Sort by - 2 (Logical address).
In order to get a correctly formatted printout, your printer device must
be configured and able to accommodate a 132 character width line.
__ Step 6. Press Enter.
__ Step 7. Return to the OS/400 command line.
__ Step 8. Use the Work with Spooled Files (WRKSPLF) command to work with
the spooled file you just created (QPCSMPRT).
__ Step 9. Print the file.

2. Getting started
__ Step 1. You should read Prerequisites to creating logical partitions before
continuing to the next step.
__ Step 2. If you have created practice partitions, delete them by referring to the
Information Center page, Deleting all of your logical partitions.
__ Step 3. If your system is active, power down your system by typing
PWRDWNSYS OPTION (*CNTRLD) DELAY (600) and pressing Enter
at command line.
__ Step 4. Using the system resource print out, fill in the system resource
worksheet to record your resources. This worksheet will help you keep
track of the resource you remove from the primary partition and add
to the secondary partitions.
__ Step 5. Change the control panel to Manual mode.
Once you begin in Manual mode, remain there.
__ Step 6. Press the power on button.

3. Remove resources from the primary partition

In this section you will remove resources to create partitions. It will simplify
partition setup if you remove sufficient resources now to create all partitions.
__ Step 1. At the IPL or Install the System display, select option 3 (Use
Dedicated Service Tools (DST)).
__ Step 2. At the sign on display, sign on to DST with service tools authority
such as QSECOFR.
__ Step 3. Verify that all the logical hardware resources on the System Bus are
operational by doing the following steps:
a. Select option 7 (Start a service tool).
b. Select option 4 (Hardware service manager).

Chapter 30. Creating Logical Partitions 751

 c. Select option 2 (Logical hardware resources).
d. Select option 1 (System bus resources).
e. Do these steps for each IOP on the System Bus Resources display:
1) Use the cursor to select the first IOP that appears on the
display.
2) Select option 9 (Resources associated with IOP) and press
Enter.
3) Page down the display and verify that the Status field says
Operational for all devices that are associated with the IOP.
4) Press F12 to go back to the Logical Hardware Resource on
System Bus display.
5) Use the cursor to select the next IOP that appears on the
display and repeat steps 3e1 through 3e4.
f. If the status of any device is not Operational, take corrective
actions or contact your Authorized Service Representative.
g. If the status of all devices are operational, press F3 twice to return
to the Use Dedicated Service Tools (DST) display.
__ Step 4. Select option 11 (Work with system partitions) and press Enter.
__ Step 5. Select option 3 (Work with partition configuration), and press Enter.
__ Step 6. At the Work with Partition Configuration display, type option 2
(Change partition processing resources) on the line beside PRIMARY
(partition), and press Enter. If this is the first time you are creating
logical partitions on a new system, you should only see PRIMARY
displayed. If you had previously partitioned your system, other
partition names may be visible. Refer to your system resource
printout. Select the primary partition.
__ Step 7. At the Change Partition Processing Resources display, press F9 for a
limits list. From the system resource worksheet enter the following
information:
v New number of processors
– At least one processor must remain in each partition after
resources have been removed for the creation of new partitions.
v New size of main storage (MB)
– 256 MB of memory must remain in the primary logical partition
after removal of resources.
– A minimum of 64 MB of memory must be allocated to each
secondary logical partition created.
v Specify the percent of New interactive performance.
v A 1 (Yes), or 2 (No) beside the Connect to system Virtual OptiConnect
option.
You need one of the following licenses in order to use virtual
OptiConnect between partitions:
– A license to option 23 of the operating system (5769-SS1
OptiConnect).
– A license to Program Request for Price Quotation (PRPQ)
5799-FWQ Optimover.
v Press Enter. If you receive a message, minimum and maximum is
out of range, return to that selection and adjust your selection.
Press Enter to go to the next display.

752 OS/400 Backup and Recovery V5R1

__ Step 8. At the Confirm Changed Partition display, review your selections.
Press Enter to confirm your choices, or press F12 to go to the previous
display and change your selections.
__ Step 9. At the Work with Partition Configuration display, select option 4
(Remove I/O resources), and press Enter.
__ Step 10. Refer to your system resource worksheet, enter 1 (Remove resources), or
2 (Remove and clear hardware resource) beside each hardware resource
you want to select.
When you remove a bus, you also remove all resources that are
owned by that bus.
When removing a hardware resource from the system do the
following:
v Use option 1 (Remove resources) if the resource is to be returned to
the original partition at a later date and you want the same
resource names and device descriptors used.
v Use option 2 (Remove and clear hardware resource) if the resource is to
be permanently removed.
__ Step 11. Press Enter.
__ Step 12. At the Confirm Remove I/O Resources display, review your
selections.
__ Step 13. Press Enter to confirm your choices or F12 to go to the previous
display and change your selections.
If you receive a Logical Partitioning Error Report, Stop! Review the
error message entry, and take appropriate action to correct the
problems at once. Refer to Managing Logical Partitions in the
Information Center for system management concerns. Do not select
F10 or option 1 (Accept problem and continue). There will be no
follow-up warning, and no way to determine if you are completing
each step successfully. For further assistance, contact your IBM Service
specialist to determine the error and recovery.
__ Step 14. If the resource removal was successful, a message to that effect will
appear at the bottom of the display.
__ Step 15. Press F12 twice, or until you reach the Work with System Partitions
display.

4. Add resources

Beginning with this step, create a partition and add previously removed resources
to it. You create partitions one at a time. For each partition you want to create,
repeat this section.
__ Step 1. At the Work with System Partitions display, select option 5 (Create a
new partition).
__ Step 2. Press F9 for a list of limits.
__ Step 3. Provide the required information for:
v Partition name. (A partition name can be any word (except
PRIMARY), not exceeding 8 characters (A - Z and 0 - 9)).
v Number of partition processors.
v Size of main storage (MB).
v Partition interactive performance workload.
v Connect to system Virtual OptiConnect.
__ Step 4. Press Enter.

Chapter 30. Creating Logical Partitions 753

 If you have entered a value that is out of range, you will receive a
message to change your numbers. You can also adjust limits for the
following:
v Processors.
v Size of main storage (MB).
v Partition interactive performance.
__ Step 5. If you have had to make adjustments, press Enter after you finish.
The Add I/O Resources display is next.
__ Step 6. At the Add I/O Resources display, type either a 1 (Own dedicated), 2
(Own bus shared), or 3 (Use bus shared) beside your selected bus
resource. Type 1 beside your selected IOP resources on shared buses.
__ Step 7. Press Enter to go to the Confirm Add I/O Resources display.
If the expected resources are not on the list, return to the Add I/O
Resources display. Make sure that you have added the correct
resources.
__ Step 8. Press Enter to confirm your selection and go to Select Load Source
Resource display. Press F12 to return to the previous display if you
wish to change your choices. If your selection was successful, a
confirmation notice will appear at the bottom of the display.
__ Step 9. At the Select Load Source Resource display, type 1 (Select IOP) next to
the desired resource.
If the expected resource is not on the list, return to the Add I/O
Resources display. Make sure that you have added the correct
resources. If the expected resource is already on the Add I/O
Resource display list, consult the iSeries Logical Partitions Hardware
Planning Guide for hardware requirements.
__ Step 10. Press Enter to go to the Confirm Load Source Resource display and
review your selection. Press Enter again to confirm your selection, or
press F12 to return to the previous display and change your choice. If
your selection was successful, a confirmation notice will appear at the
bottom of the display.
__ Step 11. At the Select Console Resources display, select your
workstation-capable IOP. Type 1 (Select console IOP), or 2 (Select
alternate console IOP) .
If the expected resource is not on the list, return to the Add I/O
Resource display. Make sure that you have added the correct
resources.
If you are using Operations Console as your secondary partition’s
console, use the F9 option to disable automatic filtering. Select the
appropriate asynchronous communications IOP as the console.
__ Step 12. Press Enter to go to Confirm Console Resource display. Review your
selection.
__ Step 13. Press Enter again to confirm your selection and go to the Confirm
New Partition display.
__ Step 14. If the partition does not have a dedicated alternate IPL device, skip
ahead to step 20 on page 755.
__ Step 15. Press F11 (Select alternate IPL resource).
__ Step 16. Type 1 (Select IOP) next to the desired alternate IPL device.
If the expected resource is not on the list, return to Add I/O Resource
display. Make sure that you have added the correct resources.
__ Step 17. Press Enter to go to Confirm Alternate IPL Resource display.

754 OS/400 Backup and Recovery V5R1

__ Step 18. Review your selections.
__ Step 19. Press Enter to confirm and return to Confirm New Partition display.
__ Step 20. Review your selections.
(An asterisk will appear beside the load source IOP you have selected.
Other special characters will appear identifying previously selected
resources, < for the console, % for the alternate IPL).
__ Step 21. Press Enter to go to Work with System Partitions display. If your
selection was successful, a confirmation notice will appear at the
bottom of the display.
__ Step 22. Mark the console IOP as the electronic customer support IOP if you
are using Operations Console on your partition. Use option 3 (Work
with partition configuration). Select the new partition by using option
9 (Select default electronic customer support resource). Choose the
appropriate IOP by entering a 1. A confirmation message should
appear. Now use F12 to return to the Work with System Partitions
display
__ Step 23. Here you have the following options:
v If this is the last partition you intend to create, proceed to Restart
the system below. You must perform a system restart to activate all
changes for your new partition.
v If you want to create additional partitions, and have removed
sufficient resources, select option 5 (Create a new partition). Repeat
the Add resources section. After you have completed the creation of
your last partition, proceed to Restart the system below, and
activate all changes for your new partitions.
v If you want to create additional partitions, and have not kept
sufficient resources for the creation process, see the Removing I/O
resources in the Information Center for instructions. You must
return to a partition (primary or secondary), with sufficient
resources and change, or remove some before you can create a new
partition. Once you have available resources, repeat the Add
resources section. After you have completed the creation of your
last partition, proceed to Restart the system below, and activate all
changes for your new partitions.

5. Restart the system

Beginning with this step, perform a system restart (IPL) and activate all changes
for your new partitions. If there are any existing secondary partitions active, make
sure they are powered down prior to performing a system restart.

To perform a system-wide restart, press F10 and Enter to confirm your choice.

All newly created secondary partitions will be in the powered off state, IPL source
set to D, IPL Mode set to Manual, and System IPL Action set to hold.

You must now install your Licensed Internal Code and your OS/400 operating
system on your secondary partitions:
__ Step 1. Sign on to DST or SST with service authority such as QSECOFR.
__ Step 2. Place the Licensed Internal Code media in the alternate IPL device for
the new secondary partition.
__ Step 3. Perform a D mode IPL for the new secondary partition:
a. From the DST menu select option 11 (Work with system partitions).

Chapter 30. Creating Logical Partitions 755

 b. Selection option 2 (Work with partition status).
c. At the Work with Partition Status display, use your cursor to select
the secondary partition that you want to IPL.
d. Select option 1(Power on).

Refer to the Software Installation book for instructions on installing software.

Refer to Managing Logical Partitions for system management.

6. Print the new logical partition configuration

__ Step 1. From an OS/400 command line in the primary partition, type STRSST.
__ Step 2. Select option 5 (Work with system partitions).
__ Step 3. Select option 1 (Display partition information).
__ Step 4. Select option 3 (Display allocated I/O resources).
__ Step 5. Press F6 to print.
__ Step 6. At the Print Format options display, select either Format - 1 (132
characters wide), or 2 (80 characters wide), depending on your default
option.
__ Step 7. Press Enter to print to a spooled file.
__ Step 8. Press F12 to return to the Display Partition Information display.
__ Step 9. Select option 2 (Display partition processing configuration) and press
Enter.
__ Step 10. Press F6 to print.
__ Step 11. At the Print Format options display, select either Format - 1 (132
characters wide), or 2 (80 characters wide), depending on your default
option.
__ Step 12. Press Enter to print to a spooled file.
__ Step 13. Return to an OS/400 command line and print these two spooled files.

756 OS/400 Backup and Recovery V5R1

Part 9. Backup and Recovery Tools and Techniques
Chapter 31. Techniques and Programming
Examples for Backup and Recovery . . . . . 759
Job Recovery Considerations . . . . . . . . 759
Interactive Job Recovery . . . . . . . . . 759
Batch Job Recovery . . . . . . . . . . 760
Considerations for Using Save Files . . . . . . 761
Using Control Language (CL) Commands for
Save Files . . . . . . . . . . . . . 762
Save File Security . . . . . . . . . . . 762
Opening a Save File . . . . . . . . . . 763
Input and Output Operations on a Save File . . 763
File-Dependent Attributes for a Save File . . . 764
Damage to a Save File . . . . . . . . . 764
Clearing a Save File . . . . . . . . . . 764
Sending Network Files . . . . . . . . . 764
Programming Examples for Backup and Recovery 765
Retrieving the Device Name from Save
Completion Messages . . . . . . . . . 765
Displaying Status Messages When Saving . . . 766
Using the Retrieve Journal Entry (RTVJRNE)
Command in a Program . . . . . . . . . 766
CL Program to Handle Escape Conditions . . . 767
Writing output to save media using the receive
journal entry command . . . . . . . . . 769
Considerations When Writing to Tape . . . 770
Journal Entries Written to an ICF File . . . 772

© Copyright IBM Corp. 1997, 2000, 2001 757

758 OS/400 Backup and Recovery V5R1
 Chapter 31. Techniques and Programming Examples for
Backup and Recovery
This chapter includes several different techniques that you can use to assist with
and manage your backup and recovery processes.

Job Recovery Considerations

Job recovery and starting again should be a basic part of application design.
Applications should be designed to handle:
v Unexpected data problems, such as alphabetic data occurring where numeric
data is expected
v Operator problems, such as operators taking the wrong option or canceling the
job
v Equipment problems, such as workstation, disk unit, and communication line
failures

Job recovery procedures should ensure the integrity of the user’s data and allow
for easy starting of the interrupted application. Journaling and commitment control
can be used in application design to help in job recovery. Recovery procedures
should be transparent to the end users.

Interactive Job Recovery

If you are running a data entry job or one that updates a single file, it is unlikely
that you may need to plan an extensive recovery strategy. The operators can
inquire against the file to determine which record was last updated and then
continue from that point.

To recover from inquire-only jobs, the workstation operators simply start where
they left off. When using update transactions for many files, consider using a
journal or commitment control. The system automatically recovers journaled files
during the initial program load (IPL) following an abnormal end of the system. In
addition, the journal can be used for user-controlled forward or backward file
| recovery. There are other object types in addition to database physical files that
| you can protect with journaling. See “Chapter 19. Planning and Setting Up
| Journaling” on page 383 for more information.

Commitment control, using the file changes recorded in the journal, provides
automatic transaction and file synchronization. During job end, the system
automatically rolls back file updates to the beginning of the transaction. In
addition, the commitment control notify object can assist you in restarting the
transaction.

When designing an interactive application, consider the possibility that you can
experience equipment problems with your workstations and communications lines.
For example, suppose your computer system loses power. If you have an
uninterruptible power supply installed to maintain power to the processing unit
and disk units, your system remains active. However, in this example, your
workstations lost power. When your programs attempt to read or write to the

© Copyright IBM Corp. 1997, 2000, 2001 759

 workstations, an error indication is returned to the program. If the application is
not designed to handle these errors, the system can spend all its time in
workstation error recovery.

You should design your interactive applications to look at error feedback areas and
handle any errors indicated. If the application handles the errors and stops, the
system resource is not used to do nonproductive error recovery. Examples of using
error feedback areas and error recovery routines can be found in the programming
languages reference manuals.

Batch Job Recovery

Print-only batch jobs normally do not need special recovery to start again. Running
the program again may be adequate.

Batch jobs that perform file updates (add, change, or delete actions) present
additional considerations for starting again and recovery. One approach to starting
again is to use an update code within the record. As a record is updated, the code
for that record can also be updated to show that processing for that record is
complete. If the job is started again, the batch program positions itself (as a result
of the update code) to the first record that it had not processed. The program then
continues processing from that point in the file.

Another way to start batch processing again is to save or copy the file before
starting the job. You can use one of the following commands to save or copy the
file:
v Save Object (SAVOBJ)
v Copy File (CPYF)

Then, if you have to start again, restore or copy the file to its original condition
and run the job again. With this approach, you need to ensure that no other job is
changing the files. One way to ensure this is to get an exclusive lock on the file
while the job is running. A variation of this approach is to use the journal. For
example, if starting again is required, you could issue the Remove Journal Change
(RMVJRNCHG) command to remove changes to the files. Then, run the job again
against the files.

If your batch job consists of a complex input stream, you probably want to design
a strategy for starting again into the input stream. Then, if the batch job needs to
be started again, the job determines from what point the stream continues.

Commitment control also can be used for batch job recovery. However, if you plan
to use commitment control for batch jobs, consider that the maximum number of
record locks allowed in a commit cycle is 4 000 000. Therefore, you may need to
divide the batch job into logical transactions. For example, if your batch program
updates a master file record followed by several detail records in another file, each
of those sets of updates can represent a logical transaction and can be committed
separately. Locks are held on all records changed within a commit cycle. Therefore,
changed data is made available more quickly if your batch job is divided into
small, logical transactions.

| Journaling can also be used to assist in batch job recovery just as it can be for
| interactive jobs. See “Chapter 19. Planning and Setting Up Journaling” on page 383
| for more information.

760 OS/400 Backup and Recovery V5R1

 Considerations for Using Save Files
Using a save file allows you to save and restore objects without first placing save
media into your save media device. You can also use a save file to send objects
from one iSeries server to another over communications lines. You can use the save
file as an online container to save the contents of a single library to run overnight.
The next day, save the contents of the save file to storage media with the Save
Save File Data (SAVSAVFDTA) command. Objects saved to a save file can be
restored directly from save media, using the RSTLIB or RSTOBJ command.

Among the items to consider for using save files are:

v Performance
v Storage capacity
v Preparing save files for use

You cannot save or send a save file that is larger than the target release allows.

Performance When Using Save Files: Performance can vary, depending on other
disk activity. Save files can be created on or moved to an ASP for improved
performance and additional protection from system disk device failures.

| Save File Storage Capacity: The maximum capacity of a save file is about one
| terabyte. You can specify the maximum size of the save file on the Create Save File
(CRTSAVF) command.

Specify data compression on the save commands to reduce the space for the save
file and the amount of media needed for the SAVSAVFDTA command. (Data
compression is not an option on the SAVSAVFDTA command.)

Preparing Save Files for Use: When saving to a save file that already contains data,
use the Clear Save File (CLRSAVF) command or specify CLEAR(*ALL) on the save
command, or reply to an inquiry message sent during the save operation.

Saving the Save File Data: There are two ways to save the save file data:
v With the SAVSAVFDTA command, only the data is saved. The description of the
save file object is not saved. The save date and time of the save file are not
updated.
v When you use the Save Object (SAVOBJ) or the Save Library (SAVLIB) command
with SAVFDTA(*YES) specified, both the object description and the data are
saved. The save date and time are updated for the save file.

| Note: While you are saving save file data, other jobs cannot use the save file
| until the save operation completes unless you are using the
| save-while-active function.

Determining the Contents of a Save File: You can use the Display Save File
(DSPSAVF) command or the List Save File API to determine the contents of a save
file.

The DSPSAVF command displays the contents of a save file. The information
includes a description of each object saved and summary information.

The List Save File (QSRLSAVF) API returns the contents of the save file in a user
space. The contents of the save file is returned at a user-selected level of library
information, object information, or member information. The QSRLSAVF API

Chapter 31. Techniques and Programming Examples for Backup and Recovery 761
 returns the same information that is shown on a DSPSAVF command. In addition,
when you specify the SAVF0200 format, the system includes the following:
v The serial number of the system on which the save operation was performed.
v The ASP from which the object was saved.
For more information about the QSRLSAVF API, click the Programming topic in
the Information Center at the following Web site:
http://www.ibm.com/eserver/iseries/infocenter

| The QSYSINC library provides structures for the SAVF0100, SAVF0200, and
| SAVF0300 formats in C, COBOL, and RPG.

Using Control Language (CL) Commands for Save Files

Use the following CL commands with save files:
v The Create Save File (CRTSAVF) command creates a save file that can be used
with save and restore commands to store data. The save file stores data that
would otherwise be written to save media. A save file can also be used as a
container to send objects to another iSeries user on the systems network
architecture distribution services (SNADS) network.
v The Change Save File (CHGSAVF) command changes one or more of the
attributes of a save file, such as the maximum number of records.
v The Override Save File (OVRSAVF) command overrides or replaces certain
attributes of a save file, or overrides any file with a save file.
v The Display File Description (DSPFD) command displays the attributes of the
save file.
v The Clear Save File (CLRSAVF) command clears the contents of a save file.
v The Display Save File (DSPSAVF) command displays the save and restore
information in a save file, or the contents of the save file.
v You can use the Save Object (SAVOBJ) or the Save Library (SAVLIB) command
to save the description of the save file. You can also save the data to tape,
diskette, optical media, or another save file in a different library.
v The Save Save File Data (SAVSAVFDTA) command writes the contents of a save
file to either tape, optical media, or diskette.

Several commands used for save and restore operations also apply to save files.

Save File Security

The authority you grant for a save file is the same as for any file. Be careful when
granting authority for save files. The authority you grant to the save file allows
access to objects in the save file. For example, the same file can be read from and
written to by a high-level language program. The authority you grant for a
particular save file should depend on what objects are in the file.

Consider the following factors when granting authorities to save files:

v A user with use (*USE) authority can read records and restore objects from the
save file. This user can save the contents of the save file to tape, optical media,
or diskette.
v A user with use (*USE) and add (*ADD) authority can write records and save
objects in a save file.
v A user with object operational (*OBJOPR) and object management (*OBJMGT)
authority can clear the contents of a save file using the CLRSAVF command. The
clear operation is required first when replacing existing records in a save file.

762 OS/400 Backup and Recovery V5R1

 v A user with either save system (*SAVSYS) special authority or object existence
(*OBJEXIST) authority for the file can save the description and contents.

| Digital signature for a save file

| Starting with V5R1M0 you can put a digital signature on a save file. The system
| verifies any digital signatures present on the save file each time you display the
| save file or use the save file in a restore operation. If the signature is not valid you
| cannot display or use the save file in a restore operation. The Verify Object on
| Restore (QVFYOBJRST) system value does not affect the verification of save files.
| Therefore the system verifies the signature every time you display the save file or
| use the save file in a restore operation.

| For more information about digital signatures, click the Security topic in the
| Information Center at the following Web site:
| http://www.ibm.com/eserver/iseries/infocenter

| Opening a Save File

The following considerations apply when opening a save file:
v Parameters specified in the file are overridden with parameters specified in the
program. Program-specified parameters are overridden with parameters on the
OVRSAVF command.
v The fewer parameters specified in the program and on the OVRSAVF command,
the faster the file is opened.
v If no record length is specified in the high-level language program that opens
the file, a length of 528 bytes is assumed. If the program specifies a record
length value, it must be 528.
v For an input file, if the relative record number specified for the parameter does
not exist in the file, an error message is sent and the file is not opened.
v If you do not specify EXTEND(*YES) for a save file opened for output, the save
file must be empty. If the save file is not empty, you receive an inquiry message
to clear the file or cancel the operation. The file can only be cleared if the job
trying to open the file has operational and object management authority. You can
use the CLRSAVF command before the file is opened to ensure that it is empty
and to avoid the inquiry message.
v If a save file is open for output or is being used in a save or restore operation,
another program or job cannot open it.
v A save file can be opened by more than one program or job for input. While it is
in use, the save file cannot be opened for output.

Input and Output Operations on a Save File

The following considerations apply to input and output operations on a save file:
v Records are always read and written sequentially. The records read from a save
file contain sequence and parity information that is validated when the records
are written into another save file. This information ensures that the records are
processed in sequence and have not been changed.
You cannot write a record that has changed since it was retrieved from another
save file. You cannot write a record that is not the next record in sequence. If
you attempt either or these, an escape message is sent to report the error.
v A read of records from the save file can be done only if the entire file has been
written.
v The force-end-of-data (FEOD) function is valid for both input and output.

Chapter 31. Techniques and Programming Examples for Backup and Recovery 763
 For an input file, FEOD signals end-of-file to the program that does the
operation.
To ensure buffered output records are not lost after an FEOD operation
completes, they are written to the file. For an output file, buffered output
records are not lost even if the job or system fails.

File-Dependent Attributes for a Save File

Following are the file-dependent attributes for a save file:
v The following file-dependent attributes apply when the save file is open:
– For input operations, the first record returned for a read operation is the one
specified by the parameter POSITION when the file is opened. After the first
record is read, all remaining records are returned sequentially to the end of
the file.
– For output operations, new records can be added to the end of records
already in the file (specified using the EXTEND parameter). Each save file
record contains sequencing information used by the system to ensure that a
record is not skipped or written more than once.
– If no record length is specified in the high-level language program that opens
the file, a length of 528 bytes is assumed. If the program specifies a record
length value, it must be 528 bytes.
v No file-dependent parameters (such as format name) can be specified for read or
write operations with a save file. Any file-dependent parameters specified are
ignored.

Damage to a Save File

A save file is marked partially damaged if an attempt to read a record or restore an
object from the file encounters an auxiliary storage error. You can restore objects
from a partially damaged save file other than the objects on the damaged part of
auxiliary storage. The objects on the damaged portion of the auxiliary storage
within the save file cannot be restored. When a file is marked partially damaged,
you cannot add more records to it until it is cleared.

Partial damage of the save file itself can occur that is unrelated to auxiliary storage
errors. Sometimes a partial damage message is issued during a SAVSAVFDTA
when the system is very busy. This can happen because an internal operation did
not complete within a given time interval. It is most often seen when the
SAVSAVFDTA job is running at a low priority and there is a heavy interactive load
on the system. Although a SAVSAVFDTA can no longer be done from that save
file, the objects in the SAVF can be restored to the system using RSTOBJ.

Clearing a Save File

Clearing a save file releases the auxiliary storage allocation to reduce the size of
the object, and resets the partial damage mark on the file if it was previously
marked damaged. A save file is cleared in one of two ways:
v By using the CLRSAVF command prior to the save operation.
v By specifying CLEAR(*ALL) as the first part of a save operation to the save file.
If CLEAR(*ALL) is not specified and emptied, a message is sent asking whether
or not you want to clear the file.

Sending Network Files

The only objects you can send with the Send Network (SNDNETF) command are
database file members or save files. The SNDNETF command creates a save file

764 OS/400 Backup and Recovery V5R1

 and copies the information into it. Once the file has been received using the
Receive Network File (RCVNETF) command, the copy on the source system is not
saved. Consider backing up the information on the remote system.

Other objects (such as programs or commands) must be saved in a save file before
they can be sent using the SNDNETF command.

Note: Do not use save files to save objects on a system at the current release to
distribute them to a system at a previous release unless TGTRLS(*PRV) is
specified on the save command. You may also specify TGTRLS(VxRxMx) on
the save command, where (VxRxMx) is the previous-release-value. The
current release to previous release rules still apply.

Programming Examples for Backup and Recovery

Following are several examples of programs for working with backup and
recovery.

Retrieving the Device Name from Save Completion Messages

The CL program in Figure 86 retrieves the device name from the CPC3701 message
(found in positions 126 through 135 of the message data) and uses the information
to determine which device is used by the next save command.

SEQNBR *... ... 1 ... ... 2 ... ... 3 ... ... 4 ... ... 5 ... ... 6 ... ... 7

1.00 PGM
2.00 DCL &MSGDATA *CHAR LEN(250)
3.00 DCL &MSGID *CHAR LEN(7)
4.00 DCL &DEV *CHAR LEN(10)
5.00 DCL &DEV1 *CHAR LEN(10) VALUE(TAP01)
6.00 DCL &DEV2 *CHAR LEN(10) VALUE(TAP02)
7.00 SAVLIB LIB(LIB1) DEV(&DEV1 &DEV2) ENDOPT(*LEAVE)
8.00 L00P: RCVMSG RMV(*NO) MSGDTA(&MSGDATA) MSGID(&MSGID)
9.00 IF (&MSGID *NE CPC3701) GOTO L00P /* Compltn */
10.00 CHGVAR &DEV %SST(&MSGDATA 126 10) /* Device name */
11.00 IF (&DEV *EQ 'TAP01') DO /* Last was TAP01 */
12.00 CHGVAR &DEV1 'TAP01' /* Set for first device */
13.00 CHGVAR &DEV2 'TAP02' /* Set for second device */
14.00 ENDDO /* Last was TAP01 */
15.00 ELSE DO /* Last was not TAP01 */
16.00 CHGVAR &DEV1 'TAP02' /* Set for first device */
17.00 CHGVAR &DEV2 'TAP01' /* Set for second device */
18.00 ENDDO /* Last was not TAP01 */
19.00 SAVLIB LIB(LIB2) DEV(&DEV1 &DEV2) /* Save Lib 2 */
20.00 ENDPGM

Figure 86. Program for Save Completion Messages

If any objects cannot be saved, the operation attempts to save remaining objects
and sends an escape message (CPF3771 for single libraries, CPF3751/CPF3778 for
more than one library, and CPF3701 for save operations to save files) stating how
many objects were saved and how many were not. To continue with the next
library, the Monitor Message (MONMSG) command must be used to handle the
escape condition. The format of the message data for the CPF3771 message is
similar to the CPC3701 message and also identifies the last device used.

The SAVCHGOBJ command operates in a similar manner, but uses CPC3704 as a

completion message, CPF3774 as an escape message for single libraries, and

Chapter 31. Techniques and Programming Examples for Backup and Recovery 765
 CPC3721 or CPF3751 for multiple libraries. For save operations to save files, these
messages are CPC3723 as a completion message and CPF3702 as an escape
message. These messages also contain the last device or save file used in the
message data.

Displaying Status Messages When Saving

Figure 87 shows a program that sends a message to the external (*EXT) program
message queue if any objects cannot be saved.

PGM /* SAVE SOURCE */

SAVLIB LIB(SRCLIB) DEV(TAPE01) PRECHK(*YES)
MONMSG MSGID(CPF0000) EXEC(DO)

SNDPGMMSG MSG('Objects were not saved - Look at the job +

log for messages') TOPGMQ(*EXT)
SNDPGMMSG MSG('SRCLIB library was not backed up') +
TOPGMQ(xxxx)
RETURN
ENDDO
ENDPGM

Figure 87. Program for Saving Source Files

Using the Retrieve Journal Entry (RTVJRNE) Command in a

Program
Use the Retrieve Journal Entry (RTVJRNE) command in a control language
program to retrieve a journal entry and place it in variables in the program. You
can retrieve the following:
v Sequence number
v Journal code
v Entry type
v Journal receiver name
v Library name for the journal receiver
v Journal entry-specific data

For example, you can use this command to automate your recovery procedures or
to change the journal receivers and then save them.

In Figure 88 on page 767, the RTVJRNE command determines when job

000666/QPGMR/WORKST01 last opened file ORDENTP:

766 OS/400 Backup and Recovery V5R1

 PGM
DCL &SEQ#; TYPE(*DEC) LEN(10 0)
DCL &JRNE TYPE(*CHAR) LEN(200)
DCL &DATE TYPE(*CHAR) LEN(6)
DCL &TIME TYPE(*CHAR) LEN(6)
RTVJRNE JRN(DSTJRN/JRNLA) FILE(DSTPRODLIB/ORDENTP) +
RCVRNG(DSTJRN/RCV30 DSTJRN/RCV27) FROMENT (*LAST) +
TOENT(*FIRST) SEARCH(*DESCEND) +
JRNCDE(F) ENTTYP(OP) JOB(000666/QPGMR/WORKST01) +
RTNSEQNBR(&SEQ#); RTNJRNE(&JRNE)
CHGVAR &DATE (%SST(&JRNE 19 6))
CHGVAR &TIME (%SST(&JRNE 25 6))
ENDPGM

Figure 88. Program for Retrieving Journal Entries

“Appendix D. Journal Entry Information” on page 805 describes the contents of the
journal entry types.

CL Program to Handle Escape Conditions

| You normally use the APYJRNCHG command to perform file recovery. However,
| usable journal receivers are needed to successfully use this command. If usable
| journal receivers are not found, an escape message is sent. Figure 89 on page 768
demonstrates how this escape condition can be handled in a CL program by
| prompting for restoration of the required receiver. This example uses database files.
| You can extend this example to use the APYJRNCHG command with all object
| types that support journaling.

Chapter 31. Techniques and Programming Examples for Backup and Recovery 767
FILERECOV: PGM
.
.
APYJRNCHG JRN(JRNLIB/JRNA) FILE((LIBA/FILEA)) +
RCVRNG(RCVLIB/RCV1 *CURRENT)
MONMSG MSGID(CPF7053 CPF9801) +
EXEC(CALL PGM(FIXLIB/RSTRCV) PARM(FILERECOV))
.
.
ENDPGM
.
.
RSTRCV: PGM PARM(&PGMNM)
/* Recover a nonexistent or unusable receiver */
/* in RCVRNG by prompting for a restore of */
/* receiver.
DCL *PGMNM TYPE(*CHAR) LEN(10) /* name of program */
/* calling RSTRCV */
/* that received */
/* CPF7053 or */
/* CPF9801 */
DCL &MSGDATA TYPE(*CHAR) LEN(22) /* variable for */
/* CPF7053 or */
/* CPF9801 */
DCL &MSGDID TYPE(*CHAR) LEN(7) /* escape message */
/* ID */
DCL &RCVNAME TYPE(*CHAR) LEN(10) /* name of */
/* receiver to */
/* restore */
DCL &RCVLIB TYPE(*CHAR) LEN(10) /* library name */
/* of receiver to */
/* restore */
DCL &RCODE TYPE(*CHAR) LEN(2) VALUE(x'0001')
/* reason code 1 of CPF7053 */
RCVMSG PGMQ(*SAME &PGMNM) MSGTYPE(*EXCP) WAIT(0) +
RMV(*NO) MSGDTA(&MSGDATA) MSGID(&MSGID)

Figure 89. Example Program Prompts for Restoring the Required Receiver For an APYJRNCHG (Part 1 of 2)

768 OS/400 Backup and Recovery V5R1

 IF COND(&MSGID *EQ 'CPF9801') THEN(DO) /* CPF9801 occurred */
CHGVAR &RCVNAME(%SST(&MSGDATA 1 10)) /* get receiver */
/* from message */
/* data */
CHGVAR &RCVLIB (%SST(&MSGDATA 11 10)) /* get library */
/* name from */
/* message data */
? RSTOBJ OBJ(&RCVNAME) SAVLIB(&RCVLIB) OBJTYPE(*JRNRCV)
/* display RSTOBJ prompt */
ENDDO
ELSE DO
IF COND((&MSGID *EQ 'CPF7053') & (%SST(&MSGDATA 1 2) +
*EQ &RCODE)) THEN(DO) /*CPF7053 RC(1) occurred*/
CHGVAR &RCVNAME (%SST(&MSGDATA 3 10)) /* get receiver */
/* name from */
/* message data */
CHGVAR &RCVLIB (%SST(&MSGDATA 13 10)) /* get library */
/* name from */
/* message data */
? RSTOBJ OBJ(&RCVNAME) SAVLIB(&RCVLIB) OBJTYPE(*JRNRCV)
/* display restore prompt */
ENDDO
ELSE
.
.
ENDDO
ENDPGM

Figure 89. Example Program Prompts for Restoring the Required Receiver For an APYJRNCHG (Part 2 of 2)

Writing output to save media using the receive journal entry

command
| Note: You may want to consider using the remote journal function to transfer your
| journal receiver data to another server instead of the following. See
| “Chapter 21. Remote Journal Function” on page 467 for more information.

Figure 90 on page 770 shows an RPG program that is being used as the exit
program for the Receive Journal Entry (RCVJRNE) command. This example writes
output onto tape media. See “Journal Entries Written to an ICF File” on page 772
for a discussion of changing the sample to write output to an OS/400-ICF file. See
“Using the Receive Journal Entry Command” on page 432 for a discussion of how
to use the RCVJRNE command.

Chapter 31. Techniques and Programming Examples for Backup and Recovery 769
SEQNBR *... ... 1 ... ... 2 ... ... 3 ... ... 4 ... ... 5 ... ... 6 ... ... 7

1.00 FTAPE O F 300 SEQ

2.00 IJRNENT DS 300
3.00 I 1 5OJOENTL
4.00 C *ENTRY PLIST
5.00 C PARM JRNENT
6.00 C PARM CALLCD 1
7.00 C CALLCD IFEQ '1' Entry rcvd
8.00 C* Ensure journal entry is not being truncated
9.00 C JOENTL CABGT300 RETURN H1 If GT output
10.00 C ADD 1 OUTRCD 70 Bump ctr
11.00 C EXCPTOUTPUT Output
12.00 C END Entry rcvd
13.00 C CALLCD IFEQ 'O' Rdy to wait
14.00 C EXSR FORCE Force out
15.00 C END Rdy to wait
16.00 C SHTDN 31 Test shtdwn
17.00 C 31 DO If shtdwn
18.00 C EXSR FORCE Force out
19.00 C MOVE '9' CALLCD Set to end
20.00 C SETON LR Set LR
21.00 C END If shtdwn
22.00 C RETURN TAG Return tag
23.00 C RETRN Return
24.00 C FORCE BEGSR Force out
25.00 C OUTRCD IFNE *ZERO If rcds
26.00 C FEOD TAPE FEOD
27.00 C Z-ADDO OUTRCD Reset
28.00 C END If rcds
29.00 C ENDSR End subr
30.00 OTAPE E OUTPUT
31.00 0 JRNENT

Figure 90. Program for writing RCVJRNE output to save media

Considerations When Writing to Tape

A separate job must be in continuous operation and dedicated to converting entries
to tape. Before issuing the RCVJRNE command, your job should issue an
OVRTAPF command, specifying fixed-length blocked records, to direct the RPG file
TAPE to a tape device.

You should not consider this approach with a streaming tape device. A user
auxiliary storage pool (ASP) is a preferred solution instead of a tape. However, this
approach is similar to writing journal entries to a communications line.

The RPG program is written assuming that the largest journal entry to be passed is
300 bytes. This is the size that is given to the data structure JRNENT. It allows a
record size of 175 bytes plus the 125 bytes of journal entry identifier information
and qualifier information. A check is made in the program to ensure that the
record image is not being truncated:
v If a code of 1 is passed from the RCVJRNE command, the program ensures that
the journal entry does not exceed 300 bytes. If it does, the program sets on the
H1 indicator and returns. The program adds 1 to the counter and writes the
record to the tape output file. Because this is an output-only file, RPG
automatically blocks the records within the RPG program.
When full, the block is passed to tape data management, where additional
blocking can occur and double-buffering to the tape device is provided. This
ensures that the tape performance is optimal. Because the records are not written

770 OS/400 Backup and Recovery V5R1

 directly to tape when the program requests the output, there can be some
interval of time before the records are written to the external media.
v When a code of 0 is passed from the RCVJRNE command, no more entries exist
in the journal. On the return to the RCVJRNE command, the DELAY parameter
value specified on the RCVJRNE command is used to wait before checking for
additional entries. To avoid keeping the records in the various buffers while the
delay occurs, the program forces the records to the tape device by using the
force end-of-data operation (FEOD).
This causes all records in either the RPG or tape data management buffers to be
written to the tape device, and a device completion notification to be received
before proceeding to the next instruction. If there is less than a full block of
records, a short block is written to tape. Tape data management correctly
handles the short block if the tape is read in a subsequent program. When the
return occurs to the RCVJRNE command, the delay time occurs whether or not
any journal entries have arrived since the last time the exit program was called.

The RPG program increments a counter each time a record is written and resets it
when the FEOD operation is used. The program issues the FEOD operation only if
a record has been written which avoids calling tape data management when there
are no records to be written. (If tape data management has no records in its buffers
when the FEOD operation occurs, no empty block is written, but system overhead
occurs.)

The RPG program uses the SHTDN operation code to check for requests to end the
job from external functions such as an End Job (ENDJOB) or End Subsystem
(ENDSBS) command with OPTION(*CNTRLD). If end-of-job is requested, the
program forces the records from the buffers, sets the counter to 9 (which tells the
RCVJRNE command to complete normally, and sets the LR indicator on). The
RETRN operation is then issued and:
v If LR is on, the program’s working storage is returned to the system.
v If LR is off, the program remains active and waits to be called again by the
RCVJRNE command.

Writing to tape occurs either by the buffers being full or when the FEOD operation
is used. This trade-off allows good performance when many journal entries are
written and minimizes the number of times the FEOD operation is used to ensure
that the entries are actually on the tape. With the sample program, the value of the
DELAY parameter and the work management specifications for your job (for
example, pool size and priority) are the major factors controlling the frequency
with which the entries are written and the performance implications on the system
for this function.

If the system ends abnormally while the job is running, so that a successful
end-of-file indication is not written, the subsequent reading of the tape can
produce results that cannot be predicted. Blocks that were successfully written can
be correctly read. The last block and any subsequent data that is on the tape from
a previous use can produce results that cannot be predicted. Copy the tape to a
database file and examine the contents before using the data.

The journal sequence numbers are in ascending sequence (unless they have been
reset) and can be used to determine where the logical end-of-file exists. To avoid
confusion, delete the tapes used for this type of approach.

Assume, for example, that the largest record size journaled was 175 bytes and the
tape record size 300 bytes, as in Figure 90 on page 770. If you need to increase the

Chapter 31. Techniques and Programming Examples for Backup and Recovery 771
 tape record size, change the value of 300 in the RPG file description specification,
the input specification, and factor 2 of the CABGT operation code. If there are
some significantly larger records being journaled, consider how much excess media
is used. An alternative would be to examine the individual fields (JOENTL) and
write two or more small records for each large record.

Journal Entries Written to an ICF File

This topic discusses the differences in programming when you use an ICF file
instead of a tape file as output for the RCVJRNE command. Refer to the program
in Figure 90 on page 770.

If you use an ICF file to transmit journal entries to another system, the FEOD
operation does not apply. Instead, there are data description specifications (DDS)
words (for example, FRCDTA) to force records from the buffers.

Usually the number of blocks transmitted to tape by records less than 175 bytes is
a minimal performance consideration. On communications lines, however, this
number can be significant. To avoid sending unnecessary trailing blanks, consider
decreasing the length of the record being transmitted by the variable length
function (VARLEN DDS keyword). For a discussion of the variable length function,
see the Intrasystem Communications Programming.

If binary synchronous equivalence link (BSCEL) is used, trailing blanks will be

truncated automatically if the TRUNC parameter is specified on the Add ICF
Device Entry (ADDICFDEVE) or the Override ICF Device Entry (OVRICFDEVE)
command. Refer to the BSC Equivalence Link Programming book for more
information about the function of the TRUNC parameter.

772 OS/400 Backup and Recovery V5R1

Part 10. Appendixes

© Copyright IBM Corp. 1997, 2000, 2001 773

774 OS/400 Backup and Recovery V5R1
Appendix A. Licensed Internal Code Installation Error Screens
One of the following three screens may be displayed if you chose option 1 (restore)
on the install selection menu and the selected disk is not currently a load source
disk. In this case, the restore cannot be done. If the selected disk is the correct one
to install to, return to the selection menu and choose the correct install option 2, 3,
4, or 5 to initialize the disk and perform an installation. If the selected disk is not
the correct one, or if an existing load source disk should have been found, follow
the appropriate procedures to determine why the correct disk did not report in, or
was not recognized.

If there is information about a missing disk(s) (second or third of the following

three screens), it indicates what the last load source disk was on this system. If that
disk should still exist (it has not been removed or replaced), determine why it was
not found. If that disk has been removed or replaced, then this is just informational
and may not indicate an error.
Restore Licensed Internal Code

Disk selected to write the Licensed Internal Code to:

Serial Number Type Model I/O Bus Controller Device
__________ ____ ___ ____ ____ ____

The disk selected has not previously been a load source. The
restore of the Licensed Internal Code cannot be done.

Press Enter to return to the Install Licensed Internal Code

screen.

Restore Licensed Internal Code

Disk selected to write the Licensed Internal Code to:

Serial Number Type Model I/O Bus Controller Device
__________ ____ ___ ____ ____ ____

The load source disk could not be found (see disk information
below).

Missing load source disk:

Serial Number Type Model I/O Bus Controller Device
__________ ____ ___ ____ ____ ____

The disk selected has not previously been a load source. The
restore of the Licensed Internal Code cannot be done.

Press Enter to return to the Install Licensed Internal Code

screen.

© Copyright IBM Corp. 1997, 2000, 2001 775

 Restore Licensed Internal Code

Disk selected to write the Licensed Internal Code to:

Serial Number Type Model I/O Bus Controller Device
__________ ____ ___ ____ ____ ____

The load source disk and its mirrored pair could not be found
(see disk information below).

Missing load source disk:

Serial Number Type Model I/O Bus Controller Device
__________ ____ ___ ____ ____ ____
__________ ____ ___ ____ ____ ____

The disk selected has not previously been a load source. The
restore of the Licensed Internal Code cannot be done.

Press Enter to return to the Install Licensed Internal Code

screen.

The following screen can be displayed if you chose option 1 (restore) on the install
selection menu, but the release level of the Licensed Internal Code on the install
media cannot be restored over the current release level on disk. Verify that you
have the correct install media (version/release/modification level). If the level is
correct, then you must do an initialize and install to get the new LIC installed over
the existing LIC on the disk.
Restore Licensed Internal Code

Disk selected to write the Licensed Internal Code to:

Serial Number Type Model I/O Bus Controller Device
__________ ____ ___ ____ ____ ____

The release level of the Licensed Internal Code on the distribution

media cannot be restored over the existing release level on the
selected disk.

Press Enter to return to the Install Licensed Internal Code

screen.

The following screen can be displayed if you chose option 1 (restore) on the install
selection menu and the selected disk is currently a load source disk, but the
pertinent data on the disk cannot be read and, therefore, a restore cannot be done.
You must do an initialize and installation to install the new LIC on this disk.

Restore Licensed Internal Code

Disk selected to write the Licensed Internal Code to:

Serial Number Type Model I/O Bus Controller Device
__________ ____ ___ ____ ____ ____

The selected disk cannot be read. The restore of the Licensed

Internal Code cannot be done.

Press Enter to return to the Install Licensed Internal Code

screen.

776 OS/400 Backup and Recovery V5R1

The following screen can be displayed if two load source disks are detected on
your system. The best disk was selected for the restore or installation. The data
about the other disk is informational so you can verify that the correct disk was
selected. If the correct disk was not selected, follow the appropriate procedures to
disable or remove the selected disk, so that the other disk is selected when you
rerun the task.
Install Licensed Internal Code - Warning

Disk selected to write the Licensed Internal Code to:

Serial Number Type Model I/O Bus Controller Device
__________ ____ ___ ____ ____ ____

Warning:
Another load source disk has also been found on this system.
If you continue the restore or install, the disk listed
above will be used.

Additional load source disk:

Serial Number Type Model I/O Bus Controller Device
__________ ____ ___ ____ ____ ____

Press Enter to continue the restore or install on the selected

disk.

