Samba3 Howto 1
Samba3 Howto 1
Reference Guide
v
vi Attribution
105
<mailto:[email protected]>
106
<mailto:[email protected]>
107
<mailto:[email protected]>
108
<mailto:[email protected]>
109
<mailto:[email protected]>
110
<mailto:[email protected]>
111
<mailto:[email protected]>
PREFACE
The editors wish to thank you for your decision to purchase this book. The
Official Samba-3 HOWTO and Reference Guide is the result of many years
of accumulation of information, feedback, tips, hints, and happy solutions.
Please note that this book is a living document, the contents of which are
constantly being updated. We encourage you to contribute your tips, tech-
niques, helpful hints, and your special insight into the Windows networking
world to help make the next generation of this book even more valuable to
Samba users.
We have made a concerted effort to document more comprehensively than
has been done previously the information that may help you to better deploy
Samba and to gain more contented network users.
This book provides example configurations, it documents key aspects of Mi-
crosoft Windows networking, provides in-depth insight into the important
configuration of Samba-3, and helps to put all of these into a useful frame-
work.
The most recent electronic versions of this document can be found at <http:
//www.samba.org/> on the “Documentation” page.
Updates, patches and corrections are most welcome. Please email your
contributions to any one of the following:
Jelmer Vernooij ([email protected])112
John H. Terpstra ([email protected])113
Gerald (Jerry) Carter ([email protected])114
We wish to advise that only original and unencumbered material can be
published. Please do not submit content that is not your own work unless
proof of consent from the copyright holder accompanies your submission.
Conventions Used
xv
xvi Preface
When John first asked me to write an introductory piece for his latest book,
I was somewhat mystified as to why he chose me. A conversation with John
provided some of the rationale, and he left it to me to fill in the rest of
the story. So, if you are willing to endure a little bit of background, I will
provide the part of the story that John wouldn’t provide.
I am the Director of Corporate Standards at Sun Microsystems, and man-
age Sun’s standards portfolio. Before that, I was the Director of Standards
at Netscape, which was when I met John. Before Sun, there was Digital
Equipment Corporation, also standards. I’ve written several books on stan-
dards, and tend to observe (and occasionally help) the technical and business
trends that drive standardization as a discipline. I tend to see standardiza-
tion as a management tool, not as a technical discipline and this is part of
the rationale that John provided.
The book that you have before you focuses on a particular standardized
way of doing something hence, it is a book about a standard. The most
important thing to keep in mind about a standard is the rationale for its
creation. Standards are created not for technical reasons, not for business
reasons, but for a deeper and much more compelling reason. Standards
are created and used to allow people to communicate in a meaningful way.
Every standard, if it is a true standard, has as its entire (and only) goal set
the increasing of relevant communication between people.
This primary goal cannot be met however, unless the standard is docu-
mented. I have been involved in too many standardization efforts when it
became apparent that everybody knows was the dominant emotion of those
providing documentation. They of the ever present they say and they know
are the bane of good standards. If they know, why are you doing a standard?
A good standard survives because people know how to use it. People know
how to use a standard when it is so transparent, so obvious, and so easy that
it become invisible. And a standard becomes invisible only when the doc-
umentation describing how to deploy it is clear, unambiguous, and correct.
These three elements must be present for a standard to be useful, allowing
communication and interaction between two separate and distinct entities
xvii
xviii Foreword
to occur without obvious effort. As you read this book, look for the evidence
of these three characteristics and notice how they are seamlessly woven into
John’s text. Clarity and unambiguity without correctness provide a tech-
nical nightmare. Correctness and clarity with ambiguity create maybe bits,
and correctness and unambiguity without clarity provide a muddle through
scenario.
And this is the rest of the story that John couldn’t (or wouldn’t) bring him-
self to state. This book provides a clear, concise, unambiguous, and tech-
nically valid presentation of Samba to make it useful to a user to someone
who wants to use the standard to increase communication and the capability
for communication between two or more entities whether person-machine,
machine-machine, or person-person. The intent of this book is not to con-
vince anyone of any agenda political, technical, or social. The intent is to
provide documentation for users who need to know about Samba, how to
use it, and how to get on with their primary responsibilities. While there
is pride on John’s part because of the tremendous success of the Samba
documentation, he writes for the person who needs a tool to accomplish a
particular job, and who has selected Samba to be that tool.
The book is a monument to John’s perseverance and dedication to Samba
and in my opinion to the goal of standardization. By writing this book, John
has provided the users of Samba those that want to deploy it to make things
better a clear, easy, and ultimately valuable resource. Additionally, he has
increased the understanding and utility of a highly useful standard, and for
this, as much as for the documentation, he is owed a debt of gratitude by
those of us who rely on standards to make our lives more manageable.
Carl Cargill, Senior Director
Corporate Standardization, The Office of the CTO
Sun Microsystems
CONTENTS
Contents
ATTRIBUTION v
PREFACE xv
FOREWORD xvii
xix
xx Contents
SAMBA 620
28.1 Features and Benefits 620
28.2 Background Information 621
28.3 Name Resolution in a Pure UNIX/Linux World 621
28.3.1 /etc/hosts 622
28.3.2 /etc/resolv.conf 623
28.3.3 /etc/host.conf 623
28.3.4 /etc/nsswitch.conf 624
28.4 Name Resolution as Used within MS Windows Networking 625
28.4.1 The NetBIOS Name Cache 627
28.4.2 The LMHOSTS File 627
28.4.3 HOSTS File 629
28.4.4 DNS Lookup 629
28.4.5 WINS Lookup 630
28.5 Common Errors 630
28.5.1 Pinging Works Only One Way 631
28.5.2 Very Slow Network Connections 631
28.5.3 Samba Server Name-Change Problem 631
PDC 666
34.1 Planning and Getting Started 666
34.1.1 Objectives 666
34.1.1.1 Domain Layout 668
34.1.1.2 Server Share and Directory Layout 669
34.1.1.3 Logon Scripts 669
34.1.1.4 Profile Migration/Creation 670
34.1.1.5 User and Group Accounts 670
34.1.2 Steps in Migration Process 670
34.2 Migration Options 671
34.2.1 Planning for Success 672
34.2.2 Samba-3 Implementation Choices 672
GLOSSARY 749
Index 786
List of Figures
4 Domain Control
4.1 An Example Domain. 55
9 Network Browsing
9.1 Cross-Subnet Browsing Example. 169
xlv
xlvi LIST OF FIGURES
6 Domain Membership
6.1 Assumptions 102
9 Network Browsing
9.1 Browse Subnet Example 1 170
9.2 Browse Subnet Example 2 170
9.3 Browse Subnet Example 3 171
9.4 Browse Subnet Example 4 171
xlviii
LIST OF TABLES xlix
29 Unicode/Charsets
29.1 Japanese Character Sets in Samba-2.2 and Samba-3 641
38 Reporting Bugs
38.1 Debuggable Functions 707
PREFACE AND
INTRODUCTION
“A man’s gift makes room for him before great men. Gifts are like hooks
that can catch hold of the mind taking it beyond the reach of forces that
otherwise might constrain it.” — Anon.
This is a book about Samba. It is a tool, a derived work of the labors of
many and of the diligence and goodwill of more than a few. This book
contains material that has been contributed in a persistent belief that each
of us can add value to our neighbors as well as to those who will follow us.
This book is designed to meet the needs of the Microsoft network adminis-
trator. UNIX administrators will benefit from this book also, though they
may complain that it is hard to find the information they think they need.
So if you are a Microsoft certified specialist, this book should meet your
needs rather well. If you are a UNIX or Linux administrator, there is no
need to feel badly — you should have no difficulty finding answers to your
current concerns also.
What Is Samba?
Samba is a big, complex project. The Samba project is ambitious and excit-
ing. The team behind Samba is a group of some thirty individuals who are
spread the world over and come from an interesting range of backgrounds.
This team includes scientists, engineers, programmers, business people, and
students.
Team members were drawn into active participation through the desire to
help deliver an exciting level of transparent interoperability between Mi-
crosoft Windows and the non-Microsoft information technology world.
The slogan that unites the efforts behind the Samba project says: Samba,
Opening Windows to a Wider World! The goal behind the project is one of
removing barriers to interoperability.
li
lii Preface and Introduction
Samba provides file and print services for Microsoft Windows clients. These
services may be hosted off any TCP/IP-enabled platform. The original
deployment platforms were UNIX and Linux, though today it is in common
use across a broad variety of systems.
The Samba project includes not only an impressive feature set in file and
print serving capabilities, but has been extended to include client function-
ality, utilities to ease migration to Samba, tools to aid interoperability with
Microsoft Windows, and administration tools.
The real people behind Samba are users like you. You have inspired the
developers (the Samba Team) to do more than any of them imagined could
or should be done. User feedback drives Samba development. Samba-3
in particular incorporates a huge amount of work done as a result of user
requests, suggestions and direct code contributions.
There is admittedly a large number of Samba books on the market today and
each book has its place. Despite the apparent plethora of books, Samba as
a project continues to receive much criticism for failing to provide sufficient
documentation. Samba is also criticized for being too complex and too
difficult to configure. In many ways this is evidence of the success of Samba
as there would be no complaints if it was not successful.
The Samba Team members work predominantly with UNIX and Linux, so it
is hardly surprising that existing Samba documentation should reflect that
orientation. The original HOWTO text documents were intended to provide
some tips, a few golden nuggets, and if they helped anyone then that was just
wonderful. But the HOWTO documents lacked structure and context. They
were isolated snapshots of information that were written to pass information
on to someone else who might benefit. They reflected a need to transmit
more information that could be conveniently put into manual pages.
Those of you who have dedicated your labors to the production of unoffi-
cial HOWTOs, to Web page information regarding Samba, or to answering
questions on the mailing lists or elsewhere, may be aware that this is a la-
bor of love. We would like to know about your contribution and willingly
receive the precious pearls of wisdom you have collected. Please email your
contribution to John H. Terpstra ([email protected])115 . As a service to other
users we will gladly adopt material that is technically accurate.
This book, the Official Samba-3 HOWTO and Reference Guide, includes the
Samba-HOWTO-Collection.pdf that ships with Samba. These documents
have been written with a new design intent and purpose.
Over the past two years many Microsoft network administrators have adopted
Samba and have become interested in its deployment. Their information
needs are very different from that of the UNIX administrator. This book
has been arranged and the information presented from the perspective of
someone with previous Microsoft Windows network administrative training
and experience.
115
<mailto:[email protected]>
liv Preface and Introduction
Troubleshooting This short section should help you when all else fails.
Appendix Here you will find a collection of things that are either too
peripheral for most users, or are a little left of field to be included in
the main body of information.
Welcome to Samba-3 and the first published document to help you and
your users to enjoy a whole new world of interoperability between Microsoft
Windows and the rest of the world.
Part I
General Installation
PREPARING SAMBA FOR
CONFIGURATION
1
Chapter 1
Binary packages of Samba are included in almost any Linux or UNIX distri-
bution. There are also some packages available at the Samba home page1 .
Refer to the manual of your operating system for details on installing pack-
ages for your specific operating system.
If you need to compile Samba from source, check Chapter 39, “How to
Compile Samba”.
The smb.conf file uses the same syntax as the various old .ini files in
Windows 3.1: Each file consists of various sections, which are started by
putting the section name between brackets ([]) on a new line. Each contains
1
<http://samba.org/>
2
Section 1.2. Configuring Samba (smb.conf) 3
zero or more key/value pairs separated by an equality sign (=). The file is
just a plaintext file, so you can open and edit it with your favorite editing
tool.
Each section in the smb.conf file represents either a share or a meta-service
on the Samba server. The section [global] is special, since it contains
settings that apply to the whole Samba server. Samba supports a number
of meta-services, each of which serves its own purpose. For example, the
[homes] share is a meta-service that causes Samba to provide a personal
home share for each user. The [printers] share is a meta-service that
establishes print queue support and that specifies the location of the inter-
mediate spool directory into which print jobs are received from Windows
clients prior to being dispatched to the UNIX/Linux print spooler.
Each section of the smb.conf file that specifies a share, or a meta-service, is
called a stanza. The global stanza specifies settings that affect all the other
stanzas in the smb.conf file. Configuration parameters are documented in
the smb.conf man page. Some parameters can be used only in the global
stanza, some only in share or meta-service stanzas, and some can be used
globally or just within a share or meta-service stanza.
Example 1.2.1 contains a very minimal smb.conf.
In the case of Samba there are three daemons, two of which are needed as a
minimum.
nmbd This daemon handles all name registration and resolution requests.
It is the primary vehicle involved in network browsing. It handles all
UDP-based protocols. The nmbd daemon should be the first com-
mand started as part of the Samba startup process.
smbd This daemon handles all TCP/IP-based connection services for file-
and print-based operations. It also manages local authentication. It
should be started immediately following the startup of nmbd.
When Samba has been packaged by an operating system vendor, the startup
process is typically a custom feature of its integration into the platform
as a whole. Please refer to your operating system platform administration
manuals for specific information pertaining to correct management of Samba
startup.
The simplest useful configuration file would contain something like that
shown in Example 1.2.2.
Section 1.2. Configuring Samba (smb.conf) 5
This will allow connections by anyone with an account on the server, using
either their login name or homes as the service name. (Note: The workgroup
that Samba should appear in must also be set. The default workgroup name
is WORKGROUP.)
Make sure you put the smb.conf file in the correct place. Note, the correct
location of this file depends on how the binary files were built. You can
discover the correct location by executing from the directory that contains
the smbd command file:
For more information about security settings for the [homes] share, please
refer to Chapter 17, “Securing Samba”.
It’s important to validate the contents of the smb.conf file using the test-
parm program. If testparm runs correctly, it will list the loaded services. If
not, it will give an error message. Make sure it runs correctly and that the
services look reasonable before proceeding. Enter the command:
Testparm will parse your configuration file and report any unknown param-
eters or incorrect syntax. It also performs a check for common misconfigu-
rations and will issue a warning if one is found.
Always run testparm again whenever the smb.conf file is changed!
6 How to Install and Test SAMBA Chapter 1
The smb.conf file is constantly checked by the Samba daemons smbd and
every instance of itself that it spawns, nmbd and winbindd. It is good
practice to keep this file as small as possible. Many administrators prefer
to document Samba configuration settings and thus the need to keep this
file small goes against good documentation wisdom. One solution that may
be adopted is to do all documentation and configuration in a file that has
another name, such as smb.conf.master. The testparm utility can be used
to generate a fully optimized smb.conf file from this master configuration
and documtenation file as shown here:
1.2.4 SWAT
To launch SWAT, just run your favorite Web browser and point it to <http:
//localhost:901/>. Replace localhost with the name of the computer
on which Samba is running if that is a different computer than your browser.
More information about SWAT can be found in Chapter 35, “SWAT: The
Samba Web Administration Tool”.
Section 1.3. List Shares Available on the Server 7
To list shares that are available from the configured Samba server, execute
the following command:
$ smbclient -L yourhostname
You should see a list of shares available on your server. If you do not, then
something is incorrectly configured. This method can also be used to see
what shares are available on other SMB servers, such as Windows 2000.
If you choose user-level security, you may find that Samba requests a pass-
word before it will list the shares. See the smbclient man page for details.
You can force it to list the shares without a password by adding the option
-N to the command line.
$ smbclient //yourhostname/aservice
Typically yourhostname is the name of the host on which smbd has been
installed. The aservice is any service that has been defined in the smb.
conf file. Try your username if you just have a [homes] section in the smb.
conf file.
Example: If the UNIX host is called bambi and a valid login name is fred,
you would type:
$ smbclient //bambi/fred
8 How to Install and Test SAMBA Chapter 1
Now that Samba is working correctly locally, you can try to access it from
other clients. Within a few minutes, the Samba host should be listed in the
Network Neighborhood on all Windows clients of its subnet. Try browsing
the server from another client or ”mounting” it.
Mounting disks from a DOS, Windows, or OS/2 client can be done by run-
ning a command such as:
The spoolservice is the name of the printer (actually the print queue) on
the target server. This will permit all print jobs that are captured by the
lpt1: port on the Windows client to be sent to the printer that owns the
spoolservice that has been specified.
C:\> print filename
You might want to read Chapter 36, “The Samba Checklist”. If you are
still stuck, refer to Chapter 37, “Analyzing and Solving Samba Problems”.
Samba has been successfully installed at thousands of sites worldwide. It is
unlikely that your particular problem is unique, so it might be productive
to perform an Internet search to see if someone else has encountered your
problem and has found a way to overcome it.
If you are new to Samba, and particularly if you are new to Windows net-
working, or to UNIX/Linux, the book “Samba-3 by Example” will help you
to create a validated network environment. Simply choose from the first
Section 1.6. Common Errors 9
five chapters the network design that most closely matches site needs, then
follow the simple step-by-step procedure to deploy it. Later, when you have
a working network you may well want to refer back to this book for further
insight into opportunities for improvement.
The best advice under the stress of abject frustration is to cool down! That
may be challenging of itself, but while you are angry or annoyed your ability
to seek out a solution is somewhat undermined. A cool head clears the way
to finding the answer you are looking for. Just remember, every problem
has a solution — there is a good chance that someone else has found it
even though you can’t right now. That will change with time, patience and
learning.
Now that you have cooled down a bit, please refer to Chapter 36, “The
Samba Checklist” for a process that can be followed to identify the cause of
your problem.
The following questions and issues are raised repeatedly on the Samba mail-
ing list.
Samba consists of three core programs: nmbd, smbd, and winbindd. nmbd
is the name server message daemon, smbd is the server message daemon,
and winbindd is the daemon that handles communication with domain con-
trollers.
If Samba is not running as a WINS server, then there will be one single
instance of nmbd running on your system. If it is running as a WINS server,
then there will be two instances — one to handle the WINS requests.
smbd handles all connection requests. It spawns a new process for each
client connection made. That is why you may see so many of them, one per
client connection.
10 How to Install and Test SAMBA Chapter 1
An error message is observed in the log files when smbd is started: “open oplock ipc:
Failed to get local UDP socket for address 100007f. Error was Cannot assign
requested.”
Your loopback device isn’t working correctly. Make sure it is configured
correctly. The loopback device is an internal (virtual) network device with
the IP address 127.0.0.1. Read your OS documentation for details on how
to configure the loopback on your system.
When we first asked for suggestions for inclusion in the Samba HOWTO
documentation, someone wrote asking for example configurations — and lots
of them. That is remarkably difficult to do without losing a lot of value that
can be derived from presenting many extracts from working systems. That
is what the rest of this document does. It does so with extensive descriptions
of the configuration possibilities within the context of the chapter that covers
it. We hope that this chapter is the medicine that has been requested.
The information in this chapter is very sparse compared with the book
“Samba-3 by Example” that was written after the original version of this
book was nearly complete. “Samba-3 by Example” was the result of feed-
back from reviewers during the final copy editing of the first edition. It was
interesting to see that reader feedback mirrored that given by the original
reviewers. In any case, a month and a half was spent in doing basic research
to better understand what new as well as experienced network administra-
tors would best benefit from. The book “Samba-3 by Example” is the result
of that research. What is presented in the few pages of this book is covered
far more comprehensively in the second edition of “Samba-3 by Example”.
The second edition of both books will be released at the same time.
11
12 Fast Start: Cure for Impatience Chapter 2
A standalone server implies no more than the fact that it is not a domain
controller and it does not participate in domain control. It can be a simple,
workgroup-like server, or it can be a complex server that is a member of a
domain security context.
As the examples are developed, every attempt is made to progress the system
toward greater capability, just as one might expect would happen in a real
business office as that office grows in size and its needs change.
The purpose of this type of server is to make available to any user any doc-
uments or files that are placed on the shared resource. The shared resource
could be a CD-ROM drive, a CD-ROM image, or a file storage area.
• The file system share point will be /export.
• All files will be owned by a user called Jack Baumbach. Jack’s login
name will be jackb. His password will be m0r3pa1n — of course,
that’s just the example we are using; do not use this in a production
environment because all readers of this document will know it.
Installation Procedure: Read-Only Server
1. Add user to system (with creation of the user’s home directory):
14 Fast Start: Cure for Impatience Chapter 2
root# testparm
root# cd /etc/samba
Section 2.3. Worked Examples 15
# Global parameters
[global]
workgroup = MIDEARTH
netbios name = HOBBIT
security = share
[data]
comment = Data
path = /export
read only = Yes
guest only = Yes
Addition of this user to the smbpasswd file allows all files to be displayed
in the Explorer Properties boxes as belonging to jackb instead of to User
Unknown.
does print. Samba makes use of the direct printing application program
interface that is provided by CUPS. When Samba has been compiled and
linked with the CUPS libraries the default printing system will be CUPS.
By specifying that the printcap name is CUPS, Samba will use the CUPS
library API to communicate directly with CUPS for all printer functions.
It is possible to force the use of external printing commands by setting the
value of the printing to either SYSV or BSD, and thus the value of the
parameter printcap name must be set to something other than CUPS. In
such case, it could be set to the name of any file that contains a list of
printers that should be made available to Windows clients.
Note
The purpose of setting the sticky bit is to prevent who does not own
the temporary print file from being able to take control of it with the
potential for devious misuse.
Section 2.3. Worked Examples 19
Note
We progress now from simple systems to a server that is slightly more com-
plex.
Our new server will require a public data storage area in which only au-
thenticated users (i.e., those with a local account) can store files, as well
as a home directory. There will be one printer that should be available for
everyone to use.
In this hypothetical environment (no espionage was conducted to obtain this
data), the site is demanding a simple environment that is secure enough but
not too difficult to use.
Site users will be Jack Baumbach, Mary Orville, and Amed Sehkah. Each
will have a password (not shown in further examples). Mary will be the
printer administrator and will own all files in the public share.
This configuration will be based on user-level security that is the default, and
for which the default is to store Microsoft Windows-compatible encrypted
passwords in a file called /etc/samba/smbpasswd. The default smb.conf
entry that makes this happen is passdb backend = smbpasswd, guest. Since
this is the default, it is not necessary to enter it into the configuration file.
Note that the guest backend is added to the list of active passdb backends
no matter whether it specified directly in Samba configuration file or not.
Installing the Secure Office Server
1. Add all users to the operating system:
20 Fast Start: Cure for Impatience Chapter 2
4. Install printer using the CUPS Web interface. Make certain that all
printers that will be shared with Microsoft Windows clients are in-
stalled as raw printing devices.
5. Start Samba using the operating system administrative interface. Al-
ternately, this can be done manually by executing:
Server Comment
--------- -------
OLORIN Samba-3.0.20
Workgroup Master
--------- -------
MIDEARTH OLORIN
The following error message indicates that Samba was not running:
to run only off one server, there are no workstation components, and it is
run off a mapped share. The data store is in a UNIX-based SQL backend.
The UNIX gurus look after that, so this is not our problem.
The server valinor will be a member server of the company domain. Ac-
counting will have only a local server. User accounts will be on the domain
controllers, as will desktop profiles and all network policy files.
1. Do not add users to the UNIX/Linux server; all of this will run off the
central domain.
3. Join the domain. Note: Do not start Samba until this step has been
completed!
4. Make absolutely certain that you disable (shut down) the nscd daemon
on any system on which winbind is configured to run.
5. Start Samba following the normal method for your operating system
26 Fast Start: Cure for Impatience Chapter 2
6. Configure the name service switch (NSS) control file on your system
to resolve user and group names via winbind. Edit the following lines
in /etc/nsswitch.conf:
8. Validate that domain user and group credentials can be correctly re-
solved by executing:
root# wbinfo -u
MIDEARTH\maryo
MIDEARTH\jackb
MIDEARTH\ameds
...
MIDEARTH\root
root# wbinfo -g
MIDEARTH\Domain Users
MIDEARTH\Domain Admins
MIDEARTH\Domain Guests
...
MIDEARTH\Accounts
10. A final test that we have this under control might be reassuring:
root# rm /export/a_file
For the remainder of this chapter the focus is on the configuration of do-
main control. The examples that follow are for two implementation strate-
gies. Remember, our objective is to create a simple but working solution.
The remainder of this book should help to highlight opportunity for greater
functionality and the complexity that goes with it.
A domain controller configuration can be achieved with a simple configura-
tion using the new tdbsam password backend. This type of configuration
is good for small offices, but has limited scalability (cannot be replicated),
and performance can be expected to fall as the size and complexity of the
domain increases.
The use of tdbsam is best limited to sites that do not need more than
28 Fast Start: Cure for Impatience Chapter 2
Note
3. Create user accounts on the system using the appropriate tool provided
with the operating system. Make sure all user home directories are
created also. Add users to groups as required for access control on files,
directories, printers, and as required for use in the Samba environment.
4. Assign each of the UNIX groups to NT groups by executing this shell
script (You could name the script initGroups.sh):
#!/bin/bash
#### Keep this as a shell script for future re-use
Place the logon scripts that will be used (batch or cmd scripts) in this
directory.
The above configuration provides a functional PDC system to which must
be added file shares and printers as required.
30 Fast Start: Cure for Impatience Chapter 2
pidfile /var/run/slapd/slapd.pid
argsfile /var/run/slapd/slapd.args
database bdb
suffix "dc=quenya,dc=org"
rootdn "cn=Manager,dc=quenya,dc=org"
rootpw {SSHA}06qDkonA8hk6W6SSnRzWj0/pBcU3m0/P
1
<http://www.idealx.org>
Section 2.3. Worked Examples 31
directory /var/lib/ldap
index objectClass eq
index cn pres,sub,eq
index sn pres,sub,eq
index uid pres,sub,eq
index displayName pres,sub,eq
index uidNumber eq
index gidNumber eq
index memberUid eq
index sambaSID eq
index sambaPrimaryGroupSID eq
index sambaDomainName eq
index default sub
5. Start the LDAP server using the appropriate tool or method for the
operating system platform on which it is installed.
6. Install the Idealx script files in the /usr/local/sbin directory, then
configure the smbldap conf.pm file to match your system configura-
tion.
7. The smb.conf file that drives this backend can be found in example
Example 2.3.9. Add additional stanzas as required.
8. Add the LDAP password to the secrets.tdb file so Samba can update
the LDAP database:
9. Add users and groups as required. Users and groups added using
Samba tools will automatically be added to both the LDAP backend
and the operating system as required.
Samba can operate in various modes within SMB networks. This HOWTO
section contains information on configuring Samba to function as the type
of server your network requires. Please read this section carefully.
39
Chapter 3
This chapter provides information regarding the types of server that Samba
may be configured to be. A Microsoft network administrator who wishes to
migrate to or use Samba will want to know the meaning, within a Samba
context, of terms familiar to the MS Windows administrator. This means
that it is essential also to define how critical security modes function before
we get into the details of how to configure the server itself.
The chapter provides an overview of the security modes of which Samba is
capable and how they relate to MS Windows servers and clients.
A question often asked is, “Why would I want to use Samba?” Most chapters
contain a section that highlights features and benefits. We hope that the
information provided will help to answer this question. Be warned though,
we want to be fair and reasonable, so not all features are positive toward
Samba. The benefit may be on the side of our competition.
Two men were walking down a dusty road, when one suddenly kicked up
a small red stone. It hurt his toe and lodged in his sandal. He took the
stone out and cursed it with a passion and fury befitting his anguish. The
other looked at the stone and said, “This is a garnet. I can turn that into a
precious gem and some day it will make a princess very happy!”
The moral of this tale: Two men, two very different perspectives regarding
the same stone. Like it or not, Samba is like that stone. Treat it the right
40
Section 3.2. Server Types 41
way and it can bring great pleasure, but if you are forced to use it and have
no time for its secrets, then it can be a source of discomfort.
Samba started out as a project that sought to provide interoperability for
MS Windows 3.x clients with a UNIX server. It has grown up a lot since
its humble beginnings and now provides features and functionality fit for
large-scale deployment. It also has some warts. In sections like this one, we
tell of both.
So, what are the benefits of the features mentioned in this chapter?
• Samba-3 can replace an MS Windows NT4 domain controller.
• Samba-3 offers excellent interoperability with MS Windows NT4-style
domains as well as natively with Microsoft Active Directory domains.
• Samba-3 permits full NT4-style interdomain trusts.
• Samba has security modes that permit more flexible authentication
than is possible with MS Windows NT4 domain controllers.
• Samba-3 permits use of multiple concurrent account database back-
ends. (Encrypted passwords that are stored in the account database
are in formats that is unique to Windows networking).
• The account database backends can be distributed and replicated using
multiple methods. This gives Samba-3 greater flexibility than MS
Windows NT4 and in many cases a significantly higher utility than
Active Directory domains with MS Windows 200x.
In this section the function and purpose of Samba’s security modes are de-
scribed. An accurate understanding of how Samba implements each security
mode as well as how to configure MS Windows clients for each mode will
significantly reduce user complaints and administrator heartache.
Microsoft Windows networking uses a protocol that was originally called the
Server Message Block (SMB) protocol. Since some time around 1996 the
protocol has been better known as the Common Internet Filesystem (CIFS)
protocol.
In the SMB/CIFS networking world, there are only two types of security:
user-level and share level. We refer to these collectively as security levels. In
implementing these two security levels, Samba provides flexibilities that are
not available with MS Windows NT4/200x servers. In fact, Samba imple-
ments share-level security only one way, but has four ways of implementing
user-level security. Collectively, we call the Samba implementations of the
security levels security modes. They are known as share, user, domain, ADS,
and server modes. They are documented in this chapter.
An SMB server informs the client, at the time of a session setup, the security
level the server is running. There are two options: share-level and user-level.
Which of these two the client receives affects the way the client then tries
to authenticate itself. It does not directly affect (to any great extent) the
way the Samba server does security. This may sound strange, but it fits
Section 3.3. Samba Security Modes 43
1. the username/password.
It is also possible for a client to send multiple session setup requests. When
the server responds, it gives the client a uid to use as an authentication tag
for that username/password. The client can maintain multiple authentica-
tion contexts in this way (WinDD is an example of an application that does
this).
In the example shown here (not likely to be used in practice) the lookup will
check /etc/passwd and /etc/group, if not found it will check NIS, then
LDAP.
Domain security provides a mechanism for storing all user and group ac-
counts in a central, shared, account repository. The centralized account
repository is shared between domain (security) controllers. Servers that
act as domain controllers provide authentication and validation services to
all machines that participate in the security context for the domain. A pri-
mary domain controller (PDC) is a server that is responsible for maintaining
the integrity of the security account database. Backup domain controllers
(BDCs) provide only domain logon and authentication services. Usually,
BDCs will answer network logon requests more responsively than will a
PDC.
When Samba is operating in security = domain mode, the Samba server
has a domain security trust account (a machine account) and causes all
authentication requests to be passed through to the domain controllers. In
other words, this configuration makes the Samba server a domain member
server, even when it is in fact acting as a domain controller. All machines
that participate in domain security must have a machine account in the
security database.
Within the domain security environment the underlying security architec-
ture uses User-level security. Even machines that are domain members must
authenticate on startup. The machine account consists of an account entry
46 Server Types and Security Modes Chapter 3
in the accounts database, the name of which is the NetBIOS name of the
machine and of which the password is randomly generated and known to
both the domain controllers and the member machine. If the machine ac-
count can not be validated during startup, users will not be able to log onto
the domain using this machine because it can not be trusted. The machine
account is referred to as a machine trust account.
We will discuss each of these in separate chapters. For now, we are most
interested in basic DMS configuration.
In order for this method to work, the Samba server needs to join the MS
Windows NT security domain. This is done as follows:
Note
Both Samba-2.2, and Samba-3 can join an Active Directory domain using
NT4 style RPC based security. This is possible if the domain is run in native
mode. Active Directory in native mode perfectly allows NT4-style domain
members. This is contrary to popular belief.
If you are using Active Directory, starting with Samba-3 you can join as a
native AD member. Why would you want to do that? Your security policy
might prohibit the use of NT-compatible authentication protocols. All your
machines are running Windows 2000 and above and all use Kerberos. In this
case Samba, as an NT4-style domain, would still require NT-compatible au-
thentication data. Samba in AD-member mode can accept Kerberos tickets.
Sites that use Microsoft Windows active directory services (ADS) should be
aware of the significance of the terms: native mode and mixed mode ADS
operation. The term realm is used to describe a Kerberos-based security
architecture (such as is used by Microsoft ADS).
Please refer to Chapter 6, “Domain Membership”, and Section 6.4 for more
information regarding this configuration option.