The following screen is displayed if mirroring is active but one disk of the load
source mirrored pair cannot be found. The restore or installation can still continue
on the selected disk, but will not be mirrored until the missing disk becomes again
active. You may want to follow the appropriate procedures to determine why one
of the disks were not found.
Install Licensed Internal Code - Warning

Disk selected to write the Licensed Internal Code to:

Serial Number Type Model I/O Bus Controller Device
__________ ____ ___ ____ ____ ____

Warning:
The mirrored unit for this load source was not found (see
disk information below). The restore or install can continue
on the selected load source. The missing mirrored unit will
be suspended when the restore or install is complete.

Missing load source disk:

Serial Number Type Model I/O Bus Controller Device
__________ ____ ___ ____ ____ ____

Press Enter to continue the restore or install on the selected

disk.

The following two screens are displayed if the disk selected for the install to is not
the same disk that was previously the load source on this system. If the missing
disk should still exist (it has not been removed or replaced), determine why it was
not found. If the disk was removed or replaced, this data is just informational and
may not indicate an error.

Appendix A. Licensed Internal Code Installation Error Screens 777

 Install Licensed Internal Code - Warning

Disk selected to write the Licensed Internal Code to:

Serial Number Type Model I/O Bus Controller Device
__________ ____ ___ ____ ____ ____

Warning:
The load source disk could not be found (see disk information
below).

Missing load source disk:

Serial Number Type Model I/O Bus Controller Device
__________ ____ ___ ____ ____ ____

Press Enter to continue the initialize and install on the

selected disk.

Install Licensed Internal Code - Warning

Disk selected to write the Licensed Internal Code to:

Serial Number Type Model I/O Bus Controller Device
__________ ____ ___ ____ ____ ____

Warning:
The load source disk and its mirrored pair could not be found
(see disk information below).

Missing load source disk:

Serial Number Type Model I/O Bus Controller Device
__________ ____ ___ ____ ____ ____
__________ ____ ___ ____ ____ ____

Press Enter to continue the initialize and install on the

selected disk.

The following screen is displayed if mirroring is active and the active load source
disk can not be found. One unit of the load source mirrored pair was found but is
not currently active. You can install to it, but you will not be allowed to IPL past
DST with it. You may want to follow the appropriate procedures to determine why
the active load source disk can not be found.

778 OS/400 Backup and Recovery V5R1

 Install Licensed Internal Code - Warning

Disk selected to write the Licensed Internal Code to:

Serial Number Type Model I/O Bus Controller Device
__________ ____ ___ ____ ____ ____

Warning:
A load source disk could not be found (see disk information
below).
The disk selected to be the load source (see above) is
suspended. You may install to it and perform an IPL from it to
get to DST and perform DASD diagnostics. However, you will not
be able to perform an IPL past DST with it.

Missing load source disk:

Serial Number Type Model I/O Bus Controller Device
__________ ____ ___ ____ ____ ____

Press Enter to continue the restore or install on the selected

disk.

One of the following three screens is displayed if no disk can be found. That is, no
disk reported in, or was recognized by the system.

If information is supplied about a missing disk(s) (second and third of the three
screens), it indicates what was the last load source disk on this system. If that disk
should still exist (it has not been removed or replaced), then determine why it was
not found. If that disk has been removed or replaced, then this data is just
informational and may not be the reason for the error.
Install Licensed Internal Code - Error

Error:
A disk could not be selected to be the load source.
You can return to the Dedicated Service Tools display and
run diagnostics to determine why a disk could not be
selected.

Correct the problem and install the Licensed Internal Code

again.

Press Enter to return to the Dedicated Service Tools display.

Appendix A. Licensed Internal Code Installation Error Screens 779

 Install Licensed Internal Code - Error

Error:
The load source disk could not be found (see disk information
below).

Missing load source disk:

Serial Number Type Model I/O Bus Controller Device
__________ ____ ___ ____ ____ ____

A disk could not be selected to be the load source.

You can return to the Dedicated Service Tools display and
run diagnostics to determine why a disk could not be
selected.

Correct the problem and install the Licensed Internal Code

again.

Press Enter to return to the Dedicated Service Tools display.

Install Licensed Internal Code - Error

Error:
The load source disk and its mirrored pair could not be found
(see disk information below).

Missing load source disk:

Serial Number Type Model I/O Bus Controller Device
__________ ____ ___ ____ ____ ____
__________ ____ ___ ____ ____ ____

A disk could not be selected to be the load source.

You can return to the Dedicated Service Tools display and
run diagnostics to determine why a disk could not be
selected.

Correct the problem and install the Licensed Internal Code

again.

Press Enter to return to the Dedicated Service Tools display.

One of the following two screens is displayed if a disk is found, but it is not at a
valid address to be the load source.

If there is information about a missing disk(s) (the second screen), it indicates what
the last load source disk was on this system. If that disk should still exist (it has
not been removed or replaced), determine why it was not found. If it has been
removed or replaced, this is just informational and may not be the reason for the
error.

780 OS/400 Backup and Recovery V5R1

 Install Licensed Internal Code - Error

Error:
A disk was found, but it is not at a valid address to be the
load source device.

Selected disk:
Serial Number Type Model I/O Bus Controller Device
__________ ____ ___ ____ ____ ____

The install cannot be done. Press Enter to return to the Dedicated

Service Tools display.

Install Licensed Internal Code - Error

Error:
A disk was found, but it is not at a valid address to be the
load source device.

Selected disk:
Serial Number Type Model I/O Bus Controller Device
__________ ____ ___ ____ ____ ____

The following disk was a load source previously, but could not be
found.

Missing load source disk:

Serial Number Type Model I/O Bus Controller Device
__________ ____ ___ ____ ____ ____

The install cannot be done. Press Enter to return to the Dedicated

Service Tools display.

The following screen is displayed if an existing load source disk is found, but is
not at a valid address to be the load source. If it was intentionally moved,
determine why no other disk could be found to which to install. If this is the
correct disk, determine why it is not at a valid address.
Install Licensed Internal Code - Error

Error:
The following disk was a load source previously, but it is not
currently at a valid address to be the load source device.

Selected disk:
Serial Number Type Model I/O Bus Controller Device
__________ ____ ___ ____ ____ ____

The install cannot be done. Press Enter to return to the Dedicated

Service Tools display.

The following screen is displayed if an existing load source disk was found and:
v Is not at a valid address to be the load source.
v Is one unit of a mirrored pair.
v Is currently not the active load source.

Appendix A. Licensed Internal Code Installation Error Screens 781

 Information about the previously active load source is also listed. If this load
source should still be available to install to, use the appropriate procedures to
determine why it could not be found.
Install Licensed Internal Code - Error

Error:
The following disk was a load source, but it is not currently
active, and it is not at a valid address to be the load source
device.

Selected disk:
Serial Number Type Model I/O Bus Controller Device
__________ ____ ___ ____ ____ ____

The following disk was the previously active load source, but it
could not be found.

Missing load source disk:

Serial Number Type Model I/O Bus Controller Device
__________ ____ ___ ____ ____ ____

The install cannot be done. Press Enter to return to the Dedicated

Service Tools display.

782 OS/400 Backup and Recovery V5R1

Appendix B. Example Disaster Recovery Plan
The objective of a disaster recovery plan is to ensure that you can respond to a
disaster or other emergency that affects information systems and minimize the
effect on the operation of the business. This topic provides you with guidelines for
the kind of information and procedures that you need to recover from a disaster.
When you have prepared the information described in this topic, store your
document in a safe, accessible location off site.

Section 1. Major Goals of a Disaster Recovery Plan–Example

The major goals of this plan are the following:
v To minimize interruptions to the normal operations.
v To limit the extent of disruption and damage.
v To minimize the economic impact of the interruption.
v To establish alternative means of operation in advance.
v To train personnel with emergency procedures.
v To provide for smooth and rapid restoration of service.

Section 2. Personnel–Example
Data Processing Personnel
Name Position Address Telephone

Data Processing Personnel

Name Position Address Telephone

© Copyright IBM Corp. 1997, 2000, 2001 783

Data Processing Personnel
Name Position Address Telephone

Organization Chart
Include a copy of the organization chart with your plan.

Section 3. Application Profile–Example

Use the Display Software Resources (DSPSFWRSC) command to complete this
table.

Application Profile
Critical Fixed Asset
Application Name Yes / No Yes / No Manufacturer Comments

Comment Legend:
1. Runs daily.
2. Runs weekly on ______________.
3. Runs monthly on ______________.

Section 4. Inventory Profile–Example

Use the Work with Hardware Products (WRKHDWPRD) command to complete
this table:

784 OS/400 Backup and Recovery V5R1

 Application Profile
Serial Own or
Manufacturer Description Model Number Leased Cost

Notes:
1. This list should be audited every ______________ months.
2. This list should include:
Processing units System printer
Disk units Tape, optical devices, and diskette units
Models Controllers
Workstation controllers I/O Processors
Personal computers General data communication
Spare workstations Spare displays
Telephones Racks
Air conditioner or heater Humidifier or dehumidifier

Miscellaneous Inventory
Description Quantity Comments

Note: This list should include:

Tapes Diskettes
PC software (such as DOS) Emulation packages
File cabinet contents or documentation Language software (such as COBOL and RPG)
Tape vault contents Printer supplies (such as paper and forms)
Optical media

Section 5. Information Services Backup Procedures

v iSeries Server
– Daily, journals receivers are changed at ______________ and at
______________.
| – Daily, a save of changed objects in the following libraries and directories is
| done at ______________:
- ______________
- ______________
- ______________
- ______________
- ______________

Appendix B. Example Disaster Recovery Plan 785

 - ______________
- ______________
- ______________

This procedure also saves the journals and journal receivers.

– On ______________ at ______________ a complete save of the system is done.
– All save media is stored off-site in a vault at ______________ location.
v Personal Computer
– It is recommended that all personal computers be backed up. Copies of the
personal computer files should be uploaded to the iSeries server on
______________ (date) at ______________ (time), just before a complete save of
the system is done. It is then saved with the normal system save procedure.
This provides for a more secure backup of personal computer-related systems
where a local area disaster could wipe out important personal computer
systems.

Section 6. Disaster Recovery Procedures

For any disaster recovery plan, the following three elements should be addressed.
Emergency Response Procedures
To document the appropriate emergency response to a fire, natural disaster,
or any other activity in order to protect lives and limit damage.
Backup Operations Procedures
To ensure that essential data processing operational tasks can be conducted
after the disruption.
Recovery Actions Procedures
To facilitate the rapid restoration of a data processing system following a
disaster.

Disaster Action Checklist

1. Plan Initiation
a. Notify senior management
b. Contact and set up disaster recovery team
c. Determine degree of disaster
d. Implement proper application recovery plan dependent on extent of
disaster (see Section 7. Recovery Plan–Mobile Site)
e. Monitor progress
f. Contact backup site and establish schedules
g. Contact all other necessary personnel—both user and data processing
h. Contact vendors—both hardware and software
i. Notify users of the disruption of service
2. Follow-Up Checklist
a. List teams and tasks of each
b. Obtain emergency cash and set up transportation to and from backup site,
if necessary
c. Set up living quarters, if necessary
d. Set up eating establishments, as required
e. List all personnel and their telephone numbers

786 OS/400 Backup and Recovery V5R1

 f. Establish user participation plan
g. Set up the delivery and the receipt of mail
h. Establish emergency office supplies
i. Rent or purchase equipment, as needed
j. Determine applications to be run and in what sequence
k. Identify number of workstations needed
l. Check out any off-line equipment needs for each application
m. Check on forms needed for each application
n. Check all data being taken to backup site before leaving and leave
inventory profile at home location
o. Set up primary vendors for assistance with problems incurred during
emergency
p. Plan for transportation of any additional items needed at backup site
q. Take directions (map) to backup site
r. Check for additional magnetic tapes, or optical media if required
s. Take copies of system and operational documentation and procedural
manuals.
t. Ensure that all personnel involved know their tasks
u. Notify insurance companies

Recovery Start-Up Procedures for Use after Actual Disaster

1. Notify ______________ Disaster Recovery Services of the need to utilize service
and of recovery plan selection.

Note: Guaranteed delivery time countdown begins at the time ______________

is notified of recovery plan selection.
a. Disaster notification numbers

______________or ______________

These telephone numbers are in service from ______________ am until

______________ pm Monday through Friday.
2. Disaster notification number

______________

This telephone number is in service for disaster notification after business

hours, on weekends, and during holidays. Please use this number only for the
notification of the actual disaster.
3. Provide ______________ with an equipment delivery site address (when
applicable), a contact, and an alternate contact for coordinating service and
telephone numbers at which contacts can be reached 24 hours a day.
4. Contact power and telephone service suppliers and schedule any necessary
service connections.
5. Notify ______________ immediately if any related plans should change.

Section 7. Recovery Plan–Mobile Site

1. Notify ______________ of the nature of the disaster and the need to select the
mobile site plan.

Appendix B. Example Disaster Recovery Plan 787

 2. Confirm in writing the substance of the telephone notification to
______________ within 48 hours of the telephone notification.
3. Confirm all needed backup media are available to load the backup machine.
4. Prepare a purchase order to cover the use of backup equipment.
5. Notify ______________ of plans for a trailer and its placement (on
______________ side of ______________). (See “Mobile Site Setup Plan”.)
6. Depending on communication needs, notify telephone company
(______________) of possible emergency line changes.
7. Begin setting up power and communications at ______________.
a. Power and communications are prearranged to hook into when trailer
arrives.
b. At the point where telephone lines come into the building
(______________), break the current linkage to the administration
controllers (______________). These lines are rerouted to lines going to the
mobile site. They are linked to modems at the mobile site.
The lines currently going from ______________ to ______________ would
then be linked to the mobile unit via modems.
c. This could conceivably require ______________ to redirect lines at
______________ complex to a more secure area in case of disaster.
8. When the trailer arrives, plug into power and do necessary checks.
9. Plug into the communications lines and do necessary checks.
10. Begin loading system from backups (see “Section 9. Restoring the Entire
System” on page 789).
11. Begin normal operations as soon as possible:
a. Daily jobs
b. Daily saves
c. Weekly saves
12. Plan a schedule to back up the system in order to restore on a home-base
computer when a site is available. (Use regular system backup procedures).
13. Secure mobile site and distribute keys as required.
14. Keep a maintenance log on mobile equipment.

Mobile Site Setup Plan

Attach the mobile site setup plan here.

Communication Disaster Plan

Attach the communication disaster plan, including the wiring diagrams.

Electrical Service
Attach the electrical service diagram here.

788 OS/400 Backup and Recovery V5R1

Section 8. Recovery Plan–Hot Site
The disaster recovery service provides an alternate hot site. The site has a backup
system for temporary use while the home site is being reestablished.
1. Notify ______________ of the nature of the disaster and of its desire for a hot
site.
2. Request air shipment of modems to ______________ for communications. (See
______________ for communications for the hot site.)
3. Confirm in writing the telephone notification to ______________ within 48
hours of the telephone notification.
4. Begin making necessary travel arrangements to the site for the operations
team.
5. Confirm that you have enough save media and that it is packed for shipment
to restore on the backup system.
6. Prepare a purchase order to cover the use of the backup system.
7. Review the checklist for all necessary materials before departing to the hot
site.
8. Make sure that the disaster recovery team at the disaster site has the necessary
information to begin restoring the site. (See “Section 12. Disaster Site
Rebuilding” on page 792).
9. Provide for travel expenses (cash advance).
10. After arriving at the hot site, contact home base to establish communications
procedures.
11. Review materials brought to the hot site for completeness.
12. Start to load the system from the save media.
13. Begin normal operations as soon as possible:
a. Daily jobs
b. Daily saves
c. Weekly saves
14. Plan the schedule to back up the hot-site system in order to restore on the
home-base computer.

Hot-Site System Configuration

Attach the hot-site system configuration here.

Section 9. Restoring the Entire System

To get your system back to the way it was before the disaster, use the procedures
in “Recovering your entire system after a complete system loss–Checklist 20” on
page 96.

Before You Begin: Find the following save media, equipment, and information from
the on-site tape vault or the off-site storage location:
v If you install from the alternate installation device, you need both your save
media and the CD-ROM media containing the Licensed Internal Code.
v All save media from the most recent complete save operation
v The most recent save media from saving security data (SAVSECDTA or SAVSYS)
v The most recent save media from saving your configuration, if necessary
v All save media that contains journals and journal receivers that you saved since
the most recent daily save operation

Appendix B. Example Disaster Recovery Plan 789

 v All save media from the most recent daily save operation
v PTF list (stored with the most recent complete save media, weekly save media,
or both)
v Save media list from most recent complete save operation
v Save media list from most recent weekly save operation
v Save media list from daily saves
v History log from the most recent complete save operation
v History log from the most recent weekly save operation
v History log from the daily save operations
v The Software Installation book
v The Backup and Recovery book
v Telephone directory
v Modem manual
v Tool kit

Section 10. Rebuilding Process

The management team must assess the damage and begin the reconstruction of a
new data center.

If the original site must be restored or replaced, the following are some of the
factors to consider:
v What is the projected availability of all needed computer equipment?
v Will it be more effective and efficient to upgrade the computer systems with
newer equipment?
v What is the estimated time needed for repairs or construction of the data site?
v Is there an alternative site that more readily could be upgraded for computer
purposes?

Once the decision to rebuild the data center has been made, go to “Section 12.
Disaster Site Rebuilding” on page 792.

Section 11. Testing the Disaster Recovery Plan

In successful contingency planning, it is important to test and evaluate the plan
regularly. Data processing operations are volatile in nature, resulting in frequent
changes to equipment, programs, and documentation. These actions make it critical
to consider the plan as a changing document.

Table 109 should be helpful for conducting a recovery test.

Table 109. Checklist for Testing the Disaster Recovery Plan
Not
Item Yes No Applicable Applicable Comments

Conducting a Recovery Test

790 OS/400 Backup and Recovery V5R1

Table 109. Checklist for Testing the Disaster Recovery Plan (continued)
Not
Item Yes No Applicable Applicable Comments

1. Select the purpose of the test. What aspects of the

plan are being evaluated?
2. Describe the objectives of the test. How will you
measure successful achievement of the objectives?
3. Meet with management and explain the test and
objectives. Gain their agreement and support.
4. Have management announce the test and the
expected completion time.
5. Collect test results at the end of the test period.
6. Evaluate results. Was recovery successful? Why or
why not?
7. Determine the implications of the test results. Does
successful recovery in a simple case imply successful
recovery for all critical jobs in the tolerable outage
period?
8. Make recommendations for changes. Call for
responses by a given date.
9. Notify other areas of results. Include users and
auditors.
10. Change the disaster recovery plan manual as
necessary.

Areas to be Tested

Appendix B. Example Disaster Recovery Plan 791

Table 109. Checklist for Testing the Disaster Recovery Plan (continued)
Not
Item Yes No Applicable Applicable Comments

1. Recovery of individual application systems by using

files and documentation stored off-site.
2. Reloading of system save media and performing an
IPL by using files and documentation stored off-site.
3. Ability to process on a different computer.
4. Ability of management to determine priority of
systems with limited processing.
5. Ability to recover and process successfully without
key people.
6. Ability of the plan to clarify areas of responsibility
and the chain of command.
7. Effectiveness of security measures and security
bypass procedures during the recovery period.
8. Ability to accomplish emergency evacuation and
basic first-aid responses.
9. Ability of users of real-time systems to cope with a
temporary loss of on-line information.
10. Ability of users to continue day-to-day operations
without applications or jobs that are considered
noncritical.
11. Ability to contact the key people or their designated
alternates quickly.
12. Ability of data entry personnel to provide the input
to critical systems by using alternate sites and
different input media.
13. Availability of peripheral equipment and processing,
such as printers and scanners.
14. Availability of support equipment, such as air
conditioners and dehumidifiers.
15. Availability of support: supplies, transportation,
communication.
16. Distribution of output produced at the recovery site.
17. Availability of important forms and paper stock.
18. Ability to adapt plan to lesser disasters.

Section 12. Disaster Site Rebuilding

v Floor plan of data center.
v Determine current hardware needs and possible alternatives. (See “Section 4.
Inventory Profile–Example” on page 784.)
v Data center square footage, power requirements and security requirements.
– Square footage ______________
– Power requirements ______________
– Security requirements: locked area, preferably with combination lock on one
door.
– Floor-to-ceiling studding
– Detectors for high temperature, water, smoke, fire and motion
– Raised floor

792 OS/400 Backup and Recovery V5R1

 Vendors

Floor Plan
Include a copy of the proposed floor plan here.

Section 13. Record of Plan Changes

Keep your plan current. Keep records of changes to your configuration, your
applications, and your backup schedules and procedures. For example, you can get
print a list of your current local hardware, by typing:
DSPLCLHDW OUTPUT(*PRINT)

Appendix B. Example Disaster Recovery Plan 793

794 OS/400 Backup and Recovery V5R1
Appendix C. Example of Configuration Planning for Mirrored
Protection
The procedure for starting mirrored protection is simple. However, determining
how long each part of the mirroring process takes and how to achieve the
maximum level of protection requires some planning. Information specifying
elapsed time estimates for different disk units is found in the mirror and journal
performance information chapter of AS/400 Programming: Performance Capabilities
Reference, ZC41-8166.

Two different configurations are used here to illustrate the time estimates found in
the Performance Capabilities Reference and the integration of new storage units to the
physical hardware. They are:
1. 9406 Model 500 system unit using 9336 disk units.
2. 9406 Model 500 system unit using 6602 disk units.

Assumptions: Two different assumptions regarding how and when storage is

added are incorporated to demonstrate the effect they have on each configuration
in terms of elapse time.
v Disk units are added as part of the start mirrored protection procedure using the
Dedicated Service Tools (DST) menu.
v Disk units are currently being used by applications and the database before
starting mirrored protection.

Mirrored Protection Rule for the 9406 System Unit: The level of mirrored protection
for the models of the 9406 system unit can easily be determined by using this rule:
v I/0 processor-level (or bus-level) protection is achieved if you are mirroring
three or more strings (or buses) of disk units and each string (or bus) contains
no more than half of the total number of storage units in the ASP that is being
mirrored.

The exceptions to the above rule are the load source disk unit (unit 1). The best
level of protection the load source achieves is controller-level protection.

9406 Model 500 Using the 9336 Disk Unit Configuration

In this example, the 9406 model 500 system unit has three buses.

Bus 1 in the current configuration does not contain any external disk units. Buses 2
and 3 each have three 6112 I/O processors. Each I/O processor has one 9336-020
disk unit attached.

© Copyright IBM Corp. 1997, 2000, 2001 795

 Bus 3 │────────────────────────────────────────────────────────────┬─────────────┬─────────────┬──│
Bus 2 │──────────────────┬─────────────┬─────────────┬───│ │ │ │
Bus 1 │──┬──────│ │ │ │ │ │ │
┌───┴───┐ ┌───┴───┐ ┌───┴───┐ ┌───┴───┐ ┌───┴───┐ ┌───┴───┐ ┌───┴───┐
│ MFIOP │ │FC#6112│ │FC#6112│ │FC#6112│ │FC#6112│ │FC#6112│ │FC#6112│
└───┬───┘ └┬──────┘ └┬──────┘ └┬──────┘ └┬──────┘ └┬──────┘ └┬──────┘
┌────┬─┴──┬────┐ │ │ │ │ │ │
┌─┴─┐┌─┴─┐┌─┴─┐┌─┴─┐ ┌┴─┬──┬──┬──┐ ┌┴─┬──┬──┬──┐ ┌┴─┬──┬──┬──┐ ┌┴─┬──┬──┬──┐ ┌┴─┬──┬──┬──┐ ┌┴─┬──┬──┬──┐
│L/S││ ││ ││ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │
└───┘└───┘└───┘└───┘ └──┴──┴──┴──┘ └──┴──┴──┴──┘ └──┴──┴──┴──┘ └──┴──┴──┴──┘ └──┴──┴──┴──┘ └──┴──┴──┴──┘
6602-030 9336-020 9336-020 9336-020 9336-020 9336-020 9336-020

Figure 91. 9406 Model 500 Original Configuration. The system has six 6112 I/O processors attached to Bus 2 and Bus
3. Bus 1 contains the multiple function I/O processor (MFIOP) and the 6602-030 storage units. Each I/O processor has
one 9336 model 020 disk attached.

Additional disk units are added to the existing strings to increase the disk capacity
as shown in Figure 92.

Bus 3 │────────────────────────────────────────────────────────────┬─────────────┬─────────────┬──│
Bus 2 │──────────────────┬─────────────┬─────────────┬───│ │ │ │
Bus 1 │──┬──────│ │ │ │ │ │ │
┌───┴───┐ ┌───┴───┐ ┌───┴───┐ ┌───┴───┐ ┌───┴───┐ ┌───┴───┐ ┌───┴───┐
│ MFIOP │ │FC#6112│ │FC#6112│ │FC#6112│ │FC#6112│ │FC#6112│ │FC#6112│
└───┬───┘ └┬──────┘ └┬──────┘ └┬──────┘ └┬──────┘ └┬──────┘ └┬──────┘
┌────┬─┴──┬────┐ │ │ │ │ │ │
┌─┴─┐┌─┴─┐┌─┴─┐┌─┴─┐ ┌┴─┬──┬──┬──┐ ┌┴─┬──┬──┬──┐ ┌┴─┬──┬──┬──┐ ┌┴─┬──┬──┬──┐ ┌┴─┬──┬──┬──┐ ┌┴─┬──┬──┬──┐
│L/S││ ││ ││ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │
└───┘└───┘└───┘└───┘ └┬─┴──┴──┴──┘ └┬─┴──┴──┴──┘ └┬─┴──┴──┴──┘ └┬─┴──┴──┴──┘ └┬─┴──┴──┴──┘ └┬─┴──┴──┴──┘
6602-030 ┌┴─┬──┬──┬──┐ ┌┴─┬──┬──┬──┐ ┌┴─┬──┬──┬──┐ ┌┴─┬──┬──┬──┐ ┌┴─┬──┬──┬──┐ ┌┴─┬──┬──┬──┐
│ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │
└──┴──┴──┴──┘ └──┴──┴──┴──┘ └──┴──┴──┴──┘ └──┴──┴──┴──┘ └──┴──┴──┴──┘ └──┴──┴──┴──┘
9336-020 9336-020 9336-020 9336-020 9336-020 9336-020

Figure 92. 9406 Model 500 Mirrored Configuration. The additional 9336-020 disk units are attached to the existing
strings. The data on the third and fourth 6602-030 storage units attached to the multiple function I/O processor
(MFIOP) are moved to other storage units to allow them to become the mirrored units for unit 1 (load source) and unit
2. The Licensed Internal Code is on the first 6602-030 storage unit.

The load source unit is handled by the 6602-030 storage units on the model 500.
The 6602-030 storage units create mirrored pairs among themselves when mirrored
protection starts. Because all storage units are added to the system ASP, half of the
new capacity or 20.6GB is usable for system and user data once it is protected.

Calculating the Time to Add Disk Units to the System ASP

Storage units in a 9336 disk unit are added concurrently to the ASP. It takes 14
minutes to add one 9336-020 storage unit to an ASP. For each level of the
configuration, add 2 additional minutes.

Time for one

9336-020 Disk 6112 Total
storage unit controller I/O processor Bus Time

14 minutes + 2 + 2 + 2 = 20 minutes

Because six disk controllers are working under six I/O processors on two buses,
and each 9336-020 added has four storage units attached (a total of 24), this
function should take approximately:
1 X 20(minutes) = 20 minutes (0.4 hours)

Twenty four 9336-020 storage units are added concurrently to the system ASP.

796 OS/400 Backup and Recovery V5R1

Calculating the Time to Move Extents
It is very difficult to accurately calculate the amount of time that is needed to
move extents off of a disk unit. Some the variables which affect the time are as
follows:
v The number of storage units which will have data moved from them to create
mirrored pairs of disk units
v The amount of data on each unit to move
v Contention in the hardware (I/O processors, buses, and disk units)

It takes 30 to 60 minutes to clear each storage unit depending on the amount and
type of data stored on the unit. Many of the move extents operations are processed
concurrently so that the cumulative effect is not entirely synchronous for each
storage unit.

Storage Units Added before Mirrored Protection is Started

The unit chosen as the mirrored unit for the load source unit (unit 1) may need to
be cleared. If all four 6602-030 storage units were being used in the system ASP
before mirrored protection is started, two units will be cleared. 30 minutes is used
in this example. Otherwise, there should be very little time spent in this part of the
process.

Storage Units in Use by the System before Mirrored Protection Is

Started
In this assumption, 24 storage units are cleared. This can take approximately three
and one-half hours. This estimate is based on field reports.

Determining the Time to Synchronize the Disk

Using the time specified in the Performance Capabilities Reference for multiple units
across all system unit models, this configuration should take approximately three
hours. Twenty-four pairs of storage units are synchronizing. During the
synchronization, the progress is displayed on the control panel as SRC C6xx 4205,
where xx indicates the percentage completed.

Total Estimate
The following estimates do not include time to save the ASP that is being mirrored
or any additional IPL time not attributed to starting mirrored protection.
Disk units added:

Add units to ASP ...................... 0.4 hours

Move extents .......................... .5 hours
Synchronization ....................... 3.0 hours

Total: 3.9 hours

Disk units in use before starting mirrored protection:

Move extents .......................... 3.5 hours

Synchronization ....................... 3.0 hours

Total: 6.5 hours

Appendix C. Example of Configuration Planning for Mirrored Protection 797

9406 Model 500 Using the 6602 Disk Unit Configuration
In this example, the 9406 model 500 system unit has two busses. Bus 1 in the
current configuration has two I/O processors. The multiple function I/O processor
has four 6602 disk units attached and the 6530 storage I/O processor has four 6602
disk units attached.

Bus 1 │──┬────────────────────┬─────│
┌───┴───┐ ┌────┴────┐
│ MFIOP │ │ FC#6530 │
└───┬───┘ └────┬────┘
┌────┬─┴──┬────┐ ┌────┬─┴──┬────┐
┌─┴─┐┌─┴─┐┌─┴─┐┌─┴─┐ ┌─┴─┐┌─┴─┐┌─┴─┐┌─┴─┐
│L/S││ ││ ││ │ │ ││ ││ ││ │
└───┘└───┘└───┘└───┘ └───┘└───┘└───┘└───┘
6602-030 6602-050

Figure 93. 9406 Model 500 Original Configuration. Bus 1 contains one multiple function I/O
processor (MFIOP) with four 6602 disk units and one 6530 storage I/O processor with four
6602 disk units.

To keep the same disk capacity after mirrored protection is started, eight 6602 disk
units are added. One 6530 storage I/O processor with seven 6602 disk units is
added to Bus 2 and one 6602 disk units are added to the 6530 storage I/O
processor attached to Bus 1 to obtain the optimal protection level. All the disk
units except the load source will have bus-level protection. The load source disk
units must be attached to the MFIOP and will have controller-level protection.

Bus 2 │────────────────────────────────────────────────────┬─────│
Bus 1 │──┬───────────────────┬─────│ │
┌───┴───┐ ┌────┴────┐ ┌────┴────┐
│ MFIOP │ │ FC#6530 │ │ FC#6530 │
└───┬───┘ └────┬────┘ └────┬────┘
┌────┬─┴──┬────┐ ┌────┬─┴──┬────┬────┐ ┌────┬────┬─┴──┬────┬────┬────┐
┌─┴─┐┌─┴─┐┌─┴─┐┌─┴─┐┌─┴─┐┌─┴─┐┌─┴─┐┌─┴─┐┌─┴─┐┌─┴─┐┌─┴─┐┌─┴─┐┌─┴─┐┌─┴─┐┌─┴─┐┌─┴─┐
│L/S││ ││ ││ ││ ││ ││ ││ ││ X ││ X ││ X ││ X ││ X ││ X ││ X ││ X │
└───┘└───┘└───┘└───┘└───┘└───┘└───┘└───┘└───┘└───┘└───┘└───┘└───┘└───┘└───┘└───┘
6602-030 6602-050 6602-050

Figure 94. 9406 Model 500 Mirrored Configuration. One 6602 disk unit is added to the
existing 6530 storage I/O processor. One 6530 storage I/O processor with seven 6602 disk
units is attached to Bus 2. Added disk units are marked with an X.

When mirrored protection is started, the data on one 6602-030 disk unit attached to
the multiple function I/O processor (MFIOP) are moved to other disk units to
allow it to become mirrored unit for unit 1 (load source) The Licensed Internal
Code is on the first 6602-030 disk unit.

Calculating the Time to Add Disk Units to the System

Disk units are added concurrently to the system. It takes 5 minutes to add one
6602 disk unit to an ASP. For each level of the configuration, add 2 minutes.

Time for one Disk 6112 Bus Total

6602 disk unit controller I/O processor Time

5 minutes + 2 + 2 + 2 = 11 minutes

798 OS/400 Backup and Recovery V5R1

 Eight 6602 disk controllers are attached to two I/O processors. Each I/O processor
is attached to a separate bus. A total of eight disk units are added. This function
should take approximately:
1 x 11 minutes = 11 minutes (0.2 hours)

Eight 6602 disk units are added concurrently to the system.

Calculating the Time to Move Extents

It is very difficult to accurately calculate the amount of time that is needed to
move extents off of a disk unit. Some the variables which affect the time are as
follows:
v The number of storage units which will have data moved from them to create
mirrored pairs of disk units
v The amount of data on each unit to move
v Contention in the hardware (I/O processors, buses, and disk units)

It takes 30 to 60 minutes to clear each 6602 disk unit depending on the amount
and type of data stored on the unit. Many of the move extents operations are
processed concurrently so that the cumulative effect is not entirely synchronous for
each storage unit.

Storage Units Added before Mirrored Protection is Started

The unit chosen as the mirrored unit for the load source unit (unit 1) needs to be
cleared. Otherwise, there should be very little time spent in this part of the process.
30 minutes is used in this example.

Storage Units in Use by the System before Mirrored Protection Is

Started
In this assumption, 8 storage units are cleared. This can take approximately 30
minutes. This estimate is based on field reports.

Determining the Time to Synchronize the Disk

Using the time specified in the Performance Capabilities Reference for multiple units
across all system unit models, this configuration should take approximately 80
minutes. Eight pairs of storage units need to be synchronized. During the
synchronization, the progress is displayed on the control panel as SRC C6xx 4205,
where xx indicates the percentage completed.

Total Estimate
The following estimates do not include time to save the ASP that is being mirrored
or any additional IPL time not attributed to starting mirrored protection.
Disk units added:

Add units to ASP ...................... 0.2 hours

Move extents .......................... .5 hours
Synchronization ....................... 1.4 hours

Total: 2.1 hours

Disk units in use before starting mirrored protection:

Move extents .......................... 0.5 hours

Synchronization ....................... 1.4 hours

Total: 1.9 hours

Appendix C. Example of Configuration Planning for Mirrored Protection 799

Assessment of Single-Points-of-Failure
For a multiple ASP system, it takes careful planning and analysis to determine
which storage units should be assigned to an ASP. In a mirrored protection
environment, you can significantly reduce the effective level of protection if the
storage units are chosen randomly. Consider the configuration in Figure 95.

Bus 5 │───────────────────────────────────────┬─────────────┬────│
Bus 4 │──────────┬─────────────┬─────│ │ │
┌───┴───┐ ┌───┴───┐ ┌───┴───┐ ┌───┴───┐
│FC#6112│ │FC#6112│ │FC#6112│ │FC#6112│
└┬──────┘ └┬──────┘ └┬──────┘ └┬──────┘
│ │ │ │
┌┴─┬──┬──┬──┐ ┌┴─┬──┬──┬──┐ ┌┴─┬──┬──┬──┐ ┌┴─┬──┬──┬──┐
│ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │
└──┴──┴──┴──┘ └──┴──┴──┴──┘ └──┴──┴──┴──┘ └──┴──┴──┴──┘
9336-020 9336-020 9336-020 9336-020

Bus 3 │───────────────────────────────────────────────┬─────────────┬────│
Bus 2 │──────────────────┬─────────────┬─────│ │ │
Bus 1 │──┬──────│ │ │ │ │
┌───┴───┐ ┌───┴───┐ ┌───┴───┐ ┌───┴───┐ ┌───┴───┐
│ MFIOP │ │FC#6112│ │FC#6112│ │FC#6112│ │FC#6112│
└───┬───┘ └┬──────┘ └┬──────┘ └┬──────┘ └┬──────┘
┌────┬─┴──┬────┐ │ │ │ │
┌─┴─┐┌─┴─┐┌─┴─┐┌─┴─┐ ┌┴─┬──┬──┬──┐ ┌┴─┬──┬──┬──┐ ┌┴─┬──┬──┬──┐ ┌┴─┬──┬──┬──┐
│L/S││ ││ ││ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │
└───┘└───┘└───┘└───┘ └──┴──┴──┴──┘ └──┴──┴──┴──┘ └──┴──┴──┴──┘ └──┴──┴──┴──┘
6602-030 9336-020 9336-020 9336-020 9336-020

Figure 95. 9406 Model 500 with 9336 Disk Unit Configuration. The model 500 has eight 6112
I/O processors attached to Bus 2, Bus 3, Bus 4, and Bus 5. Each 6112 I/O processor has
one 9336-020 disk unit. The multiple function I/O processor (MFIOP) has the 6602-030
storage units attached to Bus 1.

In Figure 95, to determine which storage units should be selected for a user ASP,
consider the following:

Two scenarios are used to illustrate the need for careful planning when selecting
storage units for user ASPs in a mirrored environment.
System 1
The system ASP and the user ASP have mirrored protection.
System 2
Only the system ASP has mirrored protection.

Both systems have eight 9336-020 storage units for the user ASP. When referring to
Figure 95, keep in mind that you must select the highest level of protection with
the least number of single-point-of-failures. In System 2, determine which storage
units must be selected to minimize disk subsystem failures that can cause the
system to stop processing.

System 1
In Figure 96 on page 801, one storage unit from each 9336-020 is selected for a user
ASP to achieve highest level of protection for both ASPs.

800 OS/400 Backup and Recovery V5R1

 Bus 5 │───────────────────────────────────────┬─────────────┬────│
Bus 4 │──────────┬─────────────┬─────│ │ │
┌───┴───┐ ┌───┴───┐ ┌───┴───┐ ┌───┴───┐
│FC#6112│ │FC#6112│ │FC#6112│ │FC#6112│
└┬──────┘ └┬──────┘ └┬──────┘ └┬──────┘
│ │ │ │
┌┴─┬──┬──┬──┐ ┌┴─┬──┬──┬──┐ ┌┴─┬──┬──┬──┐ ┌┴─┬──┬──┬──┐
│ │ │ │ 2│ │ │ │ │ 2│ │ │ │ │ 2│ │ │ │ │ 2│
└──┴──┴──┴──┘ └──┴──┴──┴──┘ └──┴──┴──┴──┘ └──┴──┴──┴──┘
9336-020 9336-020 9336-020 9336-020

Bus 3 │───────────────────────────────────────────────┬─────────────┬────│
Bus 2 │──────────────────┬─────────────┬─────│ │ │
Bus 1 │──┬──────│ │ │ │ │
┌───┴───┐ ┌───┴───┐ ┌───┴───┐ ┌───┴───┐ ┌───┴───┐
│ MFIOP │ │FC#6112│ │FC#6112│ │FC#6112│ │FC#6112│
└───┬───┘ └┬──────┘ └┬──────┘ └┬──────┘ └┬──────┘
┌────┬─┴──┬────┐ │ │ │ │
┌─┴─┐┌─┴─┐┌─┴─┐┌─┴─┐ ┌┴─┬──┬──┬──┐ ┌┴─┬──┬──┬──┐ ┌┴─┬──┬──┬──┐ ┌┴─┬──┬──┬──┐
│L/S││ ││ ││ │ │ │ │ │ 2│ │ │ │ │ 2│ │ │ │ │ 2│ │ │ │ │ 2│
└───┘└───┘└───┘└───┘ └──┴──┴──┴──┘ └──┴──┴──┴──┘ └──┴──┴──┴──┘ └──┴──┴──┴──┘
6602-030 9336-020 9336-020 9336-020 9336-020

Figure 96. System 1–9406 Model 500 Configuration. The multiple function I/O processor (MFIOP) and the 6602-030
storage units are attached to Bus 1. User ASP 2 is built by selecting a storage unit from each 9336-020 disk unit. Both
the system ASP and the user ASP has mirrored protection.

System 2
In Figure 97 on page 802, all eight storage units are selected from the two 9336-020
disk units attached to Bus 5 for user ASP 2. This minimizes the number of disk
subsystem failures that can cause the system to stop processing.

For the 9336 disk unit on System 2, the points-of-failure are:

1 - Bus 5
2 - 6112 I/O processor attached to Bus 5
2 - Controllers in the 9336 disk units attached to Bus 5
8 - Storage units
-----------
13 - points-of-failure

All of the 9336 storage units in the system ASP have bus-level protection.

If you selected the storage units for the non-mirrored ASP as shown in Figure 96,
there will be significantly more points-of-failure for the 9336 disk units that could
cause the system to stop processing.

Appendix C. Example of Configuration Planning for Mirrored Protection 801

 Bus 5 │───────────────────────────────────────┬─────────────┬────│
Bus 4 │──────────┬─────────────┬─────│ │ │
┌───┴───┐ ┌───┴───┐ ┌───┴───┐ ┌───┴───┐
│FC#6112│ │FC#6112│ │FC#6112│ │FC#6112│
└┬──────┘ └┬──────┘ └┬──────┘ └┬──────┘
│ │ │ │
┌┴─┬──┬──┬──┐ ┌┴─┬──┬──┬──┐ ┌┴─┬──┬──┬──┐ ┌┴─┬──┬──┬──┐
│ │ │ │ │ │ │ │ │ │ │ 2│ 2│ 2│ 2│ │ 2│ 2│ 2│ 2│
└──┴──┴──┴──┘ └──┴──┴──┴──┘ └──┴──┴──┴──┘ └──┴──┴──┴──┘
9336-020 9336-020 9336-020 9336-020

Bus 3 │───────────────────────────────────────────────┬─────────────┬────│
Bus 2 │──────────────────┬─────────────┬─────│ │ │
Bus 1 │──┬──────│ │ │ │ │
┌───┴───┐ ┌───┴───┐ ┌───┴───┐ ┌───┴───┐ ┌───┴───┐
│ MFIOP │ │FC#6112│ │FC#6112│ │FC#6112│ │FC#6112│
└───┬───┘ └┬──────┘ └┬──────┘ └┬──────┘ └┬──────┘
┌────┬─┴──┬────┐ │ │ │ │
┌─┴─┐┌─┴─┐┌─┴─┐┌─┴─┐ ┌┴─┬──┬──┬──┐ ┌┴─┬──┬──┬──┐ ┌┴─┬──┬──┬──┐ ┌┴─┬──┬──┬──┐
│L/S││ ││ ││ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │ │
└───┘└───┘└───┘└───┘ └──┴──┴──┴──┘ └──┴──┴──┴──┘ └──┴──┴──┴──┘ └──┴──┴──┴──┘
6602-030 9336-020 9336-020 9336-020 9336-020

Figure 97. System 2–9406 Model 500 Configuration. The multiple function I/O processor (MFIOP) with the 6602-030
storage units is attached to Bus 1. User ASP 2 is built from the two 9336 disk units attached to Bus 5. This user ASP
is not mirrored. The system ASP has mirrored protection.

The 9336 disk unit points-of-failure of System 1 are:

4 - Bus 2 through 5
8 - 6112 I/O processors attached to all buses
8 - Controllers in the 9336 disk unit subsystems
8 - Storage units
-----------
28 - Points-of-failure

The 9336 storage units in this scenario would really only have device-level
protection even though the mirrored image of the storage unit exists on another
bus.

When you have non-mirrored disk units on the system, you can see it is important
to consider the actual points-of-failure that can cause the system to stop
processing. In this case, 13 points-of-failure (System 2) versus 28 point-of-failure (if
System 2 uses the configuration of System 1 for ASP 2).
Notes:
1. If you look at the Work with Disk Configuration Status display for either of
these systems, you can see that 24-9336 storage units have bus-level protection.
However, when the unprotected storage units in the ASP were chosen as in
Figure 96 on page 801 effect is that the disk units have the lowest level
(device-level) of protection. Basically, device-level protection provides
protection from data-loss.
2. Although one storage unit from each 9336-020 is selected for a user ASP in
Figure 96 on page 801, it is more advantageous to select two 9336-020
subsystems that are attached to two separate IOPs on two separate buses so
that locating the storage units is easier.

802 OS/400 Backup and Recovery V5R1

Full Mirrored Protection Versus Partial Mirrored Protection
Full mirrored protection and partial mirrored protection do not provide the same
availability results. These two implementations of mirrored protection are quite
different. The scenarios of a disk unit on the iSeries server for each of these
mirroring methods requires different user responses.

It does not matter if you are using just the system ASP (ASP 1) or multiple user
ASPs (2 through 16), full mirrored protection protects all disk units in the iSeries
server. Partial mirrored protection protects only a portion of the disk units
designated by one or more ASPs. However, not all storage units in the disk
configuration are protected. Therefore, the planning of the disk unit placement and
what ASPs are selected for mirrored protection becomes more difficult (see
“Assessment of Single-Points-of-Failure” on page 800).

Besides the planning of ASPs, the significant difference between the two mirrored
protection methods regards availability. With full mirrored protection, you
maximize the availability of the iSeries server when a disk subsystem failure
occurs. With this method of mirrored protection, it does not matter which ASP has
the failure. With partial mirrored protection, the system continues to run while
reporting the failed storage unit to the system operator (QSYSOPR) message
queue. However, if the disk failure occurs in an ASP that does not have mirrored
protection, SRC A6xx 0266 is sent when that ASP is accessed by any job on the
system. Because the storage units in the ASP do not have mirrored units, the
storage management directory becomes unusable and all input and output
operations to the ASP are suspended.

The disk attention SRC does not mean that the system has ended. All input and
output operations are queued to allow the service representative to investigate the
cause of the disk failure. If the problem is not with the disk media, the failing
cards are replaced, the failed disk unit is powered on, and the system continues
from the point that the equipment error occurred. All queued input and output
operations resume. However, if a disk media failure occurs, the service
representative performs a main storage dump to minimize the time for the next
IPL to OS/400, and allows the system to end processing.

With full mirrored protection, the operation of the system is not interrupted while
diagnostics and most repairs to resolve the disk subsystem failure problem are
taking place. With I/O processor-level protection, the maximum concurrent
maintenance is possible, depending on the error. In any case, the user has complete
control over the shutdown of the system should a power-down be required to
service the disk problem; the system does not end abnormally.

Although critical data is protected with partial mirrored protection, and a restore
operation is not required for the data in the protected ASP, you do not have the
maximum availability that is provided by full mirrored protection because of the
exposure of the unprotected ASP. If your availability requirements state that your
system must be in operation within minutes following a failure or remain active
during your business hours, partial mirrored protection is not an option in most
cases.