Server security mode is left over from the time when Samba was not capable
of acting as a domain member server. It is highly recommended not to use
this feature. Server security mode has many drawbacks that include:
• Potential account lockout on MS Windows NT4/200x password servers.
• Lack of assurance that the password server is the one specified.
Section 3.3. Samba Security Modes 49
• Does not work with Winbind, which is particularly needed when stor-
ing profiles remotely.
• This mode may open connections to the password server and keep
them open for extended periods.
• Security on the Samba server breaks badly when the remote password
server suddenly shuts down.
• With this mode there is NO security account in the domain that the
password server belongs to for the Samba server.
In server security mode the Samba server reports to the client that it is
in user-level security. The client then does a session setup as described
earlier. The Samba server takes the username/password that the client
sends and attempts to log into the password server by sending exactly the
same username/password that it got from the client. If that server is in
user-level security and accepts the password, then Samba accepts the client’s
connection. This parameter allows the Samba server to use another SMB
server as the password server.
You should also note that at the start of all this, when the server tells
the client what security level it is in, it also tells the client if it supports
encryption. If it does, it supplies the client with a random cryptkey. The
client will then send all passwords in encrypted form. Samba supports this
type of encryption by default.
The parameter security = server means that Samba reports to clients that it
is running in user mode but actually passes off all authentication requests to
another user mode server. This requires an additional parameter password
server that points to the real authentication server. The real authentication
server can be another Samba server, or it can be a Windows NT server, the
latter being natively capable of encrypted password support.
50 Server Types and Security Modes Chapter 3
Note
This method involves the additions of the following parameters in the smb.
conf file:
e n c r y p t passwords = Yes
security = server
password s e r v e r = ” NetBIOS name of a DC ”
There are two ways of identifying whether or not a username and pass-
word pair is valid. One uses the reply information provided as part of the
authentication messaging process, the other uses just an error code.
username l e v e l
We all make mistakes. It is okay to make mistakes, as long as they are made
in the right places and at the right time. A mistake that causes lost pro-
ductivity is seldom tolerated; however, a mistake made in a developmental
test lab is expected.
Here we look at common mistakes and misapprehensions that have been the
subject of discussions on the Samba mailing lists. Many of these are avoid-
able by doing your homework before attempting a Samba implementation.
Some are the result of a misunderstanding of the English language, which
has many phrases that are potentially vague and may be highly confusing
to those for whom English is not their native tongue.
To some, the nature of the Samba security mode is obvious, but entirely
wrong all the same. It is assumed that security = server means that Samba
Section 3.5. Common Errors 53
will act as a server. Not so! This setting means that Samba will try to use
another SMB server as its source for user authentication alone.
Samba is a server regardless of which security mode is chosen. When Samba
is used outside of a domain security context, it is best to leave the security
mode at the default setting. By default Samba-3 uses User-mode security.
The smb.conf parameter security = domain does not really make Samba
behave as a domain controller. This setting means we want Samba to be a
domain member. See Chapter 4, “Domain Control” for more information.
Guess! So many others do. But whatever you do, do not think that security
= user makes Samba act as a domain member. Read the manufacturer’s
manual before the warranty expires. See Chapter 6, “Domain Membership”,
for more information.
“Why does server validate() simply give up rather than re-establish its con-
nection to the password server? Though I am not fluent in the SMB protocol,
perhaps the cluster server process passes along to its client workstation the
session key it receives from the password server, which means the password
hashes submitted by the client would not work on a subsequent connection
whose session key would be different. So server validate() must give up.”
Indeed. That’s why security = server is at best a nasty hack. Please use
security = domain; security = server mode is also known as pass-through
authentication.
Chapter 4
DOMAIN CONTROL
54
55
DOMAIN
Workstation A
Primary
Domain
Controller
Workstation B
Workstation C
Backup Domain
Backup Domain Controller 2
Controller 1
Where is the right place to make mistakes? Only out of harms way. If you
are going to make mistakes, then please do it on a test network, away from
users, and in such a way as to not inflict pain on others. Do your learning
on a test network.
56 Domain Control Chapter 4
Note
An LDAP passdb backend confers the benefit that the account backend
can be distributed and replicated, which is of great value because it
confers scalability and provides a high degree of reliability.
• Windows NT4 domain trusts. Samba-3 supports workstation and
server (machine) trust accounts. It also supports Windows NT4 style
interdomain trust accounts, which further assists in network scalability
and interoperability.
• Operation without NetBIOS over TCP/IP, rather using the raw SMB
over TCP/IP. Note, this is feasible only when operating as a Microsoft
active directory domain member server. When acting as a Samba
domain controller the use of NetBIOS is necessary to provide network
browsing support.
• Samba-3 provides NetBIOS name services (WINS), NetBIOS over TCP/IP
(TCP port 139) session services, SMB over TCP/IP (TCP port 445)
session services, and Microsoft compatible ONC DCE RPC services
(TCP port 135) services.
• Management of users and groups via the User Manager for Domains.
This can be done on any MS Windows client using the Nexus.exe
toolkit for Windows 9x/Me, or using the SRVTOOLS.EXE package for
MS Windows NT4/200x/XP platforms. These packages are available
from Microsoft’s Web site.
• Implements full Unicode support. This simplifies cross-locale inter-
nationalization support. It also opens up the use of protocols that
Samba-2.2.x had but could not use due to the need to fully support
Unicode.
The following functionalities are not provided by Samba-3:
• SAM replication with Windows NT4 domain controllers (i.e., a Samba
PDC and a Windows NT BDC, or vice versa). This means Samba
cannot operate as a BDC when the PDC is Microsoft-based Windows
NT PDC. Samba-3 can not participate in replication of account data
to Windows PDCs and BDCs.
• Acting as a Windows 2000 active directory domain controller (i.e.,
Kerberos and Active Directory). In point of fact, Samba-3 does have
some Active Directory domain control ability that is at this time purely
experimental. Active directory domain control is one of the features
that is being developed in Samba-4, the next generation Samba release.
58 Domain Control Chapter 4
Windows 9x/Me/XP Home clients are not true members of a domain for rea-
sons outlined in this chapter. The protocol for support of Windows 9x/Me-
style network (domain) logons is completely different from NT4/Windows
200x-type domain logons and has been officially supported for some time.
These clients use the old LanMan network logon facilities that are supported
in Samba since approximately the Samba-1.9.15 series.
There is a broadly held perception that SSO is the answer to the problem
of users having to deal with too many information system access credentials
(username/password pairs). Many elaborate schemes have been devised to
make it possible to deliver a user-friendly SSO solution. The trouble is that
if this implementation is not done correctly, the site may end up paying
dearly by way of complexity and management overheads. Simply put, many
SSO solutions are an administrative nightmare.
The initial goal of many network administrators is often to create and use
a centralized identity management system. It is often assumed that such a
centralized system will use a single authentication infrastructure that can
be used by all information systems. The Microsoft Windows NT4 secu-
rity domain architecture and the Micrsoft active directory service are often
put forward as the ideal foundation for such a system. It is conceptually
simple to install an external authentication agent on each of the disparate
infromation systems that can then use the Microsoft (NT4 domain or ads
service) for user authentication and access control. The wonderful dream of
a single centralized authentication service is commonly broken when reali-
ties are realized. The problem with legacy systems is often the inability to
externalize the authentication and access control system it uses because its
implementation will be excessively invasive from a re-engineering perspec-
tive, or because application software has built-in dependencies on particular
elements of the way user authentication and access control were designed
and built.
Over the past decade an industry has been developed around the various
methods that have been built to get around the key limitations of legacy
information technology systems. One approach that is often used involves
the use of a meta-directory. The meta-directory stores user credentials for
60 Domain Control Chapter 4
ing support in Samba for the use of LDAP and Microsoft ADS make Samba a
highly scalable and forward reaching organizational networking technology.
Microsoft ADS provides purely proprietary services that, with limitation,
can be extended to provide a centralized authentication infrastructure. Samba
plus LDAP provides a similar opportunity for extension of a centralized au-
thentication architecture, but it is the fact that the Samba Team are pro-
active in introducing the extension of authentication services, using LDAP
or otherwise, to applications such as SQUID (the open source proxy server)
through tools such as the ntlm auth utility, that does much to create sus-
tainable choice and competition in the FIM market place.
Primary domain control, if it is to be scalable to meet the needs of large sites,
must therefore be capable of using LDAP. The rapid adoption of OpenLDAP,
and Samba configurations that use it, is ample proof that the era of the
directoy has started. Samba-3 does not demand the use of LDAP, but the
demand for a mechanism by which user and group identity information can
be distributed makes it an an unavoidable option.
At this time, the use of Samba based BDCs, necessitates the use of LDAP.
The most commonly used LDAP implementation used by Samba sites is
OpenLDAP. It is possible to use any standards compliant LDAP server.
Those known to work includes those manufactured by: IBM, CA, Novell
(e-Directory), and others.
Over the years, public perceptions of what domain control really is has
taken on an almost mystical nature. Before we branch into a brief overview
of domain control, there are three basic types of domain controllers.
PDC and BDC must be manually configured, and other appropriate changes
also need to be made.
With MS Windows NT4, a decision is made at installation to determine
what type of machine the server will be. It is possible to promote a BDC
to a PDC, and vice versa. The only method Microsoft provide to convert a
Windows NT4 domain controller to a domain member server or a standalone
server is to reinstall it. The install time choices offered are:
• Primary Domain Controller — the one that seeds the domain SAM.
• Backup Domain Controller — one that obtains a copy of the domain
SAM.
• Domain Member Server — one that has no copy of the domain SAM;
rather it obtains authentication from a domain controller for all access
controls.
• Standalone Server — one that plays no part in SAM synchronization,
has its own authentication database, and plays no role in domain se-
curity.
Note
a
<http://utools.com/UPromote.asp>
There are two ways that MS Windows machines may interact with each
other, with other servers, and with domain controllers: either as standalone
systems, more commonly called workgroup members, or as full participants
in a security system, more commonly called domain members.
It should be noted that workgroup membership involves no special configu-
ration other than the machine being configured so the network configuration
has a commonly used name for its workgroup entry. It is not uncommon for
the name WORKGROUP to be used for this. With this mode of configura-
tion, there are no Machine Trust Accounts, and any concept of membership
Section 4.3. Basics of Domain Control 65
as such is limited to the fact that all machines appear in the network neigh-
borhood to be logically grouped together. Again, just to be clear: workgroup
mode does not involve security machine accounts.
Domain member machines have a machine trust account in the domain
accounts database. A special procedure must be followed on each machine
to effect domain membership. This procedure, which can be done only
by the local machine Administrator account, creates the domain machine
account (if it does not exist), and then initializes that account. When the
client first logs onto the domain, a machine trust account password change
will be automatically triggered.
Note
Note
The first step in creating a working Samba PDC is to understand the pa-
rameters necessary in smb.conf. An example smb.conf for acting as a PDC
can be found in Example 4.4.1.
The basic options shown in Example 4.4.1 are explained as follows:
passdb backend This contains all the user and group account information.
Acceptable values for a PDC are: smbpasswd, tdbsam, and ldapsam.
The “guest” entry provides default accounts and is included by default;
there is no need to add it explicitly.
Where use of BDCs is intended, the only logical choice is to use LDAP
so the passdb backend can be distributed. The tdbsam and smbpasswd
files cannot effectively be distributed and therefore should not be used.
3
See Chapter 9, “Network Browsing”.
68 Domain Control Chapter 4
drive, and logon script are environment support settings that help to
facilitate client logon operations and that help to provide automated
control facilities to ease network management overheads. Please refer
to the man page information for these parameters.
PROFILE Share This share is used to store user desktop profiles. Each
user must have a directory at the root of this share. This directory
must be write-enabled for the user and must be globally read-enabled.
Samba-3 has a VFS module called “fake permissions” that may be
installed on this share. This will allow a Samba administrator to make
the directory read-only to everyone. Of course this is useful only after
the profile has been properly created.
Note
Samba-3 is not, and cannot act as, an Active Directory server. It cannot
truly function as an Active Directory PDC. The protocols for some of the
functionality of Active Directory domain controllers has been partially im-
plemented on an experimental only basis. Please do not expect Samba-3
to support these protocols. Do not depend on any such functionality either
now or in the future. The Samba Team may remove these experimental
features or may change their behavior. This is mentioned for the benefit of
those who have discovered secret capabilities in Samba-3 and who have asked
when this functionality will be completed. The answer is maybe someday
or maybe never!
To be sure, Samba-3 is designed to provide most of the functionality that
Microsoft Windows NT4-style domain controllers have. Samba-3 does not
have all the capabilities of Windows NT4, but it does have a number of fea-
tures that Windows NT4 domain controllers do not have. In short, Samba-3
is not NT4 and it is not Windows Server 200x: it is not an Active Directory
server. We hope this is plain and simple enough for all to understand.
All domain controllers must run the netlogon service (domain logons in
Samba). One domain controller must be configured with domain master
= Yes (the PDC); on all BDCs set the parameter domain master = No.
Note
Now that this has been said, please do not ask the mailing list or email any
of the Samba Team members with your questions asking how to make this
work. It can’t be done. If it can be done, then to do so would violate your
software license agreement with Microsoft, and we recommend that you do
not do that.
A domain and a workgroup are exactly the same in terms of network brows-
ing. The difference is that a distributable authentication database is asso-
ciated with a domain, for secure login access to a network. Also, different
access rights can be granted to users if they successfully authenticate against
a domain logon server. Samba-3 does this now in the same way as MS Win-
dows NT/200x.
72 Domain Control Chapter 4
The SMB client logging on to a domain has an expectation that every other
server in the domain should accept the same authentication information.
Network browsing functionality of domains and workgroups is identical and
is explained in this documentation under the browsing discussions. It should
be noted that browsing is totally orthogonal to logon support.
Issues related to the single-logon network model are discussed in this section.
Samba supports domain logons, network logon scripts, and user profiles for
MS Windows for Workgroups and MS Windows 9x/Me clients, which are
the focus of this section.
When an SMB client in a domain wishes to log on, it broadcasts requests for
a logon server. The first one to reply gets the job and validates its password
using whatever mechanism the Samba administrator has installed. It is
possible (but ill advised) to create a domain where the user database is
not shared between servers; that is, they are effectively workgroup servers
advertising themselves as participating in a domain. This demonstrates how
authentication is quite different from but closely involved with domains.
Using these features, you can make your clients verify their logon via the
Samba server, make clients run a batch file when they log on to the network
and download their preferences, desktop, and start menu.
MS Windows XP Home edition is not able to join a domain and does not
permit the use of domain logons.
Before launching into the configuration instructions, it is worthwhile to look
at how a Windows 9x/Me client performs a logon:
1. The client broadcasts (to the IP broadcast address of the subnet it
is in) a NetLogon request. This is sent to the NetBIOS name DO-
MAIN<1C> at the NetBIOS layer. The client chooses the first re-
sponse it receives, which contains the NetBIOS name of the logon
server to use in the format of \\SERVER. The 1C name is the name
type that is registered by domain controllers (SMB/CIFS servers that
provide the netlogon service).
2. The client connects to that server, logs on (does an SMBsessetupX)
and then connects to the IPC$ share (using an SMBtconX).
3. The client does a NetWkstaUserLogon request, which retrieves the
name of the user’s logon script.
4. The client then connects to the NetLogon share and searches for said
Section 4.6. Domain and Network Logon Configuration 73
6. The client connects to the user’s home share and searches for the user’s
profile. As it turns out, you can specify the user’s home share as a
share name and path. For example, \\server\fred\.winprofile. If
the profiles are found, they are implemented.
7. The client then disconnects from the user’s home share and reconnects
to the NetLogon share and looks for CONFIG.POL, the policies file. If
this is found, it is read and implemented.
The main difference between a PDC and a Windows 9x/Me logon server
configuration is:
• Windows 9x/Me clients do not require and do not use Machine Trust
Accounts.
A Samba PDC will act as a Windows 9x/Me logon server; after all, it does
provide the network logon services that MS Windows 9x/Me expect to find.
Note
There are a few comments to make in order to tie up some loose ends. There
has been much debate over the issue of whether it is okay to configure Samba
as a domain controller that operates with security mode other than user-
mode. The only security mode that will not work due to technical reasons
is share-mode security. Domain and server mode security are really just a
variation on SMB user-level security.
Actually, this issue is also closely tied to the debate on whether Samba must
be the DMB for its workgroup when operating as a domain controller. In
a pure Microsoft Windows NT domain, the PDC wins the election to be
the DMB, and then registers the DOMAIN<1B> NetBIOS name. This is
not the name used by Windows clients to locate the domain controller, all
domain controllers register the DOMAIN<1C> name and Windows clients
locate a network logon server by seraching for the DOMAIN<1C> name. A
DMB is a Domain Master Browser — see Chapter 9, “Network Browsing”,
Section 9.4.1; Microsoft PDCs expect to win the election to become the
DMB, if it loses that election it will report a continuous and rapid sequence
of warning messages to its Windows event logger complaining that it has
lost the election to become a DMB. For this reason, in networks where a
Samba server is the PDC it is wise to configure the Samba domain controller
as the DMB.
Section 4.7. Common Errors 75
Note
The problem is only in the program used to make the entry. Once made,
it works perfectly. Create a user without the “$”. Then use vipw to edit
the entry, adding the “$”. Or create the whole entry with vipw if you like;
make sure you use a unique user login ID.
Note
The machine account must have the exact name that the
workstation has.
Note
same name as the domain you are joining (bad idea), you will get this
message. Change the workgroup name to something else — it does not
matter what — reboot, and try again.
This occurs when the domain SID stored in the secrets.tdb database is
changed. The most common cause of a change in domain SID is when
the domain name and/or the server name (NetBIOS name) is changed. The
only way to correct the problem is to restore the original domain SID or
remove the domain client from the domain and rejoin. The domain SID
may be reset using either the net or rpcclient utilities.
To reset or change the domain SID you can use the net command as follows:
Workstation Machine Trust Accounts work only with the domain (or net-
work) SID. If this SID changes, domain members (workstations) will not be
able to log onto the domain. The original domain SID can be recovered from
the secrets.tdb file. The alternative is to visit each workstation to rejoin it
to the domain.
“When I try to join the domain I get the message, ”The machine account
for this computer either does not exist or is not accessible.” What’s wrong?”
This problem is caused by the PDC not having a suitable Machine Trust
Account. If you are using the add machine script method to create accounts,
then this would indicate that it has not worked. Ensure the domain admin
user system is working.
78 Domain Control Chapter 4
Alternately, if you are creating account entries manually, then they have not
been created correctly. Make sure that you have the entry correct for the Ma-
chine Trust Account in smbpasswd file on the Samba PDC. If you added the
account using an editor rather than using the smbpasswd utility, make sure
that the account name is the machine NetBIOS name with a “$” appended
to it (i.e., computer name$). There must be an entry in both the POSIX
UNIX system account backend as well as in the SambaSAMAccount back-
end. The default backend for Samba-3 (i.e., the parameter passdb backend
is not specified in the smb.conf file, or if specified is set to smbpasswd,
are respectively the /etc/passwd and /etc/samba/smbpasswd (or /usr/
local/samba/lib/private/smbpasswd if compiled using Samba Team de-
fault settings). The use of the /etc/passwd can be overridden by alternative
settings in the NSS /etc/nsswitch.conf file.
Some people have also reported that inconsistent subnet masks between the
Samba server and the NT client can cause this problem. Make sure that
these are consistent for both client and server.
“Until a few minutes after Samba has started, clients get the error ‘Domain
Controller Unavailable’”
A domain controller has to announce its role on the network. This usually
takes a while. Be patient for up to 15 minutes, then try again.
After successfully joining the domain, user logons fail with one of two mes-
sages: one to the effect that the domain controller cannot be found; the other
claims that the account does not exist in the domain or that the password
Section 4.7. Common Errors 79
testparm -v | grep channel and looking for the value of these parameters.
Also use the MMC — Local Security Settings. This tool is available from the
Control Panel. The Policy settings are found in the Local Policies/Security
Options area and are prefixed by Secure Channel:..., and Digitally sign....
It is important that these be set consistently with the Samba-3 server set-
tings.
Chapter 5
BACKUP DOMAIN
CONTROL
Before you continue reading this section, please make sure that you are
comfortable with configuring a Samba domain controller as described in
Chapter 4, “Domain Control”.
This is one of the most difficult chapters to summarize. It does not matter
what we say here, for someone will still draw conclusions and/or approach
the Samba Team with expectations that are either not yet capable of being
delivered or that can be achieved far more effectively using a totally different
approach. In the event that you should have a persistent concern that is
not addressed in this book, please email John H. Terpstra1 clearly setting
out your requirements and/or question, and we will do our best to provide
a solution.
Samba-3 can act as a Backup Domain Controller (BDC) to another Samba
Primary Domain Controller (PDC). A Samba-3 PDC can operate with an
LDAP account backend. The LDAP backend can be either a common master
LDAP server or a slave server. The use of a slave LDAP server has the
benefit that when the master is down, clients may still be able to log onto
the network. This effectively gives Samba a high degree of scalability and is
an effective solution for large organizations. If you use an LDAP slave server
for a PDC, you will need to ensure the master’s continued availability — if
1
<mailto:[email protected]>
80
Section 5.2. Essential Background Information 81
the slave finds its master down at the wrong time, you will have stability
and operational problems.
The BDC is said to hold a read-only of the SAM from which it is able
to process network logon requests and authenticate users. The BDC can
continue to provide this service, particularly while, for example, the wide-
area network link to the PDC is down. A BDC plays a very important role
in both the maintenance of domain security as well as in network integrity.
In the event that the NT4 PDC should need to be taken out of service, or
if it dies, one of the NT4 BDCs can be promoted to a PDC. If this happens
while the original NT4 PDC is online, it is automatically demoted to an NT4
BDC. This is an important aspect of domain controller management. The
tool that is used to effect a promotion or a demotion is the Server Manager
for Domains. It should be noted that Samba-3 BDCs cannot be promoted in
this manner because reconfiguration of Samba requires changes to the smb.
conf file. It is easy enough to manuall change the smb.conf file and then
restart relevant Samba network services.
Beginning with Version 2.2, Samba officially supports domain logons for all
current Windows clients, including Windows NT4, 2003, and XP Profes-
sional. For Samba to be enabled as a PDC, some parameters in the [global]
section of the smb.conf have to be set. Refer to Example 5.2.1 for an ex-
ample of the minimum required settings.
Several other things like a [homes] and a [netlogon] share also need to be
set along with settings for the profile path, the user’s home drive, and so
on. This is not covered in this chapter; for more information please refer
to Chapter 4, “Domain Control”. Refer to Chapter 4, “Domain Control”
for specific recommendations for PDC configuration. Alternately, fully doc-
umented working example network configurations using OpenLDAP and
Samba as available in the book2 “Samba-3 by Example” that may be ob-
tained from local and on-line book stores.
2
<http://www.samba.org/samba/docs/Samba3-ByExample>
Section 5.2. Essential Background Information 85
Example 5.2.1. Minimal smb.conf for a PDC in Use with a BDC — LDAP
Server on PDC
workgroup = MIDEARTH
passdb backend = ldapsam : / / l o c a l h o s t : 3 8 9
domain master = y e s
domain l o g o n s = y e s
l d a p s u f f i x = dc=quenya , dc=o r g
l d a p u s e r s u f f i x = ou=U s e r s
l d a p group s u f f i x = ou=Groups
l d a p machine s u f f i x = ou=Computers
l d a p idmap s u f f i x = ou=Idmap
l d a p admin dn = cn=Manager , dc=quenya , dc=o r g
When configuring a master LDAP server that will have slave LDAP servers,
do not forget to configure this in the /etc/openldap/slapd.conf file. It
must be noted that the DN of a server certificate must use the CN attribute
to name the server, and the CN must carry the servers’ fully qualified domain
name. Additional alias names and wildcards may be present in the subjec-
tAltName certificate extension. More details on server certificate names are
in RFC2830.
It does not really fit within the scope of this document, but a working LDAP
installation is basic to LDAP-enabled Samba operation. When using an
OpenLDAP server with Transport Layer Security (TLS), the machine name
in /etc/ssl/certs/slapd.pem must be the same as in /etc/openldap/
sldap.conf. The Red Hat Linux startup script creates the slapd.pem file
with hostname “localhost.localdomain.” It is impossible to access this LDAP
server from a slave LDAP server (i.e., a Samba BDC) unless the certificate
is re-created with a correct hostname.
86 Backup Domain Control Chapter 5
Do not install a Samba PDC so that is uses an LDAP slave server. Joining
client machines to the domain will fail in this configuration because the
change to the machine account in the LDAP tree must take place on the
master LDAP server. This is not replicated rapidly enough to the slave
server that the PDC queries. It therefore gives an error message on the
client machine about not being able to set up account credentials. The
machine account is created on the LDAP server, but the password fields will
be empty. Unfortunately, some sites are unable to avoid such configurations,
and these sites should review the ldap replication sleep parameter, intended
to slow down Samba sufficiently for the replication to catch up. This is
a kludge, and one that the administrator must manually duplicate in any
scripts (such as the add machine script) that they use.
Possible PDC/BDC plus LDAP configurations include:
• PDC+BDC -> One Central LDAP Server.
• PDC -> LDAP master server, BDC -> LDAP slave server.
• PDC -> LDAP master, with secondary slave LDAP server.
BDC -> LDAP master, with secondary slave LDAP server.
• PDC -> LDAP master, with secondary slave LDAP server.
BDC -> LDAP slave server, with secondary master LDAP server.
In order to have a fallback configuration (secondary) LDAP server, you
would specify the secondary LDAP server in the smb.conf file as shown in
Example 5.2.2.
Active Directory server. This means that Samba-3 also cannot act as a BDC
to an Active Directory domain controller.
Every machine that is a domain controller for the domain MIDEARTH has
to register the NetBIOS group name MIDEARTH<1C> with the WINS
server and/or by broadcast on the local network. The PDC also registers
the unique NetBIOS name MIDEARTH<1B> with the WINS server. The
name type <1B> name is normally reserved for the Domain Master Browser
(DMB), a role that has nothing to do with anything related to authentica-
tion, but the Microsoft domain implementation requires the DMB to be on
the same machine as the PDC.
Where a WINS server is not used, broadcast name registrations alone must
suffice. Refer to Chapter 9, “Network Browsing”,Section 9.3 for more in-
formation regarding TCP/IP network protocols and how SMB/CIFS names
are handled.
Where NetBIOS over TCP/IP is disabled, all name resolution involves the
use of DNS, broadcast messaging over UDP, as well as Active Directory com-
munication technologies. In this type of environment all machines require
appropriate DNS entries. More information may be found in Section 9.3.3.
selected domain controller authenticate each other. After that the worksta-
tion sends the user’s credentials (name and password) to the local domain
controller for validation.
The creation of a BDC requires some steps to prepare the Samba server
before smbd is executed for the first time. These steps are as follows:
• The domain SID has to be the same on the PDC and the BDC. In
Samba versions pre-2.2.5, the domain SID was stored in the file pri-
vate/MACHINE.SID. For all versions of Samba released since 2.2.5 the
domain SID is stored in the file private/secrets.tdb. This file is
unique to each server and cannot be copied from a PDC to a BDC;
the BDC will generate a new SID at startup. It will overwrite the PDC
domain SID with the newly created BDC SID. There is a procedure
that will allow the BDC to aquire the domain SID. This is described
here.
To retrieve the domain SID from the PDC or an existing BDC and
store it in the secrets.tdb, execute:
• The UNIX user database has to be synchronized from the PDC to the
BDC. This means that both the /etc/passwd and /etc/group have
to be replicated from the PDC to the BDC. This can be done manually
whenever changes are made. Alternately, the PDC is set up as an NIS
master server and the BDC as an NIS slave server. To set up the BDC
as a mere NIS client would not be enough, as the BDC would not be
able to access its user database in case of a PDC failure. NIS is by no
means the only method to synchronize passwords. An LDAP solution
would also work.
• The Samba password database must be replicated from the PDC to the
BDC. Although it is possible to synchronize the smbpasswd file with
rsync and ssh, this method is broken and flawed, and is therefore
not recommended. A better solution is to set up slave LDAP servers
for each BDC and a master LDAP server for the PDC. The use of
rsync is inherently flawed by the fact that the data will be replicated
at timed intervals. There is no guarantee that the BDC will be oper-
ating at all times with correct and current machine and user account
information. This means that this method runs the risk of users being
inconvenienced by discontinuity of access to network services due to
inconsistent security data. It must be born in mind that Windows
workstations update (change) the machine trust account password at
regular intervals — administrators are not normally aware that this is
happening or when it takes place.
The use of LDAP for both the POSIX (UNIX user and group) accounts
and for the SambaSAMAccount data automatically ensures that all
account change information will be written to the shared directory.
This eliminates the need for any special action to synchronize account
information because LDAP will meet that requirement.
• The netlogon share has to be replicated from the PDC to the BDC.
This can be done manually whenever login scripts are changed, or it
can be done automatically using a cron job that will replicate the
directory structure in this share using a tool like rsync. The use of
rsync for replication of the netlogon data is not critical to network
security and is one that can be manually managed given that the
administrator will make all changes to the netlogon share as part of a
conscious move.
90 Backup Domain Control Chapter 5
Note
Domain control was a new area for Samba, but there are now many exam-
ples that we may refer to. Updated information will be published as they
become available and may be found in later Samba releases or from the
Samba Web site4 ; refer in particular to the WHATSNEW.txt in the Samba
release tarball. The book, “Samba-3 by Example” documents well tested
and proven configuration examples. You can obtain a copy of this book5 for
the Samba web site.
4
<http://samba.org>
5
<http://www.samba.org/samba/docs/Samba3-ByExample.pdf>
92 Backup Domain Control Chapter 5
This problem will occur when the passdb (SAM) files are copied from a
central server but the local BDC is acting as a PDC. This results in the
application of Local Machine Trust Account password updates to the local
SAM. Such updates are not copied back to the central server. The newer
machine account password is then overwritten when the SAM is recopied
from the PDC. The result is that the domain member machine on startup
will find that its passwords do not match the one now in the database, and
since the startup security check will now fail, this machine will not allow
logon attempts to proceed and the account expiry error will be reported.
The solution is to use a more robust passdb backend, such as the ldapsam
backend, setting up a slave LDAP server for each BDC and a master LDAP
server for the PDC.
No. The native NT4 SAM replication protocols have not yet been fully
implemented.
Can I get the benefits of a BDC with Samba? Yes, but only to a Samba
PDC.The main reason for implementing a BDC is availability. If the PDC
is a Samba machine, a second Samba machine can be set up to service logon
requests whenever the PDC is down.
As said a few times before, use of this method is broken and flawed. Machine
trust accounts will go out of sync, resulting in a broken domain. This method
is not recommended. Try using LDAP instead.
The simple answer is yes. Samba’s pdb ldap code supports binding to a
replica LDAP server and will also follow referrals and rebind to the master
if it ever needs to make a modification to the database. (Normally BDCs
are read-only, so this will not occur often).
Chapter 6
DOMAIN MEMBERSHIP
94
Section 6.2. MS Windows Workstation/Server Machine Trust Accounts 95
The older format of this data is the smbpasswd database that contains
the UNIX login ID, the UNIX user identifier (UID), and the LanMan
and NT-encrypted passwords. There is also some other information in
this file that we do not need to concern ourselves with here.
The two newer database types are called ldapsam and tdbsam. Both
store considerably more data than the older smbpasswd file did. The
extra information enables new user account controls to be implemented.
• Manual creation from the UNIX/Linux command line. Here, both the
Samba and corresponding UNIX account are created by hand.
root# chpass -a \
’machine_name$:*:101:100::0:0:Windows machine_name:/dev/null:/sbin/nologin’
The /etc/passwd entry will list the machine name with a “$” appended,
and will not have a password, will have a null shell and no home directory.
For example, a machine named “doppy” would have an /etc/passwd entry
like this:
doppy$:x:505:100:machine_nickname:/dev/null:/bin/false
in which machine nickname can be any descriptive name for the client,
such as BasementComputer. machine name absolutely must be the NetBIOS
name of the client to be joined to the domain. The “$” must be appended
to the NetBIOS name of the client or Samba will not recognize this as a
Machine Trust Account.
Now that the corresponding UNIX account has been created, the next step is
to create the Samba account for the client containing the well-known initial
98 Domain Membership Chapter 6
Machine Trust Account password. This can be done using the smbpasswd
command as shown here:
where machine name is the machine’s NetBIOS name. The RID of the new
machine account is generated from the UID of the corresponding UNIX
account.
should download the Nexus.exe package from the Microsoft Web site.
When executed from the target directory, it will unpack the same tools
but for use on this platform.
Further information about these tools may be obtained from the following
locations:
Knowledge Base article 1736731
Knowledge Base article 1725402
Launch the srvmgr.exe (Server Manager for Domains) and follow these
steps: Server Manager Account Machine Account Management
1. From the menu select Computer.
2. Click Select Domain.
3. Click the name of the domain you wish to administer in the Select
Domain panel and then click OK.
4. Again from the menu select Computer.
5. Select Add to Domain.
6. In the dialog box, click the radio button to Add NT Workstation of
Server, then enter the machine name in the field provided, and click
the Add button.
When the user elects to make the client a domain member, Windows 200x
prompts for an account and password that has privileges to create machine
accounts in the domain. A Samba administrator account (i.e., a Samba
account that has root privileges on the Samba server) must be entered
here; the operation will fail if an ordinary user account is given.
For security reasons, the password for this administrator account should be
set to a password that is other than that used for the root user in /etc/
passwd.
The name of the account that is used to create domain member machine
trust accounts can be anything the network administrator may choose. If it
is other than root, then this is easily mapped to root in the file named in
the smb.conf parameter username map = /etc/samba/smbusers.
The session key of the Samba administrator account acts as an encryption
key for setting the password of the machine trust account. The Machine
Trust Account will be created on-the-fly, or updated if it already exists.
as above for Windows 2000 (i.e., you must supply a Samba administrator
account when prompted).
This mode of server operation involves the Samba machine being made a
member of a domain security context. This means by definition that all
user authentication will be done from a centrally defined authentication
regime. The authentication regime may come from an NT3/4-style (old
domain technology) server, or it may be provided from an Active Directory
server (ADS) running on MS Windows 2000 or later.
Note
Table 6.1 lists names that are used in the remainder of this chapter.
First, you must edit your smb.conf file to tell Samba it should now use
domain security.
Change (or add) your security line in the [global] section of your smb.conf
to read:
s e c u r i t y = domain
Note that if the parameter security = user is used this machine would
function as a standalone server and not as a domain member server. Domain
security mode causes Samba to work within the domain security context.
Next change the workgroup line in the [global] section to read:
workgroup = MIDEARTH
These are the PDC and BDCs Samba will attempt to contact in order to au-
thenticate users. Samba will try to contact each of these servers in order, so
you may want to rearrange this list in order to spread out the authentication
load among Domain Controllers.
Section 6.3. Domain Member Server 103
This method allows Samba to use exactly the same mechanism that NT
does. The method either uses broadcast-based name resolution, performs a
WINS database lookup in order to find a domain controller against which to
authenticate, or locates the domain controller using DNS name resolution.
To join the domain, run this command:
If the -S DOMPDC argument is not given, the domain name will be obtained
from smb.conf and the NetBIOS name of the PDC will be obtained either
using a WINS lookup or via NetBIOS broadcast based name look up.
The machine is joining the domain DOM, and the PDC for that domain
(the only machine that has write access to the domain SAM database) is
DOMPDC; therefore, use the -S option. The Administrator%password is
the login name and password for an account that has the necessary privilege
to add machines to the domain. If this is successful, you will see the follow-
ing message in your terminal window. Where the older NT4-style domain
architecture is used:
Where Active Directory is used the command used to join the ADS domain
is:
Refer to the net man page and to Chapter 12, “Remote and Local Manage-
ment: The Net Command” for further information.
This process joins the server to the domain without separately having to
create the machine trust account on the PDC beforehand.
This command goes through the machine account password change protocol,
then writes the new (random) machine account password for this Samba
server into a file in the same directory in which a smbpasswd file would be
normally stored. The trust account information that is needed by the DMS
is written into the file /usr/local/samba/private/secrets.tdb or /etc/
samba/secrets.tdb.
This file is created and owned by root and is not readable by any other
user. It is the key to the domain-level security for your system and should
be treated as carefully as a shadow password file.
Finally, restart your Samba daemons and get ready for clients to begin using
domain security. The way you can restart your Samba daemons depends on
your distribution, but in most cases the following will suffice:
Currently, domain security in Samba does not free you from having to create
local UNIX users to represent the users attaching to your server. This means
that if domain user DOM\fred attaches to your domain security Samba server,
there needs to be a local UNIX user fred to represent that user in the UNIX
file system. This is similar to the older Samba security mode security =
server, where Samba would pass through the authentication request to a
Windows NT server in the same way as a Windows 95 or Windows 98
server would.
Please refer to Chapter 23, “Winbind: Use of Domain Accounts”, for in-
formation on a system to automatically assign UNIX UIDs and GIDs to
Windows NT domain users and groups.
The advantage of domain-level security is that the authentication in domain-
level security is passed down the authenticated RPC channel in exactly the
same way that an NT server would do it. This means Samba servers now
Section 6.4. Samba ADS Domain Membership 105
Note
a
<http://www.linuxworld.com>
realm = your . k e r b e r o s .REALM
s e c u r i t y = ADS
# The f o l l o w i n g parameter need o n l y be s p e c i f i e d ←-
i f present .
# The d e f a u l t s e t t i n g i f not p r e s e n t i s Yes .
e n c r y p t passwords = y e s
In case samba cannot correctly identify the appropriate ADS server using
the realm name, use the password server option in smb.conf:
password s e r v e r = your . k e r b e r o s . s e r v e r
The most common reason for which Samba may not be able to locate the
ADS domain controller is a consequence of sites maintaining some DNS
servers on UNIX systems without regard for the DNS requirements of the
ADS infrastructure. There is no harm in specifying a preferred ADS DC
using the password server.
Note
[libdefaults]
default_realm = YOUR.KERBEROS.REALM
[realms]
YOUR.KERBEROS.REALM = {
kdc = your.kerberos.server
}
[domain_realms]
.kerberos.server = YOUR.KERBEROS.REALM
When using Heimdal versions before 0.6, use the following configuration
settings:
[libdefaults]
default_realm = YOUR.KERBEROS.REALM
default_etypes = des-cbc-crc des-cbc-md5
default_etypes_des = des-cbc-crc des-cbc-md5
[realms]
3
<http://www.microsoft.com/windows2000/techinfo/planning/security/
kerbsteps.asp>
4
<http://www.ietf.org/rfc/rfc1510.txt?number=1510>
108 Domain Membership Chapter 6
YOUR.KERBEROS.REALM = {
kdc = your.kerberos.server
}
[domain_realms]
.kerberos.server = YOUR.KERBEROS.REALM
Test your config by doing a kinit USERNAME@REALM and making sure that
your password is accepted by the Win2000 KDC.
With Heimdal versions earlier than 0.6.x you can use only newly created
accounts in ADS or accounts that have had the password changed once after
migration, or in case of Administrator after installation. At the moment,
a Windows 2003 KDC can only be used with Heimdal releases later than
0.6 (and no default etypes in krb5.conf). Unfortunately, this whole area is
still in a state of flux.
Note
Note
Clock skew limits are configurable in the Kerberos protocols. The default
setting is five minutes.
You also must ensure that you can do a reverse DNS lookup on the IP
Section 6.4. Samba ADS Domain Membership 109
address of your KDC. Also, the name that this reverse lookup maps to must
either be the NetBIOS name of the KDC (i.e., the hostname with no domain
attached) or it can be the NetBIOS name followed by the realm.
The easiest way to ensure you get this right is to add a /etc/hosts entry
mapping the IP address of your KDC to its NetBIOS name. If you do not
get this correct, then you will get a local error when you try to join the
realm.
If all you want is Kerberos support in smbclient, then you can skip directly
to Section 6.4.5 now. Section 6.4.3 and Section 6.4.4 are needed only if you
want Kerberos support for smbd and winbindd.
As a user who has write permission on the Samba private directory (usually
root), run:
The Administrator account can be any account that has been designated
in the ADS domain security settings with permission to add machines to
the ADS domain. It is, of course, a good idea to use an account other
than Administrator. On the UNIX/Linux system, this command must be
executed by an account that has UID=0 (root).
When making a Windows client a member of an ADS domain within a
complex organization, you may want to create the machine trust account
within a particular organizational unit. Samba-3 permits this to be done
using the following syntax:
Your ADS manager will be able to advise what should be specified for the
”organizational unit” parameter.
For example, you may want to create the machine trust account in a con-
tainer called “Servers” under the organizational directory “Computers\BusinessUnit\Department,”
110 Domain Membership Chapter 6
like this:
This command will place the Samba server machine trust account in the
container Computers\BusinessUnit\Department\Servers. The container
should exist in the ADS directory before executing this command.
net ads join prompts for user name You need to log in to the domain
using kinit USERNAME@REALM. USERNAME must be a user who has rights
to add a machine to the domain.
If the join was successful, you will see a new computer account with the Net-
BIOS name of your Samba server in Active Directory (in the “Computers”
folder under Users and Computers.
Note
On your Samba server try to log in to a Windows 2000 server or your Samba
server using smbclient and Kerberos. Use smbclient as usual, but specify
the -k option to choose Kerberos authentication.
6.4.6 Notes
You must change the administrator password at least once after installing a
domain controller, to create the right encryption types.
Windows 200x does not seem to create the kerberos. udp and ldap. tcp
in the default DNS setup. Perhaps this will be fixed later in service packs.
Samba maps UNIX users and groups (identified by UIDs and GIDs) to
Windows users and groups (identified by SIDs). These mappings are done
by the idmap subsystem of Samba.
In some cases it is useful to share these mappings between Samba domain
members, so name->id mapping is identical on all machines. This may be
needed in particular when sharing files over both CIFS and NFS.
To use the LDAP ldap idmap suffix, set:
l d a p idmap s u f f i x = ou=Idmap , dc=quenya , dc= ←-
org
See the smb.conf man page entry for the ldap idmap suffix parameter for
further information.
112 Domain Membership Chapter 6
Do not forget to specify also the ldap admin dn and to make certain to set
the LDAP administrative password into the secrets.tdb using:
The original name is still in the NetBIOS name cache and must expire
after machine account deletion before adding that same name as a domain
member again. The best advice is to delete the old account and then add the
machine with a new name. Alternately, the name cache can be flished and
reloaded with current data using the nbtstat command on the Windows
client:
C:\> nbtstat -R
Section 6.6. Common Errors 113
Windows 2003 requires SMB signing. Client-side SMB signing has been
implemented in Samba-3.0. Set client use spnego = yes when communicating
with a Windows 2003 server. This will not interfere with other Windows
clients that do not support the more advanced security features of Windows
2003 because the client will simply negotiate a protocol tha both it and the
server suppport. This is a well-know fall-back facility that is built into the
SMB/CIFS protocols.
Chapter 7
STANDALONE SERVERS
If all that is needed is a server for read-only files, or for printers alone, it
may not make sense to effect a complex installation. For example, a drafting
office needs to store old drawings and reference standards. Noone can write
files to the server because it is legislatively important that all documents
remain unaltered. A share-mode read-only standalone server is an ideal
solution.
Another situation that warrants simplicity is an office that has many printers
that are queued off a single central server. Everyone needs to be able to print
to the printers, there is no need to effect any access controls, and no files
will be served from the print server. Again, a share-mode standalone server
makes a great solution.
114
Section 7.2. Background 115
7.2 Background
The term standalone server means that it will provide local authentication
and access control for all resources that are available from it. In general this
means that there will be a local user database. In more technical terms, it
means resources on the machine will be made available in either share mode
or in user mode.
No special action is needed other than to create user accounts. Standalone
servers do not provide network logon services. This means that machines
that use this server do not perform a domain logon to it. Whatever logon
facility the workstations are subject to is independent of this machine. It is,
however, necessary to accommodate any network user so the logon name he
or she uses will be translated (mapped) locally on the standalone server to
a locally known user name. There are several ways this can be done.
Samba tends to blur the distinction a little in defining a standalone server.
This is because the authentication database may be local or on a remote
server, even if from the SMB protocol perspective the Samba server is not
a member of a domain security context.
Through the use of Pluggable Authentication Modules (PAM) (see Chap-
ter 27, “PAM-Based Distributed Authentication”) and the name service
switcher (NSS), which maintains the UNIX-user database, the source of au-
thentication may reside on another server. We would be inclined to call
this the authentication server. This means that the Samba server may use
the local UNIX/Linux system password database (/etc/passwd or /etc/
shadow), may use a local smbpasswd file, or may use an LDAP backend, or
even via PAM and Winbind another CIFS/SMB server for authentication.
conf file. Example 7.3.1 is the smb.conf file that will do this. Assume that
all the reference documents are stored in the directory /export, and the
documents are owned by a user other than nobody. No home directories are
shared, and there are no users in the /etc/passwd UNIX system database.
This is a simple system to administer.
I would have spoken more briefly, if I’d had more time to prepare.
—Mark Twain
In Example 7.3.1, the machine name is set to GANDALF, and the workgroup
is set to the name of the local workgroup (MIDEARTH) so the machine will
appear together with systems with which users are familiar. The only pass-
word backend required is the “guest” backend to allow default unprivileged
account names to be used. As there is a WINS server on this network, we
of course make use of it.
A USAF Colonel was renowned for saying: “Better is the enemy of good
enough!” There are often sound reasons for avoiding complexity as well as
for avoiding a technically perfect solution. Unfortunately, many network
administrators still need to learn the art of doing just enough to keep out
of trouble.
Section 7.3. Example Configuration 117
Configuration of a simple print server is easy if you have all the right tools
on your system.
Assumptions
1. The print server must require no administration.
2. The print spooling and processing system on our print server will be
CUPS. (Please refer to Chapter 21, “CUPS Printing Support”, for
more information).
3. The print server will service only network printers. The network ad-
ministrator will correctly configure the CUPS environment to support
the printers.
4. All workstations will use only PostScript drivers. The printer driver
of choice is the one shipped with the Windows OS for the Apple Color
LaserWriter.
In this example our print server will spool all incoming print jobs to /var/
spool/samba until the job is ready to be submitted by Samba to the CUPS
print processor. Since all incoming connections will be as the anonymous
(guest) user, two things will be required to enable anonymous printing.
Enabling Anonymous Printing
• The UNIX/Linux system must have a guest account. The default for
this is usually the account nobody. To find the correct name to use
for your version of Samba, do the following:
Make sure that this account exists in your system password database
(/etc/passwd).
It is a good idea either to set a password on this account, or else to
lock it from UNIX use. Assuming that the guest account is called
pcguest, it can be locked by excuting:
• The directory into which Samba will spool the file must have write
access for the guest account. The following commands will ensure
that this directory is available for use:
Note
The example in Example 7.3.2 uses CUPS for direct printing via the CUPS
libarary API. This means that all printers will be exposed to Windows users
without need to configure a printcap file. If there is necessity to expose
only a sub-set of printers, or to define a special type of printer (for example,
a PDF filter) the printcap name = cups can be replaced with the entry
printcap name = /etc/samba/myprintcap. In this case the file specified
should contain a list of the printer names that should be exposed to Windows
network users.
MS WINDOWS NETWORK
CONFIGURATION GUIDE
120
Section 8.2. Technical Details 121
The builder of a house must ensure that all construction takes place on a
firm foundation. The same is true for the builder of a TCP/IP-based net-
working system. Fundamental network configuration problems will plague
all network users until they are resolved.
MS Windows workstations and servers can be configured either with fixed
IP addresses or via DHCP. The examples that follow demonstrate the use
of DHCP and make only passing reference to those situations where fixed
IP configuration settings can be effected.
It is possible to use shortcuts or abbreviated keystrokes to arrive at a par-
ticular configuration screen. The decision was made to base all examples in
this chapter on use of the Start button.
Fixed settings may be required for DNS and WINS if these settings
are not provided automatically via DHCP.
4. Click the DNS tab to add DNS server settings. The example sys-
tem uses manually configured DNS settings. When finished making
changes, click the OK to commit the settings. See Figure 8.4.
5. Click the WINS tab to add manual WINS server entries. This step
Section 8.2. Technical Details 123
There are two paths to the Windows 2000 Professional TCP/IP configura-
tion panel. Choose the access method that you prefer:
Click Start -> Control Panel -> Network and Dial-up Connections.
Alternatively, click Start, then right-click My Network Places, and select
Properties.
The following procedure steps through the Windows XP Professional TCP/IP
configuration process:
1. Right-click on Local Area Connection, then click Properties. See
Figure 8.6.
2. The Local Area Connection Properties is used to set TCP/IP protocol
124 MS Windows Network Configuration Guide Chapter 8
8.2.1.3 MS Windows Me
There are two paths to the Windows Millennium edition (Me) TCP/IP
configuration panel. Choose the access method that you prefer:
Click Start -> Control Panel -> Network Connections.
Alternatively, click on Start ->, and right click on My Network Places then
126 MS Windows Network Configuration Guide Chapter 8
select Properties.
The following procedure steps through the Windows Me TCP/IP configu-
ration process:
1. In the box labeled The following network components are installed:,
click on Internet Protocol TCP/IP, then click on the Properties but-
ton. See Figure 8.11.
2. Many network administrators will want to use DHCP to configure
all client TCP/IP protocol stack settings. (For information on how
to configure the ISC DHCP server for Windows client support see
XrefId[?the DNS and DHCP Configuration Guide?], XrefId[?DHCP
Server?]. The default setting on Windows Me workstations is for
DHCP-enabled operation (i.e., Obtain IP address automatically is
enabled). See Figure 8.12.
If it is necessary to provide a fixed IP address, click on Specify an IP
address and enter the IP Address and the subnet mask in the boxes
provided. For this example we are assuming that all network clients
Section 8.2. Technical Details 127
3. Fixed settings may be required for DNS and WINS if these settings
are not provided automatically via DHCP.
4. If necessary, click the DNS Configuration tab to add DNS server set-
tings. Click the WINS Configuration tab to add WINS server settings.
The Gateway tab allows additional gateways (router addresses) to be
added to the network interface settings. In most cases where DHCP
is used, it will not be necessary to create these manual settings.
1. Click Start.
3. The opening panel is the same one that can be reached by clicking
System on the Control Panel. See Figure 8.15.
4. Click the Computer Name tab. This panel shows the Computer De-
scription, the Full computer name, and the Workgroup or Domain
name. Clicking the Network ID button will launch the configuration
wizard. Do not use this with Samba-3. If you wish to change the
computer name or join or leave the domain, click the Change button.
See Figure 8.16.
Section 8.2. Technical Details 129
5. Click on Change. This panel shows that our example machine (TEMP-
TATION) is in a workgroup called WORKGROUP. We will join the
domain called MIDEARTH. See Figure 8.17.
6. Enter the name MIDEARTH in the field below the domain radio but-
ton. This panel shows that our example machine (TEMPTATION) is
set to join the domain called MIDEARTH. See Figure 8.18.
7. Now click the OK button. A dialog box should appear to allow you to
provide the credentials (username and password) of a domain admin-
istrative account that has the rights to add machines to the domain.
Enter the name “root” and the root password from your Samba-3
server. See Figure 8.19.
Note
Make sure that the Client for Microsoft Networks driver is installed
as shown. Click on the Client for Microsoft Networks entry in The
following network components are installed: box. Then click the
Properties button.
3. The Client for Microsoft Networks Properties panel is the correct lo-
cation to configure network logon settings. See Figure 8.21.
5. Now click the Access Control button. If you want to be able to assign
share access permissions using domain user and group accounts, it is
necessary to enable User-level access control as shown in this panel.
See Figure 8.23.
132 MS Windows Network Configuration Guide Chapter 8
The most common errors that can afflict Windows networking systems in-
clude:
• Incorrect IP address.
• Incorrect or inconsistent netmasks.
• Incorrect router address.
• Incorrect DNS server address.
• Incorrect WINS server address.
• Use of a Network Scope setting — watch out for this one!
The most common reasons for which a Windows NT/200x/XP Professional
client cannot join the Samba controlled domain are:
• smb.conf does not have correct add machine script settings.
• “root” account is not in password backend database.
Section 8.3. Common Errors 133
Advanced Configuration
VALUABLE NUTS AND
BOLTS INFORMATION
Samba has several features that you might want or might not want to use.
The chapters in this part each cover specific Samba features.
141
Chapter 9
NETWORK BROWSING
Note
Charles Dickens once referred to the past in these words: “It was the best
of times, it was the worst of times.” The more we look back, the more we
long for what was and hope it never returns.
142
Section 9.2. What Is Browsing? 143
WINS is exclusively a service that applies only to those systems that run
NetBIOS over TCP/IP. MS Windows 200x/XP have the capacity to operate
with support for NetBIOS disabled, in which case WINS is of no relevance.
Samba supports this also.
For those networks on which NetBIOS has been disabled (i.e., WINS is not
required), the use of DNS is necessary for hostname resolution.
To most people, browsing means they can see the MS Windows and Samba
servers in the Network Neighborhood, and when the computer icon for a
particular server is clicked, it opens up and shows the shares and printers
available on the target server.
144 Network Browsing Chapter 9
• wins support(*)
• wins hook
Those marked with an (*) are the only options that commonly may need to
be modified. Even if none of these parameters is set, nmbd will still do its
job.
For Samba, the WINS Server and WINS Support are mutually exclusive
options. When nmbd is started it will fail to execute if both options are
set in the smb.conf file. The nmbd understands that when it spawns an
instance of itself to run as a WINS server that it has to use its own WINS
server also.
9.3 Discussion
• b-node (type 0x01): The Windows client will use only NetBIOS broad-
cast requests using UDP broadcast.
146 Network Browsing Chapter 9
• p-node (type 0x02): The Windows client will use point-to-point (Net-
BIOS unicast) requests using UDP unicast directed to a WINS server.
• m-node (type 0x04): The Windows client will first use NetBIOS broad-
cast requests using UDP broadcast, then it will use (NetBIOS unicast)
requests using UDP unicast directed to a WINS server.
• h-node (type 0x08): The Windows client will use (NetBIOS unicast)
requests using UDP unicast directed to a WINS server, then it will
use NetBIOS broadcast requests using UDP broadcast.
The default Windows network client (or server) network configuration en-
ables NetBIOS over TCP/IP and b-node configuration. The use of WINS
makes most sense with h-node (hybrid mode) operation so that in the event
of a WINS breakdown or non-availability, the client can use broadcast-based
name resolution.
In those networks where Samba is the only SMB server technology, wherever
possible nmbd should be configured on one machine as the WINS server. This
makes it easy to manage the browsing environment. If each network segment
is configured with its own Samba WINS server, then the only way to get
cross-segment browsing to work is by using the remote announce and the
remote browse sync parameters to your smb.conf file.
If only one WINS server is used for an entire multisegment network, then the
use of the remote announce and the remote browse sync parameters should
not be necessary.
As of Samba-3, WINS replication is being worked on. The bulk of the code
has been committed, but it still needs maturation. This is not a supported
feature of the Samba-3.0.20 release. Hopefully, this will become a supported
feature of one of the Samba-3 release series. The delay is caused by the fact
that this feature has not been of sufficient significance to inspire someone to
pay a developer to complete it.
Right now Samba WINS does not support MS-WINS replication. This
means that when setting up Samba as a WINS server, there must only
be one nmbd configured as a WINS server on the network. Some sites have
used multiple Samba WINS servers for redundancy (one server per subnet)
and then used remote browse sync and remote announce to effect browse
list collation across all segments. Note that this means clients will only re-
solve local names and must be configured to use DNS to resolve names on
other subnets in order to resolve the IP addresses of the servers they can
Section 9.3. Discussion 147
Windows 200x/XP can register its hostname with a Dynamic DNS server
(DDNS). It is possible to force register with a dynamic DNS server in Win-
dows 200x/XP using ipconfig /registerdns.
Use of raw SMB over TCP/IP (No NetBIOS layer) can be done only with
Active Directory domains. Samba is not an Active Directory domain con-
troller: ergo, it is not possible to run Samba as a domain controller and
at the same time not use NetBIOS. Where Samba is used as an Active Di-
rectory domain member server (DMS) it is possible to configure Samba to
not use NetBIOS over TCP/IP. A Samba DMS can integrate fully into an
Active Directory domain, however, if NetBIOS over TCP/IP is disabled it is
necessary manually to create appropriate DNS entries for the Samba DMS
because they will not be automatically generated either by Samba, or by
the ADS environment.
• gc. tcp.quenya.org — Used to locate the global catalog server for the
top of the domain. This must list port 3268.
150 Network Browsing Chapter 9
The following records are also used by the Windows domain member client
to locate vital services on the Windows ADS domain controllers.
• ldap. tcp.pdc. msdcs.quenya.org
• ldap.gc. msdcs.quenya.org
• ldap.default-first-site-name. sites.gc. msdcs.quenya.org
• ldap.{SecID}.domains. msdcs.quenya.org
• ldap. tcp.dc. msdcs.quenya.org
• kerberos. tcp.dc. msdcs.quenya.org
• ldap.default-first-site-name. sites.dc. msdcs.quenya.org
• kerberos.default-first-site-name. sites.dc. msdcs.queyna.org
• SecID. msdcs.quenya.org
Presence of the correct DNS entries can be validated by executing:
;; QUESTION SECTION:
;_ldap._tcp.dc._msdcs.quenya.org. IN ANY
;; ANSWER SECTION:
_ldap._tcp.dc._msdcs.quenya.org. 600 IN SRV 0 100 389 frodo.quenya.org.
_ldap._tcp.dc._msdcs.quenya.org. 600 IN SRV 0 100 389 noldor.quenya.org.
;; ADDITIONAL SECTION:
frodo.quenya.org. 3600 IN A 10.1.1.16
noldor.quenya.org. 1200 IN A 10.1.1.17
Section 9.4. How Browsing Functions 151
WINS server using the name of the domain and the NetBIOS name type 1B.
e.g., DOMAIN<1B>. All LMBs register their IP address with the WINS
server, also with the name of the domain and the NetBIOS name type of 1D.
The 1B name is unique to one server within the domain security context,
and only one 1D name is registered for each network segment. Machines that
have registered the 1D name will be authoritive browse list maintainers for
the network segment they are on. The DMB is responsible for synchronizing
the browse lists it obtains from the LMBs.
Clients wishing to browse the network make use of this list but also depend
on the availability of correct name resolution to the respective IP address
or addresses.
The DMB should preferably be the LMB for its own subnet. In order to
achieve this, set the following options in the [global] section of the smb.conf
file as shown in Example 9.4.1
The DMB may be the same machine as the WINS server, if necessary.
Next, you should ensure that each of the subnets contains a machine that can
act as an LMB for the workgroup. Any MS Windows NT/200x/XP machine
should be able to do this, as will Windows 9x/Me machines (although these
tend to get rebooted more often, so it is not such a good idea to use them).
To make a Samba server an LMB, set the following options in the [global]
section of the smb.conf file as shown in Example 9.4.2
Do not do this for more than one Samba server on each subnet, or they will
war with each other over which is to be the LMB.
The local master parameter allows Samba to act as a LMB. The preferred
154 Network Browsing Chapter 9
master causes nmbd to force a browser election on startup and the os level
parameter sets Samba high enough so it should win any browser elections.
If you have an NT machine on the subnet that you wish to be the LMB, you
can disable Samba from becoming an LMB by setting the following options
in the [global] section of the smb.conf file as shown in Example 9.4.3.
If you are adding Samba servers to a Windows NT domain, then you must
not set up a Samba server as a DMB. By default, a Windows NT PDC for
a domain is also the DMB for that domain. Network browsing may break
if a Samba server other than the PDC registers the DMB NetBIOS name
(DOMAIN<1B>) with WINS.
For subnets other than the one containing the Windows NT PDC, you may
set up Samba servers as LMBs as described. To make a Samba server a
Local Master Browser, set the following options in the [global] section of the
smb.conf file as shown in Example 9.4.4
If you wish to have a Samba server fight the election with machines on the
same subnet, you may set the os level parameter to lower levels. By doing
Section 9.4. How Browsing Functions 155
this you can tune the order of machines that will become LMBs if they are
running. For more details on this, refer to Section 9.4.3.
If you have Windows NT machines that are members of the domain on all
subnets and you are sure they will always be running, you can disable Samba
from taking part in browser elections and ever becoming an LMB by setting
the following options in the [global] section of the smb.conf file as shown in
Example 9.4.5
If you want Samba to win elections, set the os level global option in smb.
conf to a higher number. It defaults to 20. Using 34 would make it win all
elections over every other system (except other Samba systems).
An os level of two would make it beat Windows for Workgroups and Win-
dows 9x/Me, but not MS Windows NT/200x Server. An MS Windows
NT/200x Server domain controller uses level 32. The maximum os level is
255.
If you want Samba to force an election on startup, set the preferred master
global option in smb.conf to yes. Samba will then have a slight advantage
over other potential master browsers that are not preferred master browsers.
Use this parameter with care, because if you have two hosts (whether they
are Windows 9x/Me or NT/200x/XP or Samba) on the same local subnet
both set with preferred master to yes, then periodically and continually
they will force an election in order to become the LMB.
If you want Samba to be a DMB, then it is recommended that you also set
preferred master to yes, because Samba will not become a DMB for the
156 Network Browsing Chapter 9
whole of your LAN or WAN if it is not also a LMB on its own broadcast
isolated subnet.
It is possible to configure two Samba servers to attempt to become the
DMB for a domain. The first server that comes up will be the DMB. All
other Samba servers will attempt to become the DMB every 5 minutes.
They will find that another Samba server is already the DMB and will
fail. This provides automatic redundancy should the current DMB fail.
The network bandwidth overhead of browser elections is relatively small,
requiring approximately four UDP packets per machine per election. The
maximum size of a UDP packet is 576 bytes.
The domain master browser is responsible for collating the browse lists of
multiple subnets so browsing can occur between subnets. You can make
Samba act as the domain master browser by setting domain master = yes
in smb.conf. By default it will not be a domain master browser.
Do not set Samba to be the domain master for a workgroup that has the
same name as an NT/200x domain. If Samba is configured to be the domain
master for a workgroup that is present on the same network as a Windows
NT/200x domain that has the same name, network browsing problems will
certainly be experienced.
When Samba is the domain master and the master browser, it will listen
for master announcements (made roughly every 12 minutes) from LMBs on
other subnets and then contact them to synchronize browse lists.
If you want Samba to be the domain master, you should also set the os level
high enough to make sure it wins elections, and set preferred master to yes,
to get Samba to force an election on startup.
All servers (including Samba) and clients should be using a WINS server
to resolve NetBIOS names. If your clients are only using broadcasting to
resolve NetBIOS names, then two things will occur:
1. LMBs will be unable to find a DMB because they will be looking only
on the local subnet.
2. If a client happens to get hold of a domain-wide browse list and a user
attempts to access a host in that list, it will be unable to resolve the
NetBIOS name of that host.
Section 9.4. How Browsing Functions 157
If, however, both Samba and your clients are using a WINS server, then:
1. LMBs will contact the WINS server and, as long as Samba has regis-
tered that it is a DMB with the WINS server, the LMB will receive
Samba’s IP address as its DMB.
2. When a client receives a domain-wide browse list and a user attempts
to access a host in that list, it will contact the WINS server to resolve
the NetBIOS name of that host. As long as that host has registered
its NetBIOS name with the same WINS server, the user will be able
to see that host.
or
remote announce = 1 9 2 . 1 6 8 . 1 2 . 2 3 /MIDEARTH ←-
[ 1 7 2 . 1 6 . 2 1 . 2 5 5 /ELVINDORF] . . .
where:
WORKGROUP is optional and can be either our own workgroup or that of the
remote network. If you use the workgroup name of the remote network,
our NetBIOS machine names will end up looking like they belong to
that workgroup. This may cause name resolution problems and should
be avoided.
remote browse sync
so WINS may never get to know about it. In any case, machines that have
not registered with a WINS server will fail name-to-address lookup attempts
by other clients and will therefore cause workstation access errors.
To configure Samba as a WINS server, just add wins support = yes to the
smb.conf file [global] section.
To configure Samba to register with a WINS server, just add wins server =
10.0.0.18 to your smb.conf file [global] section.