Appendix C. Example of Configuration Planning for Mirrored Protection 803

804 OS/400 Backup and Recovery V5R1
 Appendix D. Journal Entry Information
The system creates different types of journal entries in the journal receiver for
different kinds of activities. You cannot access the information in journal receivers
directly. Several system commands provide formatted information from a journal
receiver:
v Use the Display Journal (DSPJRN) command to display entries, print them, or
write them to an output file.
v Use the Receive Journal Entry (RCVJRNE) command to specify an exit program.
When entries are added to the journal receiver, they are also passed to the exit
program. The exit program can, for example, write entries to save media or send
them to another system.
v Use the Retrieve Journal Entry (RTVJRNE) command to retrieve journal entries
to a CL program.
v Use the Retrieve Journal Entries (QjoRetrieveJournalEntries) API to retrieve
journal entries into a receiver variable.
Journaling and the use of these commands and API are described in “Chapter 19.
Planning and Setting Up Journaling” on page 383, “Chapter 20. Working with
Journal Entries, Journals, and Journal Receivers” on page 427, and “Chapter 21.
Remote Journal Function” on page 467.

When the system formats journal entries for you with the DSPJRN and RTVJRNE
commands, it uses one of four layouts. These layouts include a fixed-length
portion and a variable-length portion. The variable-length portion includes
entry-specific data and null value indicators, if applicable. The fixed-length portion
of the journal entry appears as separate fields in these layouts. Additionally, the
RCVJRNE command also supports these four layouts, as well as the additional
value, *TYPEPTR. The layout of the journal entry data for the *TYPEPTR interface
is exactly the same as the layout of the data that is returned on the
QjoRetrieveJournalEntries API interface. For more information on this layout, click
the Programming topic in the Information Center at the following Web site:
http://www.ibm.com/eserver/iseries/infocenter.

The field descriptions are shown in the following:

| v Table 111 on page 815, Field Descriptions of the Fixed-Length Portion of a
| Journal Entry: *TYPE1
| v Table 112 on page 818, Field Descriptions of the Fixed-Length Portion of a
| Journal Entry: *TYPE2
| v Table 113 on page 819, Field Descriptions of the Fixed-Length Portion of a
| Journal Entry: *TYPE3
| v Table 114 on page 820, Field Descriptions of the Fixed-Length Portion of a
| Journal Entry: *TYPE4
Table 110 on page 807 shows all the journal codes and entry types for journal
entries. The remaining tables show specific information about many of the journal
entries.

The online command help provides more information about using the DSPJRN,
RCVJRNE, and RTVJRNE commands. To view the help, type the command name
at a command line and press F1.

© Copyright IBM Corp. 1997, 2000, 2001 805

 For more information on the QjoRetrieveJournalEntries API, look in the
Programming topic in the Information Center at the following Web site:
http://www.ibm.com/eserver/iseries/infocenter.

Journal Codes
This topic describes all the possible journal codes or categories of journal entries.

Journal Code A – System Accounting Entry: Journal entries with a journal code of
A contain information about job accounting. Refer to the Work Management book
for a detailed description of the contents of converted journal entries with journal
code A.

| Journal Code B– Integrated File System: Journal entries with a journal code of B
| contain information about changes to integrated file system objects. The only
| integrated file system objects that are supported are those with an object of type
| *STMF, *DIR or *SYMLNK. These objects must be in the Root (’/’), QOpensys, and
| User-defined file systems. See the Integrated File System Introduction book for
| more information about file systems.

Journal Code C – Commitment Control Operation: Journal entries with a journal

code of C contain commitment control information.

Journal Code D – Database File Operation: Journal entries with a journal code of
D contain file level information about changes for a physical file, not an individual
member.

| Journal Code E – Data Area Operation: Journal entries with a journal code of E
| contain information about changes to journaled data areas. See Work Management
| for more information about data areas.

Journal Code F – Database File Member Operation: Journal entries with a journal
code of F contain file level information about changes for a physical file member
that are being journaled to this journal. (If you use a logical file in a program, the
file level information reflects the physical file on which the logical file is based.)
Journal entries with journal code F can also contain file level information for access
paths that are associated with physical or logical file members that are being
journaled to this journal.

| Journal Code I – Internal Operation: Journal entries with a journal code of I

| contain information about access paths or indexes or other internal operations.
Entries with a journal code of I are displayed only if JRN(*INTSYSJRN) is specified
on the DSPJRN command.

Journal Code J – Journal or Receiver Operation: Journal entries with a journal code
of J contain information about the journal and the journal receivers.

Journal Code L – License Management: Journal entries with a journal code of L

contain information about license management, such as changes to the usage limit
and usage limit violations.

Journal Code M – Network Management Data: Journal entries with a journal code
of M contain information about Network Management, including TCP/IP. For a
description of the TCP/IP entries, see the TCP/IP Configuration and Reference. For a
description of the Network Management entries, see the Simple Network
Management Protocol (SNMP) Support book.

806 OS/400 Backup and Recovery V5R1

 Journal Code O – Object–Oriented Entry: Journal entries with a journal code of O
contain object–oriented information. These entries are reserved for future use.

Journal Code P – Performance Tuning Entry: Journal entries with a journal code of
P contain information about performance. For the description of the layout of these
entries, refer to the Work Management book.

| Journal Code Q – Data Queue Operation: Journal entries with a journal code of Q
| contain information about changes to journaled data queues. See CL Programming
| for more information about data queues.

Journal Code R – Operation on Specific Record: Journal entries with a journal code
of R contain information about a change to a specific record in the physical file
member that is being journaled to the journal. For a given physical file member,
the record-level journal entries appear in the journal in the order that the changes
were made to the file.

Journal Code S – Distributed Mail Services: Journal entries with a journal code of
S contain information about SNA distribution services (SNADS), X.400, and mail
server framework. For the description of the layout of these entries, refer to these
books:
v SNA Distribution Services
v AnyMail/400 Mail Server Framework Support
v OSI Message Services/400 Support

Journal Code T – Audit Trail Entry: Journal entries with a journal code of T
contain auditing information. For the description of the layout of audit journal
entries, see the iSeries Security Reference book.

Journal Code U – User-Generated Entry: Journal entries with a code of U are sent
to the journal receiver by the Send Journal Entry (SNDJRNE) command or by the
Send Journal Entry (QJOSJRNE) API. “Sending Your Own Journal Entries” on
page 429 provides more information.

Journal Entries by Journal Code and Type

Table 110 describes the possible entry types for journal entries:
Table 110. Journal Entries by Journal Code and Type
Journal
Code Description Entry Type Entry Type Description

A System DP Direct print information.

accounting
entry1
JB Job resource information.
SP Spooled print information.
| B Integrated File AA Change audit attribute.
System15
| AJ Start of apply.
| AT End of apply. See Table 115 on page 822.
| BD10 IFS object deleted.
| B0 Begin create.
| B1 Create summary.
| B2 Link to existing object.
| B3 Rename, move object.

Appendix D. Journal Entry Information 807

 Table 110. Journal Entries by Journal Code and Type (continued)
Journal
Code Description Entry Type Entry Type Description

| B4 Remove link (parent directory).

| B5 Remove link (link).
| CS10 IFS object closed.
| ET10 End journaling for object.
| FA IFS object attribute changed.
| FC10 IFS object forced.
| FF5, 10 Storage for object freed.
| FR5 IFS object restored. See Table 136 on page 849.
| FS5, 10 IFS object saved. See Table 137 on page 850.
| FW5, 10 Start of save for save-while-active. See Table 138 on page 851.
| JT Start journaling for object. See Table 139 on page 852.
| OA Change object authority.
| OF IFS object opened.
| OG Change primary group.
| OI10 Object in use at abnormal end. See Table 123 on page 824.
| OO Change object owner.
| TR IFS object truncated.
| WA12 Write, after-image.
C Commitment BC Start commitment control (STRCMTCTL).
control
operation
CM Set of record changes committed (COMMIT). See Table 118 on page 823.
EC10 End commitment control (ENDCMTCTL).
LW A logical unit of work (LUW) has ended. See Table 124 on page 824
through Table 130 on page 845.
| PC Prepare commit block.
| R1 Rollback started.
RB Set of record changes rolled back (ROLLBACK). See Table 135 on
page 849.
SC Commit transaction started.
| D Database file AC Add referential integrity constraint. See Table 148 on page 855.
operation
| CG Change file. See Table 148 on page 855.
| CT Create database file. See Table 148 on page 855.
| DC Remove referential integrity constraint. See Table 148 on page 855.
| DF10 File was deleted.
| DT10 Delete file. See Table 148 on page 855.
EF10 Journaling for a physical file ended (ENDJRNPF).
FM File moved to a different library (MOVOBJ or RNMOBJ OBJTYPE(*LIB)).
| See Table 131 on page 847.
| FN File renamed (RNMOBJ). See Table 131 on page 847.
| GC Change constraint. See Table 148 on page 855.
| GO Change owner. See Table 148 on page 855.
| GT Grant authority. See Table 148 on page 855.
| ID10 File in use. See Table 123 on page 824.
JF Journaling for a physical file started (STRJRNPF (JRNPF)).
| RV Revoke authority. See Table 148 on page 855.
| TC Add trigger. See Table 148 on page 855.
| TD Remove trigger. See Table 148 on page 855.
| TG Change trigger. See Table 148 on page 855.
| E Data area EA8,14 Update data area, after image. See Table 143 on page 853
| operation16

808 OS/400 Backup and Recovery V5R1

 Table 110. Journal Entries by Journal Code and Type (continued)
Journal
Code Description Entry Type Entry Type Description

| EB8,14 Update data area, before image. See Table 143 on page 853
| ED10 Data area deleted.
| EG Start journal for data area. See Table 139 on page 852.
| EH10 End journal for data area.
| EI10 Data area in use. See Table 123 on page 824.
| EL5 Data area restored. See Table 136 on page 849.
| EM Data area moved. See Table 131 on page 847.
| EN Data area renamed. See Table 131 on page 847.
| EQ Data area changes applied. See Table 115 on page 822.
| ES5, 10 Data area saved. See Table 137 on page 850.
| EU Remove journaled changes (RMVJRNCHG) command started.
| EW5, 10 Start of save for data area. See Table 138 on page 851.
| EX Data area changes removed. See Table 115 on page 822.
| EY Apply journaled changes (APYJRNCHG) command started.
F Database file AY Journaled changes applied to a physical file member (APYJRNCHG). See
member Table 115 on page 822.
operation
| CB Physical file member changed.
CE Change end of data for physical file member. See Table 116 on page 823.
CH18 Change file.
CL10 Physical file member closed (for shared files, a close entry is made for
the last close operation of the file). See Table 132 on page 847.
CR Physical file member cleared (CLRPFM).
DE Physical file member deleted record count.
| DM10 Delete member. See Table 148 on page 855.
EJ10 Journaling for a physical file member ended (ENDJRNPF).
EP10 Journaling access path for a database file member ended (ENDJRNAP).
FD10 Physical file member forced (written) to auxiliary storage. See Table 121
on page 824.
FI System-generated journal entry format information.
IU Physical file member in use at the time of abnormal system end.10 See
Table 123 on page 824.
IZ12 Physical file member initialized (INZPFM). See Table 122 on page 824.
JM Journaling for a physical file member started (STRJRNPF). See Table 139
on page 852.
JP Journaling access path for a database file member started (STRJRNAP).
| MC Create member. See Table 148 on page 855.
MD10 Physical file member deleted. This entry is created when you remove
the member (RMVM) or delete the file (DLTF) containing the member.
MF Physical file member saved with storage freed (SAVOBJ, SAVCHGOBJ,
or SAVLIB)5, 10.
MM Physical file containing the member moved to a different library
(MOVOBJ or RNMOBJ OBJTYPE(*LIB)). See Table 131 on page 847.
MN Physical file containing the member renamed (RNMM or RNMOBJ). See
Table 131 on page 847.
MR Physical file member restored (RSTOBJ or RSTLIB)5. See Table 136 on
page 849.
MS Physical file member saved (SAVOBJ, SAVLIB, or SAVCHGOBJ)5,10. See
Table 137 on page 850.
OP Physical file member opened (for shared files, an open entry is added
for the first open operation for the file). See Table 132 on page 847.

Appendix D. Journal Entry Information 809

 Table 110. Journal Entries by Journal Code and Type (continued)
Journal
Code Description Entry Type Entry Type Description

PD10 Database file member’s access path deleted (this entry is created when
you remove the member (RMVM) or delete the file (DLTF) containing
the member). See Table 119 on page 823.
PM11 The logical owner of a journaled access path was moved (MOVOBJ or
RNMOBJ OBJTYPE(*LIB)). See Table 131 on page 847.
PN11 The logical owner of a journaled access path was renamed (RNMOBJ or
RNMM). See Table 131 on page 847.
RC Journaled changes removed from a physical file member
(RMVJRNCHG). See Table 115 on page 822.
RG Physical file member reorganized (RGZPFM). See Table 134 on page 848.
| RM Member reorganized.
SA The point at which the APYJRNCHG command started running.
SR The point at which the RMVJRNCHG command started running.
SS The start of the save of a physical file member using the
save-while-active function5,10. See Table 138 on page 851.
I Internal IB Internal recovery.
operation
IC Access path recovery.
| IE Directory recovery.
| IF Access path recovery.
IH Access path recovery.
| II10 Access path in use.
IV Access path recovery.
J Journal or EZ10 End journaling for journal receiver.
receiver
operation
IA System IPL after abnormal end.10 See Table 123 on page 824.
IN System IPL after normal end.10 See Table 123 on page 824.
JR Start journaling for journal receiver.
KR Keep journal receivers for recovery.
LA10 Activate local journal
LI10 Inactivate local journal
NK Do not keep journal receivers for recovery.
NR Identifier for the next journal receiver (the receiver that was attached
when the indicated receiver was detached).10 See Table 117 on page 823.
PR Identifier for the previous journal receiver (the receiver that was
detached when the indicated receiver was attached).10 See Table 117 on
page 823.
RD10 Deletion of a journal receiver (DLTJRNRCV). See Table 120 on page 824.
RF Storage for a journal receiver freed (SAVOBJ, SAVCHGOBJ, or
SAVLIB).10 See Table 120 on page 824.
RR Restore operation for a journal receiver (RSTOBJ or RSTLIB)5. See
Table 136 on page 849.
RS Save operation for a journal receiver (SAVOBJ, SAVCHGOBJ, or
5, 10
SAVLIB) . See Table 136 on page 849.
| UA10 User independent auxiliary storage pool vary on abnormal.
| UN10 User independent auxiliary storage pool vary on normal.
XP Internal entry.
L License LK License key is not valid. See Table 140 on page 853.
management
LL Usage limit changed. See Table 141 on page 853.
LU Usage limit exceeded. See Table 142 on page 853.

810 OS/400 Backup and Recovery V5R1

 Table 110. Journal Entries by Journal Code and Type (continued)
Journal
Code Description Entry Type Entry Type Description

| M TCP/IP and MP Modification of QoS policies.

Network
management
data
SN7 Simple Network Management Protocol (SNMP) information
TF13 IP filter rules actions
TN13 IP NAT rules actions
TS13 Virtual Private Networking (VPN) information
O Object oriented AI Update, after image
entry
BI Update, before image
XA Allocate object
XB Bundled entries
XD Deallocate object
XI Index operation
XS Synchronization
XT Transaction state change
P Performance TP Performance shared pool change.
tuning entry1
| Q Data queue QB Start data queue journaling. See Table 139 on page 852.
| operation17
| QC19 Data queue cleared, no key.
| QD10 Data queue deleted.
| QE10 End data queue journaling.
| QI10 Queue in use at abnormal end. See Table 123 on page 824.
| QJ Data queue cleared, has key. See Table 144 on page 854.
| QK12 Send data queue entry, has key. See Table 145 on page 854.
| QL Receive data queue entry, has key. See Table 146 on page 854.
| QM Data queue moved. See Table 131 on page 847.
| QN Data queue renamed. See Table 131 on page 847.
| QR19 Receive data queue entry, no key.
| QS12 Send data queue entry, no key. See Table 147 on page 855.
| QX5, 10 Start of save for data queue. See Table 138 on page 851.
| QY5, 10 Data queue saved. See Table 137 on page 850.
| QZ5 Data queue restored. See Table 136 on page 849.
| VE19 Internal entry.
| VQ19 Internal entry.
R Operation on BR12, 14 Before-image of record updated for rollback operation. See Table 133 on
specific record page 848.
DL12 Record deleted in the physical file member. See Table 133 on page 848.
DR12 Record deleted for rollback operation. See Table 133 on page 848.
| IL19 Increment record limit.
PT12 Record added to a physical file member. If the file is set up to reuse
deleted records, then you may receive either a PT or PX journal entry
for the change. See Table 133 on page 848.
PX12, 14 Record added directly by RRN (relative record number) to a physical
file member. If the file is set up to reuse deleted records, then you may
receive either a PT or PX journal entry for the change. See Table 133 on
page 848.
UB 8, 12, 14 Before-image of a record that is updated in the physical file member
(this entry is present only if IMAGES(*BOTH) is specified on the
STRJRNPF command). See Table 133 on page 848.

Appendix D. Journal Entry Information 811

 Table 110. Journal Entries by Journal Code and Type (continued)
Journal
Code Description Entry Type Entry Type Description

UP 8, 12, 14 After-image of a record that is updated in the physical file member. See
Table 133 on page 848.
UR12, 14 After-image of a record that is updated for rollback information. See
Table 133 on page 848.
S Distributed mail AL SNA alert focal point information. 2
services
CF Mail configuration information. 2,6
DX X.400 process debug entry 4
ER Mail error information. 2,6
LG Mail logging table information. 2,6
MX A change was made to X.400 MTA configuration4.
NX A change was made to X.400 delivery notification4.
RT Mail routing information. 2,6
RX A change was made to X.400 route configuration4.
SY Mail system information. 2,6
UX A change was made to X.400 user or probe4.
XE DSNX error entry. 2
XL DSNX logging entry. 2
XX An error was detected by the X.400 process4.
T Audit trail AD A change was made to the auditing attribute.
entry3
AF All authority failures.
AP A change was made to program adopt.
CA Changes to object authority (authorization list or object).
CD A change was made to a command string.
CO Create object.
CV Connection verification.
CP Create, change, restore user profiles.
CQ A change was made to a change request descriptor.
CU Cluster operation
| CY Cryptographic configuration
| DI Directory services
DO All delete operations on the system.
DS DST security officer password reset.
EV Environment variable
GR General purpose audit record
GS A descriptor was given.
IP Inter-process communication event.
IR IP rules actions
IS Internet security management
JD Changes to the USER parameter of a job description.
JS A change was made to job data.
KF Key ring file name.
LD A link, unlink, or lookup operation to a directory.
ML A change was made to office services mail.
NA Changes to network attributes.
ND Directory search violations.
NE End point violations.
OM Object management change.
OR Object restored.
OW Changes to object ownership.
O1 Single optical object access.

812 OS/400 Backup and Recovery V5R1

 Table 110. Journal Entries by Journal Code and Type (continued)
Journal
Code Description Entry Type Entry Type Description

O2 Dual optical object access.

O3 Optical volume access.
PA Changes to programs (CHGPGM) that will now adopt the owner’s
authority.
PG Changes to an object’s primary group.
PO A change was made to printed output.
PS Profile swap.
PW Passwords used that are not valid.
RA Restore of objects when authority changes.
RJ Restore of job descriptions that contain user profile names.
RO Restore of objects when ownership information changes.
RP Restore of programs that adopt their owner’s authority.
RQ A change request descriptor was restored.
RU Restore of authority for user profiles.
RZ The primary group for an object was changed during a restore
operation.
SD A change was made to the system directory.
SE Changes to subsystem routing.
SF A change was made to a spooled output file.
SG Asynchronous signals
SK Secure sockets connection
SM A change was made by system management.
SO A change was made by server security.
ST A change was made by system tools.
SV Changes to system values.
VA Changes to access control list.
VC Connection started or ended.
VF Server files were closed.
VL An account limit was exceeded.
VN A logon or logoff operation on the network.
VO Actions on validation lists.
VP A network password error.
VR A network resources was accessed.
VS A server session started or ended.
VU A network profile was changed.
VV Service status was changed.
| X0 Network authentication.
X1 Reserved for future audit entry.
X2 Reserved for future audit entry.
X3 Reserved for future audit entry.
X4 Reserved for future audit entry.
X5 Reserved for future audit entry.
X6 Reserved for future audit entry.
X7 Reserved for future audit entry.
X8 Reserved for future audit entry.
X9 Reserved for future audit entry.
YC A change was made to DLO change access.
YR A change was made to DLO read access.
ZC9 A change was made to object change access.
ZM An object was accessed using a method.
ZR A change was made to Object read access.

Appendix D. Journal Entry Information 813

 Table 110. Journal Entries by Journal Code and Type (continued)
Journal
Code Description Entry Type Entry Type Description

U User generated User-specified. The Entry-specific data is the value specified on the
entry ENTDTA parameter of the SNDJRNE command or with the entry data
parameter for the QJOSJRNE API.

Notes:
1
See the Work Management book for the layout of the entry specific data.
2
See the SNA Distribution Services book for the layout of the entry specific data for entries generated by
SNADS.
3
See the iSeries Security Reference book for the layout of the entry specific data.
4
See the OSI Message Services/400 Support book for the layout of the entry specific data.
5
These entries do not indicate that they occurred as the result of a trigger program, even if a trigger
program caused the event. That information is not available at the time the entry is written to the journal.
6
See the AnyMail/400 Mail Server Framework Support book for the layout of the entry specific data.
7
See the Simple Network Management Protocol (SNMP) Support book for information about the entry specific
data for SNMP journal entries.
8
Neither the before- nor after-image is deposited into the journal if the after-image is exactly the same as the
before-image.
9
Only one entry per opened file. The member name will not be displayed in the entry specific data for
based on physical files.
10
Even if this journal is a local journal that has a journal state of *INACTIVE, meaning no journal can be
deposited, this journal entry type will still be deposited.
11
After you have installed V4R2M0 or a later release, this journal type is no longer generated.
12
This journal entry may have data which can only be accessed by using either the QjoRetrieveJournalEntries
API or the command RCVJRNE ENTFMT(*TYPEPTR). In all other interfaces, if the data is not visible the
incomplete data indicator will be on, and *POINTER will appear in the Entry Specific Data. For more
information, refer to “Working With Pointers in Journal Entries” on page 437.
13
Refer to the TCP/IP Configuration and Reference book for information about the entry specific data for
TCP/IP journal entries.
14
| This entry may have minimized entry specific data (ESD). It will have minimized ESD if its corresponding
| object type deposits minimized journal entries through the MINENTDDTA parameter for this journal or
| journal receiver.
15
| The entry-specific data for these journal entries are laid out in the QSYSINC include QPOLJRNL.H.
16
| The entry-specific data for these journal entries are laid out in the QSYSINC include QWCJRNL.H.
17
| The entry-specific data for these journal entries are laid out in the QSYSINC include QMHQJRNL.H.
18
| As of V5R1M0, the journal entry D CG is also being sent for the change file operations. IBM strongly
| recommends that you do your processing based on the D CG entry instead of the F CH entry because the
| F CH entry will be retired in a future release.
19
| These entries have entry-specific data which the system uses for internal processing.

814 OS/400 Backup and Recovery V5R1

 The Fixed-Length Portion of a Journal Entry
Table 111 shows the fields that are common for all journal entry types. These fields
are shown when you request *TYPE1 for the output file format or the entry type
format. The field names shown in parentheses are used in the system-supplied
output file QSYS/QADSPJRN.
Table 111. Field Descriptions of the Fixed-Length Portion of a Journal Entry: *TYPE1
Offset Field Format Description

1 Entry length Zoned (5,0) Specifies the length of the journal entry including the entry
(JOENTL) length field, all subsequent positions of the journal entry,
and any portion of the journal entry that was truncated if
the length of the output record is less than the length of the
record created for the journal entry.
Note: If the journal entry has the incomplete data indicator
on, then this length does not include that additional data
which could be pointed to. This length includes the length
of the data that is actually returned, which includes entry
specific data of up to 32 766 bytes.
6 Sequence number Zoned (10,0) Assigned by the system to each journal entry. It is initially
(JOSEQN) set to 1 for each new or restored journal and is incremented
until you request that it be reset when you attach a new
receiver. There are occasional gaps in the sequence numbers
because the system uses internal journal entries for control
purposes. These gaps occur if you use commitment control,
journal physical files, or journal access paths.
16 Journal code Char (1) Identifies the primary category of the journal entry:
(JOCODE)
A= System accounting entry
| B= Integrated file system operation
C= Commitment control operation
D= Database file operation
| E= Data area operation
F= Database file member operation
| I= Internal operation
J= Journal or receiver operation
L= License management
M= Network management data
O= Object oriented entry
P= Performance tuning entry
| Q= Data queue operation
R= Operation on a specific record
S= Distributed mail services
T= Audit trail entry
U= User-generated entry (added by the SNDJRNE
command or QJOSJRNE API)
The journal codes are described in more detail in “Journal
Codes” on page 806.
17 Entry type Char (2) Further identifies the type of user-created or system-created
(JOENTT) entry. See Table 110 on page 807 for descriptions of the entry
types.

Appendix D. Journal Entry Information 815

 Table 111. Field Descriptions of the Fixed-Length Portion of a Journal Entry: *TYPE1 (continued)
Offset Field Format Description

19 Date stamp Char (6) Specifies the system date when the entry was added and is
(JODATE) in the format of the job attribute DATFMT. The system
cannot assure that the date stamp is always in ascending
order for sequential journal entries because you can change
the value of the system date.
25 Time stamp Zoned (6,0) Corresponds to the system time (in the format hhmmss)
(JOTIME) when the entry was added. The system cannot assure that
the time stamp is always in ascending order for sequential
journal entries because you can change the value of the
system time.
31 Job name (JOJOB) Char (10) Specifies the name of the job that added the entry.
Notes:
1. If RCVSIZOPT(*MINFIXLEN) was in effect for the
journal when the journal receiver that contains this
journal entry was attached, *OMITTED is given for the
job name.
2. If the job name was not available when the journal entry
was deposited, then *NONE is written for the job name.
41 User name Char (10) Specifies the user profile name of the user that started the
(JOUSER) job.
Note: If the RCVSIZOPT(*MINFIXLEN) was in effect for the
journal when the journal receiver that contains the journal
entry was attached, then blanks are written for the user
name.
51 Job number Zoned (6,0) Specifies the job number of the user that started the job.
(JONBR) Note: If the RCVSIZOPT(*MINFIXLEN) was in effect for the
journal when the journal receiver that contains the journal
entry was attached, then zeroes are written for the job
number.
57 Program name Char (10) Specifies the name of the program that added the entry. If an
(JOPGM) application or CL program did not add the entry, the field
contains the name of a system-supplied program such as
QCMD or QPGMMENU. If the program name is the special
value *NONE, then one of the following is true:
v The program name does not apply to this journal entry.
v The program name was not available when the journal
entry was made.

For example, the program name is not available if the

program was destroyed.
Notes:
1. If the program that deposited the journal entry is an
original program model program, this data will be
complete. Otherwise, this data will be unpredictable.
2. If RCVSIZOPT(*MINFIXLEN) was in effect for the
journal when the journal receiver that contains this
journal entry was attached, *OMITTED is given for the
program name.
67 Object name Char (10) Specifies the name of the object for which the journal entry
(JOOBJ) was added.1 This is blank for some entries.

| If the journaled object is an integrated file system object,

| then this field is the first 10 bytes of the file identifier (FID).

816 OS/400 Backup and Recovery V5R1

 Table 111. Field Descriptions of the Fixed-Length Portion of a Journal Entry: *TYPE1 (continued)
Offset Field Format Description

77 Library name Char (10) Specifies the name of the library containing the object1.
(JOLIB)
| If the journaled object is an integrated file system object,
| then the first 6 characters of this field are the last 6 bytes of
| the FID.
87 Member name Char (10) Specifies the name of the physical file member or is blank if
(JOMBR) the object is not a physical file1.
97 Count/relative Zoned (10,0) Contains either the relative record number (RRN) of the
record number record that caused the journal entry or a count that is
(JOCTRR) pertinent to the specific type of journal entry. Table 115 on
page 822 through Table 142 on page 853 show specific values
for this field, if applicable.
107 Indicator flag Char (1) Contains an indicator for the operation. Table 115 on
(JOFLAG) page 822 through Table 142 on page 853 show specific values
for this field, if applicable.
108 Commit cycle Zoned (10,0) Contains a number that identifies the commit cycle. A
identifier (JOCCID) commit cycle is from one commit or rollback operation to
another.

The commit cycle identifier is found in every journal entry

that is associated with a commitment transaction. If the
journal entry was not made as part of a commitment
transaction, this field is zero.
118 Incomplete Data Char (1) Indicates whether this entry has data that is not being
(JOINCDAT) retrieved for one of the following reasons:
v The length of the entry-specific data exceeds 32 766 bytes.
v The entry is associated with a database file that has one or
more fields of data type BLOB (binary large object), CLOB
(character large object), or DBCLOB (double-byte character
large object).
The possible values are:
0 = This entry has all possible data.
1 = This entry has incomplete data.

Any data which is marked as incomplete, can only be

viewed by using either the QjoRetrieveJournalEntries API, or
the command RCVJRNE ENTFMT(*TYPEPTR).
| 119 Minimized entry Char (1) Indicates whether this entry has minimized entry specific
| specific data data.
| (JOMINESD)
| The possible values are:
| 0 = This entry has complete entry specific data.
| 1 = This entry has minimized entry specific data.
|
| 120 Reserved field Char (6) Always contains zeros. Contains hexadecimal zeros in the
(JORES) output file.

Appendix D. Journal Entry Information 817

 Table 111. Field Descriptions of the Fixed-Length Portion of a Journal Entry: *TYPE1 (continued)
Offset Field Format Description

Notes:
1
If the journal receiver was attached prior to installing V4R2M0 on your system, then the following items
are true:
v If *ALLFILE is specified for the FILE parameter on the DSPJRN, RCVJRNE, or RTVJRNE command, then
the fully qualified name is the most recent name of the file when the newest receiver in the receiver
range was the attached receiver and when the file was still being journaled.
v If a file name is specified or if library *ALL is specified on the FILE parameter, the current fully qualified
name of the file appears in the converted journal entry.

If the journal receiver was attached while V4R2M0 or a later release was running on the system, the fully
qualified name is the name of the object at the time the journal entry was deposited.

If OUTFILFMT(*TYPE2) is requested on the DSPJRN command, or

ENTFMT(*TYPE2) is requested on the RCVJRNE or RTVJRNE command, then the
fixed-length portion of each converted journal entry is the same as the format in
Table 111 on page 815, except for the information that follows the commit cycle
identifier field. The fields of the prefix that follow the commit cycle identifier are
shown in Table 112. The field names shown in parentheses in the table are the
names of the fields in the system-supplied output file QSYS/QADSPJR2.
Table 112. Field Descriptions of the Fixed-Length Portion of a Journal Entry: *TYPE2
Offset Field Format Description

1 117 The same as *TYPE1. See Table 111 on page 815.

118 User profile Char (10) Specifies the name of the user profile under which the job was
(JOUSPF) running when the entry was created.
Note: If RCVSIZOPT(*MINFIXLEN) was in effect for the journal
when the journal receiver that contains this journal entry was
attached, *OMITTED is given for the user profile.
128 System name Char (8) Specifies the name of the system on which the entry is being
(JOSYNM) displayed, printed, retrieved, or received if the journal receiver
was attached prior to installing V4R2M0 on the system. If the
journal receiver was attached while the system was running
V4R2M0 or a later release, the system name is the system where
the journal entry was actually deposited.
136 Incomplete data Char (1) The same as *TYPE1. See Table 111 on page 815.
(JOINCDAT)
| 137 Minimized entry Char (1) The same as *TYPE1. See Table 111 on page 815.
| specific data
| (JOMINESD)
138 Reserved field Char (18) Always contains zeros. Contains hexadecimal zeros in the output
(JORES) file.

A third value, *TYPE3, is supported on the OUTFILFMT parameter for the DSPJRN
command, and the ENTFMT parameter for the RCVJRNE and RTVJRNE
commands. If either OUTFILFMT(*TYPE3) is specified on the DSPJRN command
or ENTFMT(*TYPE3) is specified on the RCVJRNE or RTVJRNE command, the
information in the prefix portion of a converted journal entry is shown in Table 113
on page 819. The field names shown in parentheses in the table are the names of
the fields in the system-supplied output file QSYS/QADSPJR3. *TYPE3 has the
same information as the *TYPE1 and *TYPE2 formats, except that it has a different
date format and a null-values indicator.

818 OS/400 Backup and Recovery V5R1

 Table 113. Field Descriptions of the Fixed-Length Portion of a Journal Entry: *TYPE3
Offset Field Format Description

1 Entry length Zoned (5,0) See Table 111.

(JOENTL)
6 Sequence Zoned decimal See Table 111.
number (10,0)
(JOSEQN)
16 Journal code Char (1) See Table 111.
(JOCODE)
17 Entry type Char (2) See Table 111.
(JOENTT)
19 Time stamp Char (26) Corresponds to the system date and time when the journal entry
(JOTMST) was added in the journal receiver. The time stamp is in SAA
format. The system cannot assure that the time stamp is always in
ascending order for sequential journal entries because you can
change the value of the system time.
45 Job name Char (10) See Table 111.
(JOJOB)
55 User name Char (10) See Table 111.
(JOUSER)
65 Job number Zoned (6,0) See Table 111.
(JONBR)
71 Program name Char (10) See Table 111.
(JOPGM)
81 Object name Char (10) See Table 111.
(JOOBJ)
91 Library name Char (10) See Table 111.
(JOLIB)
101 Member name Char (10) See Table 111.
(JOMBR)
111 Count/relative Zoned (10,0) See Table 111.
record number
(JOCTRR)
121 Indicator flag Char (1) See Table 111.
(JOFLAG)
122 Commit cycle Zoned (10,0) See Table 111.
identifier
(JOCCID)
132 User profile Char (10) See Table 112.
(JOUSPF)
142 System name Char (8) See Table 112.
(JOSYNM)
150 Incomplete data Char (1) See Table 111.
(JOINCDAT)
| 151 Minimized entry Char (1) See Table 111.
| specific data
| (JOMINESD)
| 152 Reserved field Char (18) Always contains zeros. Contains hexadecimal zeros in the output
(JORES) file. This field is shown in the output file only when the DSPJRN
command is used. It is not shown when the RCVJRNE command
or the RTVJRNE command is used.

Appendix D. Journal Entry Information 819

 Table 113. Field Descriptions of the Fixed-Length Portion of a Journal Entry: *TYPE3 (continued)
Offset Field Format Description

Notes:
1
If the journal receiver was attached prior to installing V4R2M0 on your system, then the following items
are true:
v If *ALLFILE is specified for the FILE parameter on the DSPJRN, RCVJRNE, or RTVJRNE command, then
the fully qualified name is the most recent name of the file when the newest receiver in the receiver
range was the attached receiver and when the file was still being journaled.
v If a file name is specified or if library *ALL is specified on the FILE parameter, the current fully qualified
name of the file appears in the converted journal entry.

If the journal receiver was attached while V4R2M0 or a later release was running on the system, the fully
qualified name is the name of the object at the time the journal entry was deposited.

A fourth value, *TYPE4, is supported on the OUTFILFMT parameter for the

DSPJRN command and the ENTFMT parameter for the RCVJRNE and RTVJRNE
commands. If either OUTFILFMT(*TYPE4) is specified on the DSPJRN command
or ENTFMT(*TYPE4) is specified on the RCVJRNE or RTVJRNE command, the
information in the prefix portion of a converted journal entry is shown in
Table 114. The field names shown in parentheses in the table are the names of the
| fields in the system-supplied output file QSYS/QADSPJR4. *TYPE4 output
| includes all of the *TYPE3 information, plus information about journal identifiers,
| triggers, and referential constraints and entries which will be ignored by the
| APYJRNCHG or RMVJRNCHG commands.
Table 114. Field Descriptions of the Fixed-Length Portion of a Journal Entry: *TYPE4
Offset Field Format Description

1 149 These fields are the same as for *TYPE3. See Table 113.
150 Journal Char(10) Specifies the journal identifier (JID) for the object. When journaling
identifier is started for an object, the system assigns a unique JID to that
(JOJID) object. The JID remains constant even if the object is renamed or
moved. However, if journaling is stopped, there is no guarantee
that the JID will be the same if journaling is started again for the
same object.

If no JID is associated with the entry, this field has hex zeros.
160 Referential Char(1) Indicates whether this entry was recorded for actions that occurred
constraint on records that are part of a referential constraint.
(JORCST)
0 = This entry was not created as part of a referential
constraint.
1 = This entry was created as part of a referential constraint.

161 Trigger (JOTGR) Char(1) Indicates whether this entry was created as result of a trigger
program.
0 = This entry was not created as the result of a trigger
program.
1 = This entry was created as the result of a trigger
program.

162 Incomplete data Char (1) See Table 111.

(JOINCDAT)

820 OS/400 Backup and Recovery V5R1

 Table 114. Field Descriptions of the Fixed-Length Portion of a Journal Entry: *TYPE4 (continued)
Offset Field Format Description

| 163 Ignored by Char (1) Indicates whether this journal entry will be ignored by the
| APYJRNCHG or execution of the APYJRNCHG or RMVJRNCHG commands, even
| RMVJRNCHG though normally this journal entry type has an effect during those
| (JOIGNAPY) command invocations.

| Possible invocations are as follows:

| 0 = This entry is not ignored by the APYJRNCHG or
| RMVJRNCHG commands.
| 1 = This entry is ignored by the APYJRNCHG or
| RMVJRNCHG commands.
|
| 164 Minimized entry Char (1) See Table 111.
| specific data
| (JOMINESD)
| 165 Reserved area Char (5) Always contains zeros. Contains hexadecimal zeros in the output
(JORES) file.

Variable-Length Portion of a Journal Entry

For output formats *TYPE1 and *TYPE2, the variable length portion of the journal
entry includes just the Entry-specific data field. The contents of the Entry-specific data
field depends on the journal entry code and entry type.

For output formats *TYPE3, *TYPE4, and *TYPEPTR, the variable-length portion of
the converted journal entry potentially has two fields.

| Note: For the layout of the output format *TYPEPTR, look under the
| QjoRetrieveJournalEntries API information in the CL and APIs topics on the
| Information Center at the following Web site:
| http://www.ibm.com/eserver/iseries/infocenter.

The first field is the Null value indicators, and the second field is the Entry-specific
data. The Null Value Indicators field contains relevant information only for entries
| with journal code R. Null value indicators are present in journal entries for record
| level operations as follows:
| v The corresponding physical file has null capable fields.
| v The record image has been minimized in the entry specific data.
| Otherwise, it contains blanks. If the record image has not been minimized in the
| entry specific data, the Null Value Indicators field is a character string with one
| character for each field in the physical file that has record images appearing in the
| journal. Each character has the following interpretation:
v 0 = corresponding field in the record is not NULL.
v 1 = corresponding field in the record is NULL.

The second field in the variable-length portion of the journal entry is the
Entry-specific data in the journal entry.

The Null Value Indicators and Entry-Specific Data fields have been defined as
variable-length character fields in the system-supplied output files
QSYS/QADSPJR3 and QSYS/QADSPRJ4. See the DSPJRN, RCVJRNE, and
RTVJRNE commands in the CL Programming book for additional details regarding

Appendix D. Journal Entry Information 821

| the *TYPE3 and *TYPE4 formats and the exact layout of these two fields. Look
| under the QjoRetrieveJournalEntries API information in the CL and APIs topics on
| the Information Center at the following Web site:
| http://www.ibm.com/eserver/iseries/infocenter for additional details regarding
| the *TYPEPTR format and the exact layout returned.

Table 115 through Table 142 on page 853 show the layouts for some of the journal
entry types. Some journal entry types are described in other books, as indicated in
| Table 110 on page 807. Some journal entry types are documented in QSYSINC
| library includes, as indicated in Table 110 on page 807. Some entry types do not
have entry-specific data.

These layouts include specific values for fields in the fixed-length portion of the
entry and the fields in the entry-specific portion of the entry. The offsets show the
relative offset within the Entry-specific data field. The beginning position of the
Entry-specific data field depends on the format type that you specify. See Table 111
on page 815, Table 112 on page 818, Table 113 on page 819, and Table 114 on
page 820 for the format type layouts.
Table 115. APYJRNCHG (B AJ, E EQ, F AY) and RMVJRNCHG (E EX, F RC) Journal Entries
Relative
Offset Field Format Description

Specific values for this entry type:

Count/Relative Zoned (10,0) Contains the number of journal entries applied or removed.
Record Number
(JOCTRR)
Flag (JOFLAG) Char (1) The results of the apply or remove operation:
0= Command completed normally.
1= Command completed abnormally.

Entry-specific data. This data appears as one field in the standard output formats:
1 First entry applied or Zoned (10,0) Sequence number of the first entry actually applied or
removed removed.
11 Last entry applied or Zoned (10,0) Sequence number of the last entry actually applied or
removed removed.
21 Starting Receiver Char (10) The name of the first receiver from which entries were
Name applied or removed.
31 Library Name Char (10) The library for the starting receiver.
41 Ending Receiver Char (10) The name of the last or ending receiver from which entries
Name were applied or removed.
51 Library Name Char (10) The library for the ending receiver.
61 Starting sequence Char (10) Starting sequence number for the apply or remove operation
number
71 Ending sequence Char (10) Ending sequence number for the apply or remove operation
number
81 Incomplete commit Char (1) 0 = Indicates that either CMTBDY(*NO) was specified or
transaction not CMTBDY(*YES) was specified and no partial commitment
applied or removed control transactions were found in the range specified by the
starting and ending sequence numbers

1 = Indicates that CMTBDY(*YES) was specified and one or

more partial commitment control transactions were found in
the range specified by the starting and ending sequence
numbers

822 OS/400 Backup and Recovery V5R1

Table 116. Change End of Data (F CE) Journal Entry
Relative
Offset Field Format Description
Specific values for this entry type:
Count/Relative Zoned (10,0) The relative record number of the last record kept in the
Record Number physical file member.
(JOCTRR)

Table 117. CHGJRN (J NR, J PR) Journal Entries

Relative
Offset Field Format Description

Specific values for this entry type:

Count/Relative Zoned (10,0) Contains the number of receivers attached or detached.
Record Number
(JOCTRR)
Entry-specific data. This data appears as one field in the standard output formats:
1 First Receiver Name Char (10) The name of the first receiver that is attached or detached.
11 First Receiver Library Char (10) The name of the library for the first receiver that is attached
Name or detached.
21 Dual Receiver Name Char (10) The name of the dual receiver that is attached or detached.
Blank if only one receiver is used for the journal.
31 Dual Receiver Library Char (10) The name of the library for the dual receiver that is attached
Name or detached. Blank if only one receiver is used for the
journal.

Table 118. COMMIT (C CM) Journal Entry

Relative
Offset Field Format Description

Specific values for this entry type:

Count/Relative Zoned (10,0) Contains the length of the commit identification.
Record Number
(JOCTRR)
Flag (JOFLAG) Char (1) Whether the commit operation was initiated by the system or
the user:
0= All record-level changes were committed for a
commit operation initiated by a user.
2= All record-level changes were committed for a
commit operation initiated by the operating system.

Entry-specific data. This data appears as one field in the standard output formats:
1 Commit ID Char (*) Contains the commit identification specified by the operation.
The Count field specifies the length of this field.

Table 119. Delete Access Path (R PD) Journal Entry

Relative
Offset Field Format Description
Specific values for this entry type:
Journal identifier Char(10) The JID is provided with the *TYPE4 format only. It can be
(JOJID) used with the QJORJIDI API.

Appendix D. Journal Entry Information 823

 Table 120. Delete Receiver (J RD, J RF) Journal Entries
Relative
Offset Field Format Description
Specific values for this entry type:
Journal identifier Char(10) The JID is provided with the *TYPE4 format only. It can be
(JOJID) used with the QJORJIDI API.

Table 121. Force Data to Auxiliary Storage (F FD) Journal Entry

Relative
Offset Field Format Description

Specific values for this entry type:

Job name (JOJOB) Char (10) Blank if the entry is written during IPL.
Job number (JONBR) Zoned (6,0) Zero if entry is written during IPL.
Program name Char (10) Blank if the entry is written during IPL.
(JOPGM)

Table 122. INZPFM (F IZ) Journal Entry

Relative
Offset Field Format Description

Specific values for this entry type:

Count/Relative Zoned (10,0) Contains the number of records specified on the TOTRCDS
Record Number parameter of the Initialize Physical File Member (INZPFM)
(JOCTRR) command.
Flag (JOFLAG) Char (1) Indicates the type of record initialization that was done:
0= *DFT (default)
1= *DLT (delete)

Entry-specific data. This data appears as one field in the standard output formats:
1 Entry-specific data If the member is initialized with default records, this field
contains the default record image.

| Table 123. IPL (J IA, J IN) and In-Use (B OI, D ID, E EI, F IU, Q QI) Journal Entries
| Relative
| Offset Field Format Description

| Specific values for this entry type:

| Time stamp (JOTIME) Zoned (6,0) The timestamp created at IPL is read from the
| battery-powered clock. If the battery-powered clock cannot
| be read, the time is that of the system power down, not the
| time of the IPL, because the system time has not yet been
| updated at the time the journal entry is written.
| Flag (JOFLAG) Char (1) For in-use entries, indicates whether the object was
| synchronized with the journal:
| 0= Object was synchronized with journal
| 1= Object was not synchronized with journal.
|
|
Table 124. Logical Unit of Work (C LW) Journal Entry
Relative
Offset Field Format Description

Entry-specific data. This data appears as one field in the standard output formats:

824 OS/400 Backup and Recovery V5R1

Table 124. Logical Unit of Work (C LW) Journal Entry (continued)
Relative
Offset Field Format Description

1 LUW Header Portion 416 The header portion of the entry-specific data contains general
information about the logical unit of work (LUW). Table 125
describes the contents of the header portion.
After the LUW Local Portion 80 Information about local resources that participated in the
header LUW. The entry may have 0 to n records for local locations.
portion Each local record is 48 characters long. Table 126 on page 834
describes the local record.
After the LUW API Portion 112 Information about API resources that participated in the
local LUW. The entry may have 0 to n records for API resources.
portion Each API resource record is 80 characters long. Table 127 on
page 835 describes the API record.
After the LUW DDL Portion 96 Information about DDL resources that participated in the
API LUW. The entry may have 0 to n records for DDL resources.
portion Each DDL resource record is 80 characters long. Table 128 on
page 838 describes the DDL record.
After the LUW Remote Portion 128 Information about remote locations that participated in the
DDL LUW. The entry may have 0 to n records for remote
portion locations. Each remote location record is 128 characters long.
Table 129 on page 841 describes the remote record.
After the LUW DDM Portion 96 Information about DDM resources that participated in the
remote LUW. The entry may have 0 to n records for DDM resources.
portion Each DDM resource record is 96 characters long. Table 130 on
page 845 describes the DDM record.

Table 125. Logical Unit of Work (C LW) Journal Entry–Header Record

Relative
Offset Field Format Description

1 Record Type Char (4) Type of record:

HDR Header record.

5 Record Length Bin (15) Length of record. Currently 400 for HDR record.
7 Record Position (4)1 This identifies the position in the LUW journal entry where
this record starts. It is made up of two numbers:
v Bin (15): The relative number of the journal entry that
contains this record. If the LUW journal entry is greater
than 32K-1 bytes, multiple entries are actually sent to the
journal. This number represents which of these actual
journal entries contains this record (1 for the first, 2 for the
second, and so forth). Note that this is not the actual
journal entry sequence number.
v Bin (15): The offset where this record starts within this
journal entry. This is the number of bytes past the
beginning of the entry where this record starts. For
example, 0 means the first byte in the entry. Because they
always start at the beginning of the journal entry, this
offset is always 0 for HDR records.
11 Number of Journal Bin (15) The number of actual journal entries sent for this LUW
Entries journal entry. This is 1 unless the LUW journal entry is
greater than 32K-1 bytes.

Appendix D. Journal Entry Information 825

Table 125. Logical Unit of Work (C LW) Journal Entry–Header Record (continued)
Relative
Offset Field Format Description

13 Location with No (4)1 This identifies the position in the LUW journal entry where
Journal Position the LCL record starts for the local location with no journal. It
is made up of two numbers:
v Bin (15): The relative number of the journal entry that
contains the record. If the LUW journal entry is greater
than 32K-1 bytes, multiple entries are actually sent to the
journal. This number represents which of these actual
journal entries contains the record (1 for the first, 2 for the
second, and so forth). Note that this is not the actual
journal entry sequence number.
v Bin (15): The offset where the record starts within this
journal entry. This is the number of bytes past the
beginning of the entry where the record starts. For
example, 0 means the first byte in the entry.

Position 0 0 means that there is no local location that does

not have a journal.
17 First Location with (4)1 This identifies the position in the LUW journal entry where
Journal Position the LCL record starts for the first local location with a
journal. It is made up of two numbers:
v Bin (15): The relative number of the journal entry that
contains the record. If the LUW journal entry is greater
than 32K-1 bytes, multiple entries are actually sent to the
journal. This number represents which of these actual
journal entries contains the record (1 for the first, 2 for the
second, and so on). Note that this is not the actual journal
entry sequence number.
v Bin (15): The offset where the record starts within this
journal entry. This is the number of bytes past the
beginning of the entry where the record starts. For
example, 0 means the first byte in the entry.

Position 0 0 means that there are no local locations with a

journal.
21 First Remote Location (4)1 This identifies the position in the LUW journal entry where
Position the RMT record starts for the first remote location. It is made
up of two numbers:
v Bin (15): The relative number of the journal entry that
contains the record. If the LUW journal entry is greater
than 32K-1 bytes, multiple entries are actually sent to the
journal. This number represents which of these actual
journal entries contains the record (1 for the first, 2 for the
second, and so on). Note that this is not the actual journal
entry sequence number.
v Bin (15): The offset where the record starts within this
journal entry. This is the number of bytes past the
beginning of the entry where the record starts. For
example, 0 means the first byte in the entry.

Position 0 0 means there are no remote locations.

826 OS/400 Backup and Recovery V5R1

Table 125. Logical Unit of Work (C LW) Journal Entry–Header Record (continued)
Relative
Offset Field Format Description

25 LUW Operation Char (2) The operation that was performed to end this LUW:
CM A commit operation was performed. This does not
necessarily mean that the resources were committed.
In some cases a commit operation is changed to a
rollback operation according to two-phase commit
rules.
RB A rollback operation was performed. An attempt
was made to roll back all resources.

27 Protected Logical Unit Char (41) The format for the LUWID is:
of Work Identifier v Bin (15): The total length of the LUWID not including this
(LUWID) field
v Char (0 to 8): The network ID
v Char (1): The separator character .
v Char (0 to 8): The local location name
v Char (3): The separator characters .X’
v Char (12): The hex value of the instance number converted
to character
v Char (2): The separator characters ’.
v Char (5): The hex value of the sequence number converted
to decimal
68 Unprotected Logical Char (41) The format for the LUWID for unprotected conversations is
Unit of Work the same as for protected conversations.
Identifier
109 Default Journal Bin (31) The commit cycle identifier for the default journal for this
Commit Cycle ID LUW. This is 0 if no commit cycle was started for this journal
during this LUW. This is -1 if the actual commit cycle
identifier value is larger than 2 147 483 647. The Default
Journal Commit Cycle ID Long field always contains the
correct value.
113 Commitment Char (10) The name of the commitment definition for which this LUW
Definition Name took place.
123 Commitment Char (10) The commitment definition identifier of the commitment
Definition Identifier definition. This is not useful to the end user.
133 Qualified Job Name Char (26) The job that created the commitment definition.
159 Reserved Char (1) Reserved for future use. Currently always blank.
160 Commitment Char (1) The scope of the commitment definition:
Definition Scope
A Activation group level commitment definition.
E Explicitly named commitment definition.
J JOB commitment definition.

161 Activation Group Bin (31) The activation group mark for the commitment definition:
Mark
0 This is the *JOB or an explicitly named commitment
definition.
2 This is the *DFTACTGRP commitment definition.
# The number of the activation group for this
activation group level commitment definition.

Appendix D. Journal Entry Information 827

Table 125. Logical Unit of Work (C LW) Journal Entry–Header Record (continued)
Relative
Offset Field Format Description

165 Notify Object Char (37) The notify object for the commitment definition:
v Char (10) - Object name
v Char (10) - Object library
v Char (10) - Object member (blank if object is not a file)
v Char (7) - Object type (*MSGQ, *DTAARA or *FILE)
202 Default Journal Char (20) The default journal for the commitment definition:
v Char (10): Journal name
v Char (10): Journal library
222 Initiation Type Char (1) Whether this commit or rollback operation was initiated by
the user or by the system:
E Explicit commit or rollback operation initiated by
the user.
I Implicit commit or rollback operation due to
activation group end, job end, or system end.
If the LUW was finished after a system end, this is
set to I, even if an explicit commit or rollback
operation was running at the time the system
ended.

223 LUW End Status Char (1) Indication of when this LUW ended with respect to the job
that created the commitment definition for which this LUW
took place:
N The LUW ended while the job was running
normally.
E The LUW ended during job end. This means that
the LUW was still pending when a request was
made to end the job. If the requested operation is
CM, then a commit request had started before the
request to end the job and was finished during the
job-end phase.
I The LUW ended during the IPL following a system
end. If the requested operation is CM, then a
commit request was started before the system end
and was finished during the IPL.
P The LUW ended after the IPL following a system
end. In this case, the requested operation is CM and
the LUW was prepared pending the
commit/rollback decision from the initiator or last
agent when the system ended. During the IPL, local
resources were brought back to a prepared state in a
system database server job. After resynchronization
was performed to learn the commit/rollback
decision, the LUW ended by committing or rolling
back the local resources in that same system
database server job.

828 OS/400 Backup and Recovery V5R1

Table 125. Logical Unit of Work (C LW) Journal Entry–Header Record (continued)
Relative
Offset Field Format Description

224 Sync-point Role Char (1) The sync-point role played by this location during a commit
operation:
I Initiator: the root of the sync-point tree.
C Cascaded initiator: an intermediate location in the
sync-point tree.
A Agent: a leaf location in the sync-point tree.
blank This LUW ended in a rollback request.

225 Partner Role Char (1) The partner role played by this location during a commit:
I Initator: the root of the sync-point tree.
N Not-last agent: a prepare request was sent to this
location during the prepare wave.
L Last agent: a prepare request was not sent to this
location during the prepare wave. Instead, a request
was made to this location during the committed
wave to attempt a full commit operation before
reporting results back to its initiator.
blank This LUW ended in a rollback request.

Appendix D. Journal Entry Information 829

Table 125. Logical Unit of Work (C LW) Journal Entry–Header Record (continued)
Relative
Offset Field Format Description

226 LUW Disposition Char (2) The overall disposition of the LUW:
RO This location and all downstream locations voted
read-only. These resources were not committed or
rolled back because they were not changed during
the LUW. It is not known whether the other
locations in the sync-point tree committed or rolled
back.
CM All resources committed. No errors have been
detected to this point. If the Resync In Progress
Indicator field is N, the LUW has completely
committed. Otherwise, resynchronization is still
going on to assure this location that other locations
committed completely.
CF An attempt was made to commit all resources, but
one or more errors have occurred. The job log,
QHST, and QSYSOPR *MSGQ can be checked to
determine the errors.
RB All resources rolled back successfully.
RF An attempt was made to roll back all resources, but
one or more errors have occurred. The job log,
QHST, and QSYSOPR *MSGQ can be checked to
determine the errors.
HD Heuristic damage has occurred. This means one of
two things:
1. Some of the resources at this location or
downstream locations committed while others
rolled back because an operator performed a
heuristic commit operation or rollback operation.
2. An unexpected error occurred while committing
or rolling back resources at this location or
downstream locations due to a hardware or
software problem.

When heuristic damage occurs, the following LUW journal

entry records can be checked to learn the status of the
changes made during the LUW to individual resources:
LCL The Record I/O State field indicates the status of the
record I/O performed on files journaled to the
journal related to that location.
API The API State field indicates the status of that API
Commitment Resource.
DDL The DDL State field indicates the status of that SQL
Object Change.
RMT The Resource State field indicates the status of the
resources at the remote location.

830 OS/400 Backup and Recovery V5R1

Table 125. Logical Unit of Work (C LW) Journal Entry–Header Record (continued)
Relative
Offset Field Format Description

228 Heuristic Operation Char (1) Whether a heuristic commit or rollback operation occurred at
Indicator this location while a commit request was being performed
for this LUW:
blank No heuristic operation occurred.
C A heuristic commit operation occurred.
R A heuristic rollback operation occurred.

A heuristic commit operation or rollback operation means

that the operator took explicit action (while this location was
waiting for the commit or rollback decision from the initiator
or the last agent) to commit or to roll back the resources at
this location and all prepared downstream locations.
Heuristic operations can result in some resources committing
while others roll back. The LUW Disposition field can be
checked to see if this has happened (it would be HD). The
Resync In Progress Indicator field can also be checked. If it is
Y, heuristic damage may have occurred or it may still occur
because the state of the resources at the locations where
resynchronization is still going on is unknown. Messages are
written to the history log and to the system database server
joblogs when the resynchronization processes complete to
indicate whether or not damage occurred. If damage occurs,
messages are also sent to the system operator when it is
detected.
229 Resync In Progress Char (1) Whether resync to one or more remote locations was still
Indicator ongoing when the LUW ended:
N Either no resynchronization was required during
this LUW, or it was required and completed before
the LUW ended.
O Resynchronization was going on with one or more
of the locations. This can occur only if the
WAIT_FOR_OUTCOME synchronization point
option is NO, or if the LUW was interrupted by job
or system end.

Appendix D. Journal Entry Information 831

Table 125. Logical Unit of Work (C LW) Journal Entry–Header Record (continued)
Relative
Offset Field Format Description

230 Wait for outcome Char(1) The value of the Wait for outcome commitment option. This
indicates whether to wait for resynchronization to complete if
a communication or system failure occurs during a commit
or rollback.
Y Wait for outcome.
L Wait for outcome during commits initiated by this
commitment definition or during commits initiated
at a system that does not support presumed abort.
Inherit the initiator’s wait for outcome value during
commits initiated at a system that supports
presumed abort.
N Do not wait for outcome.
U Do not wait for outcome during commits initiated
by this commitment definition or during commits
initiated at a system that does not support
presumed abort. Inherit the initiator’s wait for
outcome value during commits initiated at a system
that supports presumed abort.

231 Action if problems Char(1) The value of the Action if problems commitment option. This
indicates whether to commit or rollback when problems
occur during a two-phase commit.
R Rollback if problems occur.
C Commit if problems occur.

232 Vote read-only Char(1) The value of the Vote read-only permitted commitment
permitted option. This indicates whether this commitment definition is
allowed to return a read-only vote to a remote initiator
during a two-phase commit.
N Do not allow a read-only vote.
Y Allow a read-only vote.

233 Action if ENDJOB Char(1) The value of the Action if ENDJOB commitment option. This
indicates the action to take for changes associated with the
LUW when the job the LUW is a part of is ended.
W Wait to allow normal processing of the LUW to
complete.
R Rollback during ENDJOB.
C Commit during ENDJOB.

234 OK to leave out Char(1) The value of the OK to leave out commitment option. This
indicates whether this location is allowed to be left out
during the next commit/rollback if no activity occurred to
this location during the LUW.
N Do not leave this location out of the next commit or
rollback operation.
Y It is OK to leave this location out of the next
commit or rollback operation.

832 OS/400 Backup and Recovery V5R1

Table 125. Logical Unit of Work (C LW) Journal Entry–Header Record (continued)
Relative
Offset Field Format Description

235 Last agent permitted Char(1) The value of the Last agent permitted commitment option.
This indicates whether last agent optimization may be used.
S The system is allowed to select a last agent.
N The system is not allowed to select a last agent.

236 Accept vote reliable Char(1) The value of the Accept vote reliable commitment option.
This indicates whether the vote reliable indicator received
from agents during a commit operation is accepted by this
location. If an agent votes reliable, and this location accepts
it, control is returned to the application before the committed
wave is completed for that agent. If this location does not
accept vote reliable, control is returned to the application
only after the LUW is completely committed or rolled back.
Y Accept the vote reliable indicator from agents
during commit operations.
N Do not accept the vote reliable indicator from agents
during commit operations.

237 Resolved wait for Char(1) This indicates the actual wait for outcome value that was
outcome value used during the commit or rollback of this LUW. If the Wait
for outcome commitment option is L or U, this value may
have been inherited from this location’s initiator.
Y Wait for outcome of resynchronization.
N Do not wait for outcome of resynchronization.

238 XA Transaction Char(10) If this was an X/Open transaction, this is the name of the XA
Manager Transaction Manager that was specified on the db2xa_open
API. This field will be hex zeros if this was not an XA
transaction.
248 XID Char(140) If this was an X/Open Transaction, this is the X/Open
Transaction Identifier associated with this transaction. This
field will be hex zeros if this was not an X/Open transaction,
or if it was an X/Open local transaction. The format of this
field is as follows:
Bin(31) Format Identifier
Bin(31) Global Transaction Identifier Length
Bin(31) Branch Qualifier Length
Char(128) XID value
388 Default Journal Zoned (20,0) The commit cycle identifier for the default journal for this
Commit Cycle ID LUW. This is 0 if no commit cycle was started for this journal
Long during this LUW.
408 Reserved Char(9) Reserved for future use.

Notes:
1
The format for this field is in the description.

Appendix D. Journal Entry Information 833

Table 126. Logical Unit of Work (C LW) Journal Entry–Local Record
Relative
Offset Field Format Description

1 Record Type Char (4) Type of record:

LCL Local Location record.

5 Record Length Bin (15) Length of record. Currently 48 for LCL record.
7 Record Position (4)1 This identifies the position in the LUW journal entry where
this record starts. It is made up of two numbers:
v Bin (15): The relative number of the journal entry that
contains this record. If the LUW journal entry is greater
than 32K-1 bytes, multiple entries are actually sent to the
journal. This number represents which of these actual
journal entries contains this record (1 for the first, 2 for the
second, and so on). Note that this is not the actual journal
entry sequence number.
v Bin (15): The offset where this record starts within this
journal entry. This is the number of bytes past the
beginning of the entry where this record starts. For
example, 0 means the first byte in the entry.
11 Next Local Location (4)1 This identifies the position in the LUW journal entry where
Position the next LCL record starts. It is made up of two numbers:
v Bin (15): The relative number of the journal entry that
contains the record. If the LUW journal entry is greater
than 32K-1 bytes, multiple entries are actually sent to the
journal. This number represents which of these actual
journal entries contains the record (1 for the first, 2 for the
second, and so on). Note that this is not the actual journal
entry sequence number.
v Bin (15): The offset where the record starts within this
journal entry. This is the number of bytes past the
beginning of the entry where the record starts. For
example, 0 means the first byte in the entry.

Position 0 0 indicates that this is the last local location.

15 First Resource (4)1 This identifies the position in the LUW journal entry where
Position the first API or DDL record starts for this location. It is made
up of two numbers:
v Bin (15): The relative number of the journal entry that
contains the record. If the LUW journal entry is greater
than 32K-1 bytes, multiple entries are actually sent to the
journal. This number represents which of these actual
journal entries contains the record (1 for the first, 2 for the
second, and so on). Note that this is not the actual journal
entry sequence number.
v Bin (15): The offset where the record starts within this
journal entry. This is the number of bytes past the
beginning of the entry where the record starts. For
example, 0 means the first byte in the entry.

834 OS/400 Backup and Recovery V5R1

Table 126. Logical Unit of Work (C LW) Journal Entry–Local Record (continued)
Relative
Offset Field Format Description

19 Record I/O State Char (2) Indicates whether the record I/O performed during this
LUW for files journaled to the journal related to this location
was committed or rolled back successfully:
CS Record I/O for this location was committed
successfully.
RS Record I/O for this location was rolled back
successfully.
CF An attempt to commit record I/O for this location
failed.
RF An attempt to rollback record I/O for this location
failed.
blank This is the location with no journal so there is no
record I/O associated with it.

21 Journal Char (20) Journal related to this location:

v Char (10): Journal name (blank if this is the location with
no journal)
v Char (10): Journal library (blank if this is the location with
no journal)
41 Commit Cycle ID Bin (31) The commit cycle identifier for the journal. This is 0 for the
location with no journal. It may be 0 for the location related
to the default journal if there were no resources for that
location during this LUW. This is -1 if the actual commit
cycle identifier value is larger than 2 147 483 647. The Default
Journal Commit Cycle ID Long field always contains the
correct value.
45 Default Journal Flag Char (1) Indicates whether the journal related to this location is the
default journal:
Y It is the default journal.
N It is not the default journal.

46 Commit Cycle ID Zoned (20,0) The commit cycle identifier for the journal. This is 0 for the
Long location with no journal. It may be 0 for the location related
to the default journal if there were no resources for that
location during this LUW.
66 Reserved Char (15) Reserved for future use.

Notes:
1
The format for this field is in the description.

Table 127. Logical Unit of Work (C LW) Journal Entry–API Record

Relative
Offset Field Format Description

1 Record Type Char (4) Type of record:

API API Commitment Resource record.

5 Record Length Bin (15) Length of record. Currently 80 for API record.

Appendix D. Journal Entry Information 835

Table 127. Logical Unit of Work (C LW) Journal Entry–API Record (continued)
Relative
Offset Field Format Description

7 Record Position (4)1 This identifies the position in the LUW journal entry where
this record starts. It is made up of two numbers:
v Bin (15): The relative number of the journal entry that
contains this record. If the LUW journal entry is greater
than 32K-1 bytes, multiple entries are actually sent to the
journal. This number represents which of these actual
journal entries contains this record (1 for the first, 2 for the
second, and so on). Note that this is not the actual journal
entry sequence number.
v Bin (15): The offset where this record starts within this
journal entry. This is the number of bytes past the
beginning of the entry where this record starts. For
example, 0 means the first byte in the entry.
11 Resource Location (4)1 This identifies the position in the LUW journal entry where
Position the LCL record starts for this API resource’s location. It is
made up of two numbers:
v Bin (15): The relative number of the journal entry that
contains the record. If the LUW journal entry is greater
than 32K-1 bytes, multiple entries are actually sent to the
journal. This number represents which of these actual
journal entries contains the record (1 for the first, 2 for the
second, and so on). Note that this is not the actual journal
entry sequence number.
v Bin (15): The offset where the record starts within this
journal entry. This is the number of bytes past the
beginning of the entry where the record starts. For
example, 0 means the first byte in the entry.
15 Next Resource (4)1 This identifies the position in the LUW journal entry where
Position the next API or DDL record starts for this API resource’s
location. It is made up of two numbers:
v Bin (15): The relative number of the journal entry that
contains the record. If the LUW journal entry is greater
than 32K-1 bytes, multiple entries are actually sent to the
journal. This number represents which of these actual
journal entries contains the record (1 for the first, 2 for the
second, and so on). Note that this is not the actual journal
entry sequence number.
v Bin (15): The offset where the record starts within this
journal entry. This is the number of bytes past the
beginning of the entry where the record starts. For
example, 0 means the first byte in the entry.

Position 0 0 indicates that this is the last resource for this API
resource’s location.
19 API Resource Char (10) Name of API resource.
29 API Program Char (20) Name of the exit program for the API resource:
v Char (10): exit program name
v Char (10): exit program library
49 Journal Char (20) Journal related to the location for this resource:
v Char (10): Journal name (blank if this resource belongs to
the location with no journal)
v Char (10): Journal library (blank if this resource belongs to
the location with no journal)

836 OS/400 Backup and Recovery V5R1

Table 127. Logical Unit of Work (C LW) Journal Entry–API Record (continued)
Relative
Offset Field Format Description

69 Commit Cycle ID Bin (31) The commit cycle identifier for the journal. This is 0 if this
resource belongs to the location with no journal. This is -1 if
the actual commit cycle identifier value is larger than 2 147
483 647. The Commit Cycle ID Long field always contains
the correct value.
73 Commit Protocol Char (1) The commit protocol for this resource:
2 This is a two-phase resource (API resources are
always two-phase resources).

74 Resource Usage Char (2) The currently allowed access for this resource. The allowed
access for some resources can change from one LUW to
another depending on whether one-phase resources are
registered:
RO This resource is currently read-only. Updates were
not made during the LUW.
UP This resource is currently able to be updated.
Updates may or may not have been made during
the LUW.

76 API State Char (2) Indicates whether the API resource was committed or rolled
back successfully:
CS This resource was committed successfully.
RS This resource was rolled back successfully.
CF An attempt to commit this resource failed.
RF An attempt to rollback this resource failed.

78 API Last Agent Flag Char (1) Whether this resource is to be selected as the last agent
during all commit requests:
Y This resource is to be selected as the last agent.
N This resource is not to be selected as the last agent.

79 Allow Remote Char (1) Whether remote resources are allowed to participate in a
Resources LUW with this resource:
Y Remote resources are allowed with this resource.
N Remote resources are not allowed with this resource.

80 Save While Active Char (1) Whether this resource will hold out a save-while-active
Flag request until a commitment boundary is reached:
Y This resource will hold save-while-active requests.
N This resource will not hold save-while-active
requests.

81 Commit Cycle ID Zoned (20,0) The commit cycle identifier for the journal. This is 0 if this
Long resource belongs to the location with no journal.

101 Reserved Char(12) Reserved for future use.

Appendix D. Journal Entry Information 837

 Table 127. Logical Unit of Work (C LW) Journal Entry–API Record (continued)
Relative
Offset Field Format Description

Notes:
1
The format for this field is in the description.

Table 128. Logical Unit of Work (C LW) Journal Entry–DDL Record

Relative
Offset Field Format Description

1 Record Type Char (4) Type of record:

DDL SQL Object Change record.

| 5 Record Length Bin (15) Length of record. Currently 624 for DDL record.
7 Record Position (4)1 This identifies the position in the LUW journal entry where
this record starts. It is made up of two numbers:
v Bin (15): The relative number of the journal entry that
contains this record. If the LUW journal entry is greater
than 32K-1 bytes, multiple entries are actually sent to the
journal. This number represents which of these actual
journal entries contains this record (1 for the first, 2 for the
second, and so on). Note that this is not the actual journal
entry sequence number.
v Bin (15): The offset where this record starts within this
journal entry. This is the number of bytes past the
beginning of the entry where this record starts. For
example, 0 means the first byte in the entry.
11 Resource Location (4)1 This identifies the position in the LUW journal entry where
Position the LCL record starts for this DDL resource’s location. It is
made up of two numbers:
v Bin (15): The relative number of the journal entry that
contains the record. If the LUW journal entry is greater
than 32K-1 bytes, multiple entries are actually sent to the
journal. This number represents which of these actual
journal entries contains the record (1 for the first, 2 for the
second, and so on). Note that this is not the actual journal
entry sequence number.
v Bin (15): The offset where the record starts within this
journal entry. This is the number of bytes past the
beginning of the entry where the record starts. For
example, 0 means the first byte in the entry.

838 OS/400 Backup and Recovery V5R1

Table 128. Logical Unit of Work (C LW) Journal Entry–DDL Record (continued)
Relative
Offset Field Format Description

15 Next Resource (4)1 This identifies the position in the LUW journal entry where
Position the next API or DDL record starts for this DDL resource’s
location. It is made up of two numbers:
v Bin (15): The relative number of the journal entry that
contains the record. If the LUW journal entry is greater
than 32K-1 bytes, multiple entries are actually sent to the
journal. This number represents which of these actual
journal entries contains the record (1 for the first, 2 for the
second, and so on). Note that this is not the actual journal
entry sequence number.
v Bin (15): The offset where the record starts within this
journal entry. This is the number of bytes past the
beginning of the entry where the record starts. For
example, 0 means the first byte in the entry.

Position 0 0 indicates that this is the last resource for this

DDL resource’s location.

Appendix D. Journal Entry Information 839

 Table 128. Logical Unit of Work (C LW) Journal Entry–DDL Record (continued)
Relative
Offset Field Format Description

19 DDL Resource Char (29) Object identification and operation performed on object:
| Information v Char (10): First 10 characters of object name. The object
| name field always contains the full object name.
v Char (10): Object library name
v Char (7): Object type (*FILE, *LIB or *SQLPKG)
v Char (2): Object operation

The possible object operations and their meanings are the

following:
AC Add PF Constraint
CC Create Collection
CF Create File
CG Create Program
CM Create Member
CP Create SQL Package
CS Create Service Program
CT Create User Defined Type
DC Delete Collection
DF Delete File
DG Drop Program
DP Delete SQL Package
DS Drop Service Program
DT Drop User Defined Type
FC Change File
FR Rename File
GF Grant Files
GG Grant Program
GP Grant to SQL Package
| GR Grant Java Routine
GS Grant Service Program
GT Grant User Defined Type
LP LABEL ON SQL Package
OP COMMENT ON SQL Package
OT COMMENT User Defined Type
RC Remove PF Constraint
RG Revoke Program
RF Revoke Files
RP Revoke from SQL Package
| RR Revoke Java Routine
RS Revoke Service Program
RT Revoke User Defined Type
TA Add PF Trigger
| TC Change PF Trigger
| TR Remove PF Trigger
UL Unlink Datalink
XF Transfer Files

48 Reserved Char (1) Reserved for future use.

49 Journal Char (20) Journal related to the location for this resource:
v Char (10): Journal name (blank if this resource belongs to
the location with no journal)
v Char (10): Journal library (blank if this resource belongs to
the location with no journal)

840 OS/400 Backup and Recovery V5R1

 Table 128. Logical Unit of Work (C LW) Journal Entry–DDL Record (continued)
Relative
Offset Field Format Description

69 Commit Cycle ID Bin (31) The commit cycle identifier for the journal. This is 0 if this
resource belongs to the location with no journal. This is -1 if
the actual commit cycle identifier value is larger than 2 147
483 647. The Commit Cycle ID Long field always contains
the correct value.
73 Commit Protocol Char (1) The commit protocol for this resource:
2 This is a two-phase resource (DDL resources are
always two-phase resources).

74 DDL State Char (2) Indicates whether the DDL resource was committed or rolled
back successfully:
CS This resource was committed successfully.
RS This resource was rolled back successfully.
CF An attempt to commit this resource failed.
RF An attempt to rollback this resource failed.

76 Commit Cycle ID Zoned (20,0) The commit cycle identifier for the journal. This is 0 if this
Long resource belongs to the location with no journal.
96 Object Name Char (288) The full object name
| 384 Reserved Char (1) Reserved for future use.

Notes:
1
The format for this field is in the description.

Table 129. Logical Unit of Work (C LW) Journal Entry–RMT Record

Relative
Offset Field Format Description

1 Record Type Char (4) Type of record:

RMT Remote Location record.

5 Record Length Bin (15) Length of record. Currently 128 for RMT record.
7 Record Position (4)1 This identifies the position in the LUW journal entry where
this record starts. It is made up of two numbers:
v Bin (15): The relative number of the journal entry that
contains this record. If the LUW journal entry is greater
than 32K-1 bytes, multiple entries are actually sent to the
journal. This number represents which of these actual
journal entries contains this record (1 for the first, 2 for
the second, and so on). Note that this is not the actual
journal entry sequence number.
v Bin (15): The offset where this record starts within this
journal entry. This is the number of bytes past the
beginning of the entry where this record starts. For
example, 0 means the first byte in the entry.

Appendix D. Journal Entry Information 841

Table 129. Logical Unit of Work (C LW) Journal Entry–RMT Record (continued)
Relative
Offset Field Format Description

11 Next Remote Location (4)1 This identifies the position in the LUW journal entry where
Position the next RMT record starts. It is made up of two numbers:
v Bin (15): The relative number of the journal entry that
contains the record. If the LUW journal entry is greater
than 32K-1 bytes, multiple entries are actually sent to the
journal. This number represents which of these actual
journal entries contains the record (1 for the first, 2 for the
second, and so on). Note that this is not the actual journal
entry sequence number.
v Bin (15): The offset where the record starts within this
journal entry. This is the number of bytes past the
beginning of the entry where the record starts. For
example, 0 means the first byte in the entry.

Position 0 0 indicates that this is the last remote location.

15 First Resource (4)1 This identifies the position in the LUW journal entry where
Position the first DDM record starts for this location. It is made up of
two numbers:
v Bin (15): The relative number of the journal entry that
contains the record. If the LUW journal entry is greater
than 32K-1 bytes, multiple entries are actually sent to the
journal. This number represents which of these actual
journal entries contains the record (1 for the first, 2 for the
second, and so on). Note that this is not the actual journal
entry sequence number.
v Bin (15): The offset where the record starts within this
journal entry. This is the number of bytes past the
beginning of the entry where the record starts. For
example, 0 means the first byte in the entry.

Position 0 0 indicates that there are no DDM records for this

location.
19 Remote Location Char (54) Identification of the remote location and communication
Information information for this location:
v Char (10): Remote Location name
v Char (10): Device name
v Char (10): Mode
v Char (8): Remote network ID
v Char (8): Conversation correlator network ID
v Char (8): Transaction program name
73 Relational Database Char (18) The name of the relational database opened at this remote
Name location (blank if no relational database has been opened).
91 Conversation Char (1) Whether the conversation was deallocated because of this
Deallocation Flag LUW:
N This conversation is still active.
Y This conversation was deallocated because the
LUW committed, the system ended, a resource
failed, or an unbind was performed.

842 OS/400 Backup and Recovery V5R1

Table 129. Logical Unit of Work (C LW) Journal Entry–RMT Record (continued)
Relative
Offset Field Format Description

92 Commit Protocol Char (1) The commit protocol for the resources at this location:
1 The resources are one-phase.
2 The resources are two-phase.

93 Resource Usage Char (2) The currently allowed access for this resource. The allowed
access for some resources can change from one LUW to
another depending on whether one-phase resources are
registered:
RO This resource is currently read-only. Updates were
not made during the LUW.
UP This resource is currently able to be updated.
Updates may or may not have been made during
the LUW.
Note: This does not indicate whether updates were actually
made during the LUW. It indicates only whether updates are
allowed, given the other resources currently registered.
95 Resource State Char (2) The state of the resources at this location:
CS The resources were committed successfully.
CF An attempt to commit the resources failed. This
value is only used for one-phase locations.
RS The resources were rolled back successfully.
RF An attempt to rollback the resources failed. This
value is only used for one-phase locations.
NC The resources had no changes for the current
transaction.
FC A communications failure occurred for this location.
It is not known whether resources at the location
committed or rolled back.
HC The resources were heuristically committed.
HR The resources were heuristically rolled back.
HM Heuristic damage was detected at this location.
Some of the resources at the location, or locations
further downstream, committed while others rolled
back.
ER An unexpected error occurred while communicating
with this location. This is due to a hardware or
software problem. The state of the resources is
unknown.
RI We have not yet learned the state of the resources
because resync is still ongoing.

97 Allocator Flag Char (1) Indicates whether this is the allocator location, for example,
the location that called the transaction program running on
this system:
Y This location is the allocator.
N This location is not the allocator.

Appendix D. Journal Entry Information 843

Table 129. Logical Unit of Work (C LW) Journal Entry–RMT Record (continued)
Relative
Offset Field Format Description

98 Remote Last Agent Char (1) Indicates whether this location was selected as the last agent
Flag if a commit request was performed to end this LUW:
Y This is the last agent.
N This is not the last agent.
Note: A last agent will not be selected at this location unless
the Partner Role field in the HDR record is I or L.
99 Two-phase protocol CHAR(1) The two-phase commit protocol options supported at this
location.
0 Two-phase commit protocols are not supported.
1 Two-phase commit presumed nothing protocols are
supported.
2 Two-phase commit presumed abort protocols are
supported.

100 Resync initiator CHAR(1) If resync with this location is still ongoing (the Resource
State field is RI), this value indicates whether the local
location is initiating the resync attempts.
I The local system is initiating resync with this
remote location.
N Resync is not being performed with this remote
location.
W The local system is waiting for resync to be
initiated from this remote location.

101 Voted reliable CHAR(1) Whether this location voted reliable during the commit of
this LUW.
Y The location voted reliable.
N The location did not vote reliable.

102 OK to leave out CHAR(1) Whether this location indicated it may be left out of the next
commit or rollback operation if no communications flows
occur to that location during the next LUW.
Y The location indicated it may be left out.
N The location indicated it may not be left out.

103 Left out CHAR(1) Whether this location was left out of the LUW that was just
committed or rolled back.
Y The location was left out.
N The location was not left out.

844 OS/400 Backup and Recovery V5R1

Table 129. Logical Unit of Work (C LW) Journal Entry–RMT Record (continued)
Relative
Offset Field Format Description

104 Initiator Flag CHAR(1) Indicates whether this location is the initiator location, i.e.
the location that sent the commit or rollback request to this
system.
Y The location is the initiator.
N The location is not the initiator.
Note: The system cannot determine the initiator location if
the initiator does not support two-phase commit protocols.
This field will always be set to N for locations that do not
support two-phase commit protocols.
105 Reserved CHAR(24) Reserved for future use.

Notes:
1
The format for this field is in the description.

Table 130. Logical Unit of Work (C LW) Journal Entry–DDM Record

Relative
Offset Field Format Description

1 Record Type Char (4) Type of record:

DDM Remote Database File record.

5 Record Length Bin (15) Length of record. Currently 96 for DDM record.
7 Record Position (4)1 This identifies the position in the LUW journal entry where
this record starts. It is made up of two numbers:
v Bin (15): The relative number of the journal entry that
contains this record. If the LUW journal entry is greater
than 32K-1 bytes, multiple entries are actually sent to the
journal. This number represents which of these actual
journal entries contains this record (1 for the first, 2 for the
second, and so on). Note that this is not the actual journal
entry sequence number.
v Bin (15): The offset where this record starts within this
journal entry. This is the number of bytes past the
beginning of the entry where this record starts. For
example, 0 means the first byte in the entry.
11 Resource Location (4)1 This identifies the position in the LUW journal entry where
Position the RMT record starts for this DDM file’s location. It is made
up of two numbers:
v Bin (15): The relative number of the journal entry that
contains the record. If the LUW journal entry is greater
than 32K-1 bytes, multiple entries are actually sent to the
journal. This number represents which of these actual
journal entries contains the record (1 for the first, 2 for the
second, and so on). Note that this is not the actual journal
entry sequence number.
v Bin (15): The offset where the record starts within this
journal entry. This is the number of bytes past the
beginning of the entry where the record starts. For
example, 0 means the first byte in the entry.

Appendix D. Journal Entry Information 845

Table 130. Logical Unit of Work (C LW) Journal Entry–DDM Record (continued)
Relative
Offset Field Format Description

15 Next Resource (4)1 This identifies the position in the LUW journal entry where
Position the next DDM record starts for this DDM file’s location. It is
made up of two numbers:
v Bin (15): The relative number of the journal entry that
contains the record. If the LUW journal entry is greater
than 32K-1 bytes, multiple entries are actually sent to the
journal. This number represents which of these actual
journal entries contains the record (1 for the first, 2 for the
second, and so on). Note that this is not the actual journal
entry sequence number.
v Bin (15): The offset where the record starts within this
journal entry. This is the number of bytes past the
beginning of the entry where the record starts. For
example, 0 means the first byte in the entry.
Position 0 0 indicates that this is the last resource for this
DDM file’s location.
19 DDM File Char (20) Name of the DDM file and library for the open remote file:
v Char (10): DDM file name
v Char (10): DDM file library name
29 Remote Location Char (54) Identification of the remote location and communication
Information information for this resource’s location:
v Char (10): Remote Location name
v Char (10): Device name
v Char (10): Mode
v Char (8): Remote network ID
v Char (8): Conversation correlator network ID
v Char (8): Transaction program name
93 Open Flag Char (1) Whether the DDM file was open or closed when this LUW
ended:
O The DDM file was open.
C The DDM file was closed.

94 Commit Protocol Char (1) The commit protocol for this resource:
1 This is a one-phase resource.
2 This is a two-phase resource.

95 Resource Usage Char (2) The currently allowed access for this resource. The allowed
access for some resources can change from one LUW to
another depending on whether one-phase resources are
registered:
RO This resource is currently read-only. Updates were
not made during the LUW.
UP This resource is currently able to be updated.
Updates may or may not have been made during
the LUW.
Note: This does not indicate whether updates were actually
made during the LUW. It only indicates whether updates are
allowed, given the other resources currently registered.

846 OS/400 Backup and Recovery V5R1

 Table 130. Logical Unit of Work (C LW) Journal Entry–DDM Record (continued)
Relative
Offset Field Format Description

Notes:
1
The format for this field is in the description.

Table 131. Moving and Renaming Objects (D FM, D FN, E EM, E EN, F MM, F MN, F PM, F PN, Q QM, Q QN)
Journal Entries
Relative
Offset Field Format Description

Specific values for this entry type:

| Journal identifier Char(10) Records for the entries will have a journal identifier. The JID
(JOJID) is provided with the *TYPE4 format only. It can be used
with the QJORJIDI API.
Entry-specific data. This data appears as one field in the standard output formats:
| 1 Object Name Before Char (10) The name of the object before the object was moved or
| renamed.
11 Library Name Before Char (10) The name of the library before the object was moved or
renamed.
21 Member Name Before Char (10) The name of the member before it was moved or renamed.
| This field is blank if the object is not a physical database
| file.
| 31 Object Name After Char (10) The name of the object after the object was moved or
| renamed.
| 41 Library Name After Char (10) The name of the library after the object was moved or
| renamed.
51 Member Name After Char (10) The name of the member after it was moved or renamed.
| This field is blank if the object is not a physical database
| file.
61 Internal data Char (*) Internal system information.

Table 132. File OPEN (F OP) and File CLOSE (F CL) Journal Entries
Relative
Offset Field Format Description

Entry-specific data. This data appears as one field in the standard output formats.
1 File Name Char (10) The name of the file that was opened or closed. If a physical
file is opened, this field and the JOOBJ field are the same. If
a logical file is opened, this field contains the name of the
logical file. JOOBJ field contains the name of the physical file.
11 Library Name Char (10) The library containing the file.
21 Member Name Char (10) The file member that was opened of closed.
31 Open options Char (4) Only used for file open (entry type OP). Values of the bytes
follow:
31 Input Char (1) Whether the file was opened for input:
I= File opened for input
Blank =
Input not specified

Appendix D. Journal Entry Information 847

Table 132. File OPEN (F OP) and File CLOSE (F CL) Journal Entries (continued)
Relative
Offset Field Format Description

32 Output Char (1) Whether the file was opened for output:
O= File opened for output
Blank =
Output not specified

33 Update Char (1) Whether the file was opened for update:
U= File opened for update
Blank =
Update not specified

34 Delete Char (1) Whether the file was opened for delete:
D= File opened for delete
Blank =
Delete not specified

Table 133. Journal Code R, All Journal Entry Types Except IL

Relative
Offset Field Format Description

Specific values for this entry type:

Flag (JOFLAG) Char (1) Whether a before-image is present1:
0= Before-image is not present. If before-images are
being journaled, this indicates that an update
operation or delete operation is being requested for a
record that has already been deleted.
1= Before-image is present.

Journal identifier Char(10) The JID is provided with the *TYPE4 format only. It can be
(JOJID) used with the QJORJIDI API.
Entry-specific data. This data appears as one field in the standard output formats:
1 Entry-specific Data Char (*) After-image of the record for entry types PT, PX, UP, or UR.
Before-image of the record for entry types UB, DL, BR, or DR
if before-images are being journaled and the record was not
previously deleted.

Notes:
1
The flag does not apply to these entry types: PT, PX, UP, and UR.

Table 134. RGZPFM (F RG) Journal Entry

Relative
Offset Field Format Description

Entry-specific data. This data appears as one field in the standard output formats:
1 File Name Char (10) The name of the file specified for the KEYFILE parameter on
the RGZPFM command. If KEYFILE(*NONE) was specified,
this field is blank.

848 OS/400 Backup and Recovery V5R1

Table 134. RGZPFM (F RG) Journal Entry (continued)
Relative
Offset Field Format Description

11 Library Name Char (10) The name of the library specified in the KEYFILE parameter
of the RGZPFM command. If KEYFILE(*NONE) was
specified, this field is blank.
21 Member Name Char (10) The name of the member specified in the KEYFILE
parameter of the RGZPFM command. If KEYFILE(*NONE)
was specified, this field is blank.

Table 135. ROLLBACK (C RB) Journal Entry

Relative
Offset Field Format Description

Specific values for this entry type:

Job name (JOJOB) Char (10) Blank if the entry was added during an IPL.
Program name Char (10) Blank if the entry was added during an IPL.
(JOPGM)
Flag (JOFLAG) Char (1) How the rollback operation was initiated and whether it was
successful:
0= All record-level changes were rolled back for a
rollback operation initiated by a user.
1= Not all record-level changes were successfully
rolled back for a rollback operation initiated by a
user.
2= All record-level changes were rolled back for a
rollback operation initiated by the operating system.
3= Not all record-level changes were rolled back for a
rollback operation initiated by the operating system.

Table 136. Object Restored (B FR, E EL, F MR, J RR, Q QZ) and Receiver Saved (J RS) Journal Entries
Relative
Offset Field Format Description

Specific values for this entry type:

Journal identifier Char(10) Records for the entries will have a journal identifier. The JID
(JOJID) is provided with the *TYPE4 format only. It can be used with
the QJORJIDI API.
Entry-specific data. This data appears as one field in the standard output formats:
1 Media type Char (3) The type of media used for the save or restore operation:
DKT = Diskette
OPT = Optical
SAV = Save file
TAP = Tape

4 First Volume ID Char (6) The ID of the first volume used. The optical volume ID may
contain up to 32 characters of which the first six characters
are displayed.
10 Start Save or Restore Char (6)1 The date the save or restore operation was started. The date
Date is in the format of the DATFMT attribute of the job that
performed the save or restore operation.
16 Start Save or Restore Zoned (6,0) The time the save or restore operation was started.
Time

Appendix D. Journal Entry Information 849

 Table 136. Object Restored (B FR, E EL, F MR, J RR, Q QZ) and Receiver Saved (J RS) Journal
Entries (continued)
Relative
Offset Field Format Description

22 Update History Char (1) Whether the save history is updated:

0= UPDHST(*NO) specified on save command.
1= UPDHST(*YES) specified on save command.

23 Save File Name Char (10) The name of the save file used for the operation. This field is
blank if a save file was not used.
33 Save File Library Char (10) The name of the library for the save file. This field is blank if
a save file was not used.
| 43 Media FID Char (16) File identifier for the IFS object on the media. This applies
| only to B FR entries.
| 59 Restored FID Char (16) File identifier for the restored IFS object. This applies only to
| B FR entries.
| 75 Restored over FID Char (16) File identifier for the IFS object that was restored over. This
| applies only to B FR entries.

Notes:
1
Refer to the fixed length portion of the journal entry for any information pertaining to the century of this
date.

Table 137. Object Saved (B FS, E ES, F MS, Q QY) Journal Entries
Relative
Offset Field Format Description

Entry-specific data. This data appears as one field in the standard output formats:
| 1 Media type Char (3) The type of media used to save the object:
DKT = Diskette
OPT = Optical
SAV = Save file
TAP = Tape

| 4 First Volume ID Char (6) The ID of the first volume used to save the object The optical
volume ID may contain up to 32 characters of which the first
six characters are displayed.
| 10 Start Save Date Char (6)1 The date the save operation was started. The date is in the
| format of the DATFMT attribute of the job that saved the
| object.
16 Start Save Time Zoned (6,0) The time the save operation was started.
22 Update History Char (1) Whether the save history is updated:
0= UPDHST(*NO) specified on save command.
1= UPDHST(*YES) specified on save command.

23 Save File Name Char (10) The name of the save file used for the operation. This field is
blank if a save file was not used.
33 Save File Library Char (10) The name of the library for the save file. This field is blank if
a save file was not used.
| 43 Save Active Value Char (10) The value specified for the SAVACT parameter on the
| SAVOBJ, SAVCHGOBJ, SAV, or SAVLIB command.

850 OS/400 Backup and Recovery V5R1

 Table 137. Object Saved (B FS, E ES, F MS, Q QY) Journal Entries (continued)
Relative
Offset Field Format Description

| 53 Start Save Active Date Char (6)1 For a save-while-active operation, this is the date when
| checkpoint processing was completed for the object. For a
normal save operation, this is the same as the start date.
| 59 Start Save Active Zoned (6,0) For a save-while-active operation, this is the time when
| Time checkpoint processing was completed for the object. For a
normal save operation, this is the same as the start time.
| 65 Primary Receiver Char (10) The name of the first of dual receivers that contains the
| Name start-of-save entry.
75 Primary Receiver Char (10) The name of the library containing the primary receiver.
Library
| 85 Dual Receiver Name Char (10) The name of the second of dual receivers that contains the
| start-of-save entry. This entry is blank if only a single
| receiver was used when the start-of-save entry was added.
| 95 Dual Receiver Library Char (10) The name of the library containing the dual receiver. This
| entry is blank if only a single receiver was used when the
| start-of-save entry was added.
| 105 Sequence number of Zoned (10, 0) For a save-while-active operation, the sequence number of
| matching start-of-save the corresponding start-of-save entry. For a normal save
| entry operation, this is the sequence number of the current object
| saved entry.

| 115 File ID of object Char (16) The file identifier (FID) of the object. This applies only to B
| FS entries.

Notes:
| If an object was saved using the save-while-active function, the saved copy of the object includes all of the changes
| found in the journal entries up to the corresponding object start of save-while-active entry, Table 138.

| If an object was NOT saved using the save-while-active function, the saved copy of the object includes all of the
| changes found in the journal entries up to the corresponding object saved entry, Table 137.
1
Refer to the fixed length portion of the journal entry for any information pertaining to the century of this
date.