Important
Versions of Samba prior to 1.9.17 had this parameter default to yes. If you
have any older versions of Samba on your network, it is strongly suggested
you upgrade to a recent version, or at the very least set the parameter to
“no” on all these machines.
Machines configured with wins support = yes will keep a list of all NetBIOS
names registered with them, acting as a DNS for NetBIOS names.
It is strongly recommended to set up only one WINS server. Do not set the
wins support = yes option on more than one Samba server on a network.
To configure Windows NT/200x Server as a WINS server, install and con-
figure the WINS service. See the Windows NT/200x documentation for
Section 9.5. WINS: The Windows Internetworking Name Server 161
details. Windows NT/200x WINS servers can replicate to each other, al-
lowing more than one to be set up in a complex subnet environment. Because
Microsoft refuses to document the replication protocols, Samba cannot cur-
rently participate in these replications. It is possible in the future that a
Samba-to-Samba WINS replication protocol may be defined, in which case
more than one Samba machine could be set up as a WINS server. Currently
only one Samba server should have the wins support = yes parameter set.
After the WINS server has been configured, you must ensure that all ma-
chines participating on the network are configured with the address of this
WINS server. If your WINS server is a Samba machine, fill in the Samba
machine IP address in the Primary WINS Server field of the Control Panel-
>Network->Protocols->TCP->WINS Server dialogs in Windows 9x/Me
or Windows NT/200x. To tell a Samba server the IP address of the WINS
server, add the following line to the [global] section of all smb.conf files:
wins s e r v e r = <name o r IP a d d r e s s >
where <name or IP address> is either the DNS name of the WINS server
machine or its IP address.
This line must not be set in the smb.conf file of the Samba server acting as
the WINS server itself. If you set both the wins support = yes option and
the wins server = <name> option then nmbd will fail to start.
There are two possible scenarios for setting up cross-subnet browsing. The
first details setting up cross-subnet browsing on a network containing Win-
dows 9x/Me, Samba, and Windows NT/200x machines that are not con-
figured as part of a Windows NT domain. The second details setting up
cross-subnet browsing on networks that contain NT domains.
Samba-3 permits WINS replication through the use of the wrepld utility.
This tool is not currently in use because it is still in development and has
not been worked on for a long time. As soon as this tool becomes moder-
ately functional, we will prepare man pages and enhance this section of the
documentation to provide usage and technical details.
162 Network Browsing Chapter 9
Adding static entries to your Samba WINS server is actually fairly easy. All
you have to do is add a line to wins.dat, typically located in /usr/local/
samba/var/locks or /var/run/samba.
where NAME is the NetBIOS name, TYPE is the NetBIOS type, TTL is
the time-to-live as an absolute time in seconds, ADDRESS+ is one or more
addresses corresponding to the registration, and FLAGS are the NetBIOS
flags for the registration.
To make a NetBIOS name static (permanent), simply set the TTL to 0, like
this:
Though this method works with early Samba-3 versions, there is a possibility
that it may change in future versions if WINS replication is added.
Section 9.6. Helpful Hints 163
The following hints should be carefully considered because they are stum-
bling points for many new network administrators.
Warning
Every NetBIOS machine takes part in a process of electing the LMB (and
DMB) every 15 minutes. A set of election criteria is used to determine the
order of precedence for winning this election process. A machine running
Samba or Windows NT will be biased, so the most suitable machine will
predictably win and thus retain its role.
The election process is fought out, so to speak over every NetBIOS network
interface. In the case of a Windows 9x/Me machine that has both TCP/IP
and IPX installed and has NetBIOS enabled over both protocols, the election
will be decided over both protocols. As often happens, if the Windows
9x/Me machine is the only one with both protocols, then the LMB may be
won on the NetBIOS interface over the IPX protocol. Samba will then lose
the LMB role because Windows 9x/Me will insist it knows who the LMB is.
Samba will then cease to function as an LMB, and browse list operation on
all TCP/IP-only machines will therefore fail.
Windows 95, 98, 98se, and Me are referred to generically as Windows 9x/Me.
The Windows NT4, 200x, and XP use common protocols. These are roughly
referred to as the Windows NT family, but it should be recognized that 2000
and XP/2003 introduce new protocol extensions that cause them to behave
differently from MS Windows NT4. Generally, where a server does not
164 Network Browsing Chapter 9
support the newer or extended protocol, these will fall back to the NT4
protocols.
The safest rule of all to follow is: Use only one protocol!
or
name r e s o l v e o r d e r = wins l m h o s t s ( ←-
e l i m i n a t e s b c a s t and h o s t )
where “host” refers to the native methods used by the UNIX system to
implement the gethostbyname() function call. This is normally controlled
by /etc/host.conf, /etc/nsswitch.conf and /etc/resolv.conf.
Section 9.7. Technical Overview of Browsing 165
MS Windows 2000 and later versions, as with Samba-3 and later versions,
can be configured to not use NetBIOS over TCP/IP. When configured this
way, it is imperative that name resolution (using DNS/LDAP/ADS) be cor-
rectly configured and operative. Browsing will not work if name resolution
from SMB machine names to IP addresses does not function correctly.
Samba can also act as a DMB for a workgroup. This means that it will col-
late lists from LMBs into a wide-area network server list. In order for browse
clients to resolve the names they may find in this list, it is recommended
that both Samba and your clients use a WINS server.
Do not set Samba to be the domain master for a workgroup that has the
same name as an NT Domain. On each wide-area network, you must only
ever have one DMB per workgroup, regardless of whether it is NT, Samba,
or any other type of domain master that is providing this service.
166 Network Browsing Chapter 9
Note
To get browsing to work, you need to run nmbd as usual, but must use the
workgroup option in smb.conf to control what workgroup Samba becomes
a part of.
Samba also has a useful option for a Samba server to offer itself for browsing
on another subnet. It is recommended that this option is used only for
“unusual” purposes: announcements over the Internet, for example. See
remote announce in the smb.conf man page.
If something does not work, the log.nmbd file will help to track down the
problem. Try a log level of 2 or 3 for finding problems. Also note that the
current browse list usually gets stored in text form in a file called browse.
dat.
If it does not work, you should still be able to type the server name as
\\SERVER in filemanager, then press enter, and filemanager should dis-
play the list of available shares.
Some people find browsing fails because they do not have the global guest
account set to a valid account. Remember that the IPC$ connection that
lists the shares is done as guest and so you must have a valid guest account.
Section 9.7. Technical Overview of Browsing 167
Note
MS Windows 2000 and later (as with Samba) can be configured to disallow
anonymous (i.e., guest account) access to the IPC$ share. In that case, the
MS Windows 2000/XP/2003 machine acting as an SMB/CIFS client will
use the name of the currently logged-in user to query the IPC$ share. MS
Windows 9x/Me clients are not able to do this and thus will not be able to
browse server resources.
The other big problem people have is that their broadcast address, netmask,
or IP address is wrong (specified with the interfaces option in smb.conf)
Since the release of Samba 1.9.17 (alpha1), Samba has supported the repli-
cation of browse lists across subnet boundaries. This section describes how
to set this feature up in different settings.
To see browse lists that span TCP/IP subnets (i.e., networks separated
by routers that do not pass broadcast traffic), you must set up at least
one WINS server. The WINS server acts as a DNS for NetBIOS names.
This will allow NetBIOS name-to-IP address translation to be completed
by a direct query of the WINS server. This is done via a directed UDP
packet on port 137 to the WINS server machine. The WINS server avoids
the necessity of default NetBIOS name-to-IP address translation, which is
done using UDP broadcasts from the querying machine. This means that
machines on one subnet will not be able to resolve the names of machines
on another subnet without using a WINS server. The Samba hacks, remote
168 Network Browsing Chapter 9
browse sync, and remote announce are designed to get around the natural
limitations that provent UDP broadcast propagation. The hacks are not a
universal solution and they should not be used in place of WINS, they are
considered last resort methods.
Remember, for browsing across subnets to work correctly, all machines, be
they Windows 95, Windows NT, or Samba servers, must have the IP ad-
dress of a WINS server given to them by a DHCP server or by manual
configuration: for Windows 9x/Me and Windows NT/200x/XP, this is in
the TCP/IP Properties, under Network settings; for Samba, this is in the
smb.conf file.
It is possible to operate Samba-3 without NetBIOS over TCP/IP. If you
do this, be warned that if used outside of MS ADS, this will forgo network
browsing support. ADS permits network browsing support through DNS,
providing appropriate DNS records are inserted for all Samba servers.
Subnet 1
Router 1 Router 2
Subnet 2 Subnet 3
services will broadcast that they are offering these services. The LMB on
each subnet will receive these broadcasts and keep a record of the fact that
the machine is offering a service. This list of records is the basis of the
browse list. For this case, assume that all the machines are configured to
offer services, so all machines will be on the browse list.
For each network, the LMB on that network is considered authoritative for
all the names it receives via local broadcast. This is because a machine seen
by the LMB via a local broadcast must be on the same network as the Local
Master Browser and thus is a trusted and verifiable resource. Machines on
other networks that the LMBs learn about when collating their browse lists
have not been directly seen. These records are called non-authoritative.
At this point the browse lists appear as shown in Table 9.1 (these are the
machines you would see in your network neighborhood if you looked in it on
a particular network right now).
At this point all the subnets are separate, and no machine is seen across any
of the subnets.
Now examine subnet 2 in Table 9.2. As soon as N2 B has become the LMB,
it looks for a DMB with which to synchronize its browse list. It does this
170 Network Browsing Chapter 9
by querying the WINS server (N2 D) for the IP address associated with the
NetBIOS name WORKGROUP<1B>. This name was registered by the
DMB (N1 C) with the WINS server as soon as it was started.
Once N2 B knows the address of the DMB, it tells it that is the LMB for sub-
net 2 by sending a MasterAnnouncement packet as a UDP port 138 packet.
It then synchronizes with it by doing a NetServerEnum2 call. This tells the
DMB to send it all the server names it knows about. Once the DMB receives
the MasterAnnouncement packet, it schedules a synchronization request to
the sender of that packet. After both synchronizations are complete, the
browse lists look like those in Table 9.2
The same sequence of events that occurred for N2 B now occurs for the
LMB on subnet 3 (N3 D). When it synchronizes browse lists with the DMB
(N1 A) it gets both the server entries on subnet 1 and those on subnet 2.
After N3 D has synchronized with N1 C and vica versa, the browse lists will
appear as shown in Table 9.3
Many questions are asked on the mailing lists regarding browsing. The ma-
jority of browsing problems originate from incorrect configuration of Net-
BIOS name resolution. Some are of particular note.
9.8.1 How Can One Flush the Samba NetBIOS Name Cache
without Restarting Samba?
Samba’s nmbd process controls all browse list handling. Under normal
circumstances it is safe to restart nmbd. This will effectively flush the
Samba NetBIOS name cache and cause it to be rebuilt. This does not make
certain that a rogue machine name will not reappear in the browse list.
When nmbd is taken out of service, another machine on the network will
become the browse master. This new list may still have the rogue entry in it.
If you really want to clear a rogue machine from the list, every machine on
the network must be shut down and restarted after all machines are down.
Failing a complete restart, the only other thing you can do is wait until the
entry times out and is then flushed from the list. This may take a long time
on some networks (perhaps months).
“My Client Reports ”‘This server is not configured to list shared resources.”’”
Your guest account is probably invalid for some reason. Samba uses the
guest account for browsing in smbd. Check that your guest account is
valid.
Section 9.8. Common Errors 173
• You cannot log onto the machine that is the LMB. Can you log on to
it as a guest user?
“ There are only two machines on a test network. One is a Samba server, the
other a Windows XP machine. Authentication and logons work perfectly,
but when I try to explore shares on the Samba server, the Windows XP client
becomes unresponsive. Sometimes it does not respond for some minutes.
Eventually, Windows Explorer will respond and displays files and directories
without problem.”
ACCOUNT INFORMATION
DATABASES
The three passdb backends that are fully maintained (actively supported)
by the Samba Team are: smbpasswd (being obsoleted), tdbsam (a tdb based
binary file format), and ldapsam (LDAP directory). Of these, only the
ldapsam backend stores both POSIX (UNIX) and Samba user and group
account information in a single repository. The smbpasswd and tdbsam
backends store only Samba user accounts.
In a strict sense, there are three supported account storage and access sys-
tems. One of these is considered obsolete (smbpasswd). It is recommended
to use tdbsam method for all simple systems. Use the ldapsam for larger
and more complex networks.
In a strict and literal sense, the passdb backends are account storage mech-
anisms (or methods) alone. The choice of terminology can be misleading,
however we are stuck with this choice of wording. This chapter documents
the nature of the account storage system with a focus on user and trust ac-
counts. Trust accounts have two forms, machine trust accounts (computer
accounts) and interdomain trust accounts. These are all treated as user-like
entities.
175
176 Account Information Databases Chapter 10
Plaintext This isn’t really a backend at all, but is listed here for simplicity.
Samba can be configured to pass plaintext authentication requests
to the traditional UNIX/Linux /etc/passwd and /etc/shadow-style
subsystems. On systems that have Pluggable Authentication Modules
(PAM) support, all PAM modules are supported. The behavior is
just as it was with Samba-2.2.x, and the protocol limitations imposed
by MS Windows clients apply likewise. Please refer to Section 10.2,
for more information regarding the limitations of plaintext password
usage.
smbpasswd This option allows continued use of the smbpasswd file that
maintains a plain ASCII (text) layout that includes the MS Win-
dows LanMan and NT-encrypted passwords as well as a field that
stores some account information. This form of password backend does
not store any of the MS Windows NT/200x SAM (Security Account
Manager) information required to provide the extended controls that
are needed for more comprehensive interoperation with MS Windows
NT4/200x servers.
tdbsam This backend provides a rich database backend for local servers.
This backend is not suitable for multiple domain controllers (i.e., PDC
+ one or more BDC) installations.
The tdbsam password backend stores the old smbpasswd information
plus the extended MS Windows NT/200x SAM information into a bi-
nary format TDB (trivial database) file. The inclusion of the extended
information makes it possible for Samba-3 to implement the same ac-
count and system access controls that are possible with MS Windows
NT4/200x-based systems.
The inclusion of the tdbsam capability is a direct response to user
requests to allow simple site operation without the overhead of the
complexities of running OpenLDAP. It is recommended to use this
only for sites that have fewer than 250 users. For larger sites or imple-
mentations, the use of OpenLDAP or of Active Directory integration
is strongly recommended.
Old Windows clients send plaintext passwords over the wire. Samba can
check these passwords by encrypting them and comparing them to the hash
stored in the UNIX user database.
Newer Windows clients send encrypted passwords (LanMan and NT hashes)
instead of plaintext passwords over the wire. The newest clients will send
only encrypted passwords and refuse to send plaintext passwords unless their
registry is tweaked.
Many people ask why Samba can not simply use the UNIX password database.
Windows requires passwords that are encrypted in its own format. The
UNIX passwords can’t be converted to UNIX-style encrypted passwords.
Because of that, you can’t use the standard UNIX user database, and you
have to store the LanMan and NT hashes somewhere else.
In addition to differently encrypted passwords, Windows also stores certain
data for each user that is not stored in a UNIX user database: for example,
workstations the user may logon from, the location where the user’s profile
1
<http://www.brics.dk/~amoeller/XML/schemas/>
Section 10.2. Technical Information 179
is stored, and so on. Samba retrieves and stores this information using a
passdb backend. Commonly available backends are LDAP, tdbsam, plain
text file, and MySQL. For more information, see the man page for smb.
conf regarding the passdb backend parameter.
SID
guest
smbpasswd
Yes tdbsam
Our Domain? PassDB
ldapsam
ldapsam_compat
No
winbindd_idmap.tdb
Winbind
ldapsam
No Yes
Fail Found? UID
The UNIX and SMB password encryption techniques seem similar on the
surface. This similarity is, however, only skin deep. The UNIX scheme
typically sends clear-text passwords over the network when logging in. This
is bad. The SMB encryption scheme never sends the clear-text password over
the network, but it does store the 16-byte hashed values on disk. This is also
bad. Why? Because the 16 byte hashed values are a “password equivalent.”
You cannot derive the user’s password from them, but they could potentially
180 Account Information Databases Chapter 10
UID
guest
smbpasswd
PassDB tdbsam
ldapsam
ldapsam_compat
No winbindd_idmap.tdb
Found? Winbind
ldapsam
Yes
No
Found? Fail
Yes
SID
• MS DOS Network client 3.0 with the basic network redirector installed.
• Windows Me.
Note
• Windows NT 3.5x.
• Windows NT 4.0.
• Windows XP Professional.
MS Windows clients will cache the encrypted password alone. Where plain-
text passwords are re-enabled through the appropriate registry change, the
plaintext password is never cached. This means that in the event that a
network connections should become disconnected (broken), only the cached
(encrypted) password will be sent to the resource server to effect an auto-
reconnect. If the resource server does not support encrypted passwords, the
auto-reconnect will fail. Use of encrypted passwords is strongly advised.
• Plaintext passwords are not passed across the network. Someone us-
ing a network sniffer cannot just record passwords going to the SMB
server.
• Windows NT does not like talking to a server that does not support
encrypted passwords. It will refuse to browse the server if the server
is also in user-level security mode. It will insist on prompting the user
for the password on each connection, which is very annoying. The only
thing you can do to stop this is to use SMB encryption.
• Plaintext passwords are not kept on disk and are not cached in mem-
ory.
• Plaintext passwords use the same password file as other UNIX services,
such as Login and FTP.
• Use of other services (such as Telnet and FTP) that send plaintext
passwords over the network makes sending them for SMB is not such
a big deal.
Section 10.2. Technical Information 183
systems. This tool can be used for centralized storage and retrieval of
UIDs and GIDs.
• pam ldap: A PAM module that provides LDAP integration for UNIX/Linux
system access authentication.
utilities makes it possible for all user access to the UNIX platform to be man-
aged from a central environment and yet distributed to wherever the point
of need may be physically located. Applications that benefit from this in-
frastructure include: UNIX login shells, mail and messaging systems, quota
controls, printing systems, DNS servers, DHCP servers, and also Samba.
Many sites are installing LDAP for the first time in order to provide a
scalable passdb backend for Samba. Others are faced with the need to
adapt an existing LDAP directory to new uses such as for the Samba SAM
backend. Whatever your particular need and attraction to Samba may be,
decisions made in respect of the design of the LDAP directory structure
and its implementation are of a durable nature for the site. These have far-
reaching implications that affect long term information systems management
costs.
Do not rush into an LDAP deployment. Take the time to understand how
the design of the Directory Information Tree (DIT) may impact current and
future site needs, as well as the ability to meet them. The way that Samba
SAM information should be stored within the DIT varies from site to site and
with each implementation new experience is gained. It is well understood
by LDAP veterans that first implementation create awakening, second im-
plementations of LDAP create fear, and third-generation deployments bring
peace and tranquility.
Samba provides two tools for management of user and machine accounts:
smbpasswd and pdbedit.
Some people are confused when reference is made to smbpasswd because the
name refers to a storage mechanism for SambaSAMAccount information,
but it is also the name of a utility tool. That tool is destined to eventually
be replaced by new functionality that is being added to the net toolset (see
Chapter 12, “Remote and Local Management: The Net Command”.
$ smbpasswd
Old SMB password: secret
For secret, type the old value here or press return if there is no old pass-
word.
If the old value does not match the current value stored for that user, or
the two new values do not match each other, then the password will not be
changed.
When invoked by an ordinary user, the command will allow only the user
to change his or her own SMB password.
smbpasswd is designed to work in the way familiar to UNIX users who use
the passwd or yppasswd commands. While designed for administrative
use, this tool provides essential user-level password change capabilities.
For more details on using smbpasswd, refer to the man page (the definitive
reference).
Section 10.3. Account Management Tools 189
pdbedit is a tool that can be used only by root. It is used to manage the
passdb backend. pdbedit can be used to:
• add, remove, or modify user accounts.
• list user accounts.
• migrate user accounts.
The pdbedit tool is the only one that can manage the account security and
policy settings. It is capable of all operations that smbpasswd can do as
well as a superset of them.
One particularly important purpose of the pdbedit is to allow the migra-
tion of account information from one passdb backend to another. See the
Section 10.4.6 password backend section of this chapter.
The following is an example of the user account information that is stored
in a tdbsam password backend. This listing was produced by running:
10.4.1 Plaintext
Older versions of Samba retrieved user information from the UNIX user
database and eventually some other fields from the file /etc/samba/smbpasswd
or /etc/smbpasswd. When password encryption is disabled, no SMB-specific
data is stored at all. Instead, all operations are conducted via the way that
the Samba host OS will access its /etc/passwd database. On most Linux
systems, for example, all user and group resolution is done via PAM.
Section 10.4. Password Backends 191
10.4.3 tdbsam
Samba can store user and machine account data in a “TDB” (trivial database).
Using this backend does not require any additional configuration. This back-
end is recommended for new installations that do not require LDAP.
As a general guide, the Samba Team does not recommend using the tdb-
sam backend for sites that have 250 or more users. Additionally, tdbsam is
192 Account Information Databases Chapter 10
not capable of scaling for use in sites that require PDB/BDC implementa-
tions that require replication of the account database. Clearly, for reason of
scalability, the use of ldapsam should be encouraged.
The recommendation of a 250-user limit is purely based on the notion that
this would generally involve a site that has routed networks, possibly spread
across more than one physical location. The Samba Team has not at this
time established the performance-based scalability limits of the tdbsam ar-
chitecture.
There are sites that have thousands of users and yet require only one server.
One site recently reported having 4,500 user accounts on one UNIX sys-
tem and reported excellent performance with the tdbsam passdb backend.
The limitation of where the tdbsam passdb backend can be used is not one
pertaining to a limitation in the TDB storage system, it is based only on
the need for a reliable distribution mechanism for the SambaSAMAccount
backend.
10.4.4 ldapsam
There are a few points to stress that the ldapsam does not provide. The
LDAP support referred to in this documentation does not include:
• A means of retrieving user account information from a Windows 200x
Active Directory server.
• A means of replacing /etc/passwd.
The second item can be accomplished by using LDAP NSS and PAM mod-
ules. LGPL versions of these libraries can be obtained from PADL Software3 .
More information about the configuration of these packages may be found
in LDAP, System Administration by Gerald Carter, Chapter 6, Replacing
NIS”4 .
This document describes how to use an LDAP directory for storing Samba
user account information traditionally stored in the smbpasswd(5) file. It is
assumed that the reader already has a basic understanding of LDAP con-
cepts and has a working directory server already installed. For more infor-
mation on LDAP architectures and directories, please refer to the following
sites:
3
<http://www.padl.com/>
4
<http://safari.oreilly.com/?XmlId=1-56592-491-6>
Section 10.4. Password Backends 193
• OpenLDAP5
• Sun One Directory Server6
• Novell eDirectory7
• IBM Tivoli Directory Server8
• Red Hat Directory Server9
• Fedora Directory Server10
Two additional Samba resources that may prove to be helpful are:
• The Samba-PDC-LDAP-HOWTO11 maintained by Ignacio Coupeau.
• The NT migration scripts from IDEALX12 that are geared to manage
users and groups in such a Samba-LDAP domain controller configu-
ration. Idealx also produced the smbldap-tools and the Interactive
Console Management tool.
The LDAP ldapsam code was developed and tested using the OpenLDAP
2.x server and client libraries. The same code should work with Netscape’s
Directory Server and client SDK. However, there are bound to be compile
errors and bugs. These should not be hard to fix. Please submit fixes via
the process outlined in Chapter 38, “Reporting Bugs”.
Samba is capable of working with any standards compliant LDAP server.
Samba-3.0 includes the necessary schema file for OpenLDAP 2.x in exam-
ples/LDAP/samba.schema directory of the source code distribution tarball.
The schema entry for the sambaSamAccount ObjectClass is shown here:
5
<http://www.openldap.org/>
6
<http://www.sun.com/software/products/directory_srvr_ee/index.xml>
7
<http://www.novell.com/products/edirectory/>
8
<http://www-306.ibm.com/software/tivoli/products/directory-server/>
9
<http://www.redhat.com/software/rha/directory/>
10
<http://www.linuxsecurity.com/content/view/119229>
11
<http://www.unav.es/cti/ldap-smb/ldap-smb-3-howto.html>
12
<http://samba.idealx.org/>
194 Account Information Databases Chapter 10
The samba.schema file has been formatted for OpenLDAP 2.0/2.1. The
Samba Team owns the OID space used by the above schema and recommends
its use. If you translate the schema to be used with Netscape DS, please
submit the modified schema file as a patch to [email protected] .
Just as the smbpasswd file is meant to store information that provides in-
formation additional to a user’s /etc/passwd entry, so is the sambaSamAc-
count object meant to supplement the UNIX user account information. A
sambaSamAccount is an AUXILIARY ObjectClass, so it can be used to aug-
ment existing user account information in the LDAP directory, thus pro-
viding information needed for Samba account handling. However, there are
several fields (e.g., uid) that overlap with the posixAccount ObjectClass
outlined in RFC 2307. This is by design.
In order to store all user account information (UNIX and Samba) in the
directory, it is necessary to use the sambaSamAccount and posixAccount
ObjectClasses in combination. However, smbd will still obtain the user’s
UNIX account information via the standard C library calls, such as get-
pwnam(). This means that the Samba server must also have the LDAP
NSS library installed and functioning correctly. This division of informa-
tion makes it possible to store all Samba account information in LDAP,
but still maintain UNIX account information in NIS while the network is
transitioning to a full LDAP infrastructure.
13
<mailto:[email protected]>
Section 10.4. Password Backends 195
## /etc/openldap/slapd.conf
It is recommended that you maintain some indices on some of the most useful
attributes, as in the following example, to speed up searches made on sam-
baSamAccount ObjectClasses (and possibly posixAccount and posixGroup
as well):
# Indices to maintain
## required by OpenLDAP
index objectclass eq
index cn pres,sub,eq
196 Account Information Databases Chapter 10
index sn pres,sub,eq
## required to support pdb_getsampwnam
index uid pres,sub,eq
## required to support pdb_getsambapwrid()
index displayName pres,sub,eq
index sambaSID eq
index sambaPrimaryGroupSID eq
index sambaDomainName eq
index default sub
Before you can add accounts to the LDAP database, you must create the
account containers that they will be stored in. The following LDIF file
should be modified to match your needs (DNS entries, and so on):
objectclass: organizationalUnit
ou: Computers
$ slapadd -v -l initldap.dif
Do not forget to secure your LDAP server with an adequate access control
list as well as an admin password.
Note
LDAP support if the LDAP libraries are found. The best method to verify
that Samba was built with LDAP support is:
If the build of the smbd command you are using does not produce output
that includes HAVE LDAP H it is necessary to discover why the LDAP headers
and libraries were not found during compilation.
These are described in the smb.conf man page and so are not repeated
here. However, an example for use with an LDAP directory is shown in
Example 10.4.1
200 Account Information Databases Chapter 10
There are two important points to remember when discussing the security
of sambaSAMAccount entries in the directory.
These password hashes are clear-text equivalents and can be used to imper-
sonate the user without deriving the original clear-text strings. For more
information on the details of LM/NT password hashes, refer to Chapter 10,
“Account Information Databases”.
To remedy the first security issue, the ldap ssl smb.conf parameter defaults
to require an encrypted session (ldap ssl = on) using the default port of 636
when contacting the directory server. When using an OpenLDAP server, it
is possible to use the StartTLS LDAP extended operation in the place of
LDAPS. In either case, you are strongly encouraged to use secure commu-
nications protocols (so do not set ldap ssl = off).
Section 10.4. Password Backends 201
Note that the LDAPS protocol is deprecated in favor of the LDAPv3 Start-
TLS extended operation. However, the OpenLDAP library still provides
support for the older method of securing communication between clients
and servers.
## allow the "ldap admin dn" access, but deny everyone else
access to attrs=SambaLMPassword,SambaNTPassword
by dn="cn=Samba Admin,ou=People,dc=quenya,dc=org" write
by * none
The majority of these parameters are only used when Samba is acting as a
PDC of a domain (refer to Chapter 4, “Domain Control”, for details on how
to configure Samba as a PDC). The following four attributes are only stored
with the sambaSamAccount entry if the values are non-default values:
• sambaHomePath
• sambaLogonScript
• sambaProfilePath
• sambaHomeDrive
These attributes are only stored with the sambaSamAccount entry if the
values are non-default values. For example, assume MORIA has now been
configured as a PDC and that logon home = \\%L\%u was defined in its
smb.conf file. When a user named “becky” logs on to the domain, the logon
home string is expanded to \\MORIA\becky. If the smbHome attribute
exists in the entry “uid=becky,ou=People,dc=samba,dc=org”, this value is
used. However, if this attribute does not exist, then the value of the logon
home parameter is used in its place. Samba will only write the attribute
202 Account Information Databases Chapter 10
value to the directory entry if the value is something other than the default
(e.g., \\MOBY\becky).
The following is a working LDIF that demonstrates the use of the Sam-
baSamAccount ObjectClass:
The following is an LDIF entry for using both the sambaSamAccount and
posixAccount ObjectClasses:
loginShell: /bin/bash
logoffTime: 2147483647
gidNumber: 100
sambaKickoffTime: 2147483647
sambaPwdLastSet: 1010179230
sambaSID: S-1-5-21-2447931902-1787058256-3961074038-5004
homeDirectory: /home/moria/gcarter
sambaPwdCanChange: 0
sambaPwdMustChange: 2147483647
sambaNTPassword: 878D8014606CDA29677A44EFA1353FC7
Samba-3 and later can update the non-Samba (LDAP) password stored with
an account. When using pam ldap, this allows changing both UNIX and
Windows passwords at once.
The ldap passwd sync options can have the values shown in Table 10.3.
More information can be found in the smb.conf man page.
10.4.5 MySQL
Every so often someone comes along with what seems to them like a great
new idea. Storing user accounts in a SQL backend is one of them. Those
who want to do this are in the best position to know what the specific
benefits are to them. This may sound like a cop-out, but in truth we cannot
document every little detail of why certain things of marginal utility to the
bulk of Samba users might make sense to the rest. In any case, the following
instructions should help the determined SQL user to implement a working
system. These account storage methods are not actively maintained by the
Samba Team.
You can set up your own table and specify the field names to pdb mysql
(see Table 10.5 for the column names) or use the default table. The file
examples/pdb/mysql/mysql.dump contains the correct queries to create the
required tables. Use the command:
204 Account Information Databases Chapter 10
10.4.5.2 Configuring
This plug-in lacks some good documentation, but here is some brief infor-
mation. Add the following to the passdb backend variable in your smb.
conf:
passdb backend = [ o t h e r −p l u g i n s ] mysql : ←-
i d e n t i f i e r [ o t h e r −p l u g i n s ]
The identifier can be any string you like, as long as it does not collide
with the identifiers of other plugins or other instances of pdb mysql. If you
specify multiple pdb mysql.so entries in passdb backend, you also need to
use different identifiers.
Additional options can be given through the smb.conf file in the [global]
section. Refer to Table 10.4.
Warning
Names of the columns are given in Table 10.5. The default column names
can be found in the example table dump.
You can put a colon (:) after the name of each column, which should specify
the column to update when updating the table. You can also specify nothing
behind the colon, in which case the field data will not be updated. Setting
a column name to NULL means the field should not be used.
Example 10.4.2 is shown in Example 10.4.2.
Section 10.5. Common Errors 205
I strongly discourage the use of plaintext passwords; however, you can use
them.
If you would like to use plaintext passwords, set ‘identifier:lanman pass
column’ and ‘identifier:nt pass column’ to ‘NULL’ (without the quotes) and
‘identifier:plain pass column’ to the name of the column containing the plain-
text passwords.
If you use encrypted passwords, set the ’identifier:plain pass column’ to
’NULL’ (without the quotes). This is the default.
It is possible to have not all data in the database by making some ”constant.”
For example, you can set ‘identifier:fullname column’ to something like
CONCAT(Firstname,’ ’,Surname)
Or, set ‘identifier:workstations column’ to: NULL . See the MySQL docu-
mentation for more language constructs.
10.4.6 XML
“I’ve installed Samba, but now I can’t log on with my UNIX account!”
Make sure your user has been added to the current Samba passdb backend.
Read the Section 10.3 for details.
206 Account Information Databases Chapter 10
A few complaints have been received from users who just moved to Samba-
3. The following smb.conf file entries were causing problems: new accounts
were being added to the old smbpasswd file, not to the tdbsam passdb.tdb
file:
[ global ]
... passdb backend = smbpasswd , tdbsam
...