Table 138. Start of Save-While-Active (B FW, E EW, F SS, Q QX) Journal Entries
Relative
Offset Field Format Description

Entry-specific data. This data appears as one field in the standard output formats:
1 Media type Char (3) The type of media used to save the object:
DKT = Diskette
OPT = Optical
SAV = Save file
TAP = Tape

4 First Volume ID Char (6) The ID of the first volume used to save the object. The
optical volume ID may contain up to 32 characters of which
the first six characters are displayed.
| 10 Start Save Date Char (6)1 The date the save operation was started. The date is in the
| format of the DATFMT attribute of the job that saved the
| object.
16 Start Save Time Zoned (6,0) The time the save operation was started.

Appendix D. Journal Entry Information 851

 Table 138. Start of Save-While-Active (B FW, E EW, F SS, Q QX) Journal Entries (continued)
Relative
Offset Field Format Description

22 Update History Char (1) Whether the save history is updated:

0= UPDHST(*NO) specified on the save command.
1= UPDHST(*YES) specified on the save command.

23 Save File Name Char (10) The name of the save file used for the operation. This field is
blank if a save file was not used.
33 Save File Library Char (10) The name of the library for the save file. This field is blank if
a save file was not used.
| 43 Save Active Value Char (10) The value specified for the SAVACT parameter on the
| SAVOBJ, SAVCHGOBJ, SAV, or SAVLIB command.
| 53 Save Active Date Char (6)1 For a save-while-active operation, this is the date when
| checkpoint processing was completed for the object. For a
| normal save operation, this is the same as the start date.
| 59 Save Active Time Char (6) For a save-while-active operation, this is the time when
| checkpoint processing was completed for the object. For a
| normal save operation, this is the same as the start time.
| 65 Object File ID Char (16) The file identifier (FID) of the IFS object. This applies only to
| B FW entries.

Notes:
| If an object was saved using the save-while-active function, the saved copy of the object includes all of the changes
| found in the journal entries up to the corresponding object start of save-while-active entry, Table 138.

| If an object was NOT saved using the save-while-active function, the saved copy of the object includes all of the
| changes found in the journal entries up to the corresponding object saved entry, Table 137.
1
Refer to the fixed length portion of the journal entry for any information pertaining to the century of this
date.

Table 139. Start Journal (B JT, E EG, F JM, Q QB) Journal Entries
Relative
Offset Field Format Description

Specific values for this entry type:

Flag (JOFLAG) Char (1) Indicates the type of images selected:
0= After images are journaled.
1= Before and after images are journaled.

Entry-specific data. This data appears as one field in the standard output formats:
1 Omit journal entry Char (1) Indicates the value of the OMTJRNE parameter on the Start
Journal command.
0= No entries are omitted from journaling.
| 1= Open and Close (*FILE), or Open, Close, and Force
| (*DIR or *STMF) entries are not journaled.

| 2 File identifier Char (16) The file identifier (FID) for the IFS object. This only applies
| to B JT entries.
| 18 Path name Char (*) The path name information optionally follows the FID. This
| only applies to B JT entries.

852 OS/400 Backup and Recovery V5R1

 Table 140. License Key Not Valid (L LK) Journal Entry
Relative
Offset Field Format Description

Entry-specific data. This data appears as one field in the standard output formats:
1 Product ID Char (7) The ID of the product whose license key was not valid.
8 License Term Char (6) The term of the license.
14 Feature Char (4) The product feature code.
18 Usage Limit Zoned (6,0) The usage limit for the product.
24 License Key Char (18) The license key for the product.
42 Expiration Date Char (7) The expiration date for the license key.
49 Vendor Data Char (8) Data placed in the entry by the product vendor.
57 Processor Group Char (3) The processor group for the license key.

Table 141. Usage Limit Changed (L LL) Journal Entry

Relative
Offset Field Format Description
Entry-specific data. This data appears as one field in the standard output formats:
1 Product ID Char (7) The ID of the product whose usage limit was changed.
8 License Term Char (6) The term of the license.
14 Feature Char (4) The product feature code.
18 Previous Usage Limit Zoned (6,0) The usage limit before the change.
24 Current Usage Limit Zoned (6,0) The usage limit after the change.
30 Old Expiration Date. Char (7) The expiration date before the change.
37 New Expiration Date. Char (7) The expiration date after the change.

Table 142. Usage Limit Exceeded (L LU) Journal Entries

Relative
Offset Field Format Description

Entry-specific data. This data appears as one field in the standard output formats:
1 Product ID Char (7) The ID of the product whose usage limit was exceeded.
8 License Term Char (6) The term of the license.
14 Feature Char (4) The product feature code.
18 Usage Limit Zoned (6,0) The usage limit for the product.
24 Request Flag Char (1) Whether the request was successful:
0= License request was successful.
1= License request was not successful.

25 Number of Licensed Zoned (6,0) The number of users currently licensed for the product.
Users
31 Licensed User Name Char (26) x 100 The names of up to 100 users who are licensed for the
product.

| Table 143. Update Data Area (E EA, E EB) Journal Entries

| Relative
| Offset Field Format Description

| Entry-specific data. This data appears as one field in the standard output formats:
| 1 Starting position Bin (32) Starting position of change as specified by the user (1 for
| decimal).
| 33 Length of change Bin (32) Length of change to be applied as specified by the user.
| 65 Number Bin (32) Number of decimal positions as specified by the user.
| 97 Offset to change Bin (32) Offset to change value field from the beginning of the
| entry-specific data (ESD).

Appendix D. Journal Entry Information 853

| Table 143. Update Data Area (E EA, E EB) Journal Entries (continued)
| Relative
| Offset Field Format Description

| 129 Type Char (10) Type of data area. Data area types are *CHAR, *DEC, and
| *LGL.
| Padding for Char(*) Padding to align fields.
| alignment
| Offset to Change value Char (*) Value of the change.
| change
|
| Table 144. Data Queue Cleared, has Key (Q QJ) Journal Entry
| Relative
| Offset Field Format Description

| Entry-specific data. This data appears as one field in the standard output formats:
| 1 Reserved Char (2) Reserved for future use.
| 3 Key length Bin (16) The number of characters in the key.
| 19 Key order Char (2) The Key order is as follows:
| GT Greater than
| LT Less than
| NE Not equal
| EQ Equal
| GE Greater than or equal
| LE Less than or equal
|
| 21 Key Char (*) The data to be used to remove a message from the data
| queue.
|
| Table 145. Send Data Queue, has Key (Q QK) Journal Entry
| Relative
| Offset Field Format Description

| Entry-specific data. This data appears as one field in the standard output formats:
| 1 Data Length Bin (32) The length of the data sent to the data queue
| 33 Offset to Data Bin (32) Offset to the data placed on the data queue. The offset is
| calculated from the beginning of the entry-specific data
| (ESD).
| 65 Reserved Char (2) Reserved for future use.
| 67 Key Length Bin (16) The number of characters in a key.
| 83 Reserved Char (4) Reserved for future use.
| 87 Key Char (*) A prefix added to an entry by its sender.
| Reserved Char (*) Padding to align fields.
| Offset to Data Char (*) The first 16 bytes of the data entry are API information
| data required by the QSNDDTAQ API and are not placed on the
| data queue. The remainder is the data that was placed on
| the data queue.
|
| Table 146. Received Data Queue, has Key (Q QL) Journal Entry
| Relative
| Offset Field Format Description

| Entry-specific data. This data appears as one field in the standard output formats:
| 1 Reserved Char (18) Reserved for future use.
| 19 Key length Bin (16) The number of characters in the key.

854 OS/400 Backup and Recovery V5R1

| Table 146. Received Data Queue, has Key (Q QL) Journal Entry (continued)
| Relative
| Offset Field Format Description

| 35 Key order Char (2) The Key Order is as follows:

| GT Greater than
| LT Less than
| NE Not equal
| EQ Equal
| GE Greater than or equal
| LE Less than or equal
|
| 37 Key Char (*) The data to be used to receive a message from the data
| queue.
|
| Table 147. Send Data Queue, No Key (Q QS) Journal Entry
| Relative
| Offset Field Format Description

| Entry-specific data. This data appears as one field in the standard output formats:
| 1 Reserved Char (28) Reserved for future use.
| 29 Data length Bin (32) The length of the data placed on the data queue.
| 61 Data Char (*) The first 16 bytes of the data are API information required
| by the QSNDDTAQ API and are not placed on the data
| queue. The remainder is the data placed on the data queue.
|
| Table 148. Object Level (D AC, D CG, D CT, D DC, D DT, D GC, D GO, D GT, D RV, D TC, D TD, D TG, F DM, F
| MC) Journal Entries
| Relative
| Offset Field Format Description

| Entry-specific data. This data appears as one field in the standard output formats:
| 1 Object name Char (10) The name of the object that was operated on.
| 11 Library Name Char (10) The name of the library of the object that was operated on.
| 21 Member Name Char (10) The name of the member that was operated on, if
| applicable. This field is blank if it does not apply.
| 31 Reserved Char (30) Reserved.
| 61 Internal data Char (*) Internal system information.
|

Appendix D. Journal Entry Information 855

856 OS/400 Backup and Recovery V5R1
Appendix E. How to Create and Use Output from the Save and
Restore Commands
When you use the Save (SAV) command or the Restore (RST) command, you can
direct output to a stream file or a user space. This topic describes the output
information that is created by these commands. If data already exists in the stream
file or user space that you specify, the command writes over that data. It does not
append the new data to any existing data.

To specify a stream file, you must have *W authority to the stream file and *R
authority to the directory for the stream file.

To specify a user space, you must have *CHANGE authority to the user space and
*USE authority to the library. The system needs an *EXCLRD lock on the user
space.

Format of the Output

The output for the SAV command and the RST command consists of the following
formats:
v Command information format
v Directory information format
v Object link information format
v Trailer information format

Figure 98 shows the sequence of entries in the output when you specify
OUTPUT(*ALL) or OUTPUT(*ERR):

┌────────────────────────────────────────────────┐
│ Command information │
├────────────────────────────────────────────────┤
│ Directory information for directory 1 │
│ Object link information for object link 1 │
│ ... │
│ Object link information for object link N │
├────────────────────────────────────────────────┤
│ Directory information for directory 2 │
│ Object link information for object link 1 │
│ ... │
│ Object link information for object link N │
├────────────────────────────────────────────────┤
│ Directory information for directory N │
│ Object link Information for object link 1 │
│ . . . │
│ Object link information for object link N │
├────────────────────────────────────────────────┤
│ Trailer information │
└────────────────────────────────────────────────┘

Figure 98. Output Sequence 1–SAV and RST Commands

When you specify OUTPUT(*ALL), the output contains an object link entry for all
object links (both successful and unsuccessful). When you specify OUTPUT(*ERR),
the output contains an object link entry only for unsuccessful links.

Figure 99 on page 858 shows the sequence of entries in the output when you
specify OUTPUT(*SUMMARY):

© Copyright IBM Corp. 1997, 2000, 2001 857

 ┌────────────────────────────────────────────────┐
│ Command information │
├────────────────────────────────────────────────┤
│ Directory information for directory 1 │
├────────────────────────────────────────────────┤
│ Directory information for directory 2 │
├────────────────────────────────────────────────┤
│ Directory information for directory N │
├────────────────────────────────────────────────┤
│ Trailer information │
└────────────────────────────────────────────────┘

Figure 99. Output Sequence 2–SAV and RST Commands

When you retrieve information from the output format for object links, you must
use the entry length that is returned in the header information format of each
entry. The size of each entry may include padding at the end of the entry. If you
do not use the entry length, the result may not be valid. The entry length can be
used to find the next entry. The Trailer entry is always the last entry.

Table 149 through Table 153 on page 861 show the layouts for the output formats.
“Field Descriptions” on page 861 provides more information about the fields.

After each field in the layouts is a notation that indicates how the field is set. The
field may be set:
v Only for save operations (S)
v Only for restore operations (R)
v For save operations and restore operations (S/R)
Fields that are not set contain a value of zero for numeric fields and blanks for
character fields.

For each field that specifies an offset, this offset is relative to the first field of the
header information format for each entry (the Entry type field).

Header Information Format

Table 149 shows the format for the heading information for output from the SAV
and RST commands.
Table 149. Heading Output Information–SAV and RST Commands
Offset
Decimal Hex Type Field
0 0 BINARY(4) Entry type (S/R)
4 4 BINARY(4) Entry length (S/R)

Command Information
Table 150 shows the format for the command information for output from the SAV
and RST commands.
Table 150. Command Information Output–SAV and RST Commands
Offset

Decimal Hex Type Field

0 0 Everything from the header information format

8 8 BINARY(4) Device identifier offset (S/R)
12 C BINARY(4) File label offset (S/R)
16 10 BINARY(4) Sequence number (S/R)

858 OS/400 Backup and Recovery V5R1

 Table 150. Command Information Output–SAV and RST Commands (continued)
Offset

Decimal Hex Type Field

20 14 BINARY(4) Save active (S/R)

24 18 BINARY(4) CCSID of data (S/R)
28 1C BINARY(4) Number of records (S/R)
32 20 CHAR(10) Command (S/R)
42 2A CHAR(10) Expiration date (S/R)
52 34 CHAR(8) Save date/time (S/R)
60 3C CHAR(10) Start change date (S/R)
70 46 CHAR(10) Start change time (S/R)
80 50 CHAR(10) End change date (S/R)
90 5A CHAR(10) End change time (S/R)
100 64 CHAR(6) Save release level (S/R)
106 6A CHAR(6) Target release level (S/R)
112 70 CHAR(1) Information type (S/R)
113 71 CHAR(1) Data compressed (S/R)
114 72 CHAR(1) Data compacted (S/R)
115 73 CHAR(8) Save System serial number (S/R)
123 7B CHAR(8) Restore date/time (R)
131 83 CHAR(6) Restore release level (R)
137 89 CHAR(8) Restore system serial number (R)
145 91 CHAR(10) Save active option (S/R)
Note: Format of file label. The following fields are not repeated. The start of the file label
can be found by using the file label offset field.
* * BINARY(4) Length of file label (S/R)
* * CHAR(*) File Label (S/R)
Note: Format of device identifier. The device name length and device name are repeated for
each device identifier. The first entry is found by using the device identifier offset field to
get to the Number of device identifiers field and then moving to the first device identifier. Each
device identifier consists of a length followed by the name.
* * BINARY(4) Number of device identifiers
* * BINARY(4) Device name length (S/R)
* * CHAR(*) Device name (S/R)

Directory Information
Table 151 shows the format for the directory information for output from the SAV
and RST commands.
Table 151. Directory Information Output–SAV and RST Commands
Offset

Decimal Hex Type Field

0 0 Everything from the header information format

8 8 BINARY(4) Directory identifier offset (S/R)
12 C BINARY(4) Number of object links processed successfully in
directory (S/R)
16 10 BINARY(4) Number of object links processed unsuccessful in
directory (S/R)
20 14 BINARY(4) Starting volume identifier offset (S/R)
Note: Format of directory identifier. The following fields are not repeated. The start of the
directory identifier can be found by using thedirectory identifier offset field. The directory
identifier consists of a length followed by the directory name.
* * BINARY(4) Length of directory name (S/R)

Appendix E. How to Create and Use Output from the Save and Restore Commands 859
 Table 151. Directory Information Output–SAV and RST Commands (continued)
Offset

Decimal Hex Type Field

* * CHAR(*) Directory name (S/R)

Note: Format of starting volume identifier. The following fields are not repeated. The first
entry is found by using the starting volume identifier offset field. The volume identifier
consists of a length followed by the volume name. The directory name will be stored in
UNICODE. For information on converting this name, see the documentation for the iconv
API in the Programming topic on the Information Center at the following Website:
http://www.ibm.com/eserver/iseries/infocenter.
* * BINARY(4) Length of starting volume identifier (S/R)
* * CHAR(*) Starting volume identifier (S/R)

Object Link Information

Table 152 shows the format for the object link information for output from the SAV
and RST commands.
Table 152. Object Link Information–Output from SAV and RST Commands
Offset

Decimal Hex Type Field

0 0 Everything from the header information format

8 8 BINARY(4) Object link identifier offset (S/R)
12 C BINARY(4) Object link identifier after restore offset (R)
16 10 BINARY(4) Starting volume identifier offset (S/R)
20 14 BINARY(4) Object link error message replacement identifier
offset (S/R)
24 18 BINARY(4) Object link size (S/R)
28 1C BINARY(4) Object link size multiplier (S/R)
32 20 BINARY(4) ASP at time of save operation (S/R)
36 24 BINARY(4) ASP after restore operation (R)
40 28 CHAR(10) Object link type (S/R)
50 32 CHAR(8) Save active date/time (S/R)
58 3A CHAR(10) Object link owner at time of save (S/R)
68 44 CHAR(10) Object link owner after restore (R)
78 4E CHAR(50) Object link text (S/R)
128 80 CHAR(1) Object link security message (R)
129 81 CHAR(1) Object link status (S/R)
130 82 CHAR(7) Object link error message ID (S/R)
137 89 CHAR(1) Object link data (S/R)
138 8A BIN(8) Reserved
146 92 CHAR(1) ALWCKPWRT (S/R)
Note: Format of object link identifier. The following fields are not repeated. The start of the
object link identifier can be found by using the Object link identifier offset field. An object link
identifier will consist of a length followed by the object link name.
* * BINARY(4) Length of object link name (S/R)
* * CHAR(*) Object link name (S/R)
Note: Format of object link identifier after restore operation. The following fields are not
repeated. The start of the object link identifier after the restore operation can be found by
using the Object link identifier after restore offset field. An object link identifier will consist of a
length followed by the object link name. The object link name will be stored in UNICODE.
| For information on converting this name, see the documentation for the iconv API under the
| Programming topic on the Information Center at the following URL:
| http://www.ibm.com/eserver/iseries/infocenter.

860 OS/400 Backup and Recovery V5R1

 Table 152. Object Link Information–Output from SAV and RST Commands (continued)
Offset

Decimal Hex Type Field

* * BINARY(4) Length of object link name after restore operation

(S/R)
* * CHAR(*) Object link name after restore operation (R)
Note: Format of object link error message replacement identifier. The following fields are
not repeated. The start of the object link error message replacement identifier can be found
by using the object link error message replacement identifier offset field. An error message
will consist of a length followed by the object link error message replacement data.
* * BINARY(4) Length of object link error message replacement data
(S/R)
* * CHAR(*) Object link error message replacement data (S/R)
Note: Format of starting volume identifier. The following fields are not repeated. The first
entry is found by using the Starting volume identifier offset field. The volume identifier will
consist of a length followed by the volume name.
* * BINARY(4) Length of starting volume identifier (S/R)
* * CHAR(*) Starting volume identifier (S/R)

Trailer Information
Table 153 shows the format for the trailer information format for output from the
SAV and RST commands.
Table 153. Trailer Information–Output from SAV and RST Commands
Offset

Decimal Hex Type Field

0 0 Everything from the header information format

8 8 BINARY(4) Volume identifier offset (S/R)
12 C BINARY(4) Complete data (S/R)
16 10 BINARY(4) Number of object links processed successfully (S/R)
20 14 BINARY(4) Number of object links processed unsuccessfully (S/R)
Note: Format of volume identifier. The volume identifier length and the volume identifier fields are repeated for
each volume identifier. The first entry is found by using the volume name offset field to get to the Number of volume
identifiers field and then moving to the first volume identifier. A volume identifier consists of a length followed by the
volume name.
* * BINARY(4) Number of volume identifiers
* * BINARY(4) Volume identifier length (S/R)
* * CHAR(*) Volume identifier (S/R)

Field Descriptions
ALWCKPWRT. Indicates whether an object was saved while updates to the object may have occurred. The possible
values are:
0 No updates occurred to the object while the object was being saved.
1 The system saved the object with the SAVACTOPT(*ALWCKPWRT) parameter, and it set the corresponding
system attribute for the object. Updates to the object may have occurred while the system saved the object.
See the Information Center at the following Website: http://www.ibm.com/eserver/iseries/infocenter for
more information.

ASP after restore. The auxiliary storage pool (ASP) of the object link when it was restored. The possible value is:
1 System ASP

Appendix E. How to Create and Use Output from the Save and Restore Commands 861
ASP at time of save. The auxiliary storage pool (ASP) of the object link when it was saved. The possible value is:
1 System ASP

Command. The command that was used when the operation was performed.
The possible values are:
SAV Save operation
RST Restore operation

Complete data. Indicates whether all of the information for the save or restore operation is contained in this object
link.
The possible values are:
0 The data is not complete. One or more directory information or object link information formats were not
written to the user space or byte stream file. This can occur when a user space object link is used and more
than 16MB of information about the save or restore operation is generated. This situation occurs only when
the save or restore operation processes a very large number of object links. If this situation occurs, you
should consider using a stream file to store your output information.
1 The data is complete. All of the information about the save or restore operation is contained in the output.

CCSID of data. The CCSID of the data that is stored in this output entry.

Data compacted. Indicates whether the data was stored in compacted format.
The possible values are:
’0’ The data is not compacted.
’1’ The data is compacted.

Data compressed. Indicates whether the data was stored in compressed format.
The possible values are:
’0’ The data is not compressed.
’1’ The data is compressed.

Device name. The name of a device used to perform the save or restore operation. The field contains either the
name of a device or the name of the save file that was used to perform the operation.

Device name length. The length of the Device name field.

Device name offset. The offset to the Device name field.

Directory name. The name of the directory that the object was saved from or where the object was restored.

Directory name length. The length of the directory name field.

Directory name offset. The offset to the directory name field.

End change date. The value that was specified for the end change date when the save operation was performed.
The possible values are:
*ALL No end change date was specified.
end date
The end change date that was specified on the save operation. The date is in YYMMDD format, is left
justified, and is padded with blanks.

End change time. The value that was specified for the end change time when the save operation was performed.
The possible values are:
*ALL No end change time was specified

862 OS/400 Backup and Recovery V5R1

end time
The end change time that was specified on the save operation. The time is in HHMMSS formats, is left
justified, and is padded with blanks.

Entry length. The length of this list entry.

Entry type. Indicates the type of data that is contained in this list entry.
The possible values are:
1 This list entry contains command level information. Use the command information format to map out the
data for this list entry.
2 This list entry contains directory-level information. Use the directory information format to map out the data
for this list entry.
3 This list entry contains link level information. Use the object link information format to map out the data for
this list entry.
4 This list entry contains trailer information. Use the trailer information format to map out the data for this
list entry.

Expiration date. The expiration date of the media.

The possible values are:
*PERM The data is permanent.
expiration date
The expiration date that was specified on the save operation. The date is in YYMMDD format, is left
justified, and is padded with blanks.

File label. The file label of the media file the save or restore operation is using. For a save or restore that uses a
save file, this field is blank.

File label length. The length of the File label field.

File label offset. The offset to the File label field.

Information type. Shows you the type of information that was saved with this operation. (INFTYPE parameter on
SAV command).
The possible values are:
’1’ Summary information and information about each object link that was processed was saved (*ALL).
’2’ Summary information and information about object links that were not successfully saved or restored was
saved (*ERR).
’3’ Only summary information was saved (*SUMMARY).

Number of device identifiers. The number of Device identifier fields.

Number of object links processed successfully in directory. The number of object links that were successfully
saved or restored for this directory.

Number of object links processed unsuccessfully in directory. The number of object links that were unsuccessfully
saved or restored for this directory.

Number of object links that are processed successfully (S/R). The total number of object links saved or restored
successfully.

Number of object links that are processed unsuccessfully (S/R). The total number of object links that were not
saved or restored.

Number of volume identifiers. The number of Volume identifier fields.

Object link data. Indicates whether the data for this object was saved with the object.
The possible values are:

Appendix E. How to Create and Use Output from the Save and Restore Commands 863
’0’ The object’s description was saved, but the object’s data was not saved.
’1’ The object’s description and the object’s data was saved.

Object link error message ID. The message ID of an error message that was issued for this link.

Object link error message replacement data. The error message replacement text from the link error message.

Object link error message replacement data length. The length of the error message replacement text for the object
link error message.

Object link error message replacement identifier offset. The offset to the error message replacement identifier for
the object link error message.

Object link identifier after restore offset. The offset to the Object link name after restore field.

Object link identifier offset. The offset of the object link name identifier.

Object link name. For a save operation, the name of the object link that was saved. For a restore operation, the
qualified object link name that was saved (including the directory and object link name).

Object link name length. The length of the Object link name field.

Object link name after restore. The name of the object link after it was restored.

Object link name after restore length. The length of the Object link name after restore field.

Object link owner after restore. The name of the object link owner’s user profile when the object link was restored.

Object link owner at time of save. The name of the object link owner’s user profile when the object link was saved.

Object link security message. Indicates whether a security message was issued for this object link during a restore
operation.
The possible values are:
’0’ No security messages were issued.
’1’ One or more security messages were issued.

Object link size. The size of the object link in units of the size multiplier. The true object link size is equal to or
smaller than the object link size multiplied by the object link size multiplier.

Object link size multiplier. The value to multiply the object link size by to get the true size. The value is 1 if the
object link is smaller than 1 000 000 000 bytes, 1024 if it is between 1 000 000 000 and 4 294 967 295 bytes (inclusive).
The value is 4096 if the object link is larger than 4 294 967 295 bytes.

Object link status. Indicates whether the object link was successfully processed.
The possible values are:
’0’ The object link was not successfully saved or restored.
’1’ The object link was successfully saved or restored.

Object link text. The text description of the object link.

Object link type. The type of the object link.

Restore date/time. The time at which the object links were restored in system timestamp format. See the Convert
Date and Time Form at (QWCCVTDT) API for information on converting this timestamp.

Restore system serial number. The serial number of the system on which the restore operation was performed.

Restore release level. The release level of the operating system on which the object links were restored. This field
has a VvRrMm format, containing the following:
Vv The character V followed by a 1-character version number

864 OS/400 Backup and Recovery V5R1

 Rr The character R followed by a 1-character release number
Mm The character M followed by a 1-character modification number

Save active. Indicates whether object links were allowed to be updated while they were being saved.
The possible values are:
0 SAVACT(*NO)—Object links were not allowed to be saved while they were in use by another job.
1 SAVACT(*YES)—Object links were allowed to be saved while they were in use by another job. Object links
in the save may have reached a checkpoint at different times and may not be in a consistent state in
relationship to each other.
-1 SAVACT(*SYNC)—Object links were allowed to be saved while they were in use by another job. All of the
object links and all of the directories in the save operation reached a checkpoint together and were saved in
a consistent state in relationship to each other.

Save active date/time. The time at which the object link was saved while active in system timestamp format. See the
Convert Date and Time Format (QWCCVTDT) API for information on converting this timestamp.

Save active option. Indicates which options were used with save-while-active. The possible values are:
*NONE
SAVACTOPT(*NONE) was specified. No special save-while-active options were used.
*ALWCKPWRT
SAVACTOPT(*ALWCKPWRT) was specified. This enabled the system to save objects while the system
updated them if you set the corresponding system attribute. See the Information Center at the following
URL: http://www.ibm.com/eserver/iseries/infocenter for more information.

| Save date/time. The time at which the object links were saved in system timestamp format. See the Convert Date
| and Time Format (QWCCVTDT) API information under the Programming topic on the Information Center at the
| following URL: http://www.ibm.com/eserver/iseries/infocenter for information on converting this timestamp.
Save release level. The release level of the operating system on which the object links were saved. This field has a
VvRrMm format, containing the following:
Vv The character V is followed by a 1-character version number.
Rr The character R is followed by a 1-character release number.
Mm The character M is followed by a 1-character modification number.

Save system serial number. The serial number of the system on which the save operation was performed.

Sequence number. The sequence number of the file on media. The value will be 0 if the save media is not tape.

Start change date. The value that was specified for the start change date when the save operation was performed.
The possible values are:
*LASTSAVE
The save includes object links that have changed since the last time time they were saved with
UPDHST(*YES) specified on the save operation.
*ALL No start change date was specified.
start date
The start change date that was specified on the save operation. The date is in YYMMDD format, is left
justified, and is padded with blanks.

Start change time. The value that was specified for the start change time when the save operation was performed.
The possible values are:
*ALL No start change time was specified.
start time
The start change time that was specified on the save operation. The time is in HHMMSS format, is left
justified, and is padded with blanks.

Appendix E. How to Create and Use Output from the Save and Restore Commands 865
Starting volume identifier. The starting volume identifier on which this object link was saved. This field is a
variable-length field.

Starting volume identifier length. The length of this Starting volume identifier field.

Starting volume identifier offset. The offset to the starting volume identifier field.

Target Release level. The earliest release level of the operating system on which the object links can be restored.
This field has a VvRrMm format, containing the following:
Vv The character V is followed by a 1-character version number.
Rr The character R is followed by a 1-character release number.
Mm The character M is followed by a 1-character modification number.

Volume identifier. The list of volume identifiers that are used during this save or restore operation. The list can
contain from one to 75 volumes. See ″number of volume identifiers″ to tell how many volume identifiers are in the
list. This field is a variable-length field.

Volume identifier length. The length of this Volume identifier field.

Volume identifiers offset. The offset to the Volume identifier field.

866 OS/400 Backup and Recovery V5R1

Appendix F. Procedures for Recovering the Text Index
Table 154 shows problems that can occur with the text-index database files and
associated libraries. The table also provides solutions and procedures to correct the
problems.
Table 154. Recovery for Search Index Services Files
Problem and Cause Solution Procedure
Problem: Restore all files from save media. 1. RSTLIB SAVLIB(QUSRSYS)
After the restore operation is
2. RCLDLO DLO(*DOCDTL)
All text index files are gone. complete, run the reclaim
Cause: operation2,3. After the reclaim 3. STRUPDIDX
operation is complete, start updating
Library QUSRSYS was cleared or index requests.
deleted1.
Problem: Restore the scheduling queue file and 1. RSTOBJ OBJ(QABBIQTB)
optionally clear the scheduling SAVLIB(QUSRSYS)
The scheduling queue file is not queue.
2. WRKTXTIDX
usable.
Cause: 3. Select option 2 (Change details)
on the index that has the
Scheduling queue was damaged or scheduling queue to be cleared.
deleted. 4. Under the clear parameter, select
a Y to clear the scheduling queue.
Problem: Delete all the text index files from 1. DLTF OBJ(QABB*) LIB(QUSRSYS)
QUSRSYS and then restore all the
2. RSTOBJ OBJ(QABB*)
The text index files are usable. Their files from save media4. After the
SAVLIB(QUSRSYS)
content is not usable. restore operation is complete, run the
Cause: reclaim operation2,3. After the reclaim 3. RCLDLO DLO(*DOCDTL)
operation is complete, start updating 4. STRUPDIDX
Text index files that were not all index requests.
saved at the same time were restored.
Problem: Restore text index files and then the See Table 8 on page 64 for the
document library5. See cases 1 procedure to restore the system ASP.
The document library and text index through 4 in Table 155 on page 869 The procedure restores QUSRSYS
files are lost. about the types of index request (text index files) and documents.
Cause: created when using the RSTDLO
command.
A disk unit failure occurred that
resulted in data loss on the system
ASP, or function code 24 was selected
to install the Licensed Internal Code.

© Copyright IBM Corp. 1997, 2000, 2001 867

Table 154. Recovery for Search Index Services Files (continued)
Problem and Cause Solution Procedure
6
Problem: Clear the scheduling queue , restore 1. WRKTXTIDX
the document library and start
2. Select option 2 (Change details)
The document library is lost. The text updating index requests. After the
on the index that has the
index files are not lost. update operation has completed7,
scheduling queue to be cleared.
Cause: you may optionally want to run the
reclaim operation, and then start 3. Under the clear parameter, select
Library QDOC or library QDOCnnnn updating index requests again2. See a Y to clear the scheduling queue.
was cleared or deleted. Table 155 on page 869 for rules about 4. RSTDLO DLO(*ALL)
the types of index request created SAVFLR(*ANY)
when using the RSTDLO command. 5. STRUPDIDX
6. RCLDLO DLO(*DOCDTL)
7. STRUPDIDX
6
Problem: Start updating index requests . After 1. STRUPDIDX
the update operation has completed8,
2. RSTDLO DLO(document-name
One or more documents are lost. restore the lost documents. Start
SAVFLR(folder-name)
Cause: updating index requests again. See
cases 1 through 4 in Table 155 on 3. STRUPDIDX
One or more documents were page 869 about the types of index
damaged or deleted. requests created when using the
RSTDLO command.
Problem: Create a save file for the undamaged 1. CRTSAVF
One or more text index files are lost. scheduling queue. Save the FILE(QUSRSYS/SAVTXTQ)
Cause: scheduling queue to the save file. TEXT(’Save of Text Search
Delete all the text index files from Scheduling Queue’)
One or more QABB* files (except QUSRSYS. Restore all the files from
2. SAVOBJ OBJ(QABBIQTB)
(QABBIQTB)) were damaged or save media5.. Delete the restored
LIB(QUSRSYS) DEV(*SAVF)
deleted. Library QDOC is not lost. scheduling queue and then restore it
SAVF(QUSRSYS/SAVTXTQ)
from the save file5. After the restore
ACCPTH(*YES)
operation is complete, run the
reclaim operation2,3. When the 3. DLTF FILE(QUSRSYS/QABB*)
reclaim operation is complete, start 4. RSTOBJ OBJ(QABB*)
updating text index requests. When SAVLIB(QUSRSYS)
the update index requests are 5. DLTF
complete, delete the save file. FILE(QUSRSYS/QABBIQTB)
6. RSTOBJ OBJ(QABBIQTB)
SAVLIB(QUSRSYS) DEV(*SAVF)
SAVF(QUSRSYS/SAVTXTQ)
7. RCLDLO DLO(*ALL)
8. STRUPDIDX
9. DLTF FILE(QUSRSYS/SAVTXTQ)
6
Problem: Start updating index requests . After 1. STRUPDIDX
the update operation has completed7,
2. RCLDLO DLO(*ALL)
The last-version index dates for run the reclaim operation2,3.
documents do not match, or 3. STRUPDIDX
documents you expected to find in a
search are not found.
Cause:

STRUPDIDX command was

interrupted, encountered errors, or
was not started.

868 OS/400 Backup and Recovery V5R1

Table 154. Recovery for Search Index Services Files (continued)
Problem and Cause Solution Procedure
Notes:
1
Only library QUSRSYS is lost. The recovery for this problem is not used if you are recovering the system
using function code 24 to install the Licensed Internal Code.
2
The RCLDLO command deletes from the index all document content that does not have a corresponding
document on the system. Documents that are stored in the index but not on the system take up space in
the index, although they are never found in the search.
3
The RCLDLO command looks at each index entry to see if the index date for both the entry and the
corresponding document match. If the document date does not match, the date is stored in the index. If the
document has a last-version index date with no corresponding entry in the index, the RCLDLO command
creates an add request for the document to bring the index up to date.
4
If the damaged files only are deleted and then restored without all other index files, the following occurs:
v The restore operation clears the text index
v The reclaim operation (RCLDLO command) creates an add index request for every document on the
system that is marked as being indexed before.
v The documents should correctly indicate whether they are or are not in the index.
5
If the scheduling queue contains requests, these requests will be processed the next time the STRUPDIDX
command is run.
6
Pending requests on the scheduling queue could cause unpredictable results if the queue is not processed
or cleared prior to performing the rest of the recovery operation.
7
While the update operation is in progress, the Update Started status on the Work with Text Index display is
set to Yes. When the update operation has completed, this status is set to No. A job completion message is
also sent to the user who ran the STRUPDIDX command.

Table 155. Type of Index Request Created when Using the Restore Document Library Object (RSTDLO) Command
NEWOBJ(*NEW) Is Document Was Index Entry Exists for Document Type of Index Request Created
Specified Indexed when Saved in QABB* Files
(1) No No No None
(2) No No Yes Remove
(3) No Yes No Add
(4) No Yes Yes Create an add request if version
indexed dates are different.

None if version indexed dates

are equal.
(5) Yes No N/A None
(6) Yes Yes N/A Add

Appendix F. Procedures for Recovering the Text Index 869

870 OS/400 Backup and Recovery V5R1
Appendix G. Recovering your system
This section provides instructions to completely recover your entire system. Use
these steps if you need to recover your system to the same system (restore to the
same system with the same serial number). Use these recovery instructions ONLY
if you saved your entire system by using either of the following options:
v Save menu Option 21
v Equivalent Save Commands as Save menu Option 21:

- SAVSYS
- SAVLIB LIB(*NONSYS) ACCPTH(*YES)
- SAVDLO DLO(*ALL) SAVFLR(*ANY)
- SAV DEV(’/QSYS.LIB/tape-device-name.DEVD’) OBJ((’/*’)
(’/QSYS.LIB’ *OMIT) (’/QDLS’ *OMIT)) UPDHST(*YES)

Important
Use “Recovering your entire system after a complete system loss–Checklist
20” on page 96 for any of the following cases. You can find the checklist in
“Chapter 4. Selecting the Right Recovery Strategy” on page 61.
v System has Logical Partitions.
v System uses the Alternate Installation Device Setup feature that you can
define through DST for D-IPL.5
v System has mounted User-Defined File Systems prior to save.
v You are recovering to a different system (system with a different serial
number).

Check off each item as you complete the task.

__ 1. IPL the system from the first SAVSYS media.
__ a. Mount the first SAVSYS media on the alternate IPL device. Wait for
the READY status.
__ b. At the CPU control panel, place the system in MANUAL mode.
__ c. Press the Function Select switch (or buttons) to display 02 (IPL) in
the Function display.
__ d. Press Enter.
__ e. Press the Function Select switch (or buttons) to display D (IPL from
tape or CD-ROM) in the Data display.
__ f. Press Enter.
__ g. If the system is powered down, press the power button on the
system to power the system on. Skip to Step 2 on page 872.
Otherwise, continue with Step 1h.
__ h. If the system is powered on, press the Function Select switch (or
buttons) to display 03 (continue the IPL) in the Function display.
__ i. Press Enter.

5. Refer to Chapter 22, ″Using an Alternate Installation Device″.

© Copyright IBM Corp. 1997, 2000, 2001 871

 __ 2. On the Install Licensed Internal Code screen, select 1, Install Licensed
Internal Code.

Install Licensed Internal Code

Select one of the following:

1. Install Licensed Internal Code

2. Work with Dedicated Service Tools (DST)
3. Define alternate installation device

Selection
1

__ 3. On the Install Licensed Internal Code (LIC) screen, select 2, Install Licensed
Internal Code and Initialize System, to start a ″Scratch Install″ of the
Licensed Internal Code.

Install Licensed Internal Code (LIC)

Disk selected to write the Licensed Internal Code to:

Serial Number Type Model I/O Bus Controller Device

xx-xxxxxxx xxxx xxx x x x

Select one of the following:

1. Restore Licensed Internal Code

=>2. Install Licensed Internal Code and Initialize System
3. Install Licensed Internal Code and Recover Configuration
4. Install Licensed Internal Code and Restore Disk Unit Data
5. Install Licensed Internal Code and Upgrade Load Source

Selection
2

__ 4. On the Install LIC and Initialize System - Confirmation screen, press F10 to
confirm the initialization and to continue the install.

Install LIC and Initialize System - Configuration

Warning:

All data on this system will be destroyed and the Licensed

Internal Code will be written to the selected disk if you
choose to continue the initialize and install.

Return to the install selection screen and choose one of the

other options if you want to perform some type of recovery
after the install of the Licensed Internal Code is complete.

Press F10 to continue the install.

Press F12 (Cancel) to return to the previous screen.
Press F3 (Exit) to return to the install selection screen.

__ a. The Initialize the Disk - Status screen appears.

872 OS/400 Backup and Recovery V5R1

 Initialize the Disk - Status

The load source disk is being initialized.

Estimated time to initialize in minutes: xx

Elapsed time in minutes . . . . . . . .: 0.0

__ b. The Install Licensed Internal Code - Status screen appears.

Install Licensed Internal Code - Status

Install the Licensed Internal Code in progress

Percent Complete . . . . . 0% 8.5 minutes

__ 5. On the Disk Configuration Attention Report screen, press F10 to accept any
problems and to continue.

Disk Configuration Attention Report

Type option, press Enter

5=Display Detailed Report

Press F10 to accept all the following problems and continue.

The system will attempt to correct them.

OPT Problem
_ New disk configuration

F3=Exit F10=Accept the problems and continue

__ 6. On the IPL or Install the System screen, select 3, Use Dedicated Service
Tools (DST).

IPL or Install the System

Select one of the following:

1. Perform an IPL
2. Install the operating system
3. Use Dedicated Service Tools (DST)
4. Perform automatic installation of the operating system
5. Save Licensed Internal Code

Selection
3

__ 7. Sign on to DST as DST user, QSECOFR, with the password QSECOFR.

Appendix G. Recovering your system 873

 Dedicated Service Tools (DST) Sign On

Type choices, press Enter.

DST User . . . . . . . . QSECOFR

DST password . . . . . . QSECOFR

__ a. Select option 4, Work with Disk Units.

__ b. Select option 1, Work with Disk Configuration.
__ c. Select option 3, Work with ASP Configuration.
__ d. Select option 3, Add Units to ASPs.
__ 8. On the Specify ASPs to Add Units to screen, enter ″1″ for each unit that
needs to be in the System ASP (ASP 1).
__ a. If you require more than one ASP, enter the corresponding ASP
number on each selected unit.

Specify ASPs to Add Units to

Specify the ASP to add each unit to.

Specify Serial Resource

ASP Number Type Model Capacity Name

1 00-0103706 6602 030 1031 DD031

1 00-1000341 9337 211 542 DD012
1 00-5000341 9337 211 542 DD015
1 00-7000341 9337 211 542 DD011
1 00-3000341 9337 211 542 DD014
1 00-2000341 9337 211 542 DD013
1 00-61300 6603 074 1475 DD006
1 00-52262 6606 074 1475 DD008
1 00-86978 6606 050 1967 DD009
2 00-95744 6603 074 1475 DD005
2 00-47657 6606 074 1475 DD007
- 00-0238703 6602 074 773 DD051
- 00-0128350 6602 074 773 DD051

__ b. After you complete all units, press Enter.

__ c. If the list of units is correct, press Enter to start initializing the units.
__ 9. On the Problem Report screen, press F10, Ignore problems and continue.

Problem Report

Note: Some action for the problems listed below may need to
be taken. Please select a problem to display more detailed
information about the problem and to see what possible
action may be taken to correct the problem.

Type option, press Enter.

5=Display Detailed Report

OPT Problem
_ Unit possibly configured for Power PC AS

F3=Exit F10=Ignore problems and continue F12=Cancel

874 OS/400 Backup and Recovery V5R1

__ 10. On the Confirm Add Units screen, press Enter to confirm the selected units.

Confirm Add Units

Add will take several minutes for each unit. The system will
have the displayed protection after the unit(s) are added.

Press Enter to confirm your choice for 1=Add units.

Press F9=Capacity Information to display the resulting capacity.
Press F12=Cancel to return and change your choice.

Serial Resource
ASP Unit Number Type Model Name Protection

1 Unprotected
1 00-0103706 6602 030 DD031 Unprotected
2 00-1000341 9337 211 DD012 Unprotected
3 00-5000341 9337 211 DD015 Unprotected
4 00-7000341 9337 211 DD011 Unprotected
5 00-3000341 9337 211 DD014 Device Parity
6 00-2000341 9337 211 DD013 Device Parity
7 00-61300 6603 074 DD006 Device Parity
8 00-52262 6606 074 DD008 Device Parity
9 00-86978 6606 050 DD009 Device Parity
2 Unprotected
10 00-95744 6603 074 DD005 Device Parity
11 00-47657 6606 074 DD007 Device Parity

__ a. The Function Status screen displays the percentage of completion.

__ b. The ″Selected units have been added successfully″ message appears
when the system completes the Add Units process.
__ c. Press F12 to return to the Work with Disk Configuration screen.
__ d. If your system requires mirrored protection, continue to Step 10e. If
your system does not require mirrored protection, press F3 until you
see the Exit Dedicated Service Tools (DST) screen. Select 1 to exit DST
and press Enter and continue to Step 10f.
__ e. To start mirrored protection for your system, follow these
instructions:
__ 1) On the Work with Disk Configuration screen, select 4, Work
with Mirrored Protection.
__ 2) On the Work with Mirrored Protection screen, select 2, Start
Mirrored Protection.
__ 3) Select an ASP by placing a ″1″ next to it. Press Enter to start
mirrored protection.
__ 4) On the Confirm Continuation screen, press Enter.
__ 5) On the Confirm Start Mirrored Protection screen, press Enter.
__ 6) The Function Status screen displays the completion status of
the Start Mirrored Protection request.
__ 7) The Start mirroring completed successfully message appears on
the Disk Configuration Information Report screen.
__ 8) Press Enter to continue.
__ f. If you use Operations console, follow these instructions to switch
working from ’locale console’ to ’operations console’:
__ 1) On the IPL or Install the System screen, select 3, Use Dedicated
Service Tools (DST). Press Enter to continue.

Appendix G. Recovering your system 875

 __ 2) Sign on to DST as DST User, QSECOFR, with the password
QSECOFR.
__ 3) On the Use Dedicated Service Tools (DST) screen, select 5, Work
with DST environment. Press Enter to continue.
__ 4) On the Work with DST Environment screen, select 2, System
Devices. Press Enter to continue.
__ 5) On the Work with System Devices screen, select 6, Console
Mode. Press Enter to continue.
__ 6) On the Select Console Type screen, select 2, Operations Console.
Press Enter to continue.
__ 7) Press F3 or F12 to get back to the IPL or Install the System
screen.
__ 11. On the IPL or Install the System screen, select 2, Install the Operating
System.

IPL or Install the System

Select one of the following:

1. Perform an IPL
2. Install the Operating System
3. Use Dedicated Service Tools (DST)
4. Perform automatic installation of the Operating System
5. Save Licensed Internal Code

Selection
2

__ a. On the Confirm Install of OS/400 screen, press Enter.

__ b. The Select a Language Group screen displays the primary language
feature that is currently on your save media. Press Enter to accept
this value.

Select a Language Group

Note: The language feature shown is the language feature

installed on the system.

Attention: To keep the same primary language, ensure that the

media you use for installing the operating system
matches the language feature shown. If the operating
system media does not match what is shown, the
installation process will attempt to install the
operating system in a different language feature than
Licensed Internal Code. This is undesirable.
Type Choice, press Enter.
Language Feature . . . . . . . . . 2924

__ c. The Confirm Language Feature Selection screen appears. Press Enter to

continue.
__ 12. On the Add All Disk Units to the System screen, select option 1, Keep the
current disk configuration.

876 OS/400 Backup and Recovery V5R1

 Add All Disk Units to the System

Select one of the following:

1. Keep the current disk configuration

2. Perform disk configuration using DST
3. Add all units to the system auxiliary storage pool (ASP)
4. Add all units to the system ASP and balance data

Selection
1

Note: This screen does not appear if you selected all disk units that are
known to the system on Step 8 on page 874.
__ 13. The IPL Step in Progress screen displays the IPL progress.

IPL Step in Progress

IPL step . . . : Storage Management Recovery

Authority Recovery
Journal Recovery
Database Recovery
Journal Synchronization
Start Operating System

__ 14. On the Install the Operating System screen, select option 1, Take defaults.
Make sure that the values for Date and Time are correct. Press Enter to
continue.

Install the Operating System

Type options, press Enter.

Install
option . . . . . 1 1=Take defaults (No other
options are displayed)
2=Change install options

Date
Year . . . . . . 99 00-99
Month. . . . . . 08 01-12
Day. . . . . . . 22 01-31

Time
Hour . . . . . . 16 00-23
Minute . . . . . 45 00-59
Second . . . . . 00 00-59

__ 15. The OS/400 Installation Status screen displays the installation status of the
required OS/400 Installation Profiles and Libraries.

Appendix G. Recovering your system 877

 Message ID . . : CPI2070 OS/400 Installation Status

+---------------------------------------------------------------------------------+
Stage 1 |xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx |
+---------------------------------------------------------------------------------+
0 20 40 60 80 100

Installation Objects
Stage Completed Restored

1 Creating needed profiles and libraries . . .:

2 Restoring programs to library QSYS . . . . .:
3 Restoring language objects to library QSYS .:
4 Updating program table . . . . . . . . . . .:
5 Installing database files . . . . . . . . . :
6 Completing OS/400 installation . . . . . . .:

__ 16. The system installs the remaining OS/400 objects.

Message ID . . : CPI2070 OS/400 Installation Status

+---------------------------------------------------------------------------------+
Stage 1 |xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx |
+---------------------------------------------------------------------------------+
0 20 40 60 80 100

Installation Objects
Stage Completed Restored

1 Creating needed profiles and libraries . . .: x

2 Restoring programs to library QSYS . . . . .: x 09
3 Restoring language objects to library QSYS .: x 03
4 Updating program table . . . . . . . . . . .: x
5 Installing database files . . . . . . . . . : x
6 Completing OS/400 installation . . . . . . .:

__ 17. On the Sign On screen, logon as user QSECOFR. You do not need to enter a
password at this time.
__ 18. On the IPL options screen, enter correct values for Date and Time. Only the
following options should be set to Y:
v Start system to restricted state
v Set major system options
v Define or change the system at IPL

IPL Options

Type choices, press Enter.

System date . . . . . . . . . . . . . . . 08/22/99 MM/DD/Y

System time . . . . . . . . . . . . . . . 16:58:00 HH:MM:S
Clear job queues . . . . . . . . . . . . . N Y=Yes, N=No
Clear output queues . . . . . . . . . . . N Y=Yes, N=No
Clear incomplete job logs . . . . . . . . N Y=Yes, N=No
Start print writers . . . . . . . . . . . N Y=Yes, N=No
Start system to restricted state . . . . . Y Y=Yes, N=No

Set major system options . . . . . . . . . Y Y=Yes, N=No

Define or change system at IPL . . . . . . Y Y=Yes, N=No

878 OS/400 Backup and Recovery V5R1

 __ a. On the Set Major System Options screen, select N to disable automatic
configuration.

Set Major System Options

Type choices, press Enter.

Enable automatic configuration . . . . N Y=Yes, N=No

Device configuration naming . . . . . NORMAL *NORMAL, *S36
*DEVADR
Default special environment . . . . . *NONE *NONE, *S36

__ 19. The Define or Change the System at IPL screen appears.

__ a. Select 3, System Value Commands.
__ b. On the Change System Value Commands screen, select 3, Work with
System Values.
__ c. On the Work with System Values screen, select the System Value that
you plan to change by placing a ″2″ next to it. Press Enter ONLY
after you select all the values.
Update the following System Values. Write down the existing values
so you can update them after the recovery, if necessary.
v Change QALWOBJRST to *ALL
v Change QJOBMSGQFL to *PRTWRAP
v Change QJOBMSGQMX size to a minimum value of 30
v Change QPFRADJ to 2
v Change QIPLTYPE type to 2
| v Change QVFYOBJRST to 1
__ d. After the system changes the system values, press F3 twice to return
to the Define or Change the System at IPL screen.
__ e. On the Define or Change the System at IPL screen, press F3 to exit and
continue the IPL.
__ 20. On the Change Password screen, type QSECOFR as the current password.
Enter a new password. Re-enter the password to verify and press Enter.
(New password cannot be QSECOFR.)

Change Password

Password last changed . . . . . . xx/xx/xx

Type choices, press Enter.

Current password . . . . . . . . . QSECOFR

New password . . . . . . . . . . . _______
New password (to verify) . . . . . _______

| __ 21. To configure 3422, 3430, 3480, or 3490 tape units, follow these instructions.
| If you have a 3490 Model E or F or to configure other types of tape units,
| go to step 22 on page 881.
a. Use the Work with Hardware Resource (WRKHDWRSC) command to
determine the location of the tape controller.
WRKHDWRSC TYPE(*STG)
b. Create the controller description for the tape controller by doing the
following:

Appendix G. Recovering your system 879

 1) Locate the resource name for the tape controller on the Work with
Storage Resources display. The value 34xx is displayed in the Type
column.
2) Write down the name of the resource.
3) Type a 9 (Work with resource) in the Opt column next to name of
the tape controller and press the Enter key. You see the Work with
Storage Resources display.

Note: If the resource is not listed on the display, you need to select
other resources, such as disk storage controllers. For some
server models, resources are now attached through
combined-function IOPs. Look through the resources until
you find the device you want.
4) Type 5 (Work with controller descriptions) in the Opt column in
front of the tape controller. You see the Work with Controller
Description display.
5) Type 1 (Create) in the Opt column on the top line.
6) Type the controller name (such as TAPCTL01) in the description
field and press the Enter key. You see the Create Controller
Description display.
7) If necessary, type additional information on the display. Then press
the Enter key. You return to the Work with Controller Descriptions
display.
8) If the controller description that you created does not appear, press
F5 (Refresh).
c. To create device descriptions for the tape units that are attached to the
controller, do the following:
1) On the work with Controller Descriptions display, press F23 (More
options). The list of options changes.
2) Type 9 (Work with associated descriptions) in the Opt column in
front of the new tape controller. You see the Work with Associated
Descriptions display.
3) Locate the resource for the tape unit. Because no device description
exists, the description says *NONE.
4) Write down the name of the tape resource.
5) Type a 1 (Create) in the Opt column next to the description of
*NONE and press the Enter key. You see the Create Device Desc
(Tape) (CRTDEVTAP).
6) In the Device description field, type a name such as TAP01.
7) In the Resource name prompt, type the name that you wrote down
in step 21c4. (If you did not write it down, press F12 to return to
the display. Repeat steps 21c4 through 21c7.)
8) Press the Enter key.
9) Additional parameters appear on the display.
10) If necessary, type additional information on the display. Then press
the Enter key. You return to the Work with Associated Descriptions
display.
11) Press F5 (Refresh). The name of the description that you created
should now be associated with the resource.

880 OS/400 Backup and Recovery V5R1

 12) Type 8 (Work with configuration status) in front of the controller
description and the device description. You see the Work with
Configuration Status display.
13) Type 1 (Vary on or Make available) in front of the controller and
the devices.
d. Press F3 until your return to your original menu.
| __ 22. To configure tape units that are 3490 Model E or F, or that are not models
| 34xx, use the following instructions:
a. Use the Work with Hardware Resource (WRKHDWRSC) command to
determine tape controller name.
WRKHDWRSC TYPE(*STG)
b. Locate the tape controller on the Work with Hardware Resources
display.
c. Type a 9 (Work with resource) next to tape controller name and press
the Enter key.

Note: If the tape controller is not listed on the display, you need to
select other resources, such as disk storage controllers. For some
server models, tape units are now attached through
combined-function IOPs. Look through the resources until you
find the tape unit you want.
d. Locate the resource name for the tape unit (for example, TAP01).
e. Enter a 5 (Work with Configuration Descriptions) in the Opt column
next to the tape resource name and press the Enter key.
You are shown the Work with Configuration Descriptions display.
f. Type a 1 (Create) in the Opt field and a tape device description name
(for example, TAP01) in the Description field. Press the Enter key. You
are shown the Create Device Description (Tape) display.
g. Change any values that you want to change, then press the Enter key
(twice) to create the device description. You are shown the Work with
Configuration Descriptions display again. The device that you created
should appear on the display.
h. Type an 8 (Work with configuration status) in front of the new device
description. You are shown the Work with Configuration Status display.
i. Type a 1 (Vary on or Make available) in front of the new device. If the
status does not change to Varied On or Available, wait a few minutes.
Then press F5 (Refresh). If the status still does not change to Varied On
or Available, follow normal problem analysis for the device.
j. Press F3 until you return to the main menu.

Appendix G. Recovering your system 881

 AS/400 Main Menu

Select one of the following:

1. User tasks
2. Office tasks
3. General system tasks
4. Files, libraries, and folders
5. Programming
6. Communications
7. Define or change the system
8. Problem handling
9. Display a menu
11. Client Access/400 tasks
90. Sign off

Selection or command
=>

__ 23. On the AS/400 Main Menu screen, type the command, WRKRPYLE, and check
to see if CPA3709 is there. If it is not, determine an available sequence
number and then Press F6 to add MSGID(CPA3709) RPY(G) by using the
available sequence number. Press F5 to Refresh and verify that you added
CPA3709.
__ a. Type the command CHGJOB INQMSGRPY(*SYSRPYL) to update the
current job to use the system reply list for inquiry messages.
__ 24. On the AS/400 Main Menu screen, type GO RESTORE to access the AS/400
Restore screen.
__ a. On the AS/400 Restore screen, select option 21, Restore System and
User Data.
__ b. Press Enter to continue.
__ 25. On the Specify Command Defaults screen, enter the name of the tape drive
that you are using for the restore.
__ a. Set Prompt for command to N.
__ b. Set Message queue delivery to *NOTIFY.

Specify Command Defaults

Type choices, press Enter.

Tape devices . . . . . . . . . . . . TAP01 Names

Prompt for commands. . . . . . . . . N Y=Yes, N=No
Message queue delivery . . . . . . . *NOTIFY *BREAK, *NOTIFY
Restore to different system. . . . . N Y=Yes, N=No

__ c. Press Enter to continue ending the subsystems. The restore process

starts to run unattended. The restore process stops only if the system
requires a tape mount for the restore process to continue. The restore
process should run to completion.
__ 26. Reapply any PTFs that you applied since since the last time that you saved
your system data.
__ a. Locate the most recent cumulative PTF (program temporary fix)
tape.
__ b. From a command-line, enter GO PTF to access the PTF menu.
__ c. Select 8, Install program temporary fix package, on the PTF menu.

882 OS/400 Backup and Recovery V5R1

 This installs all of the PTFs in the cumulative PTF package for the
licensed programs that are installed on your system. Refer to the
AS/400 System PTF Shipping Information Letter for required special
instructions.
Notes:
1) If you want to restore individual PTFs, see the Systems Operation
publication for more information about applying individual
PTFs.
2) If you do not have the PTFs that you need, order them and
apply them later.
__ 27. If necessary, change the QALWOBJRST system value back to its original
value by using the WRKSYSVAL command.
__ 28. If necessary, change the QJOBMSGQFL system value back to its original
value by using the WRKSYSVAL command.
__ 29. If necessary, change the QJOBMSGQMX system value back to its original
value by using the WRKSYSVAL command.
__ 30. If necessary, change the QIPLTYPE system value back to its original value
by using the WRKSYSVAL command.
__ 31. If you do not know the password for the restored QSECOFR profile,
change the password before signing off. Type the following command:
CHGUSRPRF USRPRF(QSECOFR) PASSWORD(new-password)
__ 32. To place scheduled jobs on hold, type WRKJOBSCDE and select option 3 to
hold any scheduled jobs. You need to release the jobs in Step 39 on
page 884 after you complete the restore.
__ 33. Type the command, SIGNOFF *LIST or DSPJOBLOG * *PRINT. Check the job
log to ensure that the job restored all objects. To verify if the job restored
all objects, spool the job log for printing along with any remaining spooled
job output. Check for any error messages. Correct the errors and restore
those objects from media.
__ 34. Perform an IPL of the system.
__ a. On the CPU control panel, select the NORMAL setting.
__ b. Type the following command:

PWRDWNSYS OPTION(*IMMED) RESTART(*YES

*FULL) IPLSRC(B)
| __ 35. If you installed Content Manager OnDemand for iSeries (5769-RD1) on
| your system, restart the journaling for Content Manager OnDemand for
| iSeries (5769-RD1) by typing the following commands:

CALL QRDARS/QRLCSTRJ PARM('RLC')

CALL QRDARS/QRLCSTRJ PARM('RLR')
CALL QRDARS/QRLCSTRJ PARM('RLO')
__ 36. If you installed AS/400 Integration for Windows Server (5769-WSV) on
your system and you saved with your Netfinity Servers in a VARIED ON
setting, perform the following steps:
__ a. Vary off any Netfinity Servers that are varied on by using the
WRKCFGSTS *NWS command. Select option 2.
__ b. Create any needed Network Storages by using the CRTNWSSTG
command.
__ c. Add the storage links by using the ADDNWSSTGL command.

Appendix G. Recovering your system 883

 __ d. Vary on your Netfinity Servers by using the WRKCFGSTS *NWS
command. Select option 1.
__ e. Restore AS/400 Integration for Windows Server (5769-WSV) data
by typing RST OBJ('/QNTC').
__ f. Press Enter.
__ g. Follow the instructions in Step 33 on page 883 to check for error
messages, verify your restore, and correct any errors.
__ 37. If you installed AS/400 Integration for Windows Server (5769-WSV) on
your system and you saved with your Netfinity Server in a VARIED OFF
setting, perform the following steps:
__ a. Add the links for the server descriptions. Type the following for
each server description:

ADDNWSSTGL NWSSTG(Storage_Name) NWSD(Server_Description)

__ b. Vary on your Netfinity Servers by typing WRKCFGSTS *NWS and by
selecting option 1 to vary on each Netfinity Server.
__ 38. If you installed any of the following CCA Cryptographic Service Provider
on your system, reinstall the CCA Cryptographic Service Provider by
using option 11 on the Work with Licensed Programs menu.
v Cryptographic Access Provider for AS/400 - 40 bit (5769-AC1)
v Cryptographic Access Provider for AS/400 - 56 bit (5769-AC2)
v Cryptographic Access Provider for AS/400 - 128 bit (5769-AC3)
__ 39. Release the jobs that you previously held in Step 32 on page 883.

884 OS/400 Backup and Recovery V5R1

Appendix H. Using Operational Assistant to Save Information
This chapter describes how to use options from the Operational Assistant* menus
to save information on your system. Operational Assistant provides a simple,
menu-driven approach to saving information on your system regularly. Figure 100
on page 886 provides a general view of how the Operational Assistant menus and
commands can be used to save information.

To see how Operational Assistant compares with other commands and menu
options for saving information, review Figure 2 on page 16.

Starting with the V3R7 release, Operational Assistant allows you to save your user
directories along with your user libraries and folders. If you choose to save your
user directories, you may not select which ones that the system saves. You must
either save all the user directories or none of the user directories. If you use
applications that have objects in directories, look in the Information Center for
additional information. The Information Center contains pages that deal with how
to plan a backup and recovery strategy to help you decide what strategy you
should use.

© Copyright IBM Corp. 1997, 2000, 2001 885

 Options from
RUNBCKUP menu
┌─────S┌───────────────────────────────┐
│ │ Licensed Internal Code │ Saved using
│ ├───────────────────────────────┤ RUNBCKUP command
│ │ OS/400 Objects in QSYS │ or RUNBCKUP menu,
│ └───────────────────────────────┘ option 1, 2, or 3
│
│ ┌───────────────────────────────┐W─────────────┐
│ │ User Profiles │ │
│ ├───────────────────────────────┤ │
│ │ Private Authorities │ │
│ └───────────────────────────────┘ Can be │
│ included │
11 ┌───────────────────────────────┐ on │
│ │ Configuration Objects │ *DAILY, │
│ └───────────────────────────────┘ *WEEKLY, │
│ or │
│ ┌──S┌───────────────────────────────┐ *MONTHLY │
│ │ │ OS/400 Optional Libraries │ backup │
│ 10 │ QHLPSYS QUSRTOOL │ options │
│ │ ├───────────────────────────────┤ │
│ │ │ Licensed Program Libraries │ │
│ │ │ QRPG QCBL Qxxxxx │ │
│ └──S└───────────────────────────────┘ │
│ │
│ ┌───────────────────────────────┐ │
│ │ IBM Libraries with User Data │ │
│ │ QGPL QUSRSYS QS36F #LIBRARY │ │
│ ├───────────────────────────────┤ │
│ │ User Libraries │ │
│ │ LIBA LIBB LIBC LIBxxx │ │
│ └───────────────────────────────┘ │
│ │
│ ┌───────────────────────────────┐ │
│ │ Documents and Folders │ │
│ ├───────────────────────────────┤ │
│ │ Distribution Objects │ │
│ └───────────────────────────────┘ │
│ │
│ ┌───────────────────────────────┐ │
│ │ User Directories │ │
│ │ DIRA DIRB DIRC DIRxxx │ │
│ ├───────────────────────────────┤W─────────────┘
│ │ IBM Directories │
└─────S└───────────────────────────────┘

Figure 100. Operational Assistant Save Options

When you use the Operational Assistant backup functions, referred to as

Operational Assistant backup, you can set up three backup definitions: daily,
weekly, and monthly. A backup definition includes information about what will be
saved, when it will be saved, and how it will be saved.

Summary of Operational Assistant Commands and Menu Options for

Backup
You can perform Operational Assistant backup tasks by using either menu options
or commands. Table 156 on page 887 provides a summary of the options available.
Online information for both the commands and the menu options provides
detailed information about the parameters that you can specify.

886 OS/400 Backup and Recovery V5R1

 Note: In V3R7, APIs were made available that perform many of the same
functions as the commands and menu options. For more information on
these APIs, refer to the Programming topic on the Information Center at the
following URL: http://www.ibm.com/eserver/iseries/infocenter
Table 156. Summary of Operational Assistant Backup Options
Menu Option Equivalent Operational Assistant Command

Backup Tasks Menu:

1. Run backup GO BACKUP
2. Display backup status DSPBCKSTS
10. Set up backup GO SETUPBCKUP
20. Initialize a tape
21. Initialize a tape set
Run Backup Menu:
1. Run daily backup RUNBCKUP BCKUPOPT(*DAILY)
2. Run weekly backup RUNBCKUP BCKUPOPT(*WEEKLY)
3. Run monthly backup RUNBCKUP BCKUPOPT(*MONTHLY)
10. Back up IBM-supplied libraries SAVLIB LIB(*IBM)
11. Back up the entire system Option 21 from Save menu
Set Up Backup Menu
1. Change the daily backup options CHGBCKUP BCKUPOPT(*DAILY)
2. Change the weekly backup options CHGBCKUP BCKUPOPT(*WEEKLY)
3. Change the monthly backup options CHGBCKUP BCKUPOPT(*MONTHLY)
10. Change the backup library list EDTBCKUPL BCKUPL(*LIB)
11. Change the backup folder list EDTBCKUPL BCKUPL(*FLR)
20. Change the backup schedule
Commands not on menus:
RTVBCKUP (Retrieve Backup Options)
DSPBCKUP (Display Backup Options)
DSPBCKUPL (Display Backup List)

Getting Started with Operational Assistant Backup

Operational Assistant backup is intended to be simple and to run without much
attention. However, you still need to plan and to monitor backup procedures
regularly to ensure that they are working as you intended. Do the following when
you are setting up your backups using Operational Assistant:
1. Start by saving the entire system by using option 11 from the Run Backup
menu. This menu option is the same as option 21 from the Save menu. It
ensures that you have a base set of tapes that has everything on it, in case you
need to restore your entire system.
2. Save all libraries, folders, and directories using Operational Assistant by doing
the following:
a. Initialize tapes, if necessary, using option 21 from the Backup Tasks menu.
To access this menu, type GO BACKUP.
b. Type GO SETUPBCKUP.
c. Select the option from the Set Up Backup menu to change the monthly
backup options.
d. Page down to the second page of the display and fill in the options as
shown in Figure 101 on page 888:

Appendix H. Using Operational Assistant to Save Information 887

 Change Monthly Backup Options

Type choices below, then press Enter.

What to back up:

User libraries . . . . . . . . . 2 1=Selected from list
2=All
3=None
Folders . . . . . . . . . . . . 2 1=Selected from list
2=All
3=None
User Directories . . . . . . . . 2 2=All
3=None
Security data . . . . . . . . . Y Y=Yes, N=No
Configuration . . . . . . . . . Y Y=Yes, N=No
OfficeVision/400 mail . . . . . Y Y=Yes, N=No
OfficeVision/400 calendars . . . Y Y=Yes, N=No

How to back up:

Save changed objects only . . . N Y=Yes, N=No
Submit backup as a batch job . . N Y=Yes, N=No
Print detailed report . . . . . N Y=Yes, N=No

Figure 101. Change Monthly Backup Options Display

e. Mount the tapes and press the Enter key.

Before you can use a strategy of saving changed objects with Operational
Assistant backup, you must save all the libraries by using Operational Assistant
backup. When you save changed objects by using Operational Assistant, the
system looks at information that is maintained by Operational Assistant. It does
not look at the Date last saved field in the object description.
3. After you run each type of backup (daily, weekly, and monthly) for the first
time using Operational Assistant, display the tapes to ensure that you are
saving what you think you are saving.

Defining What Should Be Saved

Operational Assistant maintains two lists for backup:
v All user libraries on the system
v All folders on the system

For each item in the list, you can specify when it is to be saved:
v Monthly only
v Weekly and monthly
v Daily, weekly, and monthly
v Not at all

For each of the three backup periods, you can specify whether all libraries and
folders on the list are to be saved, or only those libraries and folders that have
changed since the last backup. The topic “How to Define Backup Operations” on
page 890 describes how to specify this.

For example, you might specify that all user libraries on your system be saved
monthly. If your application programs do not change very often, you might specify
that only those libraries that contain database files be saved weekly. You can omit
the libraries that contain programs from your weekly backup. Your daily backup

888 OS/400 Backup and Recovery V5R1

 might include the same libraries, but it might specify that only changed objects are
saved. You could use a similar approach for saving folders.

To change the backup lists, do the following:

1. Type GO SETUPBCKUP.
2. From the Set Up Backup menu, select option 10 to change the library backup
list or option 11 to change the folder backup list. You are shown a list of all the
libraries or folders on your system:

Change Library Backup List

System: SYSTEM01
Find library . . . . . . __________ Starting characters

Type options below, then press Enter.

2=Change backup 5=Display library contents 8=Display details

-----------Backup----------- Last
Opt Library Daily Weekly Monthly Backup Changed
_ #LIBRARY Yes Yes Yes 01/01/9x No
_ AMES Yes Yes Yes 01/01/9x No
_ BALYLE Yes Yes Yes 01/01/9x No
_ BETH Yes Yes Yes 01/01/9x No
_ BURT Yes Yes Yes 01/01/9x No

The display shows when each library is scheduled to be saved, when it was
last saved, and whether it has changed since it was saved.
3. Type your changes on the display and press the Enter key.

How the System Saves Changed Objects Using Operational Assistant

Backup
You can use backup options displays or the CHGBCKUP command to specify that
one or more of the backups saves only changed objects. If you intend to use the
Operational Assistant support for saving changed objects, you must save all
libraries, folders, and directories regularly using Operational Assistant backup.

Following is what the system does when you use Operational Assistant to save
changed objects in libraries. The process is similar when you save changed objects
in folders and directories by using Operational Assistant backup.
1. The system retrieves the date and time when you last saved all libraries by
using Operational Assistant. This information is stored in an internal object that
is used by Operational Assistant. It is updated every time you run an
Operational Assistant backup and specify 2 (All libraries) for the User libraries
prompt on a backup options display. This becomes the reference date and time.
2. The system runs the Save Changed Object (SAVCHGOBJ) command for each
library in the backup list that you are saving. For the reference date and
reference time parameters, the system uses the date and time from step 1.

For example, assume that the monthly backup saves all libraries. The weekly
backup list saves entire libraries, not just changes. It includes the following
libraries:
v LIBA
v LIBB
v LIBC

Appendix H. Using Operational Assistant to Save Information 889

 v LIBD
The daily backup saves changes only. It includes these libraries:
v LIBA
v LIBC

The tape contents for this example look like this:

Date Procedure Contents

October 3 Monthly backup All user libraries (SAVLIB

LIB(*ALLUSR))
October 10 Weekly backup LIBA, LIBB, LIBC, LIBD
October 11 Daily backup Any objects in LIBA and LIBC
that have changed since
October 3.

How to Define Backup Operations

Each of the three schedules for Operational Assistant backup has a set of options
associated with it. These options define how that backup is to be run. For example,
you can specify whether the system saves only changed libraries or whether the
system prints a report of what was saved.

To change the backup options, do the following:

1. Type GO SETUPBCKUP. You see the Set Up Backup menu:

SETUPBCKUP Set Up Backup

To select one of the following, type its number below and press Enter:

1. Change daily backup options

2. Change weekly backup options
3. Change monthly backup options

10. Change library backup list

11. Change folder backup list

20. Change backup schedule

2. From the Set Up Backup menu, select option 1, 2, or 3 for the options that you
want to change. You see one of the change options displays, such as the
Change Daily Backup Options display:

890 OS/400 Backup and Recovery V5R1

 Change Daily Backup Options

Type choices below, then press Enter.

Where to back up:

Backup device . . . . . . . . . . . . . TAP01

Tape sets to rotate . . . . . . . . . . BBBB

Erase tape before backup . . . . . . . . Y

3. Type any changes that you want on this display and page forward to see
additional options:

Change Daily Backup Options

Type choices below, then press Enter.

What to back up:

User libraries . . . . . . . . . 1 1=Selected from list
2=All
3=None
Folders . . . . . . . . . . . . 1 1=Selected from list
2=All
3=None
User directories . . . . . . . . 2 2=All
3-None
Security data . . . . . . . . . N Y=Yes, N=No
Configuration . . . . . . . . . N Y=Yes, N=No
OfficeVision/400 mail . . . . . Y Y=Yes, N=No
OfficeVision/400 calendars . . . Y Y=Yes, N=No

4. Type any additional changes that you want on this display and page forward
again to see additional options:

Change Daily Backup Options

Type choices below, then press Enter.

How to back up:

Save changed objects only . . . Y Y=Yes, N=No
Submit backup as a batch job . . Y Y=Yes, N=No
Print detailed report . . . . . Y Y=Yes, N=No

5. Type any additional changes and press the Enter key.

Note: You can also use the Change Backup Options (CHGBCKUP) command to
change the backup options.

How to Define When Backup Operations Are Run

You can run a backup with Operational Assistant by using the Run Backup
(RUNBCKUP) command or by selecting an option from the Run Backup menu.
You can also set up a schedule for when daily, weekly, and monthly backups are to
be run. Jobs to run the backups are placed on the job scheduler to be started at

Appendix H. Using Operational Assistant to Save Information 891

 certain times and days. You can also request that the system send a message to
mount the backup tapes a specified number of hours before the job is scheduled to
begin.

To change the backup schedule, do the following:

1. Type GO SETUPBCKUP.
2. From the Set Up Backup menu, select option 20 to change the backup schedule.
You are shown the Change Backup Schedule display:

Change Backup Schedule

Type choices below, then press Enter. Press F4 for list of backups.

Run backup using this schedule . . . . . . Y Y=Yes, N=No

Backup Backup Time

Sunday . . . . . . . .
Monday . . . . . . . . *DAILY 22:00:00
Tuesday . . . . . . . . *DAILY 22:00:00
Wednesday . . . . . . . *DAILY 22:00:00
Thursday . . . . . . . *DAILY 22:00:00
Friday . . . . . . . . *DAILY 22:00:00
Saturday . . . . . . . *WEEKMONTH 22:00:00

Occurrence of day in month to run monthly

backup . . . . . . . . . . . . . . . . . 1

Hours before backup to send load tape

message . . . . . . . . . . . . . . . . 5

3. Type your choices and press the Enter key.

In the example that is shown, the system runs the daily backup on week days
(Monday through Friday). It uses the daily backup list and the daily backup
options. The backup job is on the job scheduler and is scheduled to begin at 10
p.m. A message is sent to the system operator at 5 p.m.
In the example, monthly backup is run on the first Saturday of the month.
Weekly backup is run every Saturday except the first Saturday of the month.

Note: When you use the backup schedule, the system places jobs whose names
begin with the following characters on the job scheduler:
v QEZBKTM
v QEZBKMG
Do not change these entries by using the job scheduling commands.

892 OS/400 Backup and Recovery V5R1

Appendix I. Notices
This information was developed for products and services offered in the U.S.A.
IBM may not offer the products, services, or features discussed in this document in
other countries. Consult your local IBM representative for information on the
products and services currently available in your area. Any reference to an IBM
product, program, or service is not intended to state or imply that only that IBM
product, program, or service may be used. Any functionally equivalent product,
program, or service that does not infringe any IBM intellectual property right may
be used instead. However, it is the user’s responsibility to evaluate and verify the
operation of any non-IBM product, program, or service.

IBM may have patents or pending patent applications covering subject matter
described in this document. The furnishing of this document does not give you
any license to these patents. You can send license inquiries, in writing, to:
IBM Director of Licensing
IBM Corporation
North Castle Drive
Armonk, NY 10504-1785
U.S.A.

For license inquiries regarding double-byte (DBCS) information, contact the IBM
Intellectual Property Department in your country or send inquiries, in writing, to:
IBM World Trade Asia Corporation
Licensing
2-31 Roppongi 3-chome, Minato-ku
Tokyo 106, Japan

The following paragraph does not apply to the United Kingdom or any other
country where such provisions are inconsistent with local law:
INTERNATIONAL BUSINESS MACHINES CORPORATION PROVIDES THIS
PUBLICATION “AS IS” WITHOUT WARRANTY OF ANY KIND, EITHER
EXPRESS OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
WARRANTIES OF NON-INFRINGEMENT, MERCHANTABILITY OR FITNESS
FOR A PARTICULAR PURPOSE. Some states do not allow disclaimer of express or
implied warranties in certain transactions, therefore, this statement may not apply
to you.

This information could include technical inaccuracies or typographical errors.

Changes are periodically made to the information herein; these changes will be
incorporated in new editions of the publication. IBM may make improvements
and/or changes in the product(s) and/or the program(s) described in this
publication at any time without notice.

Licensees of this program who wish to have information about it for the purpose
of enabling: (i) the exchange of information between independently created
programs and other programs (including this one) and (ii) the mutual use of the
information which has been exchanged, should contact:
IBM Corporation
Software Interoperability Coordinator
3605 Highway 52 N

© Copyright IBM Corp. 1997, 2000, 2001 893

 Rochester, MN 55901-7829
U.S.A.

Such information may be available, subject to appropriate terms and conditions,

including in some cases, payment of a fee.

The licensed program described in this information and all licensed material
available for it are provided by IBM under terms of the IBM Customer Agreement
or any equivalent agreement between us.

Any performance data contained herein was determined in a controlled

environment. Therefore, the results obtained in other operating environments may
vary significantly. Some measurements may have been made on development-level
systems and there is no guarantee that these measurements will be the same on
generally available systems. Furthermore, some measurement may have been
estimated through extrapolation. Actual results may vary. Users of this document
should verify the applicable data for their specific environment.

Information concerning non-IBM products was obtained from the suppliers of

those products, their published announcements or other publicly available sources.
IBM has not tested those products and cannot confirm the accuracy of
performance, compatibility or any other claims related to non-IBM products.
Questions on the capabilities of non-IBM products should be addressed to the
suppliers of those products.

All statements regarding IBM’s future direction or intent are subject to change or
withdrawal without notice, and represent goals and objectives only.

COPYRIGHT LICENSE:

This information contains sample application programs in source language, which

illustrates programming techniques on various operating platforms. You may copy,
modify, and distribute these sample programs in any form without payment to
IBM, for the purposes of developing, using, marketing or distributing application
programs conforming to the application programming interface for the operating
platform for which the sample programs are written. These examples have not
been thoroughly tested under all conditions. IBM, therefore, cannot guarantee or
imply reliability, serviceability, or function of these programs. You may copy,
modify, and distribute these sample programs in any form without payment to
IBM for the purposes of developing, using, marketing, or distributing application
programs conforming to IBM’s application programming interfaces.

If you are viewing this information softcopy, the photographs and color
illustrations may not appear.

Trademarks
The following terms are trademarks of International Business Machines
Corporation in the United States, or other countries, or both:

400
Application System/400
APPN
AS/400
AS/400e
C/400

894 OS/400 Backup and Recovery V5R1

CICS
COBOL
COBOL/400
DB2
DisplayWrite
Distributed Relational Database Architecture
DRDA
ECS
FORTRAN/400
IBM
IBM Payment Server
ILE
Integrated Language Environment
MQSeries
Netfinity
OfficeVision/400
Operating System/400
OS/2
OS/400
PLI
PowerPC
PS/2
RM/COBOL-85
RPG/400
RT
SAA
SQL/400
System/36
System/370
System/38
Ultimedia
VisualAge
Visual Warehouse
WebSphere

Java is a registered trademark of Sun Microsystems, Inc.

Lotus and Domino are trademarks of Lotus Development Corporation in the

United States, or other countries, or both.

Microsoft, Windows, Windows NT, and the Windows 95 logo are registered
trademarks of Microsoft Corporation.

Tivoli is a registered trademark of Tivoli Systems Inc.

UNIX is a registered trademark in the United States and other countries licensed
exclusively through X/Open Company Limited.

Other company, product, and service names may be trademarks or service marks
of others.

Appendix I. Notices 895

896 OS/400 Backup and Recovery V5R1
Bibliography
This topic lists publications that provide controller, device, mode, NetBIOS, and
additional information about topics described or class-of-service descriptions, configuration lists,
referred to in this book. and connection lists.
v ADTS/400: Application Development Manager v Hierarchical Storage Management, SC41-5351-01
User’s Guide, SC09-2133-02. This book provides an overview of hierarchical
This book provides information about how to storage management principles. It also
set up and use ADTS. describes considerations for planning for the
v AnyMail/400 Mail Server Framework Support, use of dynamic retrieval that is supplied with
SC41-5411-00. Backup Recovery and Media Services for
AS/400. The book also provides information on
This book contains information on supporting
the implementation of hierarchical storage
the mail server framework MSF, including MSF
management.
errors and MSF function. It describes how to
use the MSF commands, the MSF exit points v ILE Concepts, SC41-5606-05
and exit programs, and the QZMF journal This book describes concepts and terminology
support. for the Integrated Language Environment (ILE)
v Are You Saving the Right Stuff? , G325-6153. architecture of the OS/400 operating system.
Topics covered include module creation,
This poster is a visual reminder of the methods
binding, how to run and debug programs, and
available for saving and restoring information
exception handling.
on your system. It also highlights the software,
hardware, people, and service offerings that are v An Implementation Guide to AS/400 Security and
available to help you with backup, recovery, Auditing GG24-4200.
and availability. This Redbook describes how to implement
v Automated Tape Library Planning and security on the iSeries server. It includes
Management, SC41-5309-02. discussions of security as it relates to C2,
cryptography, communications, and PC
This book provides information about tasks
connectivity.
that can be performed with an automated tape
library (ATL). It describes recommended v iSeries Licensed Internal Code Diagnostic Aids -
methods for designing and using ATLs. It Volume 1, LY44-5900-03, and iSeries Licensed
compares ATL devices that are currently Internal Code Diagnostic Aids - Volume 2,
available. LY44-5901-03 .
v Backup Recovery and Media Services for iSeries, These books provide information about error
SC41-5345-02. logs, dumps, traces, and other service tools to
be used to determine, isolate, and solve
This book provides information about
programming problems occurring with the
developing and implementing a backup and
OS/400 and vertical Licensed Internal Code
recovery strategy using the Backup Recovery
(VLIC), and other hardware and software
and Media Services/400 licensed program. It
failures. They contain trace points and VLIC
describes how to create and maintain the
log (VLOG) information for the iSeries VLIC
policies that govern your backup strategy.
code and are to be used to help solve and
v CL Programming, SC41-5721-04. diagnose problems and submit authorized
This book provides the application programmer program analysis reports (APARs) or Licensed
or programmer with a wide range discussion of Internal Code trouble reports (LICTRs).
the iSeries server programming topics. v Local Device Configuration, SC41-5121-00.
v Communications Configuration, SC41-5401-00. This book provides the system operator or
This book contains general configuration system administrator with information on how
information, including detailed descriptions of to do an initial configuration and how to
network interfaces, network servers, line, change the configuration. This book also
contains conceptual information about device
configuration.

© Copyright IBM Corp. 1997, 2000, 2001 897

v Office Services Concepts and Programmer’s Guide, correct inefficiencies that might exist.
SC41–5805. Information about the manager and agent
This book provides information about writing feature is included.
applications that use OfficeVision functions. v Printer Device Programming, SC41-5713-04.
This book also includes an overview of This book provides information to help you
directory services, document distribution understand and control printing. It provides
services, document library services, document specific information on printing elements and
and folder save and restore and storage concepts of the iSeries server, printer file and
management, security services, word processing print spooling support for printing operations,
services, and information on finding new ways and printer connectivity.
to integrate your applications with OfficeVision.
It includes considerations for using personal
v Open Group Technical Standard, DRDA, Version2, computers, other printing functions such as
Volume 1: Distributed Relational Database Business Graphics Utility (BGU), advanced
Architecture (DRDA) function printing* (AFP*), and examples of
This document is the first of three volumes that working with the iSeries server printing
documents the Distributed Relational Database elements such as how to move spooled output
Architecture Specification, Version 2. This files from one output queue to a different
document is available from the Open Group output queue. It also includes an appendix of
web site at http://www.opengroup.org. control language (CL) commands used to
v Open Group Technical Standard, DRDA, Version2, manage printing workload.
Volume 3: Distributed Data Management v AS/400 Road Map for Changing to PowerPC
(DDM) Architecture Technology, SA41-5150-05.
This document is the third of three volumes This book describes how to move from an
that documents the Distributed Relational AS/400 server that is running Version 3 Release
Database Architecture Specification, Version 2. 1 or an earlier version of the OS/400 licensed
This document is available from the Open program to an iSeries server with a PowerPC
Group web site at http://www.opengroup.org. AS processor. It includes planning information
v OptiConnect for OS/400, SC41-5414-02 and step-by-step instructions.
This book provides information about v iSeries Security Reference, SC41-5302-04.
OptiConnect which is a combination of This book provides the programmer (or
hardware and software that allows you to someone who is assigned the responsibilities of
connect multiple high-end servers using a a security officer) with information about
high-speed, fiber-optic bus. OptiConnect allows system security concepts, planning for security,
applications to perform intersystem database and setting up security on the system. This
accesses across a high-performance interface. guide does not describe security for specific
v OptiConnect/400 Guide, SC24-5715 licensed programs, languages, and utilities.
This book provides information on v Simple Network Management Protocol (SNMP)
OptiConnect/400 and how to use this support, Support, SC41-5412-00.
including installation and problem analysis This book provides the system operator,
information. programmer, or system administrator with
v Optical Support, SC41-5310-02. information for configuring the iSeries server to
use the simple network management protocol
This book provides information on how to
(SNMP) support.
attach optical devices to your system. It
describes the commands that can be used to v SNA Distribution Services, SC41-5410-01
work with optical devices. This book provides information about the
v Performance Tools for iSeries, SC41-5340-01 function and administration of Systems
Network Architecture distribution services.
This book provides the programmer with the
information needed to collect data about the v Software Installation, SC41-5120-05.
system, job, or program performance. Other This book provides the system operator or
performance data: tips for printing and system administrator with step-by-step
analyzing performance data to identify and procedures for initially installing, installing
licensed programs, program temporary fixes
(PTFs), and secondary languages from IBM.

898 OS/400 Backup and Recovery V5R1

 This guide is also for users who already have
an iSeries server with an installed release and
want to upgrade to a new release.
v AS/400 System Availability and Recovery,
GG24-3912.
This Redbook gives examples for using the
save-while-active function, auxiliary storage
pools, and disk protection tools (device parity
protection, mirrored protection, and checksum
protection).
v System Operation, SC41-4203-00.
This book provides information about handling
messages, working with jobs and printer
output, devices communications, working with
support functions, cleaning up your system,
and so on.
v TCP/IP Configuration and Reference,
SC41-5420-03.
This book provides information for configuring
TCP/IP support and applications. The
applications included are TELNET, File Transfer
Protocol (FTP), Simple Mail Transfer Protocol
(SMTP), line printer requester (LPR), and line
printer daemon (LPD).
v Tips and Tools for Securing Your iSeries,
SC41-5300-07
This book provides a set of practical
suggestions for using the security features of
the iSeries server and for establishing operating
procedures that are security-conscious. The
recommendations in this book apply to an
installation with average security requirements
and exposures.
v Tivoli Storage Manager for AS/400 GH35-0114.
This book provides information about how to
set up and manage the Tivoli Storage Manager
for AS/400 licensed program.
v Work Management, SC41-5306-03.
This book provides the programmer with
information about how to create a work
management environment and how to change
it.
v AS/400e server 170, 250, 6xx, 7xx, and Sxx System
Installation and Upgrade, SY44-5950-05.
This book provides the service representative
with information about upgrading equipment
on the 9406 System Unit. It provides
information on an entire range of upgrades
such as simple memory card additions, device
and rack additions, and model upgrades. It is
used with the instruction packets that are
shipped with the upgrade equipment.