Samba will add new accounts to the first entry in the passdb backend pa-
rameter entry. If you want to update to the tdbsam, then change the entry
to:
[ globals ] ... passdb backend = tdbsam , smbpasswd
...
GROUP MAPPING: MS
WINDOWS AND UNIX
The new facility for mapping NT groups to UNIX system groups allows the
administrator to decide which NT domain groups are to be exposed to MS
Windows clients. Only those NT groups that map to a UNIX group that
has a value other than the default (-1) will be exposed in group selection
lists in tools that access domain users and groups.
Warning
212
Section 11.1. Features and Benefits 213
group
SID
groupmap_idmap.tdb
No winbindd_idmap.tdb
Found? Winbind
ldapsam
Yes
No
Found? Fail
Yes
GID
In both cases, when winbindd is not running, only locally resolvable groups
can be recognized. Please refer to Figure 11.1 and Figure 11.2. The net
groupmap is used to establish UNIX group to NT SID mappings as shown
214 Group Mapping: MS Windows and UNIX Chapter 11
GID
groupmap_idmap.tdb
No winbindd_idmap.tdb
Found? Winbind
ldapsam
Yes
No
Found? Fail
Yes
group
SID
in Figure 11.3.
ldapsam
GID net groupmap SID
groupmap_idmap.tdb
There are several possible workarounds for the operating system tools lim-
itation. One method is to use a script that generates a name for the
UNIX/Linux system group that fits the operating system limits and that
then just passes the UNIX/Linux group ID (GID) back to the calling Samba
interface. This will provide a dynamic workaround solution.
11.2 Discussion
The following steps describe how to make Samba PDC users members of the
Domain Admins group.
2. Add to this group the users that must be “Administrators”. For ex-
ample, if you want joe, john, and mary to be administrators, your
entry in /etc/group will look like this:
domadm:x:502:joe,john,mary
216 Group Mapping: MS Windows and UNIX Chapter 11
The quotes around “Domain Admins” are necessary due to the space
in the group name. Also make sure to leave no white space surrounding
the equal character (=).
Now joe, john, and mary are domain administrators.
It is possible to map any arbitrary UNIX group to any Windows NT4/200x
group as well as to make any UNIX group a Windows domain group. For
example, if you wanted to include a UNIX group (e.g., acct) in an ACL on
a local file or printer on a Domain Member machine, you would flag that
group as a domain group by running the following on the Samba PDC:
Windows does not permit user and group accounts to have the same name.
This has serious implications for all sites that use private group accounts.
A private group account is an administrative practice whereby users are
each given their own group account. Red Hat Linux, as well as several free
distributions of Linux, by default create private groups.
When mapping a UNIX/Linux group to a Windows group account, all con-
flict can be avoided by assuring that the Windows domain group name does
not overlap with any user account name.
Section 11.2. Discussion 217
This functionality is known as nested groups and was first added to Samba-
3.0.3.
All MS Windows products since the release of Windows NT 3.10 support
the use of nested groups. Many Windows network administrators depend
on this capability because it greatly simplifies security administration.
The nested group architecture was designed with the premise that day-
to-day user and group membership management should be performed on
the domain security database. The application of group security should be
implemented on domain member servers using only local groups. On the
domain member server, all file system security controls are then limited to
use of the local groups, which will contain domain global groups and domain
global users.
You may ask, What are the benefits of this arrangement? The answer is
obvious to those who have plumbed the dark depths of Windows networking
architecture. Consider for a moment a server on which are stored 200,000
files, each with individual domain user and domain group settings. The
company that owns the file server is bought by another company, resulting
in the server being moved to another location, and then it is made a member
of a different domain. Who would you think now owns all the files and
directories? Answer: Account Unknown.
Unraveling the file ownership mess is an unenviable administrative task that
can be avoided simply by using local groups to control all file and directory
access control. In this case, only the members of the local groups will have
been lost. The files and directories in the storage subsystem will still be
owned by the local groups. The same goes for all ACLs on them. It is
administratively much simpler to delete the Account Unknown membership
entries inside local groups with appropriate entries for domain global groups
in the new domain that the server has been made a member of.
Another prominent example of the use of nested groups involves imple-
mentation of administrative privileges on domain member workstations and
servers. Administrative privileges are given to all members of the built-in lo-
cal group Administrators on each domain member machine. To ensure that
all domain administrators have full rights on the member server or worksta-
tion, on joining the domain, the Domain Admins group is added to the local
218 Group Mapping: MS Windows and UNIX Chapter 11
UNIX/Linux has no concept of support for nested groups, and thus Samba
has for a long time not supported them either. The problem is that you
would have to enter UNIX groups as auxiliary members of a group in /
etc/group. This does not work because it was not a design requirement
at the time the UNIX file system security model was implemented. Since
Samba-2.2, the winbind daemon can provide /etc/group entries on demand
by obtaining user and group information from the domain controller that
the Samba server is a member of.
In effect, Samba supplements the /etc/group data via the dynamic lib-
nss winbind mechanism. Beginning with Samba-3.0.3, this facility is used
to provide local groups in the same manner as Windows does it. It works
by expanding the local groups on the fly as they are accessed. For exam-
ple, the Domain Users group of the domain is made a member of the local
group demo. Whenever Samba needs to resolve membership of the demo
local (alias) group, winbind asks the domain controller for demo members
of the Domain Users group. By definition, it can only contain user objects,
which can then be faked to be member of the UNIX/Linux group demo.
To enable the use of nested groups, winbindd must be used with NSS
winbind. Creation and administration of the local groups is done best via
the Windows Domain User Manager or its Samba equivalent, the utility net
rpc group. Creating the local group demo is achieved by executing:
Here the -L switch means that you want to create a local group. It may
be necessary to add -S and -U switches for accessing the correct host with
appropriate user or root privileges. Adding and removing group members
can be done via the addmem and delmem subcommands of net rpc group
command. For example, addition of “DOM\Domain Users” to the local
group demo is done by executing:
Having completed these two steps, the execution of getent group demo
will show demo members of the global Domain Users group as members of
the group demo. This also works with any local or domain user. In case
the domain DOM trusts another domain, it is also possible to add global
users and groups of the trusted domain as members of demo. The users from
the foreign domain who are members of the group that has been added to
the demo group now have the same local access permissions as local domain
users have.
Note
group.
You can list the various groups in the mapping database by executing net
groupmap list. Here is an example:
For complete details on net groupmap, refer to the net(8) man page.
Everyone needs tools. Some of us like to create our own, others prefer to
use canned tools (i.e., prepared by someone else for general use).
A script to create complying group names for use by the Samba group in-
terfaces is provided in Example 11.3.1. This script adds a temporary entry
in the /etc/group file and then renames it to the desired name. This is
an example of a method to get around operating system maintenance tool
limitations such as those present in some version of the groupadd tool.
#!/bin/bash
# Now change the name to what we want for the MS Windows networking end
cp /etc/group /etc/group.bak
cat /etc/group.bak | sed "s/^smbtmpgrp00/$1/g" > /etc/group
rm /etc/group.bak
The smb.conf entry for the above script shown in Example 11.3.2 demon-
strates how it may be used.
Section 11.3. Configuration Scripts 223
#!/bin/bash
groupadd Orks
groupadd Elves
groupadd Gnomes
Of course it is expected that the administrator will modify this to suit local
needs. For information regarding the use of the net groupmap tool please
refer to the man page.
224 Group Mapping: MS Windows and UNIX Chapter 11
At this time there are many little surprises for the unwary administrator.
In a real sense it is imperative that every step of automated control scripts
be carefully tested manually before putting it into active service.
The net command is one of the new features of Samba-3 and is an attempt
to provide a useful tool for the majority of remote management operations
necessary for common tasks. The net tool is flexible by design and is in-
tended for command-line use as well as for scripted control application.
Originally introduced with the intent to mimic the Microsoft Windows com-
mand that has the same name, the net command has morphed into a very
powerful instrument that has become an essential part of the Samba net-
work administrator’s toolbox. The Samba Team has introduced tools, such
as smbgroupedit and rpcclient, from which really useful capabilities have
been integrated into the net. The smbgroupedit command was absorbed
entirely into the net, while only some features of the rpcclient command
have been ported to it. Anyone who finds older references to these utilities
and to the functionality they provided should look at the net command
before searching elsewhere.
226
Section 12.1. Overview 227
12.1 Overview
The tasks that follow the installation of a Samba-3 server, whether stan-
dalone or domain member, of a domain controller (PDC or BDC) begins
with the need to create administrative rights. Of course, the creation of
user and group accounts is essential for both a standalone server and a
PDC. In the case of a BDC or a Domain Member server (DMS), domain
user and group accounts are obtained from the central domain authentica-
tion backend.
Regardless of the type of server being installed, local UNIX groups must be
mapped to the Windows networking domain global group accounts. Do you
ask why? Because Samba always limits its access to the resources of the
host server by way of traditional UNIX UID and GID controls. This means
that local groups must be mapped to domain global groups so that domain
users who are members of the domain global groups can be given access
rights based on UIDs and GIDs local to the server that is hosting Samba.
Such mappings are implemented using the net command.
UNIX systems that are hosting a Samba-3 server that is running as a member
(PDC, BDC, or DMS) must have a machine security account in the domain
authentication database (or directory). The creation of such security (or
trust) accounts is also handled using the net command.
The establishment of interdomain trusts is achieved using the net command
also, as may a plethora of typical administrative duties such as user manage-
ment, group management, share and printer management, file and printer
migration, security identifier management, and so on.
The overall picture should be clear now: the net command plays a central
role on the Samba-3 stage. This role will continue to be developed. The
inclusion of this chapter is evidence of its importance, one that has grown
in complexity to the point that it is no longer considered prudent to cover
its use fully in the online UNIX man pages.
The basic operations of the net command are documented here. This doc-
umentation is not exhaustive, and thus it is incomplete. Since the primary
focus is on migration from Windows servers to a Samba server, the emphasis
is on the use of the Distributed Computing Environment Remote Procedure
228 Remote and Local Management: The Net Command Chapter 12
Call (DCE RPC) mode of operation. When used against a server that is
a member of an Active Directory domain, it is preferable (and often neces-
sary) to use ADS mode operations. The net command supports both, but
not for every operation. For most operations, if the mode is not specified,
net will automatically fall back via the ads, rpc, and rap modes. Please
refer to the man page for a more comprehensive overview of the capabilities
of this utility.
As stated, the focus in most of this chapter is on use of the net rpc family
of operations that are supported by Samba. Most of them are supported by
the net ads mode when used in connection with Active Directory. The net
rap operating mode is also supported for some of these operations. RAP
protocols are used by IBM OS/2 and by several earlier SMB servers.
Samba’s net tool implements sufficient capability to permit all common
administrative tasks to be completed from the command line. In this section
each of the essential user and group management facilities are explored.
Samba-3 recognizes two types of groups: domain groups and local groups.
Domain groups can contain (have as members) only domain user accounts.
Local groups can contain local users, domain users, and domain groups as
members.
The purpose of a local group is to permit file permission to be set for a
group account that, like the usual UNIX/Linux group, is persistent across
redeployment of a Windows file server.
Samba provides file and print services to Windows clients. The file system
resources it makes available to the Windows environment must, of necessity,
be provided in a manner that is compatible with the Windows networking
environment. UNIX groups are created and deleted as required to serve
operational needs in the UNIX operating system and its file systems.
In order to make available to the Windows environment Samba has a facility
by which UNIX groups can be mapped to a logical entity, called a Windows
(or domain) group. Samba supports two types of Windows groups, local and
Section 12.3. UNIX and Windows Group Management 229
global. Global groups can contain as members, global users. This member-
ship is affected in the normal UNIX manner, but adding UNIX users to
UNIX groups. Windows user accounts consist of a mapping between a user
SambaSAMAccount (logical entity) and a UNIX user account. Therefore,
a UNIX user is mapped to a Windows user (i.e., is given a Windows user
account and password) and the UNIX groups to which that user belongs, is
mapped to a Windows group account. The result is that in the Windows ac-
count environment that user is also a member of the Windows group account
by virtue of UNIX group memberships.
The addition will result in immediate availability of the new group account
as validated by executing this command:
The following demonstrates that the use of the net command to add a group
account results in immediate mapping of the POSIX group that has been
created to the Windows group account as shown here:
All file system (file and directory) access controls, within the file system
of a UNIX/Linux server that is hosting a Samba server, are implemented
using a UID/GID identity tuple. Samba does not in any way override or
replace UNIX file system semantics. Thus it is necessary that all Windows
networking operations that access the file system provide a mechanism that
maps a Windows user to a particular UNIX/Linux group account. The user
account must also map to a locally known UID.
Samba depends on default mappings for the Domain Admins, Domain Users,
and Domain Guests global groups. Additional groups may be added as
shown in the examples just given. There are times when it is necessary to
map an existing UNIX group account to a Windows group. This operation,
in effect, creates a Windows group account as a consequence of creation of
the mapping.
The operations that are permitted include: add, modify, and delete. An
example of each operation is shown here.
Supported mapping types are ’d’ (domain global) and ’l’ (domain local). A
Windows group may be deleted, and then a new Windows group can be
mapped to the UNIX group by executing these commands:
The deletion and addition operations affected only the logical entities known
as Windows groups, or domain groups. These operations are inert to UNIX
system groups, meaning that they neither delete nor create UNIX system
groups. The mapping of a UNIX group to a Windows group makes the
UNIX group available as Windows groups so that files and folders on domain
member clients (workstations and servers) can be given domain-wide access
controls for domain users and groups.
Two types of Windows groups can be created: domain (global) and local.
In the previous examples the Windows groups created were of type domain
or global. The following command will create a Windows group of type
local.
Supported mapping types are ’d’ (domain global) and ’l’ (domain local), a
domain local group is Samba is treated as local to the individual Samba
serverr. Local groups can be used with Samba to enable multiple nested
group support.
Note
of a Windows group that has been mapped using the net groupmap com-
mand (see Chapter 11, “Group Mapping: MS Windows and UNIX”). The
following list of UNIX/Linux group membership shows that the user ajt is
a member of the UNIX/Linux group Engineers.
Given that the user ajt is already a member of the UNIX/Linux group and,
via the group mapping, a member of the Windows group, an attempt to add
this account again should fail. This is demonstrated here:
This shows that the group mapping between UNIX/Linux groups and Win-
Section 12.3. UNIX and Windows Group Management 235
To permit the user ajt to be added using the net rpc group utility, this
account must first be removed. The removal and confirmation of its effect
is shown here:
In this example both at the UNIX/Linux system level, the group no longer
has the ajt as a member. The above also shows this to be the case for
Windows group membership.
The account is now added again, using the net rpc group utility:
In this example the members of the Windows Domain Users account are
validated using the net rpc group utility. Note the this contents of the
UNIX/Linux group was shown four paragraphs earlier. The Windows (do-
main) group membership is shown here:
This express example shows that Windows group names are treated by
Samba (as with MS Windows) in a case-insensitive manner:
Note
The -L switch means create a local group. Use the -S argument to direct
the operation to a particular server. The parameters to the -U argument
should be for a user who has appropriate administrative right and privileges
on the machine.
Section 12.3. UNIX and Windows Group Management 237
Addition and removal of group members can be achieved using the addmem
and delmem subcommands of net rpc group command. For example, ad-
dition of “DOM\Domain Users” to the local group demo would be done by
executing:
Windows network administrators often ask on the Samba mailing list how it
is possible to grant everyone administrative rights on their own workstation.
This is of course a very bad practice, but commonly done to avoid user
complaints. Here is how it can be done remotely from a Samba PDC or
BDC:
This can be scripted, and can therefore be performed as a user logs onto the
domain from a Windows workstation. Here is a simple example that shows
how this can be done. Automating User Addition to the Workstation Power
Users Group
238 Remote and Local Management: The Net Command Chapter 12
#!/bin/bash
exit 0
1. Create the script shown in Example 12.3.1 and locate it in the directory
/etc/samba/scripts, named as autopoweruser.sh.
3. Modify the smb.conf file so the NETLOGON stanza contains the param-
eters shown in Example 12.3.2.
This script will be executed every time a user logs onto the network. There-
fore every user will have local Windows workstation management rights.
This could of course be assigned using a group, in which case there is little
justification for the use of this procedure. The key justification for the use
of this method is that it will guarantee that all users have appropriate rights
on the workstation.
The syntax for adding a user account via the net (according to the man
page) is shown here:
net [<method>] user ADD <name> [-c container] [-F user flags] \
[misc. options] [targets]
The account password can be set with the following methods (all show the
same operation):
Two basic user account operations are routinely used: change of password
and querying which groups a user is a member of. The change of password
operation is shown in Section 12.4.1.
The ability to query Windows group membership can be essential. Here is
how a remote server may be interrogated to find which groups a user is a
member of:
Domain Users
Domain Admins
Engineers
TorridGroup
BOP Shop
Emergency Services
In this example the Windows user account “William Parsons” will be mapped
to the UNIX user parsonsw, and the Windows user account “geeringm” will
be mapped to the UNIX user marygee.
With all versions of Samba earlier than 3.0.11 the only account on a Samba
server that could manage users, groups, shares, printers, and such was the
root account. This caused problems for some users and was a frequent
source of scorn over the necessity to hand out the credentials for the most
security-sensitive account on a UNIX/Linux system.
242 Remote and Local Management: The Net Command Chapter 12
Note
BUILTIN\Account Operators
No privileges assigned
BUILTIN\Backup Operators
No privileges assigned
BUILTIN\Server Operators
No privileges assigned
BUILTIN\Administrators
No privileges assigned
Everyone
No privileges assigned
Section 12.5. Administering User Rights and Privileges 243
The net command can be used to obtain the currently supported capabilities
for rights and privileges using this method:
Next, the domain user jht is given the privileges needed for day-to-day
administration:
BUILTIN\Print Operators
No privileges assigned
BUILTIN\Account Operators
No privileges assigned
BUILTIN\Backup Operators
No privileges assigned
BUILTIN\Server Operators
No privileges assigned
BUILTIN\Administrators
No privileges assigned
Everyone
No privileges assigned
MIDEARTH\Domain Admins
SeMachineAccountPrivilege
SePrintOperatorPrivilege
SeAddUsersPrivilege
SeRemoteShutdownPrivilege
SeDiskOperatorPrivilege
Section 12.6. Managing Trust Relationships 245
There are essentially two types of trust relationships: the first is between
domain controllers and domain member machines (network clients), the sec-
ond is between domains (called interdomain trusts). All Samba servers that
participate in domain security require a domain membership trust account,
as do like Windows NT/200x/XP workstations.
The net command looks in the smb.conf file to obtain its own configuration
settings. Thus, the following command ’know’ which domain to join from
the smb.conf file.
In the event that the ADS trust was not established, or is broken for one
reason or another, the following error message may be obtained:
246 Remote and Local Management: The Net Command Chapter 12
Note that the command-line parameter member makes this join specific. By
default the type is deduced from the smb.conf file configuration. To specif-
ically join as a PDC or BDC, the command-line parameter will be [PDC |
BDC]. For example:
It is best to let Samba figure out the domain join type from the settings in
Section 12.6. Managing Trust Relationships 247
The removal is made possible because machine accounts are just like user
accounts with a trailing $ character. The account management operations
treat user and machine accounts in like manner.
A Samba-3 server that is a Windows ADS domain member can execute the
following command to detach from the domain:
none
none
There are no interdomain trusts at this time; the following steps will create
them.
It is necessary to create a trust account in the local domain. A domain
controller in a second domain can create a trusted connection with this
account. That means that the foreign domain is being trusted to access
resources in the local domain. This command creates the local trust account:
1
<http://www.samba.org/samba/docs/Samba3-ByExample.pdf>
Section 12.6. Managing Trust Relationships 249
A trust account will always have an I in the field within the square brackets.
If the trusting domain is not capable of being reached, the following com-
mand will fail:
none
DAMNATION S-1-5-21-1385457007-882775198-1210191635
DAMNATION S-1-5-21-1385457007-882775198-1210191635
Where a trust account has been created on a foreign domain, Samba is able
to establish the trust (connect with) the foreign account. In the process
it creates a one-way trust to the resources on the remote domain. This
command achieves the objective of joining the trust relationship:
DAMNATION S-1-5-21-1385457007-882775198-1210191635
DAMNATION S-1-5-21-1385457007-882775198-1210191635
At other times it becomes necessary to remove the ability for users from
a foreign domain to be able to access resources in the local domain. The
command shown here will do that:
The basic security identifier that is used by all Windows networking oper-
ations is the Windows security identifier (SID). All Windows network ma-
chines (servers and workstations), users, and groups are identified by their
respective SID. All desktop profiles are also encoded with user and group
SIDs that are specific to the SID of the domain to which the user belongs.
It is truly prudent to store the machine and/or domain SID in a file for
safekeeping. Why? Because a change in hostname or in the domain (work-
group) name may result in a change in the SID. When you have the SID on
hand, it is a simple matter to restore it. The alternative is to suffer the pain
of having to recover user desktop profiles and perhaps rejoin all member
machines to the domain.
First, do not forget to store the local SID in a file. It is a good idea to put
Section 12.7. Managing Security Identifiers (SIDS) 251
this in the directory in which the smb.conf file is also stored. Here is a
simple action to achieve this:
Good, there is now a safe copy of the local machine SID. On a PDC/BDC
this is the domain SID also.
The following command reveals what the former one should have placed into
the file called my-sid:
If ever it becomes necessary to restore the SID that has been stored in the
my-sid file, simply copy the SID (the string of characters that begins with
S-1-5-21) to the command line shown here:
The following operation is useful only for machines that are being configured
as a PDC or a BDC. DMS and workstation clients should have their own
machine SID to avoid any potential namespace collision. Here is the way
that the BDC SID can be synchronized to that of the PDC (this is the
default NT4 domain practice also):
Usually it is not necessary to specify the target server (-S FRODO) or the
administrator account credentials (-Uroot%not24get).
252 Remote and Local Management: The Net Command Chapter 12
A share can be added using the net rpc share command capabilities. The
target machine may be local or remote and is specified by the -S option.
It must be noted that the addition and deletion of shares using this tool
depends on the availability of a suitable interface script. The interface scripts
Sambas smbd uses are called add share script and delete share script. A
set of example scripts are provided in the Samba source code tarball in the
directory ~samba/examples/scripts.
The following steps demonstrate the use of the share management capabili-
ties of the net utility. In the first step a share called Bulge is added. The
sharepoint within the file system is the directory /data. The command that
can be executed to perform the addition of this share is shown here:
A simple validation shown here demonstrates that the share has been re-
moved:
At this time the net tool cannot be used to manage ACLs on Samba shares.
In MS Windows language this is called Share Permissions.
Shares and files can be migrated in the same manner as user, machine, and
group accounts. It is possible to preserve access control settings (ACLs)
as well as security settings throughout the migration process. The net rpc
vampire facility is used to migrate accounts from a Windows NT4 (or later)
domain to a Samba server. This process preserves passwords and account
security settings and is a precursor to the migration of shares and files.
The net rpc share command may be used to migrate shares, directories,
files, printers, and all relevant data from a Windows server to a Samba
server.
A set of command-line switches permit the creation of almost direct clones
of Windows file servers. For example, when migrating a fileserver, file ACLs
and DOS file attributes from the Windows server can be included in the
migration process and will reappear, almost identically, on the Samba server
when the migration has been completed.
The migration process can be completed only with the Samba server already
being fully operational. The user and group accounts must be migrated
before attempting to migrate data share, files, and printers. The migration
of files and printer configurations involves the use of both SMB and MS DCE
RPC services. The benefit of the manner in which the migration process has
been implemented is that the possibility now exists to use a Samba server
as a man-in-middle migration service that affects a transfer of data from
one server to another. For example, if the Samba server is called MESSER,
the source Windows NT4 server is called PEPPY, and the target Samba
server is called GONZALES, the machine MESSER can be used to effect
the migration of all data (files and shares) from PEPPY to GONZALES. If
the target machine is not specified, the local server is assumed by default.
The success of server migration requires a firm understanding of the struc-
ture of the source server (or domain) as well as the processes on which the
migration is critically dependant.
There are two known limitations to the migration process:
1. The net command requires that the user credentials provided exist on
both the migration source and the migration target.
2. Printer settings may not be fully or may be incorrectly migrated. This
might in particular happen when migrating a Windows 2003 print
Section 12.8. Share Management 255
server to Samba.
The net rpc share migrate command operation permits the migration of
plain share stanzas. A stanza contains the parameters within which a file or
print share are defined. The use of this migration method will create share
stanzas that have as parameters the file system directory path, an optional
description, and simple security settings that permit write access to files.
One of the first steps necessary following migration is to review the share
stanzas to ensure that the settings are suitable for use.
The shares are created on the fly as part of the migration process. The
smbd application does this by calling on the operating system to execute
the script specified by the smb.conf parameter add share command.
There is a suitable example script for the add share command in the $SAMBA
SOURCES/examples/scripts directory. It should be noted that the account
that is used to drive the migration must, of necessity, have appropriate file
system access privileges and have the right to create shares and to set ACLs
on them. Such rights are conferred by these rights: SeAddUsersPrivilege
and SeDiskOperatorPrivilege. For more information regarding rights and
privileges please refer to Chapter 14, “User Rights and Privileges”.
The syntax of the share migration command is shown here:
This will migrate the share myshare from the server win2k to the Samba
Server using the permissions that are tied to the account administrator
with the password secret. The account that is used must be the same on
256 Remote and Local Management: The Net Command Chapter 12
both the migration source server and the target Samba server. The use of
the net rpc vampire, prior to attempting the migration of shares, will
ensure that accounts will be identical on both systems. One precaution
worth taking before commencement of migration of shares is to validate
that the migrated accounts (on the Samba server) have the needed rights
and privileges. This can be done as shown here:
The steps taken so far perform only the migration of shares. Directories and
directory contents are not migrated by the steps covered up to this point.
Everything covered to this point has been done in preparation for the migra-
tion of file and directory data. For many people preparation is potentially
boring and the real excitement only begins when file data can be used. The
next steps demonstrate the techniques that can be used to transfer (migrate)
data files using the net command.
Transfer of files from one server to another has always been a challenge
for MS Windows administrators because Windows NT and 200X servers
do not include the tools needed. The xcopy is not capable of preserving
file and directory ACLs. Microsoft does provide a utility that can copy
ACLs (security settings) called scopy, but it is provided only as part of the
Windows NT or 200X Server Resource Kit.
There are several tools, both commercial and freeware, that can be used
from a Windows server to copy files and directories with full preservation of
security settings. One of the best known of the free tools is called robocopy.
The net utility can be used to copy files and directories with full preservation
of ACLs as well as DOS file attributes. Note that including ACLs makes
sense only where the destination system will operate within the same security
context as the source system. This applies both to a DMS and to domain
controllers that result from a vampired domain. Before file and directory
migration, all shares must already exist.
Note
The ACLs that result on a Samba server will most probably not match
the originating ACLs. Windows supports the possibility of files that are
owned only by a group. Group-alone file ownership is not possible under
UNIX/Linux. Errors in migrating group-owned files can be avoided by using
the smb.conf file force unknown acl user = yes parameter. This facility will
automatically convert group-owned files into correctly user-owned files on
the Samba server.
An example for migration of files from a machine called nt4box to the Samba
server from which the process will be handled is shown here:
--attrs -U administrator%secret
This command will migrate all files and directories from all file shares on the
Windows server called nt4box to the Samba server from which migration
is initiated. Files that are group-owned will be owned by the user account
administrator.
The operating mode shown here is just a combination of the previous two.
It first migrates share definitions and then all shared files and directories:
network will take much time and frustration. Often it will not be possible
to find driver files that are currently in use, necessitating the installation
of newer drivers. Newer drivers often implement printing features that will
necessitate a change in the printer usage. Additionally, with very complex
printer configurations it becomes almost impossible to re-create the same
environment — no matter how extensively it has been documented.
The migration of an existing printing architecture involves the following:
• Establishment of print queues.
• Installation of printer drivers (both for the print server and for Win-
dows clients.
• Configuration of printing forms.
• Implementation of security settings.
• Configuration of printer settings.
The Samba net utility permits printer migration from one Windows print
server to another. When this tool is used to migrate printers to a Samba
server smbd, the application that receives the network requests to create
the necessary services must call out to the operating system in order to
create the underlying printers. The call-out is implemented by way of an
interface script that can be specified by the smb.conf file parameter . This
script is essential to the migration process. A suitable example script may
be obtained from the $SAMBA SOURCES/examples/scripts directory. Take
note that this script must be customized to suit the operating system envi-
ronment and may use its tools to create a print queue.
Each of the components listed above can be completed separately, or they
can be completed as part of an automated operation. Many network admin-
istrators prefer to deal with migration issues in a manner that gives them
the most control, particularly when things go wrong. The syntax for each
operation is now briefly described.
Printer migration from a Windows print server (NT4 or 200x) is shown.
This instruction causes the printer share to be created together with the
underlying print queue:
Printer drivers can be migrated from the Windows print server to the Samba
server using this command-line instruction:
Printer security settings (ACLs) can be migrated from the Windows server
to the Samba server using this command:
Printer configuration settings include factors such as paper size and default
paper orientation. These can be migrated from the Windows print server to
the Samba server with this command:
The man page documents the net file function suite, which provides the
tools to close open files using either RAP or RPC function calls. Please refer
to the man page for specific usage information.
Section 12.10. Session and Connection Management 261
The session management interface of the net session command uses the
old RAP method to obtain the list of connections to the Samba server, as
shown here:
If the asterisk (*) is used in place of the printer name argument, a list of all
printers will be returned.
To publish (make available) a printer to ADS, execute the following com-
mand:
A generic search (query) can also be made to locate a printer across the
entire ADS domain by executing:
Please refer to the net command man page for information regarding cache
management.
Another useful tool is the net time tool set. This tool may be used to query
the current time on the target server as shown here:
Section 12.13. Other Miscellaneous Operations 263
In the event that it is the intent to pass the time information obtained to
the UNIX /bin/time, it is a good idea to obtain the time from the target
server in a format that is ready to be passed through. This may be done by
executing:
IDENTITY MAPPING
(IDMAP)
264
Section 13.1. Samba Server Deployment Types and IDMAP 265
Samba-3 can act as a Windows NT4 PDC or BDC, thereby providing domain
control protocols that are compatible with Windows NT4. Samba-3 file and
print sharing protocols are compatible with all versions of MS Windows
products. Windows NT4, as with MS Active Directory, extensively makes
use of Windows SIDs.
1
Samba local account mode results in both DOMINICUS\FJones and FRANCISCUS\FJones
mapping to the UNIX user FJones.
266 Identity Mapping (IDMAP) Chapter 13
Samba-3 domain member servers and clients must interact correctly with
MS Windows SIDs. Incoming Windows SIDs must be translated to local
UNIX UIDs and GIDs. Outgoing information from the Samba server must
provide to MS Windows clients and servers appropriate SIDs.
Winbind is not used; users and groups are local: Where winbindd
is not used Samba (smbd) uses the underlying UNIX/Linux mech-
anisms to resolve the identity of incoming network traffic. This is
done using the LoginID (account name) in the session setup request
and passing it to the getpwnam() system function call. This call is im-
plemented using the name service switch (NSS) mechanism on modern
UNIX/Linux systems. By saying ”users and groups are local,” we are
implying that they are stored only on the local system, in the /etc/
passwd and /etc/group respectively.
Winbind is not used; users and groups resolved via NSS: In this sit-
uation user and group accounts are treated as if they are local accounts.
The only way in which this differs from having local accounts is that
the accounts are stored in a repository that can be shared. In practice
this means that they will reside in either an NIS-type database or else
in LDAP.
Winbind/NSS with the default local IDMAP table: There are many
sites that require only a simple Samba server or a single Samba server
that is a member of a Windows NT4 domain or an ADS domain. A
typical example is an appliance like file server on which no local ac-
counts are configured and winbind is used to obtain account credentials
from the domain controllers for the domain. The domain control can
be provided by Samba-3, MS Windows NT4, or MS Windows Active
Directory.