Bibliography 899
900 OS/400 Backup and Recovery V5R1
Index
Special Characters abnormal system end (continued)
recovery 462
all object (*ALLOBJ) special authority
correcting after restoring 355, 359,
*ALLOBJ (all object) special authority access intent 544 361, 363
correcting after restoring 355, 359, access path all-object (*ALLOBJ) special authority
361, 363 definition 5, 375 restoring 217
*ALLOBJ (all-object) special authority editing rebuild during IPL 171 allow object differences (ALWOBJDIF)
restoring 217 ending journaling 423 parameter
(CHGASPA) Change ASP Attribute explicit journaling authorization lists 218
command 728 overview 6 effect 43
*JRN (journal) exposed 377 purpose 43
definition 383 journaling allow object restore operation
*JRN (journal) object ending 423 (QALWOBJRST) system value 50
definition 5 starting 412 allow user domain objects
*JRNRCV (journal receiver) object why used 391 (QALWUSRDMN) system value 48
definition 5, 383 with SMAPP 376 allowing
protecting 375 restore
protection adopted authority objects 50
Numerics overview 5 sensitive objects 50
2440 Tape Unit recovery time 375 system-state programs 50
disabling high-speed feature 145 working with 378 alternate installation device 635
enabling high-speed feature 145 recovery times ALWOBJDIF (allow object difference)
2741 Input/Output Processor restoring 156 parameter
adding 650 restoring 242 database file 238
34xx tape units system-managed protection (SMAPP) member 238
creating tape configuration 165, 342, overview 5 ALWOBJDIF (allow object differences)
879 access path journaling parameter
6502 Input/Output Processor explicit authorization lists 218
adding 650 definition 391 effect 43
6502 IOP access path operation entry 806 purpose 43
adding disk 649 access path recovery time API (application programming interface)
6512 Input/Output Processor recovering 213 resource 541
adding 650 accessing APIs
6512 IOP dedicated service tools (DST) 660 Add Commitment Resource
adding disk 649 system service tools 662 (QTNADDCR) 535
6532 Input/Output Processor accounting (QACGJRN) journal 424 Add Remote Journal 474
adding 650 action Change Commitment Options
6533 Input/Output Processor APYJRNCHG or RMVJRNCHG (QTNCHGCO) 535
adding 650 command Change Journal State
6751 Input/Output Processor journal code 452 (QjoChangeJournalState) 478, 487
adding 650 mirrored protection recovery 289, Delete Pointer Handle
6754 Input/Output Processor 296 (QjoDeletePointerHandle) 386
adding 650 activation group end List Objects (QUSLOBJ) 387
9337 disk commitment control during 580 List Save File (QSRLSAVF) 761
adding 648 activation-group-level scope 534 Open List of Objects (QGYOLOBJ
existing array 646, 647 active disk unit status 668 API) 387
9337 Disk Array Subsystem Add All Disk Units to the System QsrRestore 261
starting device parity protection 695 display 152 Remove Commitment Control
Add Commitment Resource Resource (QTNRMVCR) 541
(QTNADDCR) API 535 Remove Remote Journal
A Add Remote Journal
(QjoAddRemoteJournal) API 474, 475
(QjoRemoveRemoteJournal) 479,
490
A900 2000 SRC (system reference code) Add Remote Journal API 474 Retrieve Commitment Information
recovery 164 addressability (QTNRCMTI) 564
abbreviated install recovering 182 Retrieve Journal Entries
definition 148 adopted authority object (QjoRetrieveJournalEntries) 386
abend 61 allowing restore operation 50 Retrieve Journal Identifier Information
abnormal end agent (QJORJIDI) API 386
definition 61, 167 description 526 Retrieve Object Description
restarting system 167 agent role 548 (QUSROBJD API) 387
abnormal IPL (initial program load) 167 Rollback Required (QTNRBRQD) 535
abnormal system end Send Data Queue (QSNDDTAQ) 386
commitment control 581

© Copyright IBM Corp. 1997, 2000, 2001 901

APIs (continued) ASP (auxiliary storage pool) (continued) auxiliary storage configuration
Send Journal Entry (QJOSJRNE) 386, types 393 (continued)
429, 535 user checklist (continued)
Application Development Manager adding disk units 669 adding disk units without device
project log (QLYPRJLOG) journal 424 calculating space parity protection 645
transaction log (QLYJRN) journal 424 requirements 681 deleting auxiliary storage pool
application program changing threshold 672, 673 (ASP) 654
example 615 creating 669 moving disk units 652, 653
Apply Journaled Changes (APYJRNCHG) creating document library objects new system 644
command (DLOs) 689 removing disk units 655, 656, 657,
broken receiver chain 285 creating objects 688, 692 658, 659
data recovery 446 deleting 654, 680 interpreting 667
database file member 388 displaying objects 684 auxiliary storage pool (ASP)
example 448 journal receivers 689 assigning files and journals 392
journal code actions 452 moving disk unit 676 benefits 7
overview 428 removing disk unit 678 calculating space requirements 681
referential constraints 445 transferring objects 686 changing threshold 672, 673
trigger programs 446 associate receivers creating objects 688
unbroken receiver chain 284 journal 462 definition 7
applying attribute deleting 654, 680
journaled changes save file journal 403
broken receiver chain 285 file-dependent 764 journal receiver 393, 401
description 444 audit (QAUDJRN) journal 424 moving disk unit 676
determining whether to 282 creating during restore 55 moving folder 687
overview 428 audit trail 807 moving library 687
QAOSDIAJRN journal 286 using journaling 439 overflowed
referential constraints 445 authority moving journal receivers 690
reorganize physical file journal receiver 402 resetting journal 691
member 455 private 219 overview 6
trigger programs 446 restoring 219 removing disk unit 678
unbroken receiver chain 284 authority holder status 667
APYJRNCHG (Apply Journaled Changes) restoring 217 system
command authorization list removing failed unit 92
broken receiver chain 285 restoring 217 transferring journals 687
data recovery 446 restoring link 218 transferring objects 686
database file member 388 automatic configuration types 7, 393
example 448 enabling user
journal code actions 452 during recovery 160 adding disk units 669
overview 428 automatic IPL after power restored calculating space
referential constraints 445 (QPWRRSTIPL) system value 167 requirements 681
trigger programs 446 auxiliary storage changing threshold 672, 673
unbroken receiver chain 284 journaling 395 creating 669
ASP journaling impact 397 creating document library objects
independent 138 auxiliary storage (ASP) (DLOs) 689
ASP (auxiliary storage pool) high percentage used 176 creating objects 688, 692
assigning files and journals 392 auxiliary storage configuration deleting 654, 680
benefits 7 checklist displaying objects 684
calculating space requirements 681 adding 2741 Input/Ouput journal receivers 689
changing threshold 672, 673 Processor 650 moving disk unit 676
creating objects 688 adding 6502 Input/Output removing disk unit 678
definition 7 Processor 650 transferring objects 686
deleting 654, 680 adding 6512 Input/Output availability
journal 403 Processor 650 commitment control
journal receiver 393, 401 adding 6532 Input/Output considerations and
moving disk unit 676 Processor 650 restrictions 591
moving folder 687 adding 6533 Input/Ouput introduction 3
moving library 687 Processor 650 using dual systems 9
overflowed adding 6751 Input/Ouput
moving journal receivers 690 Processor 650
resetting journal 691
overview 6
adding 6754 Input/Ouput
Processor 650
B
backup
removing disk unit 678 adding 9337 disk 646
introduction 3
status 667 adding 9337 Disk Array
options with Operational
system Subsystem 647, 648
Assistant 890
removing failed unit 92 adding disk to 6502 IOP 649
schedule with Operational
transferring journals 687 adding disk to 6512 IOP 649
Assistant 891
transferring objects 686
using Operational Assistant 885

902 OS/400 Backup and Recovery V5R1

backup lists 888 changing (continued) CMPJRNIMG (Compare Journal Images)
batch job recovery storage threshold (continued) command
strategy 760 system auxiliary storage pool overview 386, 428
before-image (ASP) 673 using 440
selecting 392 system auxiliary storage pool (ASP) CMT (committed) state 565
BLOB (binary large object) 437 storage threshold 673 CMTSCOPE (commit scope)
BRMS synchronization 364 changing system parameter 533
broken receiver chain at IPL code
applying journaled changes 285 during recovery 160 journal 807
bus failure 298 checklist command, AS/400
busy disk unit status 667 disk configuration CRTNWSSTG 271, 272
adding 2741 Input/Output RST 271, 272
Processor 650 RSTCFG 273
C adding 6502 Input/Output
Processor 650
RSTOBJ 271
command, CL 762
calculating
adding 6512 Input/Output Apply Journaled Changes
disk space 681
Processor 650 (APYJRNCHG)
calendar server (QCALSRV) subsystem
adding 6532 Input/Output actions 452
ending 45
Processor 650 broken receiver chain 285
capacity
adding 6533 Input/Output data recovery 446
save file storage 761
Processor 650 database file member 388
CD-ROM
adding 6751 Input/Output overview 428
restoring OS/400 licensed
Processor 650 referential constraints 445
program 148
adding 6754 Input/Output trigger programs 446
chain break 441
Processor 650 unbroken receiver chain 284
Change ASP Attribute (CHGASPA)
adding 9337 disk 646 APYJRNCHG (Apply Journaled
command 728
adding 9337 Disk Array Changes)
Change Commitment Options
Subsystem 647, 648 actions 452
(QTNCHGCO) API 535
adding disk to 6502 IOP 649 broken receiver chain 285
Change Journal (CHGJRN)
adding disk to 6512 IOP 649 data recovery 446
command 386, 416, 485, 487, 490, 510
adding disk units without device database file member 388
MINENTDTA (minimized
parity protection 645 overview 428
entry-specific data) parameter 409
deleting auxiliary storage pool referential constraints 445
Change Journal State
(ASP) 654 trigger programs 446
(QjoChangeJournalState) API 478, 480,
moving disk units 652, 653 unbroken receiver chain 284
485, 487
new system 644 Change Journal (CHGJRN) 386, 416,
Change Physical File (CHGPF) command
removing disk units 655, 656, 657, 485, 487, 490, 510
SHARE(*YES)
658, 659 Change Physical File (CHGPF)
impact on journal entries 411
CHGJRN (Change Journal) SHARE(*YES) impact on journal
Change Recovery for Access Paths
command 386, 416 entries 411
(CHGRCYAP) command 380
MINENTDTA (minimized Change Recovery for Access Paths
Change Remote Journal (CHGRMTJRN)
entry-specific data) parameter 409 (CHGRCYAP) 380
command 480, 485
CHGPF (Change Physical File) command Change Remote Journal
Change Save File (CHGSAVF)
SHARE(*YES) (CHGRMTJRN) 480
command 762
impact on journal entries 411 Change Remote Journal
Change Service Tools User Password
CHGRCYAP (Change Recovery for (CHGRMTJRN) command 485
display 139
Access Paths) command 380 Change Save File (CHGSAVF) 762
changed object
CHGRMTJRN (Change Remote Journal) CHGASPA (Change ASP
duplicating on another system 32
command 480, 485 Attribute) 728
restoring 32
CHGSAVF (Change Save File) CHGJRN (Change Journal) 386, 416
by library 280
command 762 MINENTDTA (minimized
by object 280
CIP (commit in progress) state 565 entry-specific data)
cumulative 280
CL command 762 parameter 409
directories 281
cleaning up CHGPF (Change Physical File)
not cumulative 280
hardware configuration 228 SHARE(*YES) impact on journal
saving 32
Clear Save File (CLRSAVF) entries 411
Operational Assistant backup 889
command 761, 762 CHGRCYAP (Change Recovery for
changing
clearing Access Paths) 380
auxiliary storage pool (ASP)
job queue CHGRMTJRN (Change Remote
storage threshold 672
during recovery 155 Journal) 480
journal 386, 416
output queue CHGRMTJRN (Change Remote
journal receivers 405
during recovery 155 Journal) command 485
physical file
save file 762, 764 CHGSAVF (Change Save File) 762
SHARE(*YES) impact on journal
CLOB (character large object) 437 Clear Save File (CLRSAVF) 762
entries 411
CLRSAVF (Clear Save File) CLRSAVF (Clear Save File) 762
storage threshold
command 762
auxiliary storage pool (ASP) 672

Index 903
command, CL 762 (continued) command, CL 762 (continued) command, CL 762 (continued)
CMPJRNIMG (Compare Journal CRTJRNRCV (Create Journal ENDJRN (End Journal) 423
Images) Receiver) (continued) ENDJRNAP (End Journal Access
overview 386, 428 THRESHOLD (threshold) Path) 387, 423
using 440 parameter 401 ENDJRNOBJ (End Journal
Compare Journal Images CRTPF (Create Physical File) Object) 423
(CMPJRNIMG) SHARE(*YES) impact on journal ENDJRNPF (End Journal Physical File
overview 386, 428 entries 411 Changes) 423
using 440 CRTSAVF (Create Save File) 761 ENDSBS (End Subsystem)
Create Journal (CRTJRN) Delete Journal (DLTJRN) 386 QCALSRV (calendar server)
ASP (auxiliary storage pool) Delete Journal Receiver subsystem 45
parameter 403 (DLTJRNRCV) 386 QSYSWRK (subsystem monitor)
DLTRCV (delete receiver) description 412 subsystem 45
parameter 406 Display Database Relations restricted state 45
JRN (journal) parameter 403 (DSPDBR) 248 using 45
JRNRCV (journal receiver) Display File Description example
parameter 404 (DSPFD) 762 Apply Journaled Changes
MNGRCV (manage receiver) Display Journal (DSPJRN) 386 (APYJRNCHG) 448
parameter 405 overview 428 APYJRNCHG (Apply Journaled
MSGQ (message queue) Display Journal Receiver Attributes Changes) 448
parameter 404 (DSPJRNRCVA) 386 CRTJRNRCV (Create Journal
overview 386 Display Object Links (DSPLNK) 387 Receiver) 409
RCVSIZOPT (receiver size option) Display Recovery for Access Paths Remove Journaled Changes
parameter 407 (DSPRCYAP) 380 (RMVJRNCHG) 451
Create Journal (CRTJRN) command Display Save File (DSPSAVF) 761, RMVJRNCHG (Remove Journaled
JRN (journal) parameter 403 762 Changes) 451
Create Journal Receiver (CRTJRNRCV) DLTJRN (Delete Journal) 386 Override with Save File
ASP (auxiliary storage pool) DLTJRNRCV (Delete Journal (OVRSAVF) 762
parameter 401 Receiver) 386 OVRSAVF (Override with Save
AUT (authority) parameter 402 DSPDBR (Display Database File) 762
JRNRCVR (journal receiver) Relations) 248 QRYDOCLIB (Query Document
parameter 399, 401 DSPFD (Display File Library) 200
overview 386 Description) 762 Query Document Library
THRESHOLD (threshold) DSPJRN (Display Journal) 386 (QRYDOCLIB) 200
parameter 401 overview 428 RCLDLO (Reclaim Document Library
Create Physical File (CRTPF) 411 DSPJRNRCVA (Display Journal Object) 257
Create Save File (CRTSAVF) 761 Receiver Attributes) 386 RCLSTG (Reclaim Storage)
Create Save File (CRTSAVF) DSPRCYAP (Display Recovery for duplicate names in QRCL 47
command 762 Access Paths) 380 object ownership 48
CRTJRN (Create Journal) DSPSAVF (Display Save File) 761, procedure 46, 183
ASP (auxiliary storage pool) 762 QALWUSRDMN (allow user
parameter 403 Edit Check Pending Constraint domain objects) system
DLTRCV (delete receiver) (EDTCPCST) 174 value 48
parameter 406 Edit Recovery for Access Paths recovering user ASP 183
JRN (journal) parameter 403 (EDTRCYAP) 378 user domain object 48
JRNRCV (journal receiver) EDTCPCST (Edit Check Pending what system does 47
parameter 404 Constraint) 174 why to run 176
MINENTDTA (minimized EDTRCYAP (Edit Recovery for Access RCVJRNE (Receive Journal Entry)
entry-specific data) Paths) 378 DEPENT (dependent entry)
parameter 409 End Commitment Control parameter 432
MNGRCV (manage receiver) (ENDCMTCTL) 579 exit program 433
parameter 405 End Journal (ENDJRN) 387 overview 386, 428
MSGQ (message queue) End Journal Access Path using 432
parameter 404 (ENDJRNAP) 387, 423 writing output to save media 769
overview 386 End Journal Object RCVNETF (Receive Network
RCVSIZOPT (receiver size option) (ENDJRNOBJ) 387 File) 765
parameter 407 End Journal Physical File Changes Receive Journal Entry (RCVJRNE)
CRTJRN (Create Journal) command (ENDJRNPF) 423 DEPENT (dependent entry)
JRN (journal) parameter 403 End Subsystem (ENDSBS) parameter 432
CRTJRNRCV (Create Journal Receiver) QCALSRV (calendar server) exit program 433
ASP (auxiliary storage pool) subsystem 45 overview 386, 428
parameter 401 QSYSWRK (subsystem monitor) using 432
AUT (authority) parameter 402 subsystem 45 writing output to save media 769
JRNRCVR (journal receiver) restricted state 45 Receive Network File
parameter 399, 401 using 45 (RCVNETF) 765
overview 386 ENDCMTCTL (End Commitment Reclaim Document Library Object
Control) 579 (RCLDLO) 257

904 OS/400 Backup and Recovery V5R1

command, CL 762 (continued) command, CL 762 (continued) command, CL 762 (continued)
Reclaim Storage (RCLSTG) Restore Library (RSTLIB) (continued) RSTDLO (Restore Document Library
duplicate names in QRCL 47 FRCOBJCVN (force object Object) (continued)
object ownership 48 conversion) parameter 254 overview 255
procedure 46, 183 media error 56 renaming document 258
QALWUSRDMN (allow user multiple concurrent 233 restoring authority 259
domain objects) system OPTION parameter 232 restoring descriptive
value 48 overview 231 information 259
recovering user ASP 183 user ASP 198 restoring ownership 259
user domain object 48 Restore Licensed Program user ASP 200
what system does 47 (RSTLICPGM) 255 RSTLIB (Restore Library)
why to run 176 Restore Object (RSTOBJ) 234 *ALLUSR libraries 232
Remove Journaled Changes FRCOBJCVN (force object *IBM libraries 232
(RMVJRNCHG) conversion) parameter 254 *NONSYS libraries 232
actions 452 multiple concurrent 234 FRCOBJCVN (force object
example 451 Restore User Profiles conversion) parameter 254
journal receivers 388 (RSTUSRPRF) 214 media error 56
overview 428 Retrieve Journal Entry (RTVJRNE) multiple concurrent 233
referential constraints 445 database recovery 386 OPTION parameter 232
trigger programs 446 description 386 overview 231
using 450 overview 428 user ASP 198
Remove Remote Journal using 436 RSTLICPGM (Restore Licensed
(RMVRMTJRN) command 490 using in program 766 Program) 255
Rename Directory Entry (RNMDIRE) RMVJRNCHG (Remove Journaled RSTOBJ (Restore Object) 234
restoring mail 260 Changes) FRCOBJCVN (force object
Rename Document Library Object actions 452 conversion) parameter 254
(RNMDLO) example 451 multiple concurrent 234
restoring documents 260 journal receivers 388 RSTUSRPRF (Restore User
Restore (RST) overview 428 Profiles) 214
changed objects 281 referential constraints 445 RTVJRNE (Retrieve Journal Entry)
command information output trigger programs 446 database recovery 386
format 858 using 450 description 386
directory information output RMVRMTJRN (Remove Remote overview 428
format 859 Journal) command 490 using 436
FRCOBJCVN (force object RNMDIRE (Rename Directory Entry) using in program 766
conversion) parameter 254 restoring mail 260 SAV (Save)
how to use 261 RNMDLO (Rename Document Library command information output
object link information output Object) format 858
format 860 restoring documents 260 directory information output
output header format 858 RST (Restore) format 859
output information 857 changed objects 281 object link information output
output sequence 857 command information output format 860
trailer information output format 858 output header format 858
format 861 directory information output output information 857
Restore (RST) command format 859 output sequence 857
restrictions 275 FRCOBJCVN (force object trailer information output
restrictions when restoring conversion) parameter 254 format 861
documents 277 how to use 261 Save (SAV)
Restore Authority (RSTAUT) 219 object link information output command information output
non-restricted state system 220 format 860 format 858
Restore Configuration (RSTCFG) 227 output header format 858 directory information output
Restore Document Library Object output information 857 format 859
(RSTDLO) output sequence 857 object link information output
maximum number of DLOs 257 trailer information output format 860
media error 57 format 861 output header format 858
output 256 RST (Restore) command output information 857
overview 255 restrictions 275 output sequence 857
renaming document 258 restrictions when restoring trailer information output
restoring authority 259 documents 277 format 861
restoring descriptive RSTAUT (Restore Authority) 219 save file 762
information 259 non-restricted state system 220 CHGSAVF (Change Save
restoring ownership 259 RSTCFG (Restore Configuration) 227 File) 762
user ASP 200 RSTDLO (Restore Document Library considerations for using 761
Restore Library (RSTLIB) Object) CRTSAVF (Create Save File) 761,
*ALLUSR libraries 232 maximum number of DLOs 257 762
*IBM libraries 232 media error 57 DSPSAVF (Display Save File) 762
*NONSYS libraries 232 output 256

Index 905
command, CL 762 (continued) command, CL 762 (continued) commit or rollback processing
OVRSAVF (Override with Save Start Journal Access Path during IPL 599
File) 762 (STRJRNAP) 387, 412 during job end 599
Save Save File Data Start Journal Object commit processing
(SAVSAVFDTA) 761 (STRJRNOBJ) 387 flow 551
Save Library (SAVLIB) using 414 commit protocol 544
determining what command was Start Journal Physical File (STRJRNPF) commit scope (CMTSCOPE)
used 307 description 387 parameter 533
Save Object (SAVOBJ) using 411 commit text (TEXT) parameter 539
media considerations 762 STRCMTCTL (Start Commitment commitment boundary
Save/Restore (SAVRST) 32 Control) description 526
Save/Restore Changed Objects commit scope (CMTSCOPE) commitment control
(SAVRSTCHG) 32 parameter 533 access intent 544
Save/Restore Configuration commit text (TEXT) agent
(SAVRSTCFG) 32 parameter 539 description 526
Save/Restore Document Library default journal (DFTJRN) agent role 548
Object (SAVRSTDLO) 32 parameter 539 allowing vote read-only 560
Save/Restore Library description 527 commit cycle 547
(SAVRSTLIB) 32 LCKLVL (lock level) commit cycle identifier 547
Save/Restore Object parameter 527 commit operation 547, 548, 549
(SAVRSTOBJ) 32 notify object (NFYOBJ) commit protocol 544
Save Save File Data parameter 530 commitment boundary
(SAVSAVFDTA) 762 omit journal entries (OMTJRNE) definition 526
save file considerations 761 parameter 540 committable change
SAVLIB (Save Library) STRJRN (Start Journal) definition 526
determining what command was using 413 committable resource
used 307 STRJRNAP (Start Journal Access DDL (Data Definition Language)
SAVOBJ (Save Object) Path) 387, 412 resource 541
media considerations 762 STRJRNPF (Start Journal Physical File) DDM (Distributed Data
SAVRST (Save/Restore) 32 description 387 Management) resource 541
SAVRSTCFG (Save and Restore using 411 definition 526
Configuration) 32 Work with Journal (WRKJRN) DRDA (distributed relational
SAVRSTCHG (Save/Restore Changed description 386 database) resource 541
Objects) 32 Option 1-Display Journal FILE (file) resource 541
SAVRSTDLO (Save/Restore Document Status 459 LU62 (protected conversation)
Library Object) 32 options 455 resource 541
SAVRSTLIB (Save/Restore recovery options 456 types 541
Library) 32 Work with Journal Attributes committable resource location 543
SAVRSTOBJ (Save/Restore (WRKJRNA) 386, 442 considerations 591
Object) 32 Work with Object Links save-while-active function 583
SAVSAVFDTA (Save Save File (WRKLNK) 387 definition 526
Data) 762 WRKJRN (Work with Journal) activation group level 534
save file considerations 761 description 386 contents 533
Send Journal Entry (SNDJRNE) 386, Option 1-Display Journal description 533
429 Status 459 job level 534
Send Network File (SNDNETF) 764 options 455 jobs with multiple 537
SNDJRNE (Send Journal Entry) 386, recovery options 456 naming 536
429 WRKJRNA (Work with Journal scope 534
SNDNETF (Send Network File) 764 Attributes) 386, 442 system functions 534
Start Commitment Control command information format description 525
(STRCMTCTL) RST (Restore) command 858 determining size of transaction 584
commit scope (CMTSCOPE) SAV (Save) command 858 during abnormal system or job
parameter 533 commit end 581
commit text (TEXT) forced 566 during activation group end 580
parameter 539 commit cycle during routing step end 581
default journal (DFTJRN) definition 547 during save-while-active
parameter 539 commit cycle identifier 547 function 583
description 527 commit identification ending 579
LCKLVL (lock level) definition 530 error 592
parameter 527 description 530 files being journaled 545
notify object (NFYOBJ) commit in progress (CIP) state 565 flow of conversation 551
parameter 530 commit operation initiator
omit journal entries (OMTJRNE) implicit 573 description 526
parameter 540 language support 548 initiator role 548
Start Journal (STRJRN) 387 overview 547 last agent role 549
using 413, 414 what the system does 549 logical unit of work (LUW)
definition 525

906 OS/400 Backup and Recovery V5R1

commitment control (continued) committable resource (continued) Confirm Language Feature Selection
resynchronization 564 types 541 display 151
states 564 committed (CMT) state 565 Confirm Move of Unit display 677
types 525 committed wave Confirmation Continuation display 677
overview 6 description 541 considerations
performance 588 two-phase commit 541 availability and recovery
placing resources 541 Compare Journal Images (CMPJRNIMG) commitment control 591
practice problem 622 command mirrored protection 297
purpose 525 overview 386, 428 using save files 761
recovery during IPL 569 using 440 console
resource manager comparing problem
definition 526 journal images during recovery 164
restrictions overview 428 console type
when using 591 using 440 changing during restore 230
roles 548 completion message constraint
rollback operation 547, 548, 550 retrieving device name from save pending
sequence of journal entries 545 operation 765 editing during IPL 173
start applications again compression recovery policy 728 contents
using notify objects 608 concurrent add journal entry 428
starting disk unit 643 conversion
commit scope (CMTSCOPE) condition program 252
parameter 533 heuristic mixed (HRM) 565 when restoring programs 253
commit text (TEXT) configuration coordinator role 548
parameter 539 cleaning up 228 CPA3388 message 256
default journal (DFTJRN) duplicating on another system 32 CPF0975 message
parameter 539 errors for mirrored protection 722 during recovery 164
notify object (NFYOBJ) restoring 32, 227 CPF7088 message 55
parameter 530 problems with SRM CPF8113 message 176
omit journal entries (OMTJRNE) information 228 CPFAD84 message 33
parameter 540 rules for mirrored protection 719 CPI0953 message 192
startup 527 saving 32 CPI0954 message 192
status 576 configuration list Create Journal (CRTJRN) command
terminology 526 recovering 213 ASP (auxiliary storage pool)
transaction manager configuration object parameter 403
definition 526 restoring to a different system 228 DLTRCV (delete receiver)
transaction program network 548 configuring parameter 406
two-phase commit disk JRN (journal) parameter 403
agent role 548 adding 2741 Input/Ouput JRNRCV (journal receiver)
committed wave 541 Processor 650 parameter 404
description 540 adding 6502 Input/Output MINENTDTA (minimized
initiator role 548 Processor 650 entry-specific data) parameter 409
last agent role 549 adding 6512 Input/Output MNGRCV (manage receiver)
prepare wave 540 Processor 650 parameter 405
roles 548 adding 6532 Input/Output MSGQ (message queue)
transaction program network 548 Processor 650 parameter 404
vote 540 adding 6533 Input/Ouput overview 386
using Wait for Outcome option 554 Processor 650 RCVSIZOPT (receiver size option)
commitment control operation entry 806 adding 6751 Input/Ouput parameter 407
commitment definition Processor 650 Create Journal Receiver (CRTJRNRCV)
allowing vote read-only 560 adding 6754 Input/Ouput command
using Wait for Outcome option 554 Processor 650 ASP (auxiliary storage pool)
committable change adding 9337 disk 646 parameter 401
description 526 adding 9337 Disk Array AUT (authority) parameter 402
committable resource Subsystem 647, 648 JRNRCVR (journal receiver)
API (application programming adding disk to 6502 IOP 649 parameter 399, 401
interface) resource 541 adding disk to 6512 IOP 649 overview 386
DDL (Data Definition Language) adding disk units without device THRESHOLD (threshold)
resource 541 parity protection 645 parameter 401
DDM (Distributed Data Management) deleting auxiliary storage pool Create Physical File (CRTPF) command
resource 541 (ASP) 654 SHARE(*YES)
description 526 moving disk units 652, 653 impact on journal entries 411
DRDA (distributed relational new system 644 Create Save File (CRTSAVF)
database) resource 541 removing disk units 655, 656, 657, command 761, 762
FILE (file) resource 541 658, 659 creating
location 543 Confirm Delete ASP Data display 204 document library objects (DLOs)
LU62 (protected conversation) Confirm Install of the Operating System user ASP 689
resource 541 display 151 journal 402

Index 907
creating (continued) damaged (continued) data area (continued)
journal receiver 399 job description 175 Start Journal Object (STRJRNOBJ
objects job queue 175 command) 414
user ASP 688, 692 journal 178 data area operation 806
physical file journal receiver 179 Data Definition Language (DDL)
SHARE(*YES) impact on journal journal receivers 444 resource 541
entries 411 journaled object 179 data queue
tape configuration object 180 ENDJRN (End Journal)
for 34xx tape units 165, 342, 879 without library 47, 176 command 423
for non-34xx tape units 166 operating system object 175 ENDJRNOBJ (End Journal Object)
user ASP 669 output queue 175 command 423
creation date QAOSS (text index) database journaled
database file files 175 restoring 235
restoring 238 damaged object restoring 235
CRTJRN (Create Journal) command recovery 174 objects being journaled 235
ASP (auxiliary storage pool) DASD configuration Start Journal (STRJRN
parameter 403 checklist command) 414
DLTRCV (delete receiver) adding 2741 Input/Ouput Start Journal Object (STRJRNOBJ
parameter 406 Processor 650 command) 414
JRN (journal) parameter 403 adding 6502 Input/Output data queue operation 807
JRNRCV (journal receiver) Processor 650 database
parameter 404 adding 6512 Input/Output referential constraint
MINENTDTA (minimized Processor 650 journaling 389
entry-specific data) parameter 409 adding 6532 Input/Output Receive Journal Entry (RCVJRNE)
MNGRCV (manage receiver) Processor 650 command 432
parameter 405 adding 6533 Input/Ouput restoring
MSGQ (message queue) Processor 650 referential constraints 245
parameter 404 adding 6751 Input/Ouput trigger program 247
overview 386 Processor 650 trigger program
RCVSIZOPT (receiver size option) adding 6754 Input/Ouput journaling 389
parameter 407 Processor 650 Receive Journal Entry (RCVJRNE)
CRTJRNRCV (Create Journal Receiver) adding 9337 disk 646 command 432
command adding 9337 Disk Array database file
ASP (auxiliary storage pool) Subsystem 647, 648 assigning to journal 391
parameter 401 adding disk to 6502 IOP 649 constraint
AUT (authority) parameter 402 adding disk to 6512 IOP 649 editing during IPL 173
JRNRCVR (journal receiver) adding disk units without device damaged 47, 176
parameter 399, 401 parity protection 645 deleting 248
overview 386 deleting auxiliary storage pool journaled
THRESHOLD (threshold) (ASP) 654 damaged 179
parameter 401 moving disk units 652, 653 not synchronized 179
CRTNWSSTG (Created NWS Storage new system 644 journaling
Space) command 271, 272 removing disk units 655, 656, 657, choosing 389
CRTPF (Create Physical File) command 658, 659 commitment control 545
SHARE(*YES) interpreting 667 ending 423
impact on journal entries 411 DASD failure starting 411
CRTSAVF (Create Save File) pump 64 journaling access paths 391
command 761 recovery strategy 63 member
current release-to-previous release recovery with device parity damaged 176
support protection 91 multiple members
installing previous release recovery with mirrored protection 90 example 236
compiler 323 DASD migration 12 output for journal entries 431
using TGTRLS (target release) data QAOSS (text index)
parameter 323 restoring save file 255 damaged 175
saving save file 761 renaming
data area during restore 238
D ENDJRN (End Journal)
command 423
restoring
access paths 242
damage
ENDJRNOBJ (End Journal Object) ALWOBJDIF (allow object
journal receiver recovery 464, 465
command 423 difference) parameter 238
save file 764
journaled considerations 236
damaged
restoring 235 creation date 238
database file 47, 176
restoring 235 different member set 241
document
objects being journaled 235 files being journaled 235
restoring 257
Start Journal (STRJRN MAXMBRS (maximum members)
folder
command) 414 parameter 240
restoring into 258
IBM-supplied user profile 175

908 OS/400 Backup and Recovery V5R1

database file (continued) dependent file disk configuration (continued)
MBROPT (member option) restoring 245 checklist (continued)
parameter 241 device adding 6502 Input/Output
member locking 238 recovering after restore Processor 650
members, list of 241 operation 228 adding 6512 Input/Output
shared formats 245 starting Processor 650
STRJRNPF (Start Journal Physical File) during recovery 160 adding 6532 Input/Output
command 411 device configuration Processor 650
database file member operation restoring 227 adding 6533 Input/Ouput
entry 806 device error Processor 650
database file operation entry 806 recovery actions 289 adding 6751 Input/Ouput
database relations unrecoverable 297 Processor 650
displaying 248 device name adding 6754 Input/Ouput
database server (QDBSRVnn) job retrieving from save completion Processor 650
commitment control recovery 569 messages 765 adding 9337 disk 646
database service (QSXJRN) journal 424 device parity protection adding 9337 Disk Array
DB2 multisystem files displaying status 708 Subsystem 647, 648
how to end journaling 423 excluding disk unit 707 adding disk to 6502 IOP 649
how to start journaling 412 including disk unit 705 adding disk to 6512 IOP 649
DBCLOB(double-byte character large overview 8 adding disk units without device
object) 437 recovery steps 91 parity protection 645
DDL (Data Definition Language) starting 695, 699 deleting auxiliary storage pool
resource 541 9337 Disk Array Subsystem 695 (ASP) 654
DDM (distributed data management) stopping 701 moving disk units 652, 653
using with commitment control 541 working with 695 new system 644
dedicated service tools (DST) DFTJRN (default journal) parameter 539 removing disk units 655, 656, 657,
definition 61 directory 658, 659
options 660 restoring definition 61
starting 660 changed objects 281 interpreting 667
stopping 662 restoring objects 261 recovering 142
Dedicated Service Tools (DST) Sign On directory, IFS DISK CONFIGURATION ATTENTION
display 139, 151 ENDJRN (End Journal) REPORT display 139
default command 423 Disk Configuration Error Report
owner (QDFTOWN) user profile Start Journal (STRJRN) display 168
restoring objects 218 command) 413 disk configuration status
default journal (DFTJRN) parameter 539 directory entry displaying 663
define or change system at IPL renaming printing 663
during recovery 160 restoring mail 260 understanding 663
Define or Change the System at IPL directory information format disk failure
menu 160 RST (Restore) command 859 pump 64
defining system SAV (Save) command 859 recovery with device parity
at IPL disabling protection 91
during recovery 160 high-speed feature on 2440 Tape recovery with mirrored protection 90
definition 526 Unit 145 disk protection
activation group level 534 disaster recovery configuration sequences 737
contents 533 sample plan 783 adding a storage controller 737
description 533 disk disk space
job level 534 adding while system active 643 calculating 681
jobs with multiple 537 concurrent add 643 disk storage
naming 536 recovering configuration 142 isolating 7
scope 534 understanding configuration segmenting 7
system functions 534 status 663 disk unit
Delete Journal (DLTJRN) command 386 disk compression 725 adding to system 645
Delete Journal Receiver (DLTJRNRCV) considerations 725 assigning to auxiliary storage pool
command 386 disk unit capacity 726 (ASP) 669
Delete Pointer Handle disk unit full considerations 728 device parity protection
(QjoDeletePointerHandle) API 386 error codes 740 including 705
delete receiver (DLTRCV) parameter 406 SRC 6xxx 7051 740 excluding in device parity
deleting SRC 6xxx 7052 741 protection 707
auxiliary storage pool (ASP) 654 restrictions 725 failure
journal 249 SRC A6xx 0277 729 load source unit before IPL 298
journal receiver 251 starting 732 moving 676
journal receivers 406, 418 stopping 735 nonconfigured status 669
journaled object 423 disk configuration reasons 152
journals 419 checklist removing
physical file 248 adding 2741 Input/Ouput from ASP 678
user ASP 680 Processor 650 resuming status 668

Index 909
disk unit (continued) distributed transaction processing (DTP) DPY/Active disk unit status 668
status 667 model DPY/Failed disk unit status 668
suspended status 668 components 566 DPY/Rebuilding disk unit status 668
unprotected status 668 distribution media DPY/Resyncing disk unit status 668
disk unit failure restoring Licensed Internal Code 123 DPY/Unknown disk unit status 668
recovery strategy 63 restoring OS/400 licensed DPY/Unprotected disk unit status 668
disk unit full program 148 DRDA (distributed relational database)
system response 728 distribution object resource 541
disk unit number restoring 258 DSNX (QDSNX) journal 424
definition 667 distribution services (QAOSDIAJRN) DSPDBR (Display Database Relations)
disk unit status journal command 248
active 668 applying journaled changes 286 DSPFD (Display File Description)
busy 667 overview 424 command 762
DPY/Active 668 DLO (document library object) DSPJRN (Display Journal)
DPY/Failed 668 creating command 386
DPY/Rebuilding 668 user ASP 689 overview 428
DPY/Resyncing 668 maximum number on RSTDLO DSPJRNRCVA (Display Journal Receiver
DPY/Unknown 668 command 257 Attributes) command 386
DPY/Unprotected 668 reclaiming 257 DSPRCYAP (Display Recovery for Access
not operational 667 renaming Paths) command 380
not ready 667 restoring documents 260 DSPSAVF (Display Save File)
operational 667 restoring command 761, 762
performance degraded 667 descriptive information 259 DST (dedicated service tools)
read/write protected 667 media error 57 definition 61
redundant failure 668 overview 255 options 660
suspended 668 renaming document 258 starting 660
write protected 667 user ASP 200 stopping 662
disk utilization using RST (Restore) dual systems
journaling 395 command 277 overview 9
display restoring authority 259 duplicating 32
printing journal entries 429 restoring ownership 259 changed objects 32
Display Access Path Status display 163, DLTJRN (Delete Journal) command 386 configuration 32
172 DLTJRNRCV (Delete Journal Receiver) document library objects 32
Display Constraint Status display 164, command 386 object 32
174 DLTRCV (delete receiver) parameter 406 object in directory 32
Display Database Relations (DSPDBR) document
command 248 restoring
Display Disk Configuration Capacity
display 193
damaged 257
overview 255
E
Edit Check Pending Constraint
Display File Description (DSPFD) document library
(EDTCPCST) command 174
command 762 querying 200
Edit Check Pending Constraints
Display Journal (DSPJRN) document library object
display 163, 173
command 386 duplicating on another system 32
edit description
overview 428 saving
recovering 213
Display Journal Receiver Attributes and restoring 32
restoring 157
(DSPJRNRCVA) command 386 document library object (DLO)
Edit Rebuild of Access Paths
Display Object Links (DSPLNK) creating
display 162, 171
command 387 user ASP 689
Edit Recovery for Access Paths
Display Recovery for Access Paths maximum number on RSTDLO
(EDTRCYAP) command 378
(DSPRCYAP) command 380 command 257
EDTCPCST (Edit Check Pending
Display Save File (DSPSAVF) reclaiming 257
Constraint) command 174
command 761, 762 renaming
EDTRCYAP (Edit Recovery for Access
displaying restoring documents 260
Paths) command 378
database relations 248 restoring
enabling
device parity protection status 708 descriptive information 259
automatic configuration
journal media error 57
during recovery 160
overview 428 overview 255
high-speed feature on 2440 Tape
journal receivers 442 renaming document 258
Unit 145
object user ASP 200
End Journal (ENDJRN) command 387,
user ASP 684 using RST (Restore)
423
save file 761 command 277
End Journal Access Path (ENDJRNAP)
status messages during save 766 restoring authority 259
command 387, 423
distributed data management (DDM) restoring ownership 259
End Journal Object (ENDJRNOBJ)
using with commitment control 541 documenting
command 387, 423
distributed mail services 807 journaling environment 415
End Journal Physical File Changes
distributed relational database (DRDA) Domino server
(ENDJRNPF) command 423
resource 541 recovering 265

910 OS/400 Backup and Recovery V5R1

end journaling evaluating file (continued)
for DB2 multisystem files 423 changes to journaling 416 using journaled changes to recover
End Subsystem (ENDSBS) command example physical file 462
QCALSRV (calendar server) application program 615 file-dependent attributes for save
subsystem 45 Apply Journaled Changes file 764
QSYSWRK (subsystem monitor) (APYJRNCHG) command 448 file system
subsystem 45 APYJRNCHG (Apply Journaled QNetWare
restricted state 45 Changes) command 448 restoring 264
using 45 configuration fixed-length portion
ending planning for mirrored journal entry 815
commitment control 579 protection 795 flow
dedicated service tools (DST) 662 creating journal 409 commit processing 551
journaling 423 creating journal receiver 409 folder
access path 423 CRTJRNRCV (Create Journal Receiver) creating
file changes 423 command 409 user ASP 689
physical file 423 database file restoring
mirrored protection 723 multiple members 236 damaged 258
subsystem handling escape conditions 767 overview 255
QCALSRV (calendar server) job with multiple commitment procedure 258
subsystem 45 definitions 537 transferring
QSYSWRK (subsystem monitor) Remove Journaled Changes different ASP 687
subsystem 45 (RMVJRNCHG) command 451 force object conversion (FRCOBJCVN)
restricted state 45 retrieving the device name from save parameter 254
using 45 completion messages 765 force object conversion on restore
system service tools 662 RMVJRNCHG (Remove Journaled (QFRCCVNRST) system value 253
ENDJRN (End Journal) command 423 Changes) command 451 force-write ratio (FRCRATIO) parameter
ENDJRNAP (End Journal Access Path) transaction logging file 602 journaling 394
command 387, 423 using Retrieve Journal Entry forced commit operation 566
ENDJRNOBJ (End Journal Object) (RTVJRNE) command in forced rollback operation 565
command 423 program 766 FRCOBJCVN (force object conversion)
ENDJRNPF (End Journal Physical File excluding parameter 254
Changes) command 423 disk unit in device parity FRCRATIO (force-write ratio) parameter
ENDSBS (End Subsystem) command protection 707 journaling 394
QSYSWRK (subsystem monitor) exit program
subsystem 45 receiving journal entries 433
restricted state 45
using 45
explicit access path journaling
definition 391
G
GIGMIG (AS/400 DASD Migration)
entire system explicit journaling
service offering 12
restore operation definition 376
unattended 209 exposed access path 377
restoring 208
entry type H
journal 807
error
F hard link
restoring 262
failure
commitment control 592 hardware
active mirrored load source 299
mirrored protection configuration
bus 298
configuration 722 cleaning up 228
I/O processor 298
permanent read 297 hardware resource
load source unit
restore operation definition 667
before IPL 298
not recoverable 56 header format
fiber optic bus
recoverable 56 RST (Restore) command 858
OptiConnect for OS/400 10
SRM (system resource management) SAV (Save) command 858
file
information 228 heuristic decision
constraint
unrecoverable device 297 and cancelling resynchronization 570
editing during IPL 173
error handling definition 540
database
mirrored protection heuristic mixed (HRM) condition 565
shared formats 245
disk 297 high-speed feature
journaled
error message 297 2440 Tape Unit
naming 403
error screens disabling 145
restoring 235
LIC 775 enabling 145
journaling 387
licensed internal code HRM (heuristic mixed) condition 565
restoring 236
installation 775 human error
logical 242
escape condition recovery strategy 63
save
handling with program 767
considerations for using 761
estimating
sequential processing (SEQONLY)
journal receiver size 396
journaling 395

Index 911
I install options
selecting
job number
resetting counter
IBM-supplied journal restoring operating system 154 during recovery 155
managing 424 Install the Operating System job queue
IBM-supplied user profile display 141, 153 clearing during recovery 155
damaged 175 installation damaged 175
ICF file abbreviated job recovery strategy
writing journal entries 772 definition 148 batch 760
identifier installation device interactive 759
commit cycle 547 alternate 635 journal
journal (JID) 414 installation error screens applying changes 444
IFS object licensed internal code 775 ASP (auxiliary storage pool) 403
directory (*DIR) Integrated File System 806 assigning files 391
ENDJRN (End Journal) integrated file system (IFS) object associate receivers 462
command 423 journaling associating with receiver 404
journaled directory (*DIR) 413 changing 416
damaged 179 starting 413 changing receivers 405
not synchronized 179 stream file (*STMF) 413 creating 402
restoring 235 symbolic link (*SYMLNK) 413 damaged 178
restoring interactive job data area 383
IFS objects being journaled 235 recovery strategy 759 data queue 383
stream file (*STMF) internal microprogramming interface definition 5, 383
ENDJRN (End Journal) (IMPI) system deleting 249, 419
command 423 restoring programs 253 displaying
symbolic link (*SYMLNK) interpreting overview 428
ENDJRN (End Journal) disk configuration 667 example 409
command 423 introduction how many 391
IMPI (internal microprogramming availability 3 IBM-supplied
interface) system backup options 3 managing 424
restoring programs 253 recovery options 3 IFS object 383
implicit commit operation 573 save options 3 journal receivers 416
implicit rollback operation 573 IPL (initial program load) library 403
including after abnormal end 167 managing receivers 405
disk unit in device parity commitment control recovery 569 message queue 404
protection 705 disk related failure of load source messages 404
independent ASP unit 298 naming 403
definition 62 editing check pending overflowed
recovering disk configuration after constraints 173 resetting 691
complete system loss 138 editing rebuild of access paths 171 receiver
indpendent ASP (auxiliary storage pool) normal 58 definition 5
recovery procedures options recovering from QRCL library 185
complete data loss 94 during recovery 160 removing changes 444
no data loss 93 performing normal 58 restoring 248, 422
some data loss 93 reducing time 5 journal receiver name 400
initial program load (IPL) resetting journal sequence saving 420
after abnormal end 167 numbers 417 threshold messages 404
commitment control recovery 569 restoring operating system 149 transferring into user ASP 687
disk related failure of load source selecting options WRKJRN (Work with Journal)
unit 298 restoring operating system 158 options 455
editing check pending IPL Options display 159, 170 journal attributes
constraints 173 IPL or Install the System display 150 working with 442
editing rebuild of access paths 171 IPL status messages journal code 807
options example display 153 actions of the APYJRNCHG or
during recovery 160 isolating RMVJRNCHG command 452
performing normal 58 disk units 7 journal code A 806
reducing time 5 journal code B 806
resetting journal sequence journal code C 806
numbers 417
restoring operating system 149 J journal code D 806
journal code E 806
selecting options JID (journal identifier) 414
journal code F 806
restoring operating system 158 job
journal code I 806
initiator multiple commitment definitions 537
journal code J 806
description 526 job description
journal code L 806
initiator role 548 damaged 175
journal code M 806
inoperable journal receivers 444 job end
journal code O 807
input commitment control 581
journal code P 807
save file 763 job-level scope 534
journal code Q 807

912 OS/400 Backup and Recovery V5R1

journal code R 807 journal images journal management (continued)
journal code S 807 comparing journal entry (continued)
journal code T 807 overview 428 journal code O 807
journal code U 807 using 440 journal code P 807
journal entry journal management journal code Q 807
access path operation 806 access path journal code R 807
audit trail 807 when used 391 journal code S 807
code 807 access path operation entry 806 journal code T 807
commitment control operation 806 audit trail 807 journal code U 807
contents 428, 815, 821 before-images 392 journal or receiver operation 806
data area operation 806 choosing which objects 389 license management 806
data queue operation 807 commitment control 545 network management data 806
database file member operation 806 commitment control operation object oriented 807
database file operation 806 entry 806 operation on specific record 807
definition 5, 383 data area operation entry 806 performance tuning 807
displaying and printing 429 data queue operation 807 system accounting 806
distributed mail services 807 database file member operation user-generated entry 807
fixed-length portion 815 entry 806 journal or receiver operation
general information 427 database file operation entry 806 entry 806
Integrated File System 806 distributed mail services 807 journal receiver
journal code A 806 ending 423 user ASP (auxiliary storage
journal code B 806 evaluating changes 416 pool) 393
journal code C 806 features that increase journal receiver keeping records 415
journal code D 806 size 397 license management entry 806
journal code E 806 files with referential constraints 389 managing 415
journal code F 806 files with trigger programs 389 maximum sequence number 417
journal code I 806 how to assign files 391 network management data 806
journal code J 806 inoperable receivers 444 number of journals 391
journal code L 806 journal code A 806 object oriented entry 807
journal code M 806 journal code B 806 objects the system does not
journal code O 807 journal code C 806 journal 390
journal code P 807 journal code D 806 operation on specific record 807
journal code Q 807 journal code E 806 output file layouts 805
journal code R 807 journal code F 806 overview 5, 383
journal code S 807 journal code I 806 performance impact 394
journal code T 807 journal code J 806 performance tuning entry 807
journal code U 807 journal code L 806 planning 388
journal or receiver operation 806 journal code M 806 receiver chains 440
license management 806 journal code O 807 receiver directory 442
methods for using 427 journal code P 807 resetting sequence numbers 417
network management data 806 journal code Q 807 sequence of commitment control
object oriented 807 journal code R 807 entries 545
operation on specific record 807 journal code S 807 sequential processing
output journal code T 807 (SEQONLY) 395
database file 431 journal code U 807 setting up 398
work station 430 journal entry storage impact 395
performance tuning 807 access path operation 806 system accounting entry 806
receiving audit trail 807 system change-journal management
DEPENT (dependent entry) commitment control resetting sequence numbers 417
parameter 432 operation 806 user ASP (auxiliary storage pool)
exit program 433 data area operation 806 journal receiver 393
overview 428 data queue operation 807 user-generated entry 807
using 432 database file member using for audit trail 439
retrieving operation 806 using force-write ratio (FRCRATIO)
overview 428 database file operation 806 parameter 394
using 436 distributed mail services 807 using with save-while-active
RTVJRNE (Retrieve Journal Entry) Integrated File System 806 function 394
command 766 journal code A 806 journal or receiver operation entry 806
sending 429 journal code B 806 journal receiver
system accounting 806 journal code C 806 ASP (auxiliary storage pool) 401
transmitting to another system 772 journal code D 806 associating with journal 404
types 807 journal code E 806 authority 402
user-created 429 journal code F 806 chain break 441
user-generated entry 807 journal code I 806 changing 405, 416
variable-length portion 821 journal code J 806 creating 399
writing to ICF file 772 journal code L 806 damaged 179
journal identifier (JID) 414 journal code M 806 recovering 464, 465