Winbind is a great convenience in this situation. All that is needed
is a range of UID numbers and GID numbers that can be defined in
the smb.conf file. The /etc/nsswitch.conf file is configured to use
winbind, which does all the difficult work of mapping incoming SIDs
to appropriate UIDs and GIDs. The SIDs are allocated a UID/GID
in the order in which winbind receives them.
This configuration is not convenient or practical in sites that have
more than one Samba server and that require the same UID or GID
for the same user or group across all servers. One of the hazards of
this method is that in the event that the winbind IDMAP file becomes
corrupted or lost, the repaired or rebuilt IDMAP file may allocate
UIDs and GIDs to different users and groups from what was there
previously with the result that MS Windows files that are stored on
the Samba server may now not belong to the rightful owners.
It is important that all LDAP IDMAP clients use only the master
LDAP server because the idmap backend facility in the smb.conf file
does not correctly handle LDAP redirects.
Winbind with NSS to resolve UNIX/Linux user and group IDs: The
use of LDAP as the passdb backend is a smart solution for PDC, BDC,
and domain member servers. It is a neat method for assuring that
UIDs, GIDs, and the matching SIDs are consistent across all servers.
The use of the LDAP-based passdb backend requires use of the PADL
nss ldap utility or an equivalent. In this situation winbind is used to
handle foreign SIDs, that is, SIDs from standalone Windows clients
(i.e., not a member of our domain) as well as SIDs from another do-
main. The foreign UID/GID is mapped from allocated ranges (idmap
uid and idmap gid) in precisely the same manner as when using win-
bind with a local IDMAP table.
The nss ldap tool set can be used to access UIDs and GIDs via LDAP
as well as via Active Directory. In order to use Active Directory,
it is necessary to modify the ADS schema by installing either the
AD4UNIX schema extension or using the Microsoft Services for UNIX
version 3.5 or later to extend the ADS schema so it maintains UNIX
account credentials. Where the ADS schema is extended, a Microsoft
Management Console (MMC) snap-in is also installed to permit the
UNIX credentials to be set and managed from the ADS User and
Computer Management tool. Each account must be separately UNIX-
enabled before the UID and GID data can be used by Samba.
Section 13.1. Samba Server Deployment Types and IDMAP 269
Microsoft Windows domain security systems generate the user and group
SID as part of the process of creation of an account. Windows does not
have a concept of the UNIX UID or a GID; rather, it has its own type of
security descriptor. When Samba is used as a domain controller, it provides a
method of producing a unique SID for each user and group. Samba generates
a machine and a domain SID to which it adds an RID that is calculated
algorithmically from a base value that can be specified in the smb.conf
file, plus twice (2x) the UID or GID. This method is called “algorithmic
mapping”.
For example, if a user has a UID of 4321, and the algorithmic RID base has
a value of 1000, the RID will be 1000 + (2 x 4321) = 9642. Thus, if the
domain SID is S-1-5-21-89238497-92787123-12341112, the resulting SID
is S-1-5-21-89238497-92787123-12341112-9642.
The foregoing type of SID is produced by Samba as an automatic function
and is either produced on the fly (as is the case when using a passdb back-
end = [tdbsam | smbpasswd]), or may be stored as a permanent part of
an account in an LDAP-based ldapsam.
ADS uses a directory schema that can be extended to accommodate ad-
ditional account attributes such as UIDs and GIDs. The installation of
Microsoft Service for UNIX 3.5 will expand the normal ADS schema to
include UNIX account attributes. These must of course be managed sepa-
rately through a snap-in module to the normal ADS account management
MMC interface.
Security identifiers used within a domain must be managed to avoid con-
flict and to preserve itegrity. In an NT4 domain context, the PDC manages
the distribution of all security credentials to the backup domain controllers
(BDCs). At this time the only passdb backend for a Samba domain con-
troller that is suitable for such information is an LDAP backend.
BDCs have read-only access to security credentials that are stored in LDAP.
Changes in user or group account information are passed by the BDC to the
PDC. Only the PDC can write changes to the directory.
IDMAP information can be written directly to the LDAP server so long as
270 Identity Mapping (IDMAP) Chapter 13
all domain controllers have access to the master (writable) LDAP server.
Samba-3 at this time does not handle LDAP redirects in the IDMAP back-
end. This means that it is is unsafe to use a slave (replicate) LDAP server
with the IDMAP facility.
Anyone who wishes to use winbind will find the following example config-
urations helpful. Remember that in the majority of cases winbind is of
primary interest for use with domain member servers (DMSs) and domain
member clients (DMCs).
The following is a simple example of an NT4 DMS smb.conf file that shows
only the global section.
#Global parameters
[global]
workgroup = MEGANET2
security = DOMAIN
idmap uid = 10000-20000
idmap gid = 10000-20000
template primary group = "Domain Users"
template shell = /bin/bash
...
passwd: files winbind
shadow: files winbind
group: files winbind
...
hosts: files [dns] wins
...
The use of DNS in the hosts entry should be made only if DNS is used on
site.
2. Execute:
The success of the join can be confirmed with the following command:
3. Start the nmbd, winbind, and smbd daemons in the order shown.
272 Identity Mapping (IDMAP) Chapter 13
The procedure for joining an ADS domain is similar to the NT4 domain
join, except the smb.conf file will have the following contents:
# Global parameters
[global]
workgroup = BUTTERNET
netbios name = GARGOYLE
realm = BUTTERNET.BIZ
security = ADS
template shell = /bin/bash
idmap uid = 500-10000000
idmap gid = 500-10000000
winbind use default domain = Yes
winbind nested groups = Yes
printer admin = "BUTTERNET\Domain Admins"
ADS DMS operation requires use of kerberos (KRB). For this to work, the
krb5.conf must be configured. The exact requirements depends on which
version of MIT or Heimdal Kerberos is being used. It is sound advice to use
only the latest version, which at this time are MIT Kerberos version 1.3.5
and Heimdal 0.61.
The creation of the DMS requires the following steps:
1. Create or install an smb.conf file with the above configuration.
2. Edit the /etc/nsswitch.conf file as shown above.
3. Execute:
The success or failure of the join can be confirmed with the following
command:
The specific error message may differ from the above because it de-
pends on the type of failure that may have occurred. Increase the log
level to 10, repeat the test, and then examine the log files produced
to identify the nature of the failure.
4. Start the nmbd, winbind, and smbd daemons in the order shown.
The idmap rid facility is a new tool that, unlike native winbind, creates a
predictable mapping of MS Windows SIDs to UNIX UIDs and GIDs. The
key benefit of this method of implementing the Samba IDMAP facility is
that it eliminates the need to store the IDMAP data in a central place. The
downside is that it can be used only within a single ADS domain and is not
compatible with trusted domain implementations.
This alternate method of SID to UID/GID mapping can be achieved using
the idmap rid plug-in. This plug-in uses the RID of the user SID to derive
the UID and GID by adding the RID to a base value specified. This utility
requires that the parameter “allow trusted domains = No” be specified, as
it is not compatible with multiple domain environments. The idmap uid
and idmap gid ranges must be specified.
The idmap rid facility can be used both for NT4/Samba-style domains and
Active Directory. To use this with an NT4 domain, do not include the realm
parameter; additionally, the method used to join the domain uses the net
rpc join process.
An example smb.conf file for and ADS domain environment is shown here:
274 Identity Mapping (IDMAP) Chapter 13
# Global parameters
[global]
workgroup = KPAK
netbios name = BIGJOE
realm = CORP.KPAK.COM
server string = Office Server
security = ADS
allow trusted domains = No
idmap backend = idmap_rid:KPAK=500-100000000
idmap uid = 500-100000000
idmap gid = 500-100000000
template shell = /bin/bash
winbind use default domain = Yes
winbind enum users = No
winbind enum groups = No
winbind nested groups = Yes
printer admin = "Domain Admins"
The use of this tool requires configuration of NSS as per the native use of
winbind. Edit the /etc/nsswitch.conf so it has the following parameters:
...
passwd: files winbind
shadow: files winbind
group: files winbind
...
hosts: files wins
...
Section 13.2. Examples of IDMAP Backend Usage 275
The specific error message may differ from the above because it de-
pends on the type of failure that may have occurred. Increase the log
level to 10, repeat the test, and then examine the log files produced
to identify the nature of the failure.
4. Start the nmbd, winbind, and smbd daemons in the order shown.
5. Validate the operation of this configuration by executing:
The storage of IDMAP information in LDAP can be used with both NT4/Samba-
3-style domains and ADS domains. OpenLDAP is a commonly used LDAP
server for this purpose, although any standards-complying LDAP server
can be used. It is therefore possible to deploy this IDMAP configuration
276 Identity Mapping (IDMAP) Chapter 13
using the Sun iPlanet LDAP server, Novell eDirectory, Microsoft ADS plus
ADAM, and so on.
The following example is for an ADS domain:
# Global parameters
[global]
workgroup = SNOWSHOW
netbios name = GOODELF
realm = SNOWSHOW.COM
server string = Samba Server
security = ADS
log level = 1 ads:10 auth:10 sam:10 rpc:10
ldap admin dn = cn=Manager,dc=SNOWSHOW,dc=COM
ldap idmap suffix = ou=Idmap
ldap suffix = dc=SNOWSHOW,dc=COM
idmap backend = ldap:ldap://ldap.snowshow.com
idmap uid = 150000-550000
idmap gid = 150000-550000
template shell = /bin/bash
winbind use default domain = Yes
In the case of an NT4 or Samba-3-style domain the realm is not used, and
the command used to join the domain is net rpc join. The above example
also demonstrates advanced error-reporting techniques that are documented
in Section 38.3.
Where MIT kerberos is installed (version 1.3.4 or later), edit the /etc/krb5.
conf file so it has the following contents:
[logging]
default = FILE:/var/log/krb5libs.log
kdc = FILE:/var/log/krb5kdc.log
admin_server = FILE:/var/log/kadmind.log
[libdefaults]
default_realm = SNOWSHOW.COM
dns_lookup_realm = false
dns_lookup_kdc = true
Section 13.2. Examples of IDMAP Backend Usage 277
[appdefaults]
pam = {
debug = false
ticket_lifetime = 36000
renew_lifetime = 36000
forwardable = true
krb4_convert = false
}
[libdefaults]
default_realm = SNOWSHOW.COM
clockskew = 300
[realms]
SNOWSHOW.COM = {
kdc = ADSDC.SHOWSHOW.COM
}
[domain_realm]
.snowshow.com = SNOWSHOW.COM
Note
Edit the NSS control file /etc/nsswitch.conf so it has the following entries:
278 Identity Mapping (IDMAP) Chapter 13
...
passwd: files ldap
shadow: files ldap
group: files ldap
...
hosts: files wins
...
You will need the PADL2 nss ldap tool set for this solution. Configure the
/etc/ldap.conf file so it has the information needed. The following is an
example of a working file:
host 192.168.2.1
base dc=snowshow,dc=com
binddn cn=Manager,dc=snowshow,dc=com
bindpw not24get
pam_password exop
nss_base_passwd ou=People,dc=snowshow,dc=com?one
nss_base_shadow ou=People,dc=snowshow,dc=com?one
nss_base_group ou=Groups,dc=snowshow,dc=com?one
ssl no
4. Download, build, and install the PADL nss ldap tool set. Configure
the /etc/ldap.conf file as shown above.
5. Configure an LDAP server and initialize the directory with the top-
level entries needed by IDMAP, shown in the following LDIF file:
2
<http://www.padl.com>
Section 13.2. Examples of IDMAP Backend Usage 279
dn: dc=snowshow,dc=com
objectClass: dcObject
objectClass: organization
dc: snowshow
o: The Greatest Snow Show in Singapore.
description: Posix and Samba LDAP Identity Database
dn: cn=Manager,dc=snowshow,dc=com
objectClass: organizationalRole
cn: Manager
description: Directory Manager
dn: ou=Idmap,dc=snowshow,dc=com
objectClass: organizationalUnit
ou: idmap
6. Execute the command to join the Samba DMS to the ADS domain as
shown here:
8. Start the nmbd, winbind, and smbd daemons in the order shown.
13.2.4 IDMAP and NSS Using LDAP from ADS with RFC2307bis
Schema Extension
The use of this method is messy. The information provided in the following
is for guidance only and is very definitely not complete. This method does
work; it is used in a number of large sites and has an acceptable level of
performance.
The following is an example smb.conf file:
# Global parameters
[global]
workgroup = BOBBY
realm = BOBBY.COM
security = ADS
idmap uid = 150000-550000
idmap gid = 150000-550000
template shell = /bin/bash
winbind cache time = 5
winbind use default domain = Yes
winbind trusted domains only = Yes
winbind nested groups = Yes
The DMS must be joined to the domain using the usual procedure. Addi-
tionally, it is necessary to build and install the PADL nss ldap tool set. Be
sure to build this tool set with the following:
...
passwd: files ldap
shadow: files ldap
group: files ldap
...
hosts: files wins
Section 13.2. Examples of IDMAP Backend Usage 281
...
The /etc/ldap.conf file must be configured also. Refer to the PADL doc-
umentation and source code for nss ldap to specific instructions.
The next step involves preparation of the ADS schema. This is briefly
discussed in the remaining part of this chapter.
The Microsoft Windows Service for UNIX (SFU) version 3.5 is available for
free download3 from the Microsoft Web site. You will need to download this
tool and install it following Microsoft instructions.
Instructions for obtaining and installing the AD4UNIX tool set can be found
from the Geekcomix4 Web site.
3
<http://www.microsoft.com/windows/sfu/>
4
<http://www.geekcomix.com/cgi-bin/classnotes/wiki.pl?LDAP01/An_
Alternative_Approach>
Chapter 14
282
Section 14.1. Rights Management Capabilities 283
Note
The creation of UNIX system accounts has traditionally been the sole right
of the system administrator, better known as the root account. It is possible
in the UNIX environment to create multiple users who have the same UID.
Any UNIX user who has a UID=0 is inherently the same as the root account
user.
All versions of Samba call system interface scripts that permit CIFS function
calls that are used to manage users, groups, and machine accounts in the
UNIX environment. All versions of Samba up to and including version 3.0.10
required the use of a Windows administrator account that unambiguously
maps to the UNIX root account to permit the execution of these interface
scripts. The requirement to do this has understandably met with some
disdain and consternation among Samba administrators, particularly where
it became necessary to permit people who should not possess root-level
access to the UNIX host system.
Samba 3.0.11 introduced support for the Windows privilege model. This
model allows certain rights to be assigned to a user or group SID. In order
to enable this feature, enable privileges = yes must be defined in the global
section of the smb.conf file.
Currently, the rights supported in Samba-3 are listed in Table 14.1. The
remainder of this chapter explains how to manage and use these privileges
284 User Rights and Privileges Chapter 14
on Samba servers.
There are two primary means of managing the rights assigned to users and
groups on a Samba server. The NT4 User Manager for Domains may
be used from any Windows NT4, 2000, or XP Professional domain member
client to connect to a Samba domain controller and view/modify the rights
assignments. This application, however, appears to have bugs when run
on a client running Windows 2000 or later; therefore, Samba provides a
command-line utility for performing the necessary administrative actions.
The net rpc rights utility in Samba 3.0.11 has three new subcommands:
list [name|accounts] When called with no arguments, net rpc list simply
lists the available rights on the server. When passed a specific user
or group name, the tool lists the privileges currently assigned to the
specified account. When invoked using the special string accounts,
net rpc rights list returns a list of all privileged accounts on the
server and the assigned rights.
grant <user> <right [right ...]> When called with no arguments, this
function is used to assign a list of rights to a specified user or group.
For example, to grant the members of the Domain Admins group on
a Samba domain controller, the capability to add client machines to
the domain, one would run:
Note
The privileges that have been implemented in Samba-3.0.11 are shown be-
low. It is possible, and likely, that additional privileges may be implemented
in later releases of Samba. It is also likely that any privileges currently imple-
mented but not used may be removed from future releases as a house-keeping
matter, so it is important that the successful as well as unsuccessful use of
these facilities should be reported on the Samba mailing lists.
The Samba Team are implementing only those privileges that are logical and
useful in the UNIX/Linux envronment. Many of the Windows 200X/XP
privileges have no direct equivalence in UNIX.
Please note that every Windows NT4 and later server requires a domain
Administrator account. Samba version commencing with 3.0.11 permit Ad-
ministrative duties to be performed via assigned rights and privileges (see
Chapter 14, “User Rights and Privileges”). An account in the server’s passdb
backend can be set to the well-known RID of the default administrator ac-
count. To obtain the domain SID on a Samba domain controller, run the
following command:
Section 14.3. Common Errors 289
You may assign the domain administrator RID to an account using the
pdbedit command as shown here:
Note
Releases of Samba version 3.0.11 and later make it possible to operate with-
out an Administrator account providing equivalent rights and privileges have
been established for a Windows user or a Windows group account.
When a Windows NT4 (or later) client joins a domain, the domain global
Domain Admins group is added to the membership of the local Administra-
tors group on the client. Any user who is a member of the domain global
Domain Admins group will have administrative rights on the Windows client.
This is often not the most desirable solution because it means that the user
will have administrative rights and privileges on domain servers also. The
290 User Rights and Privileges Chapter 14
The problem lies in the differences in how file and directory permissions and
controls work between the two environments. This difference is one that
Samba cannot completely hide, even though it does try to bridge the chasm
to a degree.
POSIX Access Control List technology has been available (along with ex-
tended attributes) for UNIX for many years, yet there is little evidence today
of any significant use. This explains to some extent the slow adoption of
ACLs into commercial Linux products. MS Windows administrators are as-
tounded at this, given that ACLs were a foundational capability of the now
decade-old MS Windows NT operating system.
The purpose of this chapter is to present each of the points of control that
are possible with Samba-3 in the hope that this will help the network ad-
ministrator to find the optimum method for delivering the best environment
for MS Windows desktop users.
291
292 File, Directory, and Share Access Controls Chapter 15
Samba offers much flexibility in file system access management. These are
the key access control facilities present in Samba today:
Samba Access Control Facilities
• UNIX File and Directory Permissions
Samba honors and implements UNIX file system access controls. Users
who access a Samba server will do so as a particular MS Windows user.
This information is passed to the Samba server as part of the logon
or connection setup process. Samba uses this user identity to validate
whether or not the user should be given access to file system resources
(files and directories). This chapter provides an overview for those
to whom the UNIX permissions and controls are a little strange or
unknown.
• Samba Share Definitions
In configuring share settings and controls in the smb.conf file, the
network administrator can exercise overrides to native file system per-
missions and behaviors. This can be handy and convenient to effect
behavior that is more like what MS Windows NT users expect, but it is
seldom the best way to achieve this. The basic options and techniques
are described herein.
• Samba Share ACLs
Just as it is possible in MS Windows NT to set ACLs on shares them-
selves, so it is possible to do in Samba. Few people make use of this
facility, yet it remains one of the easiest ways to affect access con-
trols (restrictions) and can often do so with minimum invasiveness
compared with other methods.
Section 15.2. File System Access Controls 293
Perhaps the most important recognition to be made is the simple fact that
MS Windows NT4/200x/XP implement a totally divergent file system tech-
nology from what is provided in the UNIX operating system environment.
First we consider what the most significant differences are, then we look at
how Samba helps to bridge the differences.
Samba operates on top of the UNIX file system. This means it is subject
to UNIX file system conventions and permissions. It also means that if the
MS Windows networking environment requires file system behavior, that
differs from UNIX file system behavior then somehow Samba is responsible
for emulating that in a transparent and consistent manner.
It is good news that Samba does this to a large extent, and on top of
that, provides a high degree of optional configuration to override the default
behavior. We look at some of these overrides, but for the greater part we
stay within the bounds of default behavior. Those wishing to explore the
depths of control ability should review the smb.conf man page.
The following compares file system features for UNIX with those of MS
Windows NT/200x:
UNIX file and directory names are case sensitive and case preserving.
Samba implements the MS Windows file name behavior, but it does so
as a user application. The UNIX file system provides no mechanism to
perform case-insensitive file name lookups. MS Windows does this by
default. This means that Samba has to carry the processing overhead
to provide features that are not native to the UNIX operating system
environment.
Consider the following. All are unique UNIX names but one single MS
Windows file name:
MYFILE.TXT
MyFile.txt
myfile.txt
Links and Short-Cuts MS Windows make use of links and shortcuts that
are actually special types of files that will redirect an attempt to ex-
ecute the file to the real location of the file. UNIX knows of file and
directory links, but they are entirely different from what MS Windows
users are used to.
Symbolic links are files in UNIX that contain the actual location of the
data (file or directory). An operation (like read or write) will operate
directly on the file referenced. Symbolic links are also referred to as
“soft links.” A hard link is something that MS Windows is not familiar
with. It allows one physical file to be known simultaneously by more
than one file name.
There are many other subtle differences that may cause the MS Windows
administrator some temporary discomfort in the process of becoming famil-
iar with UNIX/Linux. These are best left for a text that is dedicated to the
purpose of UNIX/Linux training and education.
There are three basic operations for managing directories: create, delete,
rename. Table 15.1 compares the commands in Windows and UNIX that
implement these operations.
296 File, Directory, and Share Access Controls Chapter 15
$ ls -la
total 632
drwxr-xr-x 13 maryo gnomes 816 2003-05-12 22:56 .
drwxrwxr-x 37 maryo gnomes 3800 2003-05-12 22:29 ..
dr-xr-xr-x 2 maryo gnomes 48 2003-05-12 22:29 muchado02
drwxrwxrwx 2 maryo gnomes 48 2003-05-12 22:29 muchado03
drw-rw-rw- 2 maryo gnomes 48 2003-05-12 22:29 muchado04
d-w--w--w- 2 maryo gnomes 48 2003-05-12 22:29 muchado05
dr--r--r-- 2 maryo gnomes 48 2003-05-12 22:29 muchado06
drwsrwsrwx 2 maryo gnomes 48 2003-05-12 22:29 muchado08
---------- 1 maryo gnomes 1242 2003-05-12 22:31 mydata00.lst
--w--w--w- 1 maryo gnomes 7754 2003-05-12 22:33 mydata02.lst
-r--r--r-- 1 maryo gnomes 21017 2003-05-12 22:32 mydata04.lst
-rw-rw-rw- 1 maryo gnomes 41105 2003-05-12 22:32 mydata06.lst
$
The columns represent (from left to right) permissions, number of hard links
to file, owner, group, size (bytes), access date, time of last modification, and
file name.
An overview of the permissions field is shown in Figure 15.1.
Section 15.2. File System Access Controls 297
Any bit flag may be unset. An unset bit flag is the equivalent of ”cannot”
and is represented as a “-” character (see Example 15.2.1)
-rwxr-x--- Means:
^^^ The owner (user) can read, write, execute
^^^ the group can read and execute
^^^ everyone else cannot do anything with it.
created within it will be owned by the user and/or group whose ‘set user or
group’ bit is set. This can be helpful in setting up directories for which it is
desired that all users who are in a group should be able to write to and read
from a file, particularly when it is undesirable for that file to be exclusively
owned by a user whose primary group is not the group that all such users
belong to.
When a directory is set d-wx--x---, the owner can read and create (write)
files in it, but because the (r) read flags are not set, files cannot be listed
(seen) in the directory by anyone. The group can read files in the directory
but cannot create new files. If files in the directory are set to be readable
and writable for the group, then group members will be able to write to (or
delete) them.
People have asked on the Samba mailing list how is it possible to protect files
or directories from deletion by users. For example, Windows NT/2K/XP
provides the capacity to set access controls on a directory into which people
can write files but not delete them. It is possible to set an ACL on a Windows
file that permits the file to be written to but not deleted. Such concepts
are foreign to the UNIX operating system file space. Within the UNIX file
system anyone who has the ability to create a file can write to it. Anyone
who has write permission on the directory that contains a file and has write
permission for it has the capability to delete it.
For the record, in the UNIX environment the ability to delete a file is con-
trolled by the permissions on the directory that the file is in. In other words,
a user can delete a file in a directory to which that user has write access,
even if that user does not own the file.
Of necessity, Samba is subject to the file system semantics of the host oper-
ating system. Samba is therefore limited in the file system capabilities that
can be made available through Windows ACLs, and therefore performs a
”best fit” translation to POSIX ACLs. Some UNIX file systems do, how-
ever support, a feature known as extended attributes. Only the Windows
concept of inheritance is implemented by Samba through the appropriate
extended attribute.
The specific semantics of the extended attributes are not consistent across
UNIX and UNIX-like systems such as Linux. For example, it is possible on
Section 15.2. File System Access Controls 299
2. Login as the root user, then set the immutibile flag on a test file as
follows:
3. Login as the user who owns the file (not root) and attempt to remove
the file as follows:
It will not be possible to delete the file if the immutible flag is correctly
honored.
On operating systems and file system types that support the immutible bit,
it is possible to create directories that cannot be deleted. Check the man
page on your particular host system to determine whether or not immutable
directories are writable. If they are not, then the entire directory and its
contents will effectively be protected from writing (file creation also) and
deletion.
300 File, Directory, and Share Access Controls Chapter 15
The following parameters in the smb.conf file sections define a share control
or affect access controls. Before using any of the following options, please
refer to the man page for smb.conf.
User- and group-based controls can prove quite useful. In some situations
it is distinctly desirable to force all file system operations as if a single user
were doing so. The use of the force user and force group behavior will
achieve this. In other situations it may be necessary to use a paranoia level
of control to ensure that only particular authorized persons will be able to
access a share or its contents. Here the use of the valid users or the invalid
users parameter may be useful.
As always, it is highly advisable to use the easiest to maintain and the least
ambiguous method for controlling access. Remember, when you leave the
scene, someone else will need to provide assistance, and if he or she finds
too great a mess or does not understand what you have done, there is risk
of Samba being removed and an alternative solution being adopted.
Table 15.2 enumerates these controls.
This section deals with how to configure Samba per-share access control
restrictions. By default, Samba sets no restrictions on the share itself. Re-
strictions on the share itself can be set on MS Windows NT4/200x/XP
302 File, Directory, and Share Access Controls Chapter 15
shares. This can be an effective way to limit who can connect to a share. In
the absence of specific restrictions, the default setting is to allow the global
user Everyone - Full Control (full control, change and read).
Section 15.4. Access Controls on Shares 303
At this time Samba does not provide a tool for configuring access control
settings on the share itself the only way to create those settings is to use
either the NT4 Server Manager or the Windows 200x Microsoft Management
Console (MMC) for Computer Management. There are currently no plans
to provide this capability in the Samba command-line tool set.
Samba stores the per-share access control settings in a file called share
info.tdb. The location of this file on your system will depend on how Samba
was compiled. The default location for Samba’s tdb files is under /usr/
local/samba/var. If the tdbdump utility has been compiled and installed
on your system, then you can examine the contents of this file by executing
tdbdump share info.tdb in the directory containing the tdb files.
The tool you need to manage share permissions on a Samba server from
a Windows NT4 Workstation or Server is the NT Server Manager. Server
Manager is shipped with Windows NT4 Server products but not with Win-
dows NT4 Workstation. You can obtain the NT Server Manager for MS
Windows NT4 Workstation from the Microsoft web site support1 section.
Instructions
1. Launch the NT4 Server Manager and click on the Samba server you
want to administer. From the menu select Computer, then click on
Shared Directories.
2. Click on the share that you wish to manage and click the Properties
tab, then click the Permissions tab. Now you can add or change access
control settings as you wish.
On MS Windows NT4/200x/XP system, ACLs on the share itself are set us-
ing native tools, usually from File Manager. For example, in Windows 200x,
1
<http://support.microsoft.com/default.aspx?scid=kb;en-us;173673>
304 File, Directory, and Share Access Controls Chapter 15
right-click on the shared folder, then select Sharing, then click on Permis-
sions. The default Windows NT4/200x permission allows ”Everyone” full
control on the share.
MS Windows 200x and later versions come with a tool called the Com-
puter Management snap-in for the MMC. This tool is located by clicking
on Control Panel -> Administrative Tools -> Computer Management.
Instructions
2. If the Samba server is not shown in the Select Computer box, type in
the name of the target Samba server in the field Name:. Now click the
on [+] next to System Tools, then on the [+] next to Shared Folders
in the left panel.
Warning
Windows NT clients can use their native security settings dialog box to view
and modify the underlying UNIX permissions.
This ability is careful not to compromise the security of the UNIX host on
which Samba is running and still obeys all the file permission rules that a
Samba administrator can set.
Samba does not attempt to go beyond POSIX ACLs, so the various finer-
grained access control options provided in Windows are actually ignored.
Note
Clicking on the Ownership button brings up a dialog box telling you who
owns the given file. The owner name will be displayed like this:
SERVER is the NetBIOS name of the Samba server, user is the username
of the UNIX user who owns the file, and (Long name) is the descriptive
string identifying the user (normally found in the GECOS field of the UNIX
password database). Click on the Close button to remove this dialog.
If the parameter nt acl support is set to false, the file owner will be shown
as the NT user Everyone.
The Take Ownership button will not allow you to change the ownership of
this file to yourself (clicking it will display a dialog box complaining that the
user as whom you are currently logged onto the NT client cannot be found).
The reason for this is that changing the ownership of a file is a privileged
operation in UNIX, available only to the root user. Because clicking on this
button causes NT to attempt to change the ownership of a file to the current
user logged into the NT client, this will not work with Samba at this time.
There is an NT chown command that will work with Samba and allow a user
with administrator privilege connected to a Samba server as root to change
the ownership of files on both a local NTFS file system or remote mounted
NTFS or Samba drive. This is available as part of the Seclib NT security
library written by Jeremy Allison of the Samba Team and is downloadable
from the main Samba FTP site.
string identifying the user (normally found in the GECOS field of the UNIX
password database).
If the parameter nt acl support is set to false, the file owner will be shown
as the NT user Everyone, and the permissions will be shown as NT Full
Control.
The permissions field is displayed differently for files and directories. Both
are discussed next.
The second set of directory permissions has no real meaning in the UNIX
permissions world and represents the inherited permissions that any file
created within this directory would inherit.
Samba synthesizes these inherited permissions for NT by returning as an
NT ACL the UNIX permission mode that a new file created by Samba on
this share would receive.
There are four parameters that control interaction with the standard Samba
create mask parameters:
• security mask
• force security mode
• directory security mask
• force directory security mode
When a user clicks on OK to apply the permissions, Samba maps the given
permissions into a user/group/world r/w/x triplet set, and then checks the
changed permissions for a file against the bits set in the security mask pa-
rameter. Any bits that were changed that are not set to 1 in this parameter
are left alone in the file permissions.
Essentially, zero bits in the security mask may be treated as a set of bits
the user is not allowed to change, and one bits are those the user is allowed
to change.
If not explicitly set, this parameter defaults to the same value as the cre-
ate mask parameter. To allow a user to modify all the user/group/world
permissions on a file, set this parameter to 0777.
Next Samba checks the changed permissions for a file against the bits set
in the force security mode parameter. Any bits that were changed that
correspond to bits set to 1 in this parameter are forced to be set.
Essentially, bits set in the force security mode parameter may be treated
as a set of bits that, when modifying security on a file, the user has always
set to be on.
If not explicitly set, this parameter defaults to the same value as the force
create mode parameter. To allow a user to modify all the user/group/world
permissions on a file with no restrictions, set this parameter to 000. The
310 File, Directory, and Share Access Controls Chapter 15
security mask and force security mode parameters are applied to the
change request in that order.
For a directory, Samba performs the same operations as described above for
a file except it uses the parameter directory security mask instead of
security mask, and force directory security mode parameter instead
of force security mode.
The directory security mask parameter by default is set to the same value as
the directory mask parameter and the force directory security mode
parameter by default is set to the same value as the force directory mode
parameter. In this way Samba enforces the permission restrictions that an
administrator can set on a Samba share, while still allowing users to modify
the permission bits within that restriction.
If you want to set up a share that allows users full control in modifying
the permission bits on their files and directories and does not force any
particular bits to be set on, then set the following parameters in the smb.
conf file in that share-specific section:
s e c u r i t y mask = 0777
f o r c e s e c u r i t y mode = 0
d i r e c t o r y s e c u r i t y mask = 0777
f o r c e d i r e c t o r y s e c u r i t y mode = 0
Note
If a file has no UNIX read access for the owner, it will show up as “read-only”
Section 15.5. MS Windows Access Control Lists and UNIX Interoperability 311
What this can mean is that if the owner changes the permissions to allow
himself or herself read access using the security dialog, clicks on OK to get
back to the standard attributes tab dialog, and clicks on OK on that dialog,
then NT will set the file permissions back to read-only (as that is what the
attributes still say in the dialog). This means that after setting permissions
and clicking on OK to get back to the attributes dialog, you should always
press Cancel rather than OK to ensure that your changes are not overridden.
Windows administrators are familiar with simple ACL controls, and they
typically consider that UNIX user/group/other (ugo) permissions are inad-
equate and not sufficiently fine-grained.
UNIX vendors differ in the manner in which POSIX ACLs are implemented.
There are a number of Linux file systems that support ACLs. Samba has to
provide a way to make transparent all the differences between the various
implementations of POSIX ACLs. The pressure for ACLs support in Samba
has noticeably increased the pressure to standardize ACLs support in the
UNIX world.
Samba has to deal with the complicated matter of handling the challenge of
the Windows ACL that implements inheritance, a concept not anticipated
by POSIX ACLs as implemented in UNIX file systems. Samba provides
support for masks that permit normal ugo and ACLs functionality to be
312 File, Directory, and Share Access Controls Chapter 15
overrided. This further complicates the way in which Windows ACLs must
be implemented.
As can be seen from the mapping table, there is no one-to-one mapping ca-
pability, and therefore Samba must make a logical mapping that will permit
Windows to operate more-or-less the way that is intended by the adminis-
trator.
The UNIX administrator can set any directory permission from within the
UNIX environment. The Windows administrator is more restricted in that
it is not possible from within Windows Explorer to remove read permission
for the file owner.
Directory permissions function in much the same way as shown for file per-
missions, but there are some notable exceptions and a few peculiarities that
the astute administrator will want to take into account in the setting up of
directory permissions.
File, directory, and share access problems are common topics on the mailing
list. The following are examples recently taken from the mailing list.
314 File, Directory, and Share Access Controls Chapter 15
“We are facing some troubles with file/directory permissions. I can log
on the domain as admin user (root), and there’s a public share on which
everyone needs to have permission to create/modify files, but only root can
change the file, no one else can. We need to constantly go to the server to
chgrp -R users * and chown -R nobody * to allow others users to change
the file.”
There are many ways to solve this problem, and here are a few hints:
2. Set the ownership to whatever public user and group you want
Note
3. Directory is /foodbar:
Note
4. Now type:
5. Now type:
$ su - jill
$ cd /foodbar
$ touch Afile
$ ls -al
You should see that the file Afile created by Jill will have ownership
and permissions of Jack, as follows:
Note
When you have a user in admin users, Samba will always do file operations
for this user as root, even if force user has been set.
Answer: Word does the following when you modify/change a Word docu-
ment: MS Word creates a new document with a temporary name. Word
then closes the old document and deletes it, then renames the new document
to the original document name. There is no mechanism by which Samba
can in any way know that the new document really should be owned by the
owners of the original file. Samba has no way of knowing that the file will
be renamed by MS Word. As far as Samba is able to tell, the file that gets
created is a new file, not one that the application (Word) is updating.
directory in which you are changing Word documents: chmod g+s ‘direc-
tory name’. This ensures that all files will be created with the group that
owns the directory. In smb.conf share declaration section set:
f o r c e c r e a t e mode = 0660
f o r c e d i r e c t o r y mode = 0770
These two settings will ensure that all directories and files that get created
in the share will be readable/writable by the owner and group set on the
directory itself.
318 File, Directory, and Share Access Controls Chapter 15
Table 15.5. How Windows File ACLs Map to UNIX POSIX File ACLs
Windows ACE File Attribute Flag
Full Control #
Traverse Folder/Execute File x
List Folder/Read Data r
Read Attributes r
Read Extended Attribures r
Create Files/Write Data w
Create Folders/Append Data w
Write Attributes w
Write Extended Attributes w
Delete Subfolders and Files w
Delete #
Read Permissions all
Change Permissions #
Take Ownership #
Chapter 16
One area that causes trouble for many network administrators is locking.
The extent of the problem is readily evident from searches over the Internet.
Samba provides all the same locking semantics that MS Windows clients
expect and that MS Windows NT4/200x servers also provide.
The term locking has exceptionally broad meaning and covers a range of
functions that are all categorized under this one term.
320
Section 16.2. Discussion 321
Note
16.2 Discussion
There are two types of locking that need to be performed by an SMB server.
The first is record locking that allows a client to lock a range of bytes in
an open file. The second is the deny modes that are specified when a file is
open.
Record locking semantics under UNIX are very different from record locking
under Windows. Versions of Samba before 2.2 have tried to use the native
fcntl() UNIX system call to implement proper record locking between dif-
ferent Samba clients. This cannot be fully correct for several reasons. The
simplest is that a Windows client is allowed to lock a byte range up to 2ˆ32
or 2ˆ64, depending on the client OS. The UNIX locking only supports byte
ranges up to 2ˆ31. So it is not possible to correctly satisfy a lock request
above 2ˆ31. There are many more differences, too many to be listed here.
Samba 2.2 and above implement record locking completely independently of
the underlying UNIX system. If a byte-range lock that the client requests
happens to fall into the range of 0 to 2ˆ31, Samba hands this request down
to the UNIX system. No other locks can be seen by UNIX, anyway.
Strictly speaking, an SMB server should check for locks before every read
and write call on a file. Unfortunately, with the way fcntl() works, this can be
slow and may overstress the rpc.lockd. This is almost always unnecessary
because clients are independently supposed to make locking calls before
reads and writes if locking is important to them. By default, Samba only
makes locking calls when explicitly asked to by a client, but if you set strict
locking = yes, it will make lock checking calls on every read and write call.
You can also disable byte-range locking completely by using locking = no.
This is useful for those shares that do not support locking or do not need it
(such as CD-ROMs). In this case, Samba fakes the return codes of locking
calls to tell clients that everything is okay.
322 File and Record Locking Chapter 16
The second class of locking is the deny modes. These are set by an applica-
tion when it opens a file to determine what types of access should be allowed
simultaneously with its open. A client may ask for DENY NONE, DENY READ,
DENY WRITE, or DENY ALL. There are also special compatibility modes called
DENY FCB and DENY DOS.
Read-ahead: The client reads the local copy of the file, eliminating network
latency.
Write caching: The client writes to the local copy of the file, eliminating
network latency.
Lock caching: The client caches application locks locally, eliminating net-
work latency.
The performance enhancement of oplocks is due to the opportunity of ex-
clusive access to the file — even if it is opened with deny-none — because
Windows monitors the file’s status for concurrent access from other pro-
cesses.
Windows Defines Four Kinds of Oplocks:
Level1 Oplock The redirector sees that the file was opened with deny none
(allowing concurrent access), verifies that no other process is accessing
the file, checks that oplocks are enabled, then grants deny-all/read-
write/exclusive access to the file. The client now performs operations
on the cached local file.
If a second process attempts to open the file, the open is deferred while
the redirector ”breaks” the original oplock. The oplock break signals
the caching client to write the local file back to the server, flush the
Section 16.2. Discussion 323
local locks, and discard read-ahead data. The break is then complete,
the deferred open is granted, and the multiple processes can enjoy
concurrent file access as dictated by mandatory or byte-range locking
options. However, if the original opening process opened the file with
a share mode other than deny-none, then the second process is granted
limited or no access, despite the oplock break.
Level2 Oplock Performs like a Level1 oplock, except caching is only oper-
ative for reads. All other operations are performed on the server disk
copy of the file.
Batch Oplock Manipulates file openings and closings and allows caching
of file attributes.
An important detail is that oplocks are invoked by the file system, not an
application API. Therefore, an application can close an oplocked file, but the
file system does not relinquish the oplock. When the oplock break is issued,
the file system then simply closes the file in preparation for the subsequent
open by the second process.
Opportunistic locking is actually an improper name for this feature. The
true benefit of this feature is client-side data caching, and oplocks is merely
a notification mechanism for writing data back to the networked storage
disk. The limitation of oplocks is the reliability of the mechanism to process
an oplock break (notification) between the server and the caching client. If
this exchange is faulty (usually due to timing out for any number of reasons),
then the client-side caching benefit is negated.
The actual decision that a user or administrator should consider is whether
it is sensible to share among multiple users data that will be cached locally
on a client. In many cases the answer is no. Deciding when to cache or
not cache data is the real question, and thus oplocks should be treated as
a toggle for client-side caching. Turn it “on” when client-side caching is
desirable and reliable. Turn it “off” when client-side caching is redundant,
unreliable, or counterproductive.
Oplocks is by default set to “on” by Samba on all configured shares, so
careful attention should be given to each case to determine if the potential
324 File and Record Locking Chapter 16
If a client session has been caching writes and reads locally due to oplocks,
it is likely that the data will be lost when the application restarts or recovers
from the TCP interrupt. When the TCP connection drops, the client state
is lost. When the file server recovers, an oplock break is not sent to the
client. In this case, the work from the prior session is lost. Observing this
scenario with oplocks disabled and with the client writing data to the file
server real-time, the failover will provide the data on disk as it existed at
the time of the disconnect.
Section 16.2. Discussion 325
Oplocks is most effective when it is confined to shares that are exclusively ac-
cessed by a single user, or by only one user at a time. Because the true value
of oplocks is the local client caching of data, any operation that interrupts
the caching mechanism will cause a delay.
Home directories are the most obvious examples of where the performance
benefit of oplocks can be safely realized.
As each additional user accesses a file in a share with oplocks enabled, the po-
tential for delays and resulting perceived poor performance increases. When
multiple users are accessing a file on a share that has oplocks enabled, the
management impact of sending and receiving oplock breaks and the result-
ing latency while other clients wait for the caching client to flush data offset
the performance gains of the caching user.
As each additional client attempts to access a file with oplocks set, the
potential performance improvement is negated and eventually results in a
performance bottleneck.
Local UNIX and NFS clients access files without a mandatory file-locking
mechanism. Thus, these client platforms are incapable of initiating an oplock
break request from the server to a Windows client that has a file cached.
Local UNIX or NFS file access can therefore write to a file that has been
cached by a Windows client, which exposes the file to likely data corruption.
If files are shared between Windows clients and either local UNIX or NFS
users, turn oplocks off.
326 File and Record Locking Chapter 16
Multiuser databases clearly pose a risk due to their very nature — they are
typically heavily accessed by numerous users at random intervals. Placing
a multiuser database on a share with oplocks enabled will likely result in a
locking management bottleneck on the Samba server. Whether the database
application is developed in-house or a commercially available product, ensure
that the share has oplocks disabled.
Samba includes an smb.conf parameter called force user that changes the
user accessing a share from the incoming user to whatever user is defined by
the smb.conf variable. If oplocks is enabled on a share, the change in user
access causes an oplock break to be sent to the client, even if the user has
not explicitly loaded a file. In cases where the network is slow or unreliable,
an oplock break can become lost without the user even accessing a file.
This can cause apparent performance degradation as the client continually
reconnects to overcome the lost oplock break.
Avoid the combination of the following:
• force user in the smb.conf share configuration.
• Slow or unreliable networks.
• Oplocks enabled.
Oplocks is a unique Windows file locking feature. It is not really file locking,
but is included in most discussions of Windows file locking, so is considered
a de facto locking feature. Oplocks is actually part of the Windows client file
caching mechanism. It is not a particularly robust or reliable feature when
implemented on the variety of customized networks that exist in enterprise
computing.
Like Windows, Samba implements oplocks as a server-side component of
the client caching mechanism. Because of the lightweight nature of the
Windows feature design, effective configuration of oplocks requires a good
understanding of its limitations, and then applying that understanding when
configuring data access for each particular customized network and client
usage state.
Oplocks essentially means that the client is allowed to download and cache a
file on its hard drive while making changes; if a second client wants to access
the file, the first client receives a break and must synchronize the file back to
the server. This can give significant performance gains in some cases; some
programs insist on synchronizing the contents of the entire file back to the
server for a single change.
Section 16.3. Samba Oplocks Control 329
Level1 Oplocks (also known as just plain “oplocks”) is another term for
opportunistic locking.
Level2 Oplocks provides opportunistic locking for a file that will be treated
as read only. Typically this is used on files that are read-only or on files that
the client has no initial intention to write to at time of opening the file.
Kernel Oplocks are essentially a method that allows the Linux kernel to
co-exist with Samba’s oplocked files, although this has provided better in-
tegration of MS Windows network file locking with the underlying OS. SGI
IRIX and Linux are the only two OSs that are oplock-aware at this time.
Unless your system supports kernel oplocks, you should disable oplocks if
you are accessing the same files from both UNIX/Linux and SMB clients.
Regardless, oplocks should always be disabled if you are sharing a database
file (e.g., Microsoft Access) between multiple clients, because any break the
first client receives will affect synchronization of the entire file (not just
the single record), which will result in a noticeable performance impairment
and, more likely, problems accessing the database in the first place. Notably,
Microsoft Outlook’s personal folders (*.pst) react quite badly to oplocks. If
in doubt, disable oplocks and tune your system from that point.
If client-side caching is desirable and reliable on your network, you will
benefit from turning on oplocks. If your network is slow and/or unreliable,
or you are sharing your files among other file sharing mechanisms (e.g.,
NFS) or across a WAN, or multiple people will be accessing the same files
frequently, you probably will not benefit from the overhead of your client
sending oplock breaks and will instead want to disable oplocks for the share.
Another factor to consider is the perceived performance of file access. If
oplocks provide no measurable speed benefit on your network, it might not
be worth the hassle of dealing with them.
[ acctdata ]
oplocks = False
l e v e l 2 oplocks = False
The default oplock type is Level1. Level2 oplocks are enabled on a per-share
basis in the smb.conf file.
Alternately, you could disable oplocks on a per-file basis within the share:
v e t o o p l o c k f i l e s = / ∗ .mdb/ ∗ .MDB/ ∗ . dbf / ∗ . ←-
DBF/
Kernel oplocks is an smb.conf parameter that notifies Samba (if the UNIX
kernel has the capability to send a Windows client an oplock break) when a
UNIX process is attempting to open the file that is cached. This parameter
addresses sharing files between UNIX and Windows with oplocks enabled
on the Samba server: the UNIX process can open the file that is Oplocked
(cached) by the Windows client and the smbd process will not send an
oplock break, which exposes the file to the risk of data corruption. If the
UNIX kernel has the ability to send an oplock break, then the kernel oplocks
parameter enables Samba to send the oplock break. Kernel oplocks are
enabled on a per-server basis in the smb.conf file.
kernel oplocks = yes
benefit of file caching without the risk of data corruption. Veto oplocks can
be enabled on a per-share basis, or globally for the entire server, in the smb.
conf file as shown in Example 16.3.1.
oplock break wait time is an smb.conf parameter that adjusts the time inter-
val for Samba to reply to an oplock break request. Samba recommends: “Do
not change this parameter unless you have read and understood the Samba
oplock code.” Oplock break wait time can only be configured globally in the
smb.conf file as shown:
o p l o c k break w a i t time = 0 ( default )
Oplock break contention limit is an smb.conf parameter that limits the re-
sponse of the Samba server to grant an oplock if the configured number
of contending clients reaches the limit specified by the parameter. Samba
recommends “Do not change this parameter unless you have read and un-
derstood the Samba oplock code.” Oplock break contention limit can be
enabled on a per-share basis, or globally for the entire server, in the smb.
conf file as shown in Example 16.3.2.
All Windows operating systems in the NT family that act as database servers
for data files (meaning that data files are stored there and accessed by other
Windows PCs) may need to have oplocks disabled in order to minimize the
risk of data file corruption. This includes Windows 9x/Me, Windows NT,
Windows 200x, and Windows XP. 1
The major difference is the location in the Windows registry where the values
for disabling oplocks are entered. Instead of the LanManServer location, the
LanManWorkstation location may be used.
You can verify (change or add, if necessary) this registry value using the
Windows Registry Editor. When you change this registry value, you will
have to reboot the PC to ensure that the new setting goes into effect.
The location of the client registry entry for oplocks has changed in Windows
2000 from the earlier location in Microsoft Windows NT.
1
Microsoft has documented this in Knowledge Base article 300216.
Section 16.4. MS Windows Oplocks and Caching Controls 333
Note
You can also deny the granting of oplocks by changing the following registry
entries:
HKEY_LOCAL_MACHINE\System\
CurrentControlSet\Services\MRXSmb\Parameters\
OplocksDisabled REG_DWORD 0 or 1
Default: 0 (not disabled)
Note
HKEY_LOCAL_MACHINE\System\
CurrentControlSet\Services\LanmanServer\Parameters
EnableOplocks REG_DWORD 0 or 1
Default: 1 (Enabled by Default)
EnableOpLockForceClose REG_DWORD 0 or 1
Default: 0 (Disabled by Default)
334 File and Record Locking Chapter 16
Note
• Since no other station has the file open, the server grants station 1
exclusive oplock.
• Since station 1 has not yet written to the file, the server asks station
1 to break to Level2 oplock.
• Station 1 informs the server that it has broken to level2 Oplock (al-
ternately, station 1 could have closed the file).
• Station 2 (or any station that has the file open) sends a write request
SMB. The server returns the write response.
• The server asks all stations that have the file open to break to none,
meaning no station holds any oplock on the file. Because the worksta-
tions can have no cached writes or locks at this point, they need not
respond to the break-to-none advisory; all they need do is invalidate
locally cashed read-ahead data.
Section 16.4. MS Windows Oplocks and Caching Controls 335
\HKEY_LOCAL_MACHINE\System\
CurrentControlSet\Services\LanmanWorkstation\Parameters
UseOpportunisticLocking REG_DWORD 0 or 1
Default: 1 (true)
This indicates whether the redirector should use oplocks performance en-
hancement. This parameter should be disabled only to isolate problems.
\HKEY_LOCAL_MACHINE\System\
CurrentControlSet\Services\LanmanServer\Parameters
EnableOplocks REG_DWORD 0 or 1
Default: 1 (true)
This specifies whether the server allows clients to use oplocks on files. Oplocks
are a significant performance enhancement, but have the potential to cause
lost cached data on some networks, particularly WANs.
This specifies the minimum link throughput allowed by the server before it
disables raw I/O and oplocks for this connection.
This specifies the maximum time allowed for a link delay. If delays exceed
this number, the server disables raw I/O and oplocks for this connection.
336 File and Record Locking Chapter 16
This specifies the time that the server waits for a client to respond to an
oplock break request. Smaller values can allow detection of crashed clients
more quickly but can potentially cause loss of cached data.
If you have applied all of the settings discussed in this chapter but data
corruption problems and other symptoms persist, here are some additional
things to check out.
We have credible reports from developers that faulty network hardware,
such as a single faulty network card, can cause symptoms similar to read
caching and data corruption. If you see persistent data corruption even after
repeated re-indexing, you may have to rebuild the data files in question. This
involves creating a new data file with the same definition as the file to be
rebuilt and transferring the data from the old file to the new one. There are
several known methods for doing this that can be found in our knowledge
base.
server and the client. Disabling of all forms of file caching on the MS
Windows client may be necessary also.
• Defective network cards, cables, or hubs/switches. This is generally
a more prevalent factor with low-cost networking hardware, although
occasionally there have also been problems with incompatibilities in
more up-market hardware.
• There have been some random reports of Samba log files being written
over data files. This has been reported by very few sites (about five
in the past 3 years) and all attempts to reproduce the problem have
failed. The Samba Team has been unable to catch this happening and
thus unable to isolate any particular cause. Considering the millions
of systems that use Samba, for the sites that have been affected by
this as well as for the Samba Team, this is a frustrating and vexing
challenge. If you see this type of thing happening, please create a bug
report on Samba Bugzilla2 without delay. Make sure that you give as
much information as you possibly can to help isolate the cause and to
allow replication of the problem (an essential step in problem isolation
and correction).
“It sometimes takes approximately 35 seconds to delete files over the network
after XP SP1 has been applied.” This is a bug in Windows XP. More
information can be found in Microsoft Knowledge Base article 8114924 .
You may want to check for an updated documentation regarding file and
record locking issues on the Microsoft Support5 web site. Additionally, a
search for the work locking on the Samba web6 site.
Section of the Microsoft MSDN Library on opportunistic locking:
Microsoft Knowledge Base, “Maintaining Transactional Integrity with OPLOCKS”,
Microsoft Corporation, April 1999, Microsoft KB Article 2249927 .
Microsoft Knowledge Base, “Configuring Opportunistic Locking in Windows
2000”, Microsoft Corporation, April 2001 Microsoft KB Article 2962648 .
Microsoft Knowledge Base, “PC Ext: Explanation of Opportunistic Locking
on Windows NT”, Microsoft Corporation, April 1995 Microsoft KB Article
1292029 .
4
<http://support.microsoft.com/?id=811492>
5
<http://support.microsoft.com/>
6
<http://www.samba.org/>
7
<http://support.microsoft.com/?id=224992>
8
<http://support.microsoft.com/?id=296264>
9
<http://support.microsoft.com/?id=129202>
Chapter 17
SECURING SAMBA
17.1 Introduction
There are three levels at which security principles must be observed in order
to render a site at least moderately secure. They are the perimeter firewall,
the configuration of the host server that is running Samba, and Samba itself.
339
340 Securing Samba Chapter 17
The key challenge of security is that protective measures suffice at best only
to close the door on known exploits and breach techniques. Never assume
that because you have followed these few measures, the Samba server is now
an impenetrable fortress! Given the history of information systems so far, it
is only a matter of time before someone will find yet another vulnerability.
In many installations of Samba, the greatest threat comes from outside your
immediate network. By default, Samba accepts connections from any host,
which means that if you run an insecure version of Samba on a host that is
directly connected to the Internet, you can be especially vulnerable.
One of the simplest fixes in this case is to use the hosts allow and hosts deny
options in the Samba smb.conf configuration file to allow access to your
server only from a specific range of hosts. An example might be:
hosts allow = 1 27 . 0. 0. 1 192.168.2.0/24 ←-
192.168.3.0/24
h o s t s deny = 0 . 0 . 0 . 0 / 0
Section 17.3. Technical Discussion of Protective Measures and Issues 341
The above will allow SMB connections only from localhost (your own
computer) and from the two private networks 192.168.2 and 192.168.3. All
other connections will be refused as soon as the client sends its first packet.
The refusal will be marked as not listening on called name error.
If you want to restrict access to your server to valid users only, then the
following method may be of use. In the smb.conf [global] section put:
v a l i d u s e r s = @smbusers , j a c k o
This restricts all server access either to the user jacko or to members of the
system group smbusers.
This tells Samba to listen for connections only on interfaces with a name
starting with eth such as eth0 or eth1, plus on the loopback interface
called lo. The name you will need to use depends on what OS you are
using. In the above, I used the common name for Ethernet adapters on
Linux.
If you use the above and someone tries to make an SMB connection to your
host over a PPP interface called ppp0, then [s]he will get a TCP connection
refused reply. In that case, no Samba code is run at all, because the operating
system has been told not to pass connections from that interface to any
Samba process. However, the refusal helps a would-be cracker by confirming
that the IP address provides valid active services.
342 Securing Samba Chapter 17
Many people use a firewall to deny access to services they do not want ex-
posed outside their network. This can be a good idea, although I recommend
using it in conjunction with the above methods so you are protected even if
your firewall is not active for some reason.
If you are setting up a firewall, you need to know what TCP and UDP ports
to allow and block. Samba uses the following:
Port 135/TCP - used by smbd
Port 137/UDP - used by nmbd
Port 138/UDP - used by nmbd
Port 139/TCP - used by smbd
Port 445/TCP - used by smbd
The last one is important because many older firewall setups may not be
aware of it, given that this port was only added to the protocol in recent
years.
When configuring a firewall, the high order ports (1024-65535) are often
used for outgoing connections and therefore should be permitted through
the firewall. It is prudent to block incoming packets on the high order ports
except for established connections.
If the above methods are not suitable, then you could also place a more spe-
cific deny on the IPC$ share that is used in the recently discovered security
hole. This allows you to offer access to other shares while denying access to
IPC$ from potentially untrustworthy hosts.
To do this you could use:
[ IPC$ ]
Section 17.3. Technical Discussion of Protective Measures and Issues 343
This instructs Samba that IPC$ connections are not allowed from anywhere
except the two listed network addresses (localhost and the 192.168.115 sub-
net). Connections to other shares are still allowed. Because the IPC$ share
is the only share that is always accessible anonymously, this provides some
level of protection against attackers who do not know a valid username/-
password for your host.
If you use this method, then clients will be given an ‘access denied’ reply
when they try to access the IPC$ share. Those clients will not be able to
browse shares and may also be unable to access some other resources. This
is not recommended unless for some reason you cannot use one of the other
methods just discussed.
[HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\Lsa]
"lmcompatibilitylevel"=dword:00000003
The value 0x00000003 means to send NTLMv2 response only. Clients will
use NTLMv2 authentication; use NTLMv2 session security if the server
supports it. Domain controllers accept LM, NTLM, and NTLMv2 authen-
tication.
[HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\Lsa\MSV1_0]
"NtlmMinClientSec"=dword:00080000
The value 0x00080000 means permit only NTLMv2 session security. If either
NtlmMinClientSec or NtlmMinServerSec is set to 0x00080000, the connec-
tion will fail if NTLMv2 session security is negotiated.
344 Securing Samba Chapter 17
“ We are unable to keep individual users from mapping to any other user’s
home directory once they have supplied a valid password! They only need to
enter their own password. I have not found any method to configure Samba
so that users may map only their own home directory.”
“User xyzzy can map his home directory. Once mapped, user xyzzy can also
map anyone else’s home directory.”
This is not a security flaw, it is by design. Samba allows users to have exactly
the same access to the UNIX file system as when they were logged onto the
UNIX box, except that it only allows such views onto the file system as are
allowed by the defined shares.
Section 17.5. Common Errors 345
If your UNIX home directories are set up so that one user can happily cd
into another user’s directory and execute ls, the UNIX security solution is
to change file permissions on the user’s home directories so that the cd and
ls are denied.
Samba tries very hard not to second guess the UNIX administrator’s security
policies and trusts the UNIX admin to set the policies and permissions he
or she desires.
Samba allows the behavior you require. Simply put the only user = %S
option in the [homes] share definition.
The only user works in conjunction with the users = list, so to get the
behavior you require, add the line:
u s e r s = %S
INTERDOMAIN TRUST
RELATIONSHIPS
The range of values specified must not overlap values used by the host
operating system and must not overlap values used in the passdb backend
for POSIX user accounts. The maximum value is limited by the upper-
most value permitted by the host operating system. This is a UNIX kernel
limited parameter. Linux kernel 2.6 based systems support a maximum
value of 4294967295 (32-bit unsigned variable).
346
Section 18.1. Features and Benefits 347
Note
access rights and privileges in another domain. The language that describes
this capability is couched in terms of trusts. Specifically, one domain will
trust the users from another domain. The domain from which users can
access another security domain is said to be a trusted domain. The domain
in which those users have assigned rights and privileges is the trusting do-
main. With NT3.x/4.0 all trust relationships are always in one direction
only, so if users in both domains are to have privileges and rights in each
others’ domain, then it is necessary to establish two relationships, one in
each direction.
Further, in an NT4-style MS security domain, all trusts are nontransitive.
This means that if there are three domains (let’s call them red, white, and
blue), where red and white have a trust relationship, and white and blue
have a trust relationship, then it holds that there is no implied trust between
the red and blue domains. Relationships are explicit and not transitive.
New to MS Windows 2000 ADS security contexts is the fact that trust
relationships are two-way by default. Also, all inter-ADS domain trusts are
transitive. In the case of the red, white, and blue domains, with Windows
2000 and ADS, the red and blue domains can trust each other. This is an
inherent feature of ADS domains. Samba-3 implements MS Windows NT4-
style interdomain trusts and interoperates with MS Windows 200x ADS
security domains in similar manner to MS Windows NT4-style domains.
For MS Windows NT4, all domain trust relationships are configured using
the Domain User Manager. This is done from the Domain User Manager
Policies entry on the menu bar. From the Policy menu, select Trust Rela-
tionships. Next to the lower box labeled Permitted to Trust this Domain
are two buttons, Add and Remove. The Add button will open a panel in
which to enter the name of the remote domain that will be able to assign
Section 18.3. Native MS Windows NT4 Trusts Configuration 349
access rights to users in your domain. You will also need to enter a pass-
word for this trust relationship, which the trusting domain will use when
authenticating users from the trusted domain. The password needs to be
typed twice (for standard confirmation).
A trust relationship will work only when the other (trusting) domain makes
the appropriate connections with the trusted domain. To consummate the
trust relationship, the administrator launches the Domain User Manager
from the menu selects Policies, then select Trust Relationships, and clicks
on the Add button next to the box that is labeled Trusted Domains. A
panel opens in which must be entered the name of the remote domain as
well as the password assigned to that trust.
A two-way trust relationship is created when two one-way trusts are created,
one in each direction. Where a one-way trust has been established between
two MS Windows NT4 domains (let’s call them DomA and DomB), the
following facilities are created:
Domain A
Domain B
Trusts
In order to set the Samba PDC to be the trusted party of the relationship,
you first need to create a special account for the domain that will be the
trusting party. To do that, you can use the smbpasswd utility. Creating
the trusted domain account is similar to creating a trusted machine account.
Suppose, your domain is called SAMBA, and the remote domain is called
RUMBA. The first step will be to issue this command from your favorite
shell:
where -a means to add a new account into the passdb database and -i
means to “create this account with the Interdomain trust flag”.
The account name will be “rumba$” (the name of the remote domain). If
this fails, you should check that the trust account has been added to the
system password database (/etc/passwd). If it has not been added, you
can add it manually and then repeat the previous step.
After issuing this command, you will be asked to enter the password for the
account. You can use any password you want, but be aware that Windows
NT will not change this password until 7 days following account creation.
After the command returns successfully, you can look at the entry for the
new account (in the standard way as appropriate for your configuration)
and see that the account’s name is really RUMBA$ and it has the “I” flag
set in the flags field. Now you are ready to confirm the trust by establishing
it from Windows NT Server.
Open User Manager for Domains and from the Policies menu, select Trust
Relationships.... Beside the Trusted domains list box, click the Add... but-
ton. You will be prompted for the trusted domain name and the relationship
password. Type in SAMBA, as this is the name of the remote domain and
the password used at the time of account creation. Click on OK and, if
everything went without incident, you will see the Trusted domain rela-
tionship successfully established message.
352 Interdomain Trust Relationships Chapter 18
This time activities are somewhat reversed. Again, we’ll assume that your
domain controlled by the Samba PDC is called SAMBA and the NT-controlled
domain is called RUMBA.
The very first step is to add an account for the SAMBA domain on RUMBA’s
PDC.
Launch the Domain User Manager, then from the menu select Policies,
Trust Relationships. Now, next to the Trusted Domains box, press the
Add button and type in the name of the trusted domain (SAMBA) and the
password to use in securing the relationship.
Using your favorite shell while logged in as root, issue this command:
You will be prompted for the password you just typed on your Windows
NT4 Server box. An error message, "NT STATUS NOLOGON INTERDOMAIN
TRUST ACCOUNT," that may be reported periodically is of no concern and
may safely be ignored. It means the password you gave is correct and the
NT4 server says the account is ready for interdomain connection and not for
ordinary connection. After that, be patient; it can take a while (especially
in large networks), but eventually you should see the Success message.
Congratulations! Your trust relationship has just been established.
Note
The system detected a possible attempt to compromise security. Please ensure that
you can contact the server that authenticated you.
354 Interdomain Trust Relationships Chapter 18
The event logs on the box I’m trying to connect to have entries regarding
group policy not being applied because it is a member of a down-level domain.
If there is a computer account in the Windows 200x domain for the machine
in question, and it is disabled, this problem can occur. If there is no com-
puter account (removed or never existed), or if that account is still intact
(i.e., you just joined it to another domain), everything seems to be fine. By
default, when you unjoin a domain (the Windows 200x domain), the com-
puter tries to automatically disable the computer account in the domain. If
you are running as an account that has privileges to do this when you unjoin
the machine, it is done; otherwise it is not done.
dn: uid=domain_name$,ou=People,dc={your-domain},dc={your-top-level-domain}
changetype: modify
sambaAcctFlags: [I ]
-D "cn=Manager,dc={your-domain},dc={your-top-level-domain}" \
-W -f /path-to/foobar
Create a single-sided trust under the NT4 Domain User Manager, then ex-
ecute:
It works with Samba-3 and NT4 domains, and also with Samba-3 and Win-
dows 200x ADS in mixed mode. Both domain controllers, Samba and NT
must have the same WINS server; otherwise, the trust will never work.