Index 913
journal receiver (continued) journaling (continued) journaling (continued)
definition 5, 383 before-images 392 journal entry (continued)
deleting 251, 406, 418 choosing which objects 389 journal code B 806
directory 442 commitment control 545 journal code C 806
correcting 251 commitment control operation journal code D 806
displaying 442 entry 806 journal code E 806
estimating size 396 data area journal code F 806
examples 409 ending 423 journal code I 806
inoperable 444 starting 414 journal code J 806
internal entries 407 data area operation entry 806 journal code L 806
library 401 data queue journal code M 806
managing 405, 416 ending 423 journal code O 807
maximum sequence number 417 starting 414 journal code P 807
maximum size 395 data queue operation 807 journal code Q 807
messages 404 database file member operation journal code R 807
moving entry 806 journal code S 807
from overflowed ASP 690 database file operation entry 806 journal code T 807
naming 399 DB2 multisystem files 412, 423 journal code U 807
placing in nonlibrary user ASP 694 definition 5 journal or receiver operation 806
placing in user ASP 689 distributed mail services 807 license management 806
recovering from QRCL library 185 ending 423 network management data 806
resetting sequence numbers 417 evaluating changes 416 object oriented 807
restoring 248, 422 explicit operation on specific record 807
saving 420 definition 376 performance tuning 807
size 401 features that increase journal receiver system accounting 806
system-generated names 399 size 397 user-generated entry 807
threshold 401 files with referential constraints 389 journal or receiver operation
journal receiver chain 440 files with trigger programs 389 entry 806
journal status how to assign files 391 journal receiver
display on the WRKJRN command IFS directories (*DIR) 413 user ASP (auxiliary storage
option 459 IFS object pool) 393
journaled change ending 423 keeping records 415
APYJRNCHG (Apply Journaled IFS stream file (*STMF) 413 license management entry 806
Changes) command 446, 448 IFS symbolic links (*SYMLNK) 413 managing 415
recovery of journaled object 462 inoperable receivers 444 maximum sequence number 417
RMVJRNCHG (Remove Journaled integrated file system (IFS) object network management data 806
Changes) command 450, 451 starting 413 number of journals 391
journaled changes journal code A 806 object oriented entry 807
applying journal code B 806 objects the system does not
broken receiver chain 285 journal code C 806 journal 390
determining whether to 282 journal code D 806 operation on specific record 807
overview 428 journal code E 806 output file layouts 805
referential constraints 445 journal code F 806 overview 5, 383
trigger programs 446 journal code I 806 performance impact 394
unbroken receiver chain 284 journal code J 806 performance tuning entry 807
removing journal code L 806 physical file
overview 428 journal code M 806 starting 411
referential constraints 445 journal code O 807 physical file changes
trigger programs 446 journal code P 807 ending 423
journaled file journal code Q 807 planning 388
naming 403 journal code R 807 receiver chains 440
restoring 235 journal code S 807 receiver directory 442
journaled IFS object journal code T 807 resetting sequence numbers 417
restoring 235 journal code U 807 sequence of commitment control
journaled object journal entry entries 545
damaged 179 access path operation 806 sequential processing
not synchronized 179 audit trail 807 (SEQONLY) 395
journaling commitment control setting up 398
access path operation 806 storage impact 395
ending 423 data area operation 806 system accounting entry 806
overview 6 data queue operation 807 system change-journal management
starting 412 database file member resetting sequence numbers 417
when used 391 operation 806 user ASP (auxiliary storage pool)
with SMAPP 376 database file operation 806 journal receiver 393
access path operation entry 806 distributed mail services 807 user-generated entry 807
applying changes 282 Integrated File System 806 using for audit trail 439
audit trail 807 journal code A 806

914 OS/400 Backup and Recovery V5R1

journaling (continued) load source unit (continued) media error (continued)
using force-write ratio (FRCRATIO) recovery procedure restoring storage 318
parameter 394 complete data loss, no user member
using with save-while-active ASP 67 damaged 176
function 394 complete data loss, user ASP not locking
overflowed 68 during restore 238
no data loss 65 renaming
L some data loss 66
LOB (large object) 437
during restore 238
restoring
LAP (last agent pending) state 565
local location ALWOBJDIF (allow object
last agent pending (LAP) state 565
definition 526 difference) parameter 238
last agent role 549
lock level (LCKLVL) parameter 527 creation date 238
LCKLVL (lock level) parameter 527
locking member option (MBROPT)
library
database members parameter 241
duplicating on another system 32
during restore 238 message
journal 403
restore processing 53 CPA3388 256
journal receiver 401
logic flow steps 631 CPF7088 55
locking during restore procedure 53
logical file CPF8113 176
moving
restoring 242 CPFAD84 33
different ASP 687
logical partitions CPI0953 192
naming
restoring 231 CPI0954 192
journaling 403
logical unit of work save completion 765
restoring 32
definition 525 message queue
*ALLUSR libraries 232
resynchronization 564 journal 404
*IBM libraries 232
states 564 journal receiver threshold 404
*NONSYS libraries 232
types 525 message reply list
media error 56
lost object 47, 176 restoring 157
OPTION parameter 232
LU62 (protected conversation) microcode
overview 231
resource 541 restoring 122
user ASP 198
MINENTDTA (minimized entry-specific
saving 32
data) parameter 409
determining what command was
used 307 M Minimized entry-specific data
(MINENTDTA) parameter 409
library list mail
minimizing locks 586
changing restoring 258
mirrored protection
during recovery 160 mail server framework
active disk unit status 668
library user ASP journal entry 807
active load source failure 299
definition 62, 393 mail server framework (QZMF)
configuration errors 722
LIC journal 424
configuration rules 719
error screens 775 main storage dump 167
device error
installation error screens 775 major system options
recovery actions 289
license management setting
disk error handling 297
QLZALOG journal 424 during recovery 160
ending 723
license management entry 806 setting during recovery 160
I/O processor or bus failure 298
licensed internal code manage receiver (MNGRCV)
missing disk units 298
installation error screens 775 parameter 405
nonconfigured unit
Licensed Internal Code managed system services (QCQJMJRN)
using for replacement 294
definition 62 journal 424
overview 7
restoring managing
permanent read error
preparation 122 IBM-supplied journals 424
recovery actions 289
starting 122 journal receivers 405, 416
recovery actions
steps 123 journaling environment 415
errors and failures 289
using distribution media 123 manual IPL (initial program load)
performed by the service
restoring using function code 23 restoring operating system 149
representative 296
SRC (system reference) codes 127 maximum journal sequence number 417
recovery steps 90
Licensed Internal Code IPL in Progress maximum members (MAXMBRS)
replacing unit 292
display 141 parameter
resuming 297
licensed programs exceeding during restore 240
resuming status 668
restoring 255 maximum size
starting 719
link journal receiver 395
stopping 723
restoring 262 MAXMBRS (maximum members)
suspended disk unit status 668
List Objects (QUSLOBJ) API 387 parameter
suspended status 668
load source exceeding during restore 240
unknown load source status 301
failure MBROPT (member option)
using spare nonconfigured unit 294
active mirrored 299 parameter 241
working with 719
unknown status 301 media error
mirrored unit
load source unit during RSTDLO procedure 57, 58
replacing 292
definition 62 during RSTLIB procedure 56

Index 915
mirrored unit (continued) nonload source unit (continued) omit journal entries (OMTJRNE)
resuming 291 recovery procedure (continued) parameter 540
suspending 290 no data loss 76 OMTJRNE (omit journal entries)
mirroring normal initial program load (IPL) 58 parameter 540
device error not operational disk unit status 667 Open List of Objects (QGYOLOBJ)
recovery actions 289 not ready disk unit status 667 API 387
overview 7 not synchronized opening
permanent read error journaled file 179 save file 763
recovery actions 289 notify object operating system
MNGRCV (manage receiver) description 530 damaged object 175
parameter 405 notify for each program 609 preventing unauthorized
moving one for all programs 614 installation 151
disk unit 676 standard commit processing restoring
disk units 652, 653 program 618 choosing procedure 148
folder standard processing program 614 manual IPL 149
different ASP 687 start applications again 608 overview 147
journal receiver updating 531 preparation 147
overflowed ASP (auxiliary storage notify object (NFYOBJ) parameter 530 reasons 147
pool) 690 selecting install options 154
library steps 149
different ASP 687
object
O using distribution media 148
operation on specific record 807
object 462
different ASP 693 Operational Assistant
creating
user profile backup
user ASP 688, 692
different system 217 lists, folders 888
damaged 180
multisystem files lists, libraries 888
displaying journaled status 442
how to end journaling 423 options 890
duplicating on another system 32
how to start journaling 412 overview 885
journaled
recovering 117
deleting 423
schedule 891
lost owner 48
N ownership
saving 885
saving changed objects 889
naming restoring 218
what to save 888
commitment definitions 536 primary group
operational disk unit status 667
journal receivers 399 restoring 218
OPM (Original Program Model) objects
journaled files 403 restore sequence 45
restoring 254
journals 403 restoring 32
optical media
libraries 403 RSTOBJ (Restore Object)
overview 4
network command 234
optical support
database 245 saving 32
overview 4
restoring 245 previous release system 323
OptiConnect for OS/400
network attribute transferring
overview 10
recovering 213 between ASPs 686
option
resetting when restoring to a different different ASP 693
recovery using the Work with Journal
system 162 user ASP
command 456
network management data 806 displaying 684
Work with Journal (WRKJRN) 455
new system without library 47, 176
order
configuring disk 644 object in directory
restoring objects 45
NFYOBJ (notify object) parameter 530 duplicating on another system 32
Original Program Model (OPM) objects
non-34xx tape units restoring 32, 261
restoring 254
creating tape configuration 166 saving 32
OS/400 Integration for Novell NetWare
nonconfigured disk unit object link information format
(QNetWare) file system
definition 669 RST (Restore) command 860
restoring 264
reasons 152 SAV (Save) command 860
OS/400 licensed program
nonconfigured unit object oriented entry 807
preventing unauthorized
mirrored protection 294 object ownership
installation 151
nonlibrary user ASP ALWOBJDIF (allow object differences)
restoring
definition 62, 393 parameter 218
choosing procedure 148
placing journal receivers 694 ObjectConnect
manual IPL 149
working with 692 communications requirements 29
overview 147
nonload source unit components 30
preparation 147
recovery procedure how system runs commands 30
reasons 147
complete data loss, no user job flow 30
selecting install options 154
ASP 78 list of commands 29
steps 149
complete data loss, user ASP not overview 10, 29
using distribution media 148
overflowed 79 problem determination 33
complete data loss, user ASP setting up 30
overflowed 82

916 OS/400 Backup and Recovery V5R1

output performance tuning (QPFRADJ) profile
journal entry journal 424 QDFTOWN (default owner)
database file 431 performance tuning entry 807 restoring objects 218
work station 430 permanent error 289 program
RST (Restore) command 857 permanent read error conversion 252
RSTDLO (Restore Document Library mirrored protection notify object 614
Object) command 256 recovery actions 289, 297 recreation 252
SAV (Save) command 857 physical file restoring 252
save file 763 changing different release 253
output file SHARE(*YES) impact on journal translation 252
journal entry entries 411 validation value 252
database file 805 creating program failure
output queue SHARE(*YES) impact on journal recovery strategy 63
clearing during recovery 155 entries 411 program temporary fix (PTF)
damaged 175 deleting 248 restoring 278
output sequence ENDJRNPF (End Journal Physical File programming example
RST (Restore) command 857 Changes) command 423 handling escape conditions 767
SAV (Save) command 857 journaled member retrieving the device name from save
overflowed object deleting 423 completion messages 765
deleting 196 recover using journaled changes 462 using Retrieve Journal Entry
overflowed status Start Journal Physical File (STRJRNPF (RTVJRNE) command in
definition 191 command) 411 program 766
overflowed user ASP physical file member protected conversation (LU62)
deleting objects 196 reorganize when applying journaled resource 541
determining status 192 changes 455 protecting
displaying 193 PIP (prepare in progress) state 565 access paths 375
recovering 191 planning protection
resetting 192, 193 journal management 388 access path
Override with Save File (OVRSAVF) power failure system-managed 375
command 762 recovery strategy 62 PRP (prepared) state 565
overview power loss PTF (program temporary fix)
access path protection 5 uninterruptible power supply 9 restoring 278
auxiliary storage pool (ASP) 7 workstation 9 pump (disk pump) 64
device parity protection 8 PowerPC based system
journal management 5 restoring programs 253
mirrored protection 7
restore operations 3
prepare in progress (PIP) state 565
prepare wave
Q
QACGJRN (accounting) journal 424
save operations 3 description 540
QALWOBJRST (allow object restore
OVRSAVF (Override with Save File) two-phase commit 540
operation) system value 50
command 762 prepared (PRP) state 565
QALWUSRDMN (allow user domain
ownership preparing
objects) system value 48
ALWOBJDIF (allow object differences) save files for use 761
QAOSDIAJRN (distribution services)
parameter 218 previous release system
journal
restoring 218 saving objects 323
applying journaled changes 286
previous release-to-current release
overview 424
support
QAPZ files
P considerations when restoring
configuration objects 331
attempting to restore 54
parallel restore operations 59 QAUDJRN (audit) journal
restoring data to current release
parameter creating during restore 55
system 331
force object conversion description 424
saving data on previous release
(FRCOBJCVN) 254 QAUDJRN journal 439
system 331
FRCOBJCVN (force object QCALSRV (calendar server) subsystem
primary group
conversion) 254 ending 45
ALWOBJDIF (allow object differences)
SAVACTWAIT (save-active wait) 583 QCMN (communications) subsystem
parameter 219
save-active wait (SAVACTWAIT) 583 ObjectConnect 30
restoring 218
parent file QCQJMJRN (managed system services)
printer writer
restoring 245 journal 424
starting
participant role 548 QDBSRVnn (database server) job
during recovery 159
pending constraint commitment control recovery 569
printing
editing during IPL 173 QDFTOWN (default owner) user profile
journal entry 429
performance assigning ownership during reclaim
private authority
commitment control 588 storage procedure 48
restoring 219
journaling 394, 540 restoring objects 218
problem determination
using save files 761 QDSNX (DSNX) journal 424
ObjectConnect 33
performance degraded disk unit QFRCCVNRST (force object conversion
processing flow 615
status 667 on restore) system value 253

Index 917
QjoAddRemoteJournal QVFYOBJRST (verify object on restore) receiver (continued)
(QjoAddRemoteJournal) API 474, 475 system value 50 journal (continued)
QjoDeletePointerHandle (Delete Pointer QZMF (mail server framework) saving 420
Handle) API 386 journal 424 size 401
QjoRetrieveJournalEntries (Retrieve system-generated names 399
Journal Entries) API 386 threshold 401
QJORJIDI (Retrieve Journal Identifier
Information) API 386
R restoring 248
receiver chain 440
RBR (rollback required) state 565
QJOSJRNE (Send Journal Entry) broken
RCLDLO (Reclaim Document Library
API 386, 429, 535 applying journaled changes 285
Object) command 257
QLYJRN (Application Development definition 251
RCLSTG (Reclaim Storage) command
Manager transaction log) journal 424 unbroken
duplicate names in QRCL 47
QLYPRJLOG (Application Development applying journaled changes 284
object ownership 48
Manager project log) journal 424 receiver chain break 441
procedure 46, 183
QLZALOG (license management) receiver directory
QALWUSRDMN (allow user domain
journal 424 correcting 251
objects) system value 48
QNetWare working with 442
recovering user ASP 183
restoring 264 receiving
user domain object 48
QNTC file system journal entry
what system does 47
restoring 263 DEPENT (dependent entry)
why to run 176
QPFRADJ (performance tuning) parameter 432
RCVJRNE (Receive Journal Entry)
journal 424 exit program 433
command
QPWRRSTIPL (automatic IPL after power overview 428
DEPENT (dependent entry)
restored) system value 167 using 432
parameter 432
QRCL (recovery) library Reclaim Document Library Object
exit program 433
duplicate names 47 (RCLDLO) command 257
overview 386, 428
journal 185 Reclaim Storage (RCLSTG) command
referential constraints 432
journal receiver 185 duplicate names in QRCL 47
trigger programs 432
using for recovery 185 object ownership 48
using 432
QRYDOCLIB (Query Document Library) procedure 46, 183
writing output to save media 769
command 200 QALWUSRDMN (allow user domain
RCVNETF (Receive Network File)
QSNADS (SNADS) journal 424 objects) system value 48
command 765
QSNMP (SNMP) journal 424 recovering user ASP 183
read error 297
QSOC (OptiConnect/400) subsystem user domain object 48
read/write protected disk unit
ObjectConnect 30 what system does 47
status 667
QSOCCT mode description why to run 176
Receive Journal Entry (RCVJRNE)
ObjectConnect 30 reclaiming
command
QSR (ObjectConnect) library 30 document library object (DLO) 257
DEPENT (dependent entry)
QSRLSAVF (List Save File) API 761 storage
parameter 432
QsrRestore API 261 duplicate names in QRCL 47
exit program 433
QSXJRN (database service) journal 424 procedure 46, 183
overview 386, 428
QSYSMSG message queue QALWUSRDMN (allow user
using 432
error messages 297 domain objects) system
writing output to save media 769
QSYSOPR message queue value 48
Receive Network File (RCVNETF)
error messages 297 recovering user ASP 183
command 765
QSYSWRK (subsystem monitor) user domain object 48
receiver
subsystem what the system does 47
damage recovery 464
ending 45 why to run 176
damaged journal recovery 465
QTNADDCR (add commitment control record lock
journal
resource) API 541 description 585
ASP (auxiliary storage pool) 401
QTNADDCR (Add Commitment duration 528
authority 402
Resource) API 535 record locking
changing 405, 416
QTNCHGCO (Change Commitment commitment control recovery 569
creating 399
Options) API 535 recording
deleting 406, 418
QTNRBRQD (Rollback Required) journaling environment 415
directory 442
API 535 recover
displaying 442
QTNRCMTI (Retrieve Commitment Windows server files 273
estimating size 396
Information) API 564 recover using journaled changes 462
examples 409
QTNRMVCR (remove commitment recoverable error
inoperable 444
control resource) API 541 restore operation 56
internal 407
Query Document Library (QRYDOCLIB) recovering
library 401
command 200 abnormal system end 462
managing 405
querying access path recovery times 156, 213
maximum sequence number 417
document library 200 addressability
messages 404
QUSER user profile user ASP 182
naming 399
ObjectConnect 30 configuration lists 213
resetting sequence numbers 417
damaged database files 176
restoring 422

918 OS/400 Backup and Recovery V5R1

recovering (continued) recovery (continued) recovery checklist (continued)
damaged journal receivers 464, 465 disaster user information (continued)
database file sample plan 783 using Restore menu options 22 and
damaged 176 mirrored protection 297 23 114
devices that will not vary on 228 strategy recovery from unreadable sectors during
disk configuration 142 batch jobs 760 disk failure 174
Domino server 265 interactive jobs 759 recovery procedure 871
edit descriptions 157, 213 jobs 759 complete system loss 96
Licensed Internal Code unreadable sectors 174 including independent ASP 98
preparation 122 recovery (QRCL) library with independent ASP and Service
starting 122 duplicate names 47 Tools Network Interface 102
steps 123 journal 185 device parity protection 91
using distribution media 123 journal receiver 185 independent ASP
message reply list 157 using for recovery 185 complete data loss 94
network attributes 213 recovery actions no data loss 93
objects using journaled changes 462 mirrored protection 289, 296 some data loss 93
OS/400 licensed program performed by the service load source unit
choosing procedure 148 representative complete data loss, no user
manual IPL 149 mirrored protection 296 ASP 67
overview 147 recovery checklist complete data loss, user ASP not
preparation 147 complete system loss 96 overflowed 68
reasons 147 including independent ASP 98 complete data loss, user ASP
selecting install options 154 with independent ASP and Service overflowed 72
steps 149 Tools Network Interface 102 no data loss 65
using distribution media 148 device parity protection 91 some data loss 66
overflowed user ASP 191, 192, 193 independent ASP mirrored protection 90
reply list entries 213 complete data loss 94 non-load source unit
restoring no data loss 93 complete data loss, no user
preparation 122 some data loss 93 ASP 78
starting 122 load source unit complete data loss, user ASP not
steps 123 complete data loss, no user overflowed 79
using distribution media 123 ASP 67 complete data loss, user ASP
service attributes 156 complete data loss, user ASP not overflowed 82
System/36 environment 230 overflowed 68 no data loss 76
system information 156, 213 complete data loss, user ASP some data loss 77
system management objects 156 overflowed 72 system ASP
system reply list 157 no data loss 65 complete data loss, no user
system values 156, 213 some data loss 66 ASP 78
tape controller 228 mirrored protection 90 complete data loss, user ASP not
text index search files 260 non-load source unit overflowed 79
text search services 260 complete data loss, no user complete data loss, user ASP
unsuccessful restore operation 56 ASP 78 overflowed 82
user ASP complete data loss, user ASP not removing failed unit 92
addressability 182 overflowed 79 some data loss 77
after system ASP 182 complete data loss, user ASP user ASP
overflowed 191, 192, 193 overflowed 82 complete data loss, not
stand-alone 196 no data loss 76 overflowed 86
user information some data loss 77 complete data loss, overflowed 88
choosing procedure 107 system ASP no data loss 76
using Operational Assistant complete data loss, no user some data loss 86
backup 117 ASP 78 user information
using Restore menu option 21 112 complete data loss, user ASP not using commands 108
using Restore menu options 22 and overflowed 79 using Restore menu option 21 112
23 114 complete data loss, user ASP using Restore menu options 22 and
using changed objects 279 overflowed 82 23 114
using journaling 279 removing failed unit 92 recovery steps 871
using SAVSTG (save storage) some data loss 77 complete system loss 96
media 311 user ASP including independent ASP 98
Windows server 270 complete data loss, not with independent ASP and Service
recovery overflowed 86 Tools Network Interface 102
commitment control complete data loss, overflowed 88 device parity protection 91
considerations and no data loss 76 independent ASP
restrictions 591 some data loss 86 complete data loss 94
common terminology 61 user information no data loss 93
considerations for mirrored using commands 108 some data loss 93
protection 297 using Restore menu option 21 112
damaged objects 174

Index 919
recovery steps 871 (continued) remote journal function (continued) Remove Journaled Changes
load source unit adding remote journal 474 (RMVJRNCHG) command (continued)
complete data loss, no user application dependent data 472 overview 428
ASP 67 asynchronous replication 481 referential constraints 445
complete data loss, user ASP not asynchronously maintained 474 trigger programs 446
overflowed 68 auxiliary storage considerations 494 two journal receivers 388
complete data loss, user ASP backup system 467 Remove Remote Journal
overflowed 72 benefits 467 (QjoRemoveRemoteJournal) API 479,
no data loss 65 catch-up 482 490
some data loss 66 caught-up 481 Remove Remote Journal (RMVRMTJRN)
mirrored protection 90 communications 470 command 490
non-load source unit communications protocols 470 removing
complete data loss, no user configuration examples 468 disk unit
ASP 78 confirmed journal entries 508 from ASP 678
complete data loss, user ASP not data replication 494 disk units 655, 656, 657, 658, 659
overflowed 79 data replication environment 494 failed disk unit 202
complete data loss, user ASP database resynchronization 497 failed unit
overflowed 82 delivery mode 472 system ASP 92
no data loss 76 displaying information 498 internal journal entries 407
some data loss 77 hot-backup 467 journaled changes 444
system ASP hot-backup application 495 overview 428
complete data loss, no user hot-backup application apply 495 referential constraints 445
ASP 78 hot-backup environment 497 trigger programs 446
complete data loss, user ASP not inactivating replication of journal journaled member 423
overflowed 79 entries 485 Rename Directory Entry (RNMDIRE)
complete data loss, user ASP introduction 467 command
overflowed 82 IPL considerations 509 restoring mail 260
removing failed unit 92 journal attributes 478 Rename Document Library Object
some data loss 77 journal entry latency 481 (RNMDLO) command
user ASP journal state 488 restoring documents 260
complete data loss, not library redirection 473, 475 renaming
overflowed 86 local journal database file
complete data loss, overflowed 88 activating 487 during restore 238
no data loss 76 inactivating 487 directory entry
some data loss 86 local system 470 restoring mail 260
user information performance considerations 492 document library object
using commands 108 planning 472 restoring documents 260
using Restore menu option 21 112 primary system 467, 470 reorganize physical file member
using Restore menu options 22 and receiver configuration 511 when applying journaled
23 114 recovery example 511 changes 455
recovery strategy remote journal reply list
disk failure 63 activating 479 restoring 157
human error 63 adding 474 reply list entry
power failure 62 inactivating 479 recovering 213
program failure 63 removing 490 reset (RST) state 565
selecting 61 remote journal network 469 resetting
system failure 63 remote journal type descriptions 476 job number counter 155
recovery time save and restore considerations 500 journal
access path 375 setting up 473 overflowed status 691
redundant failure disk unit status 668 source journal 470 overflowed user ASP 192, 193
referential constraint source system 469 resource
journaling 389 switch-back 487 placing under commitment
pending switch-over 487 control 541
editing during IPL 173 synchronous replication 480 resource, hardware
Receive Journal Entry (RCVJRNE) synchronously maintained 474 definition 667
command 432 target system 469 resource manager
restoring 245 troubleshooting 491 definition 526
release-to-release support 323 unconfirmed journal entries 508 resource not detected status
remote journal working with journal entries 505 correcting 228
activating 479 remote journal type Restore (RST) command
adding 474 descriptions 476 changed objects 281
attaching a new receiver 486 Remove Commitment Control Resource command information output
inactivating 479, 485 (QTNRMVCR) API 541 format 858
removing 490 Remove Journaled Changes directory information output
remote journal function (RMVJRNCHG) command format 859
activating replication of journal example 451 FRCOBJCVN (force object conversion)
entries 480 journal code actions 452 parameter 254

920 OS/400 Backup and Recovery V5R1

Restore (RST) command (continued) restoring restoring (continued)
how to use 261 *ALLOBJ (all object) special files being journaled 235
object link information output authority 355, 359, 361, 363 folders
format 860 *ALLOBJ (all-object) special overview 255
output header format 858 authority 217 procedure 258
output information 857 access path recovery times 156 hard link 262
output sequence 857 access paths 242 IFS objects being journaled 235
restrictions 275 allowing sensitive programs 50 into damaged folder 258
restrictions when restoring ALWOBJDIF (allow object differences) journal
documents 277 parameter 218, 219 journal receiver name 400
trailer information output format 861 authority journal receivers 248, 422
Restore Authority (RSTAUT) document library object 259 journals 248, 422
command 219 authority holders 217 libraries
non-restricted state system 220 authorization list link 218 *ALLUSR libraries 232
Restore Configuration (RSTCFG) authorization lists 217 *IBM libraries 232
command 227 changed objects *NONSYS libraries 232
Restore Document Library Object by library 280 OPTION parameter 232
(RSTDLO) command by object 280 overview 231
maximum number of DLOs 257 cumulative 280 library
media error 57 not cumulative 280 media error 56
output 256 changing console type 230 user ASP 198
overview 255 configuration 32, 227 Licensed Internal Code
renaming document 258 problems with system resource preparation 122
restoring authority 259 management (SRM) starting 122
restoring descriptive information 259 information 228 steps 123
restoring ownership 259 considerations when restoring NWSD using distribution media 123
user ASP 200 and storage spaces 270 licensed programs 255
Restore Library (RSTLIB) command correct sequence 45 link 262
*ALLUSR libraries 232 damaged document 257 list of members 241
*IBM libraries 232 database file locking objects 53
*NONSYS libraries 232 ALWOBJDIF (allow object logical files 242
FRCOBJCVN (force object conversion) difference) parameter 238 logical partitions 231
parameter 254 creation date 238 mail 258
media error 56 different member set 241 member
multiple concurrent 233 how to 236 renaming 238
OPTION parameter 232 MAXMBRS (maximum members) message reply list 157
overview 231 parameter 240 NWSD 273
user ASP 198 MBROPT (member option) object
Restore Licensed Program (RSTLICPGM) parameter 241 how to 234
command 255 member locking 238 multiple names 262
restore menu renaming 238 object in directory 32
option 21 (entire system) 208 determining tape contents 307 objects in directories 261
option 22 (system data only) 208 different system objectss being journaled 235
option 23 (all user data) 208 network attributes 162 OPM (Original Program Model)
Restore menu distribution objects 258 objects 254
commands run by menu options 208 DLO (document library object) OS/400 Enhanced Integration for
using 207 maximum number 257 Novell NetWare information 264
Restore Object (RSTOBJ) command 234 DLOs (document library objects) OS/400 licensed program
FRCOBJCVN (force object conversion) overview 255 choosing procedure 148
parameter 254 document library object manual IPL 149
multiple concurrent 234 descriptive information 259 overview 147
restore operation media error 57 preparation 147
overview 3 document library object (DLO) reasons 147
restore operations renaming document 258 selecting install options 154
parallel 59 user ASP 200 steps 149
restore sequence document library objects (DLOs) using distribution media 148
naming conventions 403 overview 255 ownership
restore strategy documents document library object 259
disk failure 63 media error 57 parts of system 40
human error 63 overview 255 predefined storage spaces created on
power failure 62 Domino server 265 pre-V4R5 systems 271
program failure 63 edit descriptions 157 predefined storage spaces created on
selecting 61 entire system 208 V4R5 and later systems 271
system failure 63 unattended 209 program
Restore User Profiles (RSTUSRPRF) error is not recoverable 56 different release 253
command 214 error is recoverable 56 program temporary fixes (PTF) 278
file with trigger program 247 programs 252

Index 921
restoring (continued) Retrieve Object Description (QUSROBJD) RSTDLO (Restore Document Library
PTF (program temporary fixes) 278 API 387 Object) command
QAPZ files 54 retrieving maximum number of DLOs 257
QGPL (general purpose) library journal entry media error 57
QAPZ files 54 overview 428 output 256
QNetWare file system 264 using 436 overview 255
QUSRSYS (user system) library RMVJRNCHG (Remove Journaled renaming document 258
QAPZ files 54 Changes) command restoring authority 259
referential constraints 245 example 451 restoring descriptive information 259
related objects 45 journal code actions 452 restoring ownership 259
save file data 255 overview 428 user ASP 200
security considerations 50 referential constraints 445 RSTLIB (Restore Library) command
security information trigger programs 446 *ALLUSR libraries 232
object authorities 219 two journal receivers 388 *IBM libraries 232
object ownership 218 RMVRMTJRN (Remove Remote Journal) *NONSYS libraries 232
ownership 218 command 490 FRCOBJCVN (force object conversion)
primary group 218 RNMDIRE (Rename Directory Entry) parameter 254
private authorities 219 command media error 56
sequence 213 restoring mail 260 multiple concurrent 233
user profiles 214 RNMDLO (Rename Document Library OPTION parameter 232
service attributes 156 Object) command overview 231
shared formats 245 restoring documents 260 user ASP 198
soft link 262 role RSTLICPGM (Restore Licensed Program)
storage agent 548 command 255
resuming 318 commit processing 548 RSTOBJ (Restore Object) command 234,
symbolic link 262 coordinator 548 271
system information 156 initiator 548 FRCOBJCVN (force object conversion)
system management objects 156 last agent 549 parameter 254
system reply list 157 participant 548 multiple concurrent 234
system values 156 rollback operation RSTUSRPRF (Restore User Profiles)
unit 299 forced 565 command 214
unsuccessful 56 implicit 573 RTVJRNE (Retrieve Journal Entry)
user-defined storage spaces 272 language support 548 command 386
user profile overview 547 overview 428
different system 217 what the system does 550 using 436
procedure 214 Rollback Required (QTNRBRQD) using in program 766
using Restore menu 207, 208 API 535
verifying success 54 rollback required (RBR) state 565
Windows server 270
restoring logical partitions 231
routing step end
commitment control during 581
S
S/36 environment
restricted state RST (reset) state 565
recovering 230
definition 45 RST (Restore) command
SAV (Save) command
starting 45 changed objects 281
command information output
restrictions command information output
format 858
availability and recovery format 858
directory information output
commitment control 591 directory information output
format 859
Resulting Capacity display 677 format 859
object link information output
resuming FRCOBJCVN (force object conversion)
format 860
mirror protection 297 parameter 254
output header format 858
mirrored unit 291 how to use 261
output information 857
restore storage 318 object link information output
output sequence 857
resuming status 668 format 860
trailer information output format 861
resynchronization output header format 858
SAVACTWAIT (save-active wait)
definition 564 output information 857
parameter
retranslation 252, 253 output sequence 857
commitment control 583
Retrieve Commitment Information restrictions 275
Save (SAV) command
(QTNRCMTI) API 564 restrictions when restoring
command information output
Retrieve Journal Entries documents 277
format 858
(QjoRetrieveJournalEntries) API 386 trailer information output format 861
directory information output
Retrieve Journal Entry (RTVJRNE) RST (Restore Object) command 271, 272
format 859
command 386 RSTAUT (Restore Authority)
object link information output
database recovery 386 command 219
format 860
overview 428 non-restricted state system 220
output header format 858
using 436 RSTCFG (Restore Configuration)
output information 857
using in program 766 command 227, 273
output sequence 857
Retrieve Journal Identifier Information
trailer information output format 861
(QJORJIDI) API 386

922 OS/400 Backup and Recovery V5R1

save-active wait (SAVACTWAIT) saving (continued) Send Journal Entry (SNDJRNE)
parameter object in directory 32 command 386, 429
commitment control 583 options with Operational Send Network File (SNDNETF)
save completion messages Assistant 890 command 764
retrieving device name 765 save file data sending
save file media considerations 761 journal entry 429
clearing 762, 764 schedule with Operational network files 764
considerations for using 761 Assistant 891 save files 764
damage 764 status messages 766 sensitive object
determining the contents 761 unit 299 allowing restore 50
displaying 761 using Operational Assistant 885 SEQONLY (sequential processing)
file-dependent attributes 764 saving and restoring journaling 395
input operations 763 document library objects 32 sequence
opening 763 SAVLIB (Save Library) command commitment control journal
output operations 763 determining what command was entries 545
performance when using 761 used 307 restoring objects 45
preparing for use 761 SAVRST (Save/Restore) command 32 sequential processing (SEQONLY)
security 762 SAVRSTCFG (Save/Restore journaling 395
sending 764 Configuration) command 32 service attribute
storage capacity 761 SAVRSTCHG (Save/Restore Changed restoring 156
using control language (CL) Objects) command 32 service representative
commands 762 SAVRSTDLO (Save/Restore Document mirrored protection recovery
save file data Library Object) command 32 action 296
restoring 255 SAVRSTLIB (Save/Restore Library) set major system options
saving 761 command 32 during recovery 160
Save Library (SAVLIB) command SAVRSTOBJ (Save/Restore Object) Set Major System Options display 160
determining what command was command 32 setting
used 307 SAVSAVFDTA (Save Save File Data) major system options
Save Object (SAVOBJ) command command during recovery 160
media considerations 762 save file considerations 761 setting up
save operation SAVSTG (save storage) media journaling 398
introduction 3 using in recovery 311 ObjectConnect 30
overview 3 SAVSYS (Save System) media size
Save/Restore (SAVRST) command 32 definition 122 journal receiver 401
Save/Restore Changed Objects scope SMAPP (system-managed access-path
(SAVRSTCHG) command 32 activation group level 534 protection)
Save/Restore Configuration commitment definition 534 overview 5
(SAVRSTCFG) command 32 definition 534 SNADS (QSNADS) journal 424
Save/Restore Document Library Object job level 534 SNDJRNE (Send Journal Entry)
(SAVRSTDLO) command 32 screens command 386, 429
Save/Restore Library (SAVRSTLIB) LIC error 775 SNDNETF (Send Network File)
command 32 licensed internal code installation command 764
Save/Restore Object (SAVRSTOBJ) error 775 SNMP (QSNMP) journal 424
command 32 sector damage 64 soft link
Save Save File Data (SAVSAVFDTA) security restoring 262
command save file 762 space, disk
description 762 security information calculating 681
save file considerations 761 components 213 spare nonconfigured unit
save storage (SAVSTG) media restoring 214 using 294
using in recovery 311 sequence 213 special authority
save-while-active function sequence *ALLOBJ (all-object)
commitment control during 583 restoring 213 restoring 217
using with journaling 394 security-relevant object Specify ASP to Move Disk Units
saving allowing restore operation 50 display 676
changed objects 32 Select ASP to Delete Data From Specify Install Options display 155
Operational Assistant backup 889 display 203 Specify Restore Options display 156
completion message Select DST Console Mode display 140 SQL server mode 535
retrieving device name 765 Select Product to Work with PTFs SRC (system reference code)
configuration 32 display 159, 170 A6xx 0277 729
determining tape contents 307 selecting A900 2000
journal receivers 420 install options recovery 164
journals 420 restoring operating system 154 restoring Licensed Internal Code
library 32 Send Data Queue (QSNDDTAQ) using 23 127
determining what command was API 386 SRM (system resource management)
used 307 Send Journal Entry (QJOSJRNE) information
object 32 API 386, 429, 535 correcting problems 228
previous release system 323

Index 923
SST (system service tools) starting system STRCMTCTL (Start Commitment Control)
definition 62 Disk Configuration Error Report command (continued)
options 660 display 168 notify object (NFYOBJ)
starting 662 Work with Current Main Storage parameter 530
stopping 662 Dump display 168 omit journal entries (OMTJRNE)
standard application program state (logical unit of work) parameter 540
example 615 CIP (commit in progress) 565 stream file
standard commit processing program CMT (committed) 565 output from RST (Restore)
notify objects 618 commit in progress (CIP) 565 command 857
Start Commitment Control committed (CMT) 565 output from SAV (Save)
(STRCMTCTL) command LAP (last agent pending) 565 command 857
commit scope (CMTSCOPE) last agent pending (LAP) 565 stream file, IFS
parameter 533 logical unit of work (LUW) 564 ENDJRN (End Journal)
commit text (TEXT) parameter 539 PIP (prepare in progress) 565 command 423
default journal (DFTJRN) prepare in progress (PIP) 565 Start Journal (STRJRN)
parameter 539 prepared (PRP) 565 command) 413
description 527 PRP (prepared) 565 STRJRN (Start Journal) command
LCKLVL (lock level) parameter 527 RBR (rollback required) 565 omitting open and close
notify object (NFYOBJ) reset (RST) 565 operations 413
parameter 530 rollback required (RBR) 565 using 413
omit journal entries (OMTJRNE) RST (reset) 565 STRJRNAP (Start Journal Access Path)
parameter 540 vote-read-only (VRO) 565 command
Start Journal (STRJRN) command 387 VRO (vote-read-only) 565 description 387, 412
omitting open and close status STRJRNPF (Start Journal Physical File)
operations 413 auxiliary storage pool (ASP) 667 command
using 413, 414 disk description 387
Start Journal Access Path (STRJRNAP) understanding 663 omitting open and close
command disk unit 667 operations 411
description 387, 412 display journal status 459 using 411
Start Journal Object (STRJRNOBJ) unknown load source 301 subsystem
command 387 stopping ending
using 414 dedicated service tools (DST) 662 QCALSRV (calendar server)
Start Journal Physical File (STRJRNPF) device parity protection 701 subsystem 45
command mirrored protection 723 QSYSWRK (subsystem monitor)
description 387 system service tools 662 subsystem 45
omitting open and close storage restricted state 45
operations 411 capacity using 45
using 411 save file 761 subsystem monitor (QSYSWRK)
start journaling reclaiming subsystem
for DB2 multisystem files 412 duplicate names in QRCL 47 ending 45
starting procedure 46, 183 suspended disk unit status 668
commitment control QALWUSRDMN (allow user suspended status 668
commit scope (CMTSCOPE) domain objects) system suspending
parameter 533 value 48 mirrored units 290
commit text (TEXT) recovering user ASP 183 symbolic link
parameter 539 user domain object 48 restoring 262
default journal (DFTJRN) what the system does 47 symbolic link, IFS
parameter 539 why to run 176 ENDJRN (End Journal)
notify object (NFYOBJ) unit command 423
parameter 530 not operational 297 Start Journal (STRJRN)
omit journal entries (OMTJRNE) storage capacity command) 413
parameter 540 save file 761 sync-point tree 548
parameters to specify 527 storage unit synchronization
dedicated service tools (DST) 660 not operational 297 BRMS 364
device strategy recovery considerations 297
during recovery 160 batch job recovery 760 synchronizing
device parity protection 695, 699 interactive job recovery 759 system
9337 Disk Array Subsystem 695 job recovery 759 methods overview 352
journaling STRCMTCTL (Start Commitment Control) planning and procedures 351
access path 412 command system
mirrored protection 719 commit scope (CMTSCOPE) parts 40
printer writer parameter 533 System/36 environment
during recovery 159 commit text (TEXT) parameter 539 during recovery 160
system default journal (DFTJRN) recovering 230
after abnormal end 167 parameter 539 system accounting journal entry 806
system service tools 662 description 527 system ASP
LCKLVL (lock level) parameter 527 definition 7

924 OS/400 Backup and Recovery V5R1

system ASP (auxiliary storage pool) system value (continued) transferring (continued)
assigning files and journals 392 QALWUSRDMN (allow user domain user profile
definition 62, 393 objects) 48 different system 217
recovery procedures QFRCCVNRST (force object translation 253
complete data loss, no user conversion on restore) 253 trigger
ASP 78 QPWRRSTIPL (automatic IPL after restoring 247
complete data loss, user ASP not power restored) 167 trigger program
overflowed 79 QVFYOBJRST (verify object on journaling 389
complete data loss, user ASP restore) 50 Receive Journal Entry (RCVJRNE)
overflowed 82 recovering 213 command 432
removing failed unit 92 restoring 156 restoring 247
some data loss 77 two-phase commit
removing failed unit 92 agent role 548
system change-journal management 405
resetting sequence numbers 417
T committed wave 541
description 540
tape
system data initiator role 548
save
restoring 208 last agent role 549
determining what command was
system-deleted journal receivers 406 prepare wave 540
used 307
system failure roles 548
tape configuration
recovery strategy 63 transaction program network 548
creating
system information vote 540
for 34xx tape units 165, 342, 879
recovering 213 type
for non-34xx tape units 166
restoring 156 journal entry 807
tape controller
system-managed access-path protection
recovering after restore 228
(SMAPP)
tape unit
access path journaling 376
benefits 376
2440 U
enabling high-speed feature 145 unattended restore operation 209
choosing which access paths 377
target release (TGTRLS) parameter unbroken receiver chain
overview 5, 375
valid values 323 applying journaled changes 284
performance 377
temporary error 289 uninterruptible power supply
working with 378
terminology overview 9
system management object
recovery 61 role in backup and recovery 9
restoring 156
TEXT (commit text) parameter 539 unit
system reference code (SRC)
text index search files mirrored
A900 2000
recovering 260 resuming 291
recovery 164
text search services suspending 290
restoring Licensed Internal Code
recovering 260 missing mirrored disk 298
using 23 127
TGTRLS (target release) parameter not operational storage 297
system reply list
valid values 323 restoring 299
restoring 157
thread scoped transactions 535 saving 299
system resource management (SRM)
threshold spare nonconfigured 294
information
auxiliary storage pool (ASP) unit number
correcting problems 228
changing 672 definition 667
system service tools (SST)
journal receiver 401, 404 unprotected status 668
definition 62
system auxiliary storage pool (ASP) unreadable sectors 64
options 660
changing 673 recovery 174
starting 662
time-out disk error 297 unrecoverable device error 297
stopping 662
trailer information format unrecoverable error
system state program
RST (Restore) command 861 restore operation 56
allowing restore operation 50
SAV (Save) command 861 unsuccessful
system status display
transaction restore operation 56
auxiliary storage
definition 525 UPS 9
high percentage used 176
determining size 584 Use Dedicated Service Tools (DST)
system value
logging file example 602 display 140
allow object restore operation
transaction manager user ASP
(QALWOBJRST) 50
definition 526 definition 7
allow user domain objects
transaction program network 548 user ASP (auxiliary storage pool)
(QALWUSRDMN) 48
transferring assigning files and journals 392
automatic IPL after power restored
existing journals into a user ASP 687 definition 62, 393
(QPWRRSTIPL) 167
folder determining overflowed status 192
changing
different ASP 687 journal receiver 393
during recovery 160
library overflowed
force object conversion on restore
different ASP 687 deleting objects 196
(QFRCCVNRST) 253
object recovering 191
QALWOBJRST (allow object restore
different ASP 693 resetting 192, 193
operation) 50
objects between ASPs 686 recovering 196

Index 925
user ASP (auxiliary storage pool) vote-read-only (VRO) state 565
(continued) VRO (vote-read-only) state 565
recovery procedure
load source unit loss, not
overflowed 68
load source unit loss,
W
Wait for Outcome option 554
overflowed 72
what to save 888
recovery procedures
Windows server
complete data loss, not
recovering 270
overflowed 86
Windows server files
complete data loss, overflowed 88
recovering 273
no data loss 76
work station
some data loss 86
output for journal entries 430
user auxiliary storage pool (ASP)
Work with Current Main Storage Dump
adding disk units 669
display 168
calculating space requirements 681
Work with Journal (WRKJRN) command
changing threshold 672, 673
display journal status option 459
creating 669
journal functions 386
creating document library objects
options 455
(DLOs) 689
recovery options 456
creating objects 688, 692
Work with Journal Attributes
deleting 654, 680
(WRKJRNA) command 386, 442
displaying objects 684
Work with Object Links (WRKLNK)
journal receivers 689
command 387
moving disk unit 676
working with
removing disk unit 678
access path recovery time 378
transferring objects 686
device parity protection 695
user-created journal entry 429
Disk Configuration Error Report
user data
display 168
restoring 208
journal attributes 442
User-Defined File Systems
mirrored protection 719
restoring 187
nonlibrary user ASPs 692
user domain object
receiver directory 442
reclaiming 48
Work with Current Main Storage
user-generated entry 807
Dump display 168
user information
write protected disk unit status 667
recovering
writing
choosing procedure 107
output using RCVJRNE (Receiver
using commands 108
Journal Entry) command 769
using Operational Assistant
WRKJRN (Work with Journal) command
backup 117
display journal status 459
user profile
journal functions 386
*ALLOBJ (all-object) special authority
options 455
restoring 217
recovery options 456
IBM-supplied
WRKJRNA (Work with Journal
damaged 175
Attributes) command 386, 442
moving to different system 217
restoring 214
user space
output from RST (Restore) X
command 857 XA transaction support 566
output from SAV (Save)
command 857
using
journal entries 427

V
validation value 252
variable-length portion
journal entry 821
verify object on restore (QVFYOBJRST)
system value 50
verifying
successful restore 54
vote
definition 540

926 OS/400 Backup and Recovery V5R1

Readers’ Comments — We’d Like to Hear from You
iSeries
Backup and Recovery
Version 5

Publication No. SC41-5304-05

Overall, how satisfied are you with the information in this book?

Very Satisfied Satisfied Neutral Dissatisfied Very

Dissatisfied
Overall satisfaction h h h h h

How satisfied are you that the information in this book is:

Very Satisfied Satisfied Neutral Dissatisfied Very

Dissatisfied
Accurate h h h h h
Complete h h h h h
Easy to find h h h h h
Easy to understand h h h h h
Well organized h h h h h
Applicable to your tasks h h h h h

Please tell us how we can improve this book:

Thank you for your responses. May we contact you? h Yes h No

When you send comments to IBM, you grant IBM a nonexclusive right to use or distribute your comments in any
way it believes appropriate without incurring any obligation to you.

Name Address

Company or Organization

Phone No.
 ___________________________________________________________________________________________________
Readers’ Comments — We’d Like to Hear from You Cut or Fold
SC41-5304-05
 Along Line

_ _ _ _ _ _ _Fold
_ _ _and
_ _ _Tape
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _Please
_ _ _ _ _do
_ _not
_ _ staple
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _Fold
_ _ _and
_ _ Tape
______

NO POSTAGE
NECESSARY
IF MAILED IN THE
UNITED STATES

BUSINESS REPLY MAIL

FIRST-CLASS MAIL PERMIT NO. 40 ARMONK, NEW YORK

POSTAGE WILL BE PAID BY ADDRESSEE

IBM CORPORATION
ATTN DEPT 542 IDCLERK
3605 HWY 52 N
ROCHESTER MN 55901-7829

_________________________________________________________________________________________
Fold and Tape Please do not staple Fold and Tape

Cut or Fold
SC41-5304-05 Along Line



Printed in the United States of America

on recycled paper containing 10%
recovered post-consumer fiber.

SC41-5304-05
Spine information:

 iSeries OS/400 Backup and Recovery V5R1 Version 5