Chapter 19
HOSTING A MICROSOFT
DISTRIBUTED FILE SYSTEM
TREE
The distributed file system (DFS) provides a means of separating the logical
view of files and directories that users see from the actual physical locations
of these resources on the network. It allows for higher availability, smoother
storage expansion, load balancing, and so on.
For information about DFS, refer to the Microsoft documentation1 . This
document explains how to host a DFS tree on a UNIX machine (for DFS-
aware clients to browse) using Samba.
A Samba server can be made a DFS server by setting the global Boolean host
msdfs parameter in the smb.conf file. You designate a share as a DFS root
using the share-level Boolean msdfs root parameter. A DFS root directory
on Samba hosts DFS links in the form of symbolic links that point to other
servers. For example, a symbolic link junction->msdfs:storage1\share1
in the share directory acts as the DFS junction. When DFS-aware clients at-
tempt to access the junction link, they are redirected to the storage location
(in this case, \\storage1\share1).
DFS trees on Samba work with all DFS-aware clients ranging from Windows
95 to 200x. Example 19.1.1 shows how to setup a DFS tree on a Samba
1
<http://www.microsoft.com/NTServer/nts/downloads/winfeatures/
NTSDistrFile/AdminGuide.asp>
356
Section 19.2. Common Errors 357
root# cd /export/dfsroot
root# chown root /export/dfsroot
root# chmod 755 /export/dfsroot
root# ln -s msdfs:storageA\\shareA linka
root# ln -s msdfs:serverB\\share,serverC\\share linkb
You should set up the permissions and ownership of the directory acting as
the DFS root so that only designated users can create, delete, or modify the
msdfs links. Also note that symlink names should be all lowercase. This
limitation exists to have Samba avoid trying all the case combinations to get
at the link name. Finally, set up the symbolic links to point to the network
shares you want and start Samba.
Users on DFS-aware clients can now browse the DFS tree on the Samba
server at \\samba\dfs. Accessing links linka or linkb (which appear as
directories to the client) takes users directly to the appropriate shares on
the network.
• For security purposes, the directory acting as the root of the DFS tree
should have ownership and permissions set so only designated users
can modify the symbolic links in the directory.
A network administrator sent advice to the Samba mailing list after long
sessions trying to determine why DFS was not working. His advice is worth
noting.
“I spent some time trying to figure out why my particular DFS root wasn’t
working. I noted in the documentation that the symlink should be in all
lowercase. It should be amended that the entire path to the symlink should
all be in lowercase as well.”
“For example, I had a share defined as such:”
[ pub ]
path = / e x p o r t /home/ S h a r e s / p u b l i c s h a r e
msdfs r o o t = y e s
“and I could not make my Windows 9x/Me (with the dfs client installed)
follow this symlink:”
CLASSICAL PRINTING
SUPPORT
Printing is often a mission-critical service for the users. Samba can pro-
vide this service reliably and seamlessly for a client network consisting of
Windows workstations.
A Samba print service may be run on a standalone or domain member server,
side by side with file serving functions, or on a dedicated print server. It can
be made as tightly or as loosely secured as needs dictate. Configurations
may be simple or complex. Available authentication schemes are essentially
the same as described for file services in previous chapters. Overall, Samba’s
printing support is now able to replace an NT or Windows 2000 print server
full-square, with additional benefits in many cases. Clients may down-
load and install drivers and printers through their familiar Point’n’Print
mechanism. Printer installations executed by Logon Scripts are no prob-
lem. Administrators can upload and manage drivers to be used by clients
through the familiar Add Printer Wizard. As an additional benefit, driver
and printer management may be run from the command line or through
scripts, making it more efficient in case of large numbers of printers. If a
central accounting of print jobs (tracking every single page and supplying
the raw data for all sorts of statistical reports) is required, this function is
best supported by the newer Common UNIX Printing System (CUPS) as
the print subsystem underneath the Samba hood.
This chapter outlines the fundamentals of Samba printing as implemented
by the more traditional UNIX BSD- and System V-style printing systems.
359
360 Classical Printing Support Chapter 20
Much of the information in this chapter applies also to CUPS. If you use
CUPS, you may be tempted to jump to the next chapter, but you will
certainly miss a few things if you do. For further information refer to Chap-
ter 21, “CUPS Printing Support”.
Note
Important
It is apparent from postings on the Samba mailing list that print configura-
tion is one of the most problematic aspects of Samba administration today.
Many new Samba administrators have the impression that Samba performs
some sort of print processing. Rest assured, Samba does not perform any
type of print processing. It does not do any form of print filtering.
Samba obtains from its clients a data stream (print job) that it spools to
a local spool area. When the entire print job has been received, Samba
invokes a local UNIX/Linux print command and passes the spooled file to
it. It is up to the local system printing subsystems to correctly process the
print job and to submit it to the printer.
The syntax for the configuration file is easy to grasp. You should know that
is not very picky about its syntax. As has been explained elsewhere in this
book, Samba tolerates some spelling errors (such as browseable instead of
browsable), and spelling is case-insensitive. It is permissible to use Yes/No
or True/False for Boolean settings. Lists of names may be separated by
commas, spaces, or tabs.
To see all (or at least most) printing-related settings in Samba, including the
implicitly used ones, try the command outlined below. This command greps
for all occurrences of lp, print, spool, driver, ports, and [ in testparm’s
output. This provides a convenient overview of the running smbd print
configuration. This command does not show individually created printer
shares or the spooling paths they may use. Here is the output of my Samba
setup, with settings shown in Example 20.3.1:
[global]
smb ports = 139 445
lpq cache time = 10
load printers = Yes
printcap name = /etc/printcap
disable spoolss = No
enumports command =
addprinter command =
deleteprinter command =
show add printer wizard = Yes
os2 driver map =
printer admin =
min print space = 0
max print jobs = 1000
printable = No
printing = bsd
print command = lpr -r -P’%p’ %s
364 Classical Printing Support Chapter 20
[homes]
[printers]
path = /var/spool/samba
printable = Yes
You can easily verify which settings were implicitly added by Samba’s de-
fault behavior. Remember: it may be important in your future dealings with
Samba.
Note
Should you need to troubleshoot at any stage, please always come back to
this point first and verify if testparm shows the parameters you expect. To
give you a warning from personal experience, try to just comment out the
load printers parameter. If your 2.2.x system behaves like mine, you’ll see
this:
I assumed that commenting out of this setting should prevent Samba from
publishing my printers, but it still did. It took some time to figure out the
reason. But I am no longer fooled ... at least not by this.
Only when the parameter is explicitly set to load printers = No would Samba
conform with my intentions. So, my strong advice is:
• Never rely on commented-out parameters.
• Always set parameters explicitly as you intend them to behave.
• Use testparm to uncover hidden settings that might not reflect your
intentions.
The following is the most minimal configuration file:
This example should show that you can use testparm to test any Samba
configuration file. Actually, we encourage you not to change your working
system (unless you know exactly what you are doing). Don’t rely on the
assumption that changes will only take effect after you restart smbd! This
is not the case. Samba rereads it every 60 seconds and on each new client
connection. You might have to face changes for your production clients that
366 Classical Printing Support Chapter 20
you didn’t intend to apply. You will now note a few more interesting things;
testparm is useful to identify what the Samba print configuration would
be if you used this minimalistic configuration. Here is what you can expect
to find:
[printers]
printable = Yes
will regard the whole of the string after the = sign as the value you want to
define. This is an invalid value that will be ignored, and a default value will
be used in its place.
The [global] section is one of four special sections (along with [homes], [print-
ers], and [print$] ). The [global] contains all parameters that apply to the
server as a whole. It is the place for parameters that have only a global
meaning. It may also contain service-level parameters that define default
Section 20.4. Extended Printing Configuration 369
settings for all other sections and shares. This way you can simplify the
configuration and avoid setting the same value repeatedly. (Within each
individual section or share, you may, however, override these globally set
share settings and specify other values).
Caution
max print jobs = 100 Sets the upper limit to 100 print jobs being active
on the Samba server at any one time. Should a client submit a job
that exceeds this number, a ”no more space available on server” type
of error message will be returned by Samba to the client. A setting of
zero (the default) means there is no limit at all.
lpq cache time = 20 Controls the cache time for the results of the lpq
command. It prevents the lpq command being called too often and
reduces the load on a heavily used print server.
use client driver = no If set to yes, only takes effect for Windows NT/200x/XP
clients (and not for Win 95/98/ME). Its default value is No (or False).
It must not be enabled on print shares (with a yes or true setting) that
have valid drivers installed on the Samba server. For more detailed
explanations, see the smb.conf man page.
Section 20.4. Extended Printing Configuration 371
The printers section is the second special section. If a section with this name
appears in the smb.conf, users are able to connect to any printer specified
in the Samba host’s printcap file, because Samba on startup then creates
a printer share for every printer name it finds in the printcap file. You
could regard this section as a convenient shortcut to share all printers with
minimal configuration. It is also a container for settings that should apply
as default to all printers. (For more details, see the smb.conf man page.)
Settings inside this container must be share-level parameters.
UNIX login. On some systems the guest account might not have the
privilege to be able to print. Test this by logging in as your guest user
using su - guest and run a system print command like:
lpr -P printername /etc/motd
read only = yes Normally (for other types of shares) prevents users from
creating or modifying files in the service’s directory. However, in a
printable service, it is always allowed to write to the directory (if user
privileges allow the connection), but only via print spooling operations.
Normal write operations are not permitted.
If a [my printer name] section appears in the smb.conf file, which in-
cludes the parameter printable = yes Samba will configure it as a printer
share. Windows 9x/Me clients may have problems with connecting or load-
ing printer drivers if the share name has more than eight characters. Do not
name a printer share with a name that may conflict with an existing user or
file share name. On client connection requests, Samba always tries to find
file shares with that name first. If it finds one, it will connect to this and
will not connect to a printer with the same name!
printer admin = kurt The printer admin definition is different for this
explicitly defined printer share from the general [printers] share. It is
not a requirement; we did it to show that it is possible.
browseable = yes This makes the printer browseable so the clients may
conveniently find it when browsing the Network Neighborhood.
reason, you may want to do something completely different than print the
file. An example is a command that just copies the print file to a tempo-
rary location for further investigation when you need to debug printing. If
you craft your own print commands (or even develop print command shell
scripts), make sure you pay attention to the need to remove the files from
the Samba spool directory. Otherwise, your hard disk may soon suffer from
shortage of free space.
You learned earlier that Samba, in most cases, uses its built-in settings for
many parameters if it cannot find an explicitly stated one in its configuration
file. The same is true for the print command. The default print command
varies depending on the printing parameter setting. In the commands listed
in Table 20.1 , you will notice some parameters of the form %X where X is
p, s, J, and so on. These letters stand for printer name, spool file, and job
ID, respectively. They are explained in more detail in Table 20.1 presents
an overview of key printing options but excludes the special case of CUPS,
is discussed in Chapter 21, “CUPS Printing Support”.
After a print job has finished spooling to a service, the print command will
be used by Samba via a system() call to process the spool file. Usually the
command specified will submit the spool file to the host’s printing subsys-
tem. But there is no requirement at all that this must be the case. The print
subsystem may not remove the spool file on its own, so whatever command
you specify, you should ensure that the spool file is deleted after it has been
processed.
There is no difficulty with using your own customized print commands with
the traditional printing systems. However, if you do not wish to roll your
own, you should be well informed about the default built-in commands that
Samba uses for each printing subsystem (see Table 20.1). In all the com-
mands listed in the last paragraphs, you see parameters of the form %X.
These are macros, or shortcuts, used as placeholders for the names of real
objects. At the time of running a command with such a placeholder, Samba
will insert the appropriate value automatically. Print commands can handle
all Samba macro substitutions. In regard to printing, the following ones do
have special relevance:
• %s, %f — the path to the spool file name.
• %p — the appropriate printer name.
• %J — the job name as transmitted by the client.
• %c — the number of printed pages of the spooled job (if known).
• %z — the size of the spooled print job (in bytes).
The print command must contain at least one occurrence of %s or %f. The
%p is optional. If no printer name is supplied, the %p will be silently removed
from the print command. In this case, the job is sent to the default printer.
376 Classical Printing Support Chapter 20
If specified in the [global] section, the print command given will be used for
any printable service that does not have its own print command specified.
If there is neither a specified print command for a printable service nor a
global print command, spool files will be created but not processed! Most
importantly, print files will not be removed, so they will consume disk space.
Printing may fail on some UNIX systems when using the nobody account.
If this happens, create an alternative guest account and give it the privilege
to print. Set up this guest account in the [global] section with the guest
account parameter.
You can form quite complex print commands. You need to realize that print
commands are just passed to a UNIX shell. The shell is able to expand the
included environment variables as usual. (The syntax to include a UNIX
environment variable $variable in the Samba print command is %$vari-
able.) To give you a working print command example, the following will log
a print job to /tmp/print.log, print the file, then remove it. The semicolon
(“;” is the usual separator for commands in shell scripts:
p r i n t command = echo P r i n t i n g %s >> /tmp/ ←-
p r i n t . l o g ; l p r −P %p %s ; rm %s
You may have to vary your own command considerably from this example
depending on how you normally print files on your system. The default
for the print command parameter varies depending on the setting of the
printing parameter. Another example is:
p r i n t command = / u s r / l o c a l /samba/ b i n / ←-
m y p r i n t s c r i p t %p %s
Prior to Samba-2.2.x, print server support for Windows clients was limited
to LanMan printing calls. This is the same protocol level as Windows 9x/Me
PCs offer when they share printers. Beginning with the 2.2.0 release, Samba
started to support the native Windows NT printing mechanisms. These are
implemented via MS-RPC (Remote Procedure Calls). MS-RPCs use the
SPOOLSS named pipe for all printing.
The additional functionality provided by the new SPOOLSS support in-
cludes:
Section 20.5. Printing Developments Since Samba-2.2 377
There is much confusion about what all this means. The question is often
asked, “Is it or is it not necessary for printer drivers to be installed on a
Samba host in order to support printing from Windows clients?” The answer
to this is no, it is not necessary.
1
<http://imprints.sourceforge.net/>
2
<http://msdn.microsoft.com/>
378 Classical Printing Support Chapter 20
Windows NT/2000 clients can, of course, also run their APW to install
drivers locally (which then connect to a Samba-served print queue). This is
the same method used by Windows 9x/Me clients. (However, a bug existed
in Samba 2.2.0 that made Windows NT/2000 clients require that the Samba
server possess a valid driver for the printer. This was fixed in Samba 2.2.1).
But it is a new capability to install the printer drivers into the [print$] share
of the Samba server, and a big convenience, too. Then all clients (including
95/98/ME) get the driver installed when they first connect to this printer
share. The uploading or depositing of the driver into this [print$] share and
the following binding of this driver to an existing Samba printer share can
be achieved by different means:
• Running the APW on an NT/200x/XP Professional client (this does
not work from 95/98/ME clients).
• Using the Imprints toolset.
• Using the smbclient and rpcclient command-line tools.
• Using cupsaddsmb (only works for the CUPS printing system, not for
LPR/LPD, LPRng, and so on).
Samba does not use these uploaded drivers in any way to process spooled
files. These drivers are utilized entirely by the clients who download and
install them via the “Point’n’Print” mechanism supported by Samba. The
clients use these drivers to generate print files in the format the printer
(or the UNIX print system) requires. Print files received by Samba are
handed over to the UNIX printing system, which is responsible for all further
processing, as needed.
guest ok = yes Gives read-only access to this share for all guest users.
Access may be granted to download and install printer drivers on
clients. The requirement for guest ok = yes depends on how your
site is configured. If users will be guaranteed to have an account on
the Samba host, then this is a non-issue.
Section 20.5. Printing Developments Since Samba-2.2 381
Note
read only = yes Because we do not want everybody to upload driver files
(or even change driver settings), we tagged this share as not writable.
write list = @ntadmin, root The [print$] was made read-only by the
previous setting so we should create a write list entry also. UNIX
groups are denoted with a leading “@” character. Users listed here are
allowed write-access (as an exception to the general public’s read-only
access), which they need to update files on the share. Normally, you
will want to name only administrative-level user account in this setting.
Check the file system permissions to make sure these accounts can copy
files to the share. If this is a non-root account, then the account should
also be mentioned in the global printer admin parameter. See the smb.
conf man page for more information on configuring file shares.
itself, the subdirectories must be exactly the names listed below (you may
leave out the subdirectories of architectures you do not need to support).
Therefore, create a directory tree below the [print$] share for each architec-
ture you wish to support like this:
[print$]--+
|--W32X86 # serves drivers to Windows NT x86
|--WIN40 # serves drivers to Windows 95/98
|--W32ALPHA # serves drivers to Windows NT Alpha_AXP
|--W32MIPS # serves drivers to Windows NT R4000
|--W32PPC # serves drivers to Windows NT PowerPC
Required Permissions
Once you have created the required [print$] service and associated sub-
directories, go to a Windows NT 4.0/200x/XP client workstation. Open
Network Neighborhood or My Network Places and browse for the Samba
host. Once you have located the server, navigate to its Printers and Faxes
folder. You should see an initial listing of printers that matches the printer
shares defined on your Samba host.
Section 20.6. Installing Drivers into [print$] 383
Have you successfully created the [print$] share in smb.conf, and have you
forced Samba to reread its smb.conf file? Good. But you are not yet ready
to use the new facility. The client driver files need to be installed into this
share. So far, it is still an empty share. Unfortunately, it is not enough
to just copy the driver files over. They need to be correctly installed so
that appropriate records for each driver will exist in the Samba internal
databases so it can provide the correct drivers as they are requested from
MS Windows clients. And that is a bit tricky, to say the least. We now
discuss two alternative ways to install the drivers into [print$] :
• Using the Samba command-line utility rpcclient with its various sub-
commands (here, adddriver and setdriver) from any UNIX work-
station.
• Running a GUI (Printer Properties and Add Printer Wizard) from
any Windows NT/200x/XP client workstation.
The latter option is probably the easier one (even if the process may seem
a little bit weird at first).
The printers initially listed in the Samba host’s Printers folder accessed
from a client’s Explorer will have no real printer driver assigned to them.
By default this driver name is set to a null string. This must be changed
now. The local Add Printer Wizard (APW), run from NT/2000/XP clients,
will help us in this task.
Installation of a valid printer driver is not straightforward. You must at-
tempt to view the printer properties for the printer to which you want the
driver assigned. Open Windows Explorer, open Network Neighborhood,
browse to the Samba host, open Samba’s Printers folder, right-click on the
printer icon, and select Properties.... You are now trying to view printer
and driver properties for a queue that has this default NULL driver assigned.
This will result in the following error message: “Device settings cannot be
displayed. The driver for the specified printer is not installed, only spooler
properties will be displayed. Do you want to install the driver now?”
Do not click on Yes! Instead, click on No in the error dialog. Now you will
be presented with the printer properties window. From here, the way to
384 Classical Printing Support Chapter 20
assign a driver to a printer is open. You now have the choice of:
• Select a driver from the pop-up list of installed drivers. Initially this
list will be empty.
• Click on New Driver to install a new printer driver (which will start
up the APW).
Once the APW is started, the procedure is exactly the same as the one
you are familiar with in Windows (we assume here that you are familiar
with the printer driver installations procedure on Windows NT). Make sure
your connection is, in fact, set up as a user with printer admin privileges
(if in doubt, use smbstatus to check for this). If you wish to install printer
drivers for client operating systems other than Windows NT x86, you will
need to use the Sharing tab of the printer properties dialog.
Assuming you have connected with an administrative (or root) account (as
named by the printer admin parameter), you will also be able to modify
other printer properties such as ACLs and default device settings using this
dialog. For the default device settings, please consider the advice given
further in Section 20.6.2.
The second way to install printer drivers into [print$] and set them up in
a valid way is to do it from the UNIX command line. This involves four
distinct steps:
1. Gather information about required driver files and collect the files.
2. Deposit the driver files into the [print$] share’s correct subdirectories
(possibly by using smbclient).
We provide detailed hints for each of these steps in the paragraphs that
follow.
Section 20.6. Installing Drivers into [print$] 385
To find out about the driver files, you have two options. You can check the
contents of the driver CDROM that came with your printer. Study the *.
inf files located on the CD-ROM. This may not be possible, since the *.inf
file might be missing. Unfortunately, vendors have now started to use their
own installation programs. These installations packages are often in some
Windows platform archive format. Additionally, the files may be re-named
during the installation process. This makes it extremely difficult to identify
the driver files required.
Then you have the second option. Install the driver locally on a Windows
client and investigate which filenames and paths it uses after they are in-
stalled. (You need to repeat this procedure for every client platform you
want to support. We show it here for the W32X86 platform only, a name
used by Microsoft for all Windows NT/200x/XP clients.)
A good method to recognize the driver files is to print the test page from
the driver’s Properties dialog (General tab). Then look at the list of driver
files named on the printout. You’ll need to recognize what Windows (and
Samba) are calling the Driver File, Data File, Config File, Help File, and
(optionally) Dependent Driver Files (this may vary slightly for Windows
NT). You need to note all filenames for the next steps.
Another method to quickly test the driver filenames and related paths is
provided by the rpcclient utility. Run it with enumdrivers or with the
getdriver subcommand, each at the 3 info level. In the following example,
TURBO XP is the name of the Windows PC (in this case it was a Windows
XP Professional laptop). I installed the driver locally to TURBO XP from
a Samba server called KDE-BITSHOP. We could run an interactive rpcclient
session; then we would get an rpcclient /> prompt and would type the sub-
commands at this prompt. This is left as a good exercise for you. For now,
we use rpcclient with the -c parameter to execute a single subcommand
line and exit again. This is the method you use if you want to create scripts
to automate the procedure for a large number of printers and drivers. Note
the different quotation marks used to overcome the different spaces between
words:
[Windows NT x86]
Printer Driver Info 3:
Version: [2]
Driver Name: [Heidelberg Digimaster 9110 (PS)]
Architecture: [Windows NT x86]
Driver Path: [C:\WINNT\System32\spool\DRIVERS\W32X86\2\HDNIS01_de.DLL]
Datafile: [C:\WINNT\System32\spool\DRIVERS\W32X86\2\Hddm91c1_de.ppd]
Configfile: [C:\WINNT\System32\spool\DRIVERS\W32X86\2\HDNIS01U_de.DLL]
Helpfile: [C:\WINNT\System32\spool\DRIVERS\W32X86\2\HDNIS01U_de.HLP]
Dependentfiles: [C:\WINNT\System32\spool\DRIVERS\W32X86\2\Hddm91c1_de.DLL]
Dependentfiles: [C:\WINNT\System32\spool\DRIVERS\W32X86\2\Hddm91c1_de.INI]
Dependentfiles: [C:\WINNT\System32\spool\DRIVERS\W32X86\2\Hddm91c1_de.dat]
Dependentfiles: [C:\WINNT\System32\spool\DRIVERS\W32X86\2\Hddm91c1_de.cat]
Dependentfiles: [C:\WINNT\System32\spool\DRIVERS\W32X86\2\Hddm91c1_de.def]
Dependentfiles: [C:\WINNT\System32\spool\DRIVERS\W32X86\2\Hddm91c1_de.hre]
Dependentfiles: [C:\WINNT\System32\spool\DRIVERS\W32X86\2\Hddm91c1_de.vnd]
Dependentfiles: [C:\WINNT\System32\spool\DRIVERS\W32X86\2\Hddm91c1_de.hlp]
Dependentfiles: [C:\WINNT\System32\spool\DRIVERS\W32X86\2\HDNIS01Aux.dll]
Dependentfiles: [C:\WINNT\System32\spool\DRIVERS\W32X86\2\HDNIS01_de.NTF]
Monitorname: []
Defaultdatatype: []
You may notice that this driver has quite a large number of Dependent files
(there are worse cases, however). Also, strangely, the Driver File is tagged
here Driver Path. We do not yet have support for the so-called WIN40 archi-
tecture installed. This name is used by Microsoft for the Windows 9x/Me
platforms. If we want to support these, we need to install the Windows
9x/Me driver files in addition to those for W32X86 (i.e., the Windows NT
2000/XP clients) onto a Windows PC. This PC can also host the Windows
9x/Me drivers, even if it runs on Windows NT, 2000, or XP.
Since the [print$] share is usually accessible through the Network Neighbor-
hood, you can also use the UNC notation from Windows Explorer to poke
at it. The Windows 9x/Me driver files will end up in subdirectory 0 of the
WIN40 directory. The full path to access them is \\WINDOWSHOST\print$\WIN40\0\.
Section 20.6. Installing Drivers into [print$] 387
Note
Now we need to collect all the driver files we identified in our previous step.
Where do we get them from? Well, why not retrieve them from the very PC
and the same [print$] share that we investigated in our last step to identify
the files? We can use smbclient to do this. We will use the paths and
names that were leaked to us by getdriver. The listing is edited to include
line breaks for readability:
After this command is complete, the files are in our current local directory.
You probably have noticed that this time we passed several commands to
388 Classical Printing Support Chapter 20
We are now going to locate the driver files into the [print$] share. Remem-
ber, the UNIX path to this share has been defined previously in your smb.
conf file. You also have created subdirectories for the different Windows
client types you want to support. If, for example, your [print$] share maps
to the UNIX path /etc/samba/drivers/, your driver files should now go
here:
• For all Windows NT, 2000, and XP clients, /etc/samba/drivers/
W32X86/ but not (yet) into the 2 subdirectory.
• For all Windows 95, 98, and Me clients, /etc/samba/drivers/WIN40/
but not (yet) into the 0 subdirectory.
We again use smbclient to transfer the driver files across the network. We
specify the same files and paths as were leaked to us by running getdriver
against the original Windows install. However, now we are going to store
the files into a Samba/UNIX print server’s [print$] share.
Whew — that was a lot of typing! Most drivers are a lot smaller — many
have only three generic PostScript driver files plus one PPD. While we did
retrieve the files from the 2 subdirectory of the W32X86 directory from the
Windows box, we do not put them (for now) in this same subdirectory of the
Samba box. This relocation will automatically be done by the adddriver
command, which we will run shortly (and do not forget to also put the files
for the Windows 9x/Me architecture into the WIN40/ subdirectory should
you need them).
For now we verify that our files are there. This can be done with smbclient,
too (but, of course, you can log in via SSH also and do this through a
standard UNIX shell access):
Notice that there are already driver files present in the 2 subdirectory (prob-
ably from a previous installation). Once the files for the new driver are there
too, you are still a few steps away from being able to use them on the clients.
The only thing you could do now is retrieve them from a client just like you
retrieve ordinary files from a file share, by opening print$ in Windows Ex-
Section 20.6. Installing Drivers into [print$] 391
plorer. But that wouldn’t install them per Point’n’Print. The reason is that
Samba does not yet know that these files are something special, namely
printer driver files, and it does not know to which print queue(s) these
driver files belong.
Next, you must tell Samba about the special category of the files you just up-
loaded into the [print$] share. This is done by the adddriver command. It
will prompt Samba to register the driver files into its internal TDB database
files. The following command and its output has been edited for readability:
After this step, the driver should be recognized by Samba on the print server.
You need to be very careful when typing the command. Don’t exchange the
order of the fields. Some changes would lead to an NT STATUS UNSUCCESSFUL
error message. These become obvious. Other changes might install the
driver files successfully but render the driver unworkable. So take care!
Hints about the syntax of the adddriver command are in the man page.
provides a more detailed description, should you need it.
392 Classical Printing Support Chapter 20
One indication for Samba’s recognition of the files as driver files is the suc-
cessfully installed message. Another one is the fact that our files have
been moved by the adddriver command into the 2 subdirectory. You can
check this again with smbclient:
Another verification is that the timestamp of the printing TDB files is now
updated (and possibly their file size has increased).
Now the driver should be registered with Samba. We can easily verify this
and will do so in a moment. However, this driver is not yet associated with
a particular printer. We may check the driver status of the files by at least
three methods:
• From any Windows client browse Network Neighborhood, find the
Samba host, and open the Samba Printers and Faxes folder. Select
any printer icon, right-click and select the printer Properties. Click
the Advanced tab. Here is a field indicating the driver for that printer.
A drop-down menu allows you to change that driver (be careful not to
do this unwittingly). You can use this list to view all drivers known
to Samba. Your new one should be among them. (Each type of client
will see only its own architecture’s list. If you do not have every driver
installed for each platform, the list will differ if you look at it from
Windows95/98/ME or Windows NT/2000/XP.)
• From a Windows 200x/XP client (not Windows NT) browse Network
Neighborhood, search for the Samba server, open the server’s Print-
ers folder, and right-click on the white background (with no printer
highlighted). Select Server Properties. On the Drivers tab you will
see the new driver listed. This view enables you to also inspect the list
of files belonging to that driver (this does not work on Windows NT,
but only on Windows 2000 and Windows XP; Windows NT does not
provide the Drivers tab). An alternative and much quicker method for
Windows 2000/XP to start this dialog is by typing into a DOS box
(you must of course adapt the name to your Samba server instead of
SAMBA-CUPS):
• From a UNIX prompt, run this command (or a variant thereof), where
SAMBA-CUPS is the name of the Samba host and xxxx represents the
actual Samba password assigned to root:
You will see a listing of all drivers Samba knows about. Your new one
should be among them. But it is only listed under the [Windows NT
x86] heading, not under [Windows 4.0], since you didn’t install that
part. Or did you? In our example it is named dm9110. Note that the
third column shows the other installed drivers twice, one time for each
supported architecture. Our new driver only shows up for Windows
NT 4.0 or 2000. To have it present for Windows 95, 98, and Me, you’ll
have to repeat the whole procedure with the WIN40 architecture and
subdirectory.
You can name the driver as you like. If you repeat the adddriver step with
the same files as before but with a different driver name, it will work the
same:
Hddm91c1_de.vnd,Hddm91c1_de.hlp,Hddm91c1KMMin.DLL, \
HDNIS01Aux.dll,HDNIS01_de.NTF,Hddm91c1_de_reg.HLP"
You will be able to bind that driver to any print queue (however, you are
responsible that you associate drivers to queues that make sense with respect
to target printers). You cannot run the rpcclient adddriver command
repeatedly. Each run consumes the files you had put into the [print$] share
by moving them into the respective subdirectories, so you must execute
an smbclient ... put command before each rpcclient ... adddriver
command.
Samba needs to know which printer owns which driver. Create a mapping
of the driver to a printer, and store this information in Samba’s memory,
the TDB files. The rpcclient setdriver command achieves exactly this:
Ah, no, I did not want to do that. Repeat, this time with the name I
intended:
Now we have done most of the work, but not all of it.
Note
As Don Quixote said, “The proof of the pudding is in the eating.” The proof
for our setup lies in the printing. So let’s install the printer driver onto the
client PCs. This is not as straightforward as it may seem. Read on.
Especially important is the installation onto the first client PC (for each
architectural platform separately). Once this is done correctly, all further
clients are easy to set up and shouldn’t need further attention. What follows
is a description for the recommended first procedure. You now work from a
client workstation. You should check that your connection is not unwittingly
mapped to bad user nobody. In a DOS box type:
net use \\SAMBA-SERVER\print$ /user:root
Replace root, if needed, by another valid printer admin user as given in the
definition. Should you already be connected as a different user, you will
get an error message. There is no easy way to get rid of that connection,
because Windows does not seem to know a concept of logging off from a share
connection (do not confuse this with logging off from the local workstation;
that is a different matter). On Windows NT/200x, you can force a logoff
from all smb/cifs connections by restarting the workstation service. You can
try to close all Windows file explorers and Internet Explorer for Windows.
Section 20.7. Client Driver Installation Procedure 397
As a last resort, you may have to reboot. Make sure there is no automatic
reconnection set up. It may be easier to go to a different workstation and
try from there. After you have made sure you are connected as a printer
admin user (you can check this with the smbstatus command on Samba),
do this from the Windows workstation:
Most likely you are tempted to try to print a test page. After all, you now
can open the printer properties, and on the General tab there is a button
offering to do just that. But chances are that you get an error message
saying ”Unable to print Test Page.” The reason might be that there is
not yet a valid device mode set for the driver or that the “printer driver
data” set is still incomplete.
You must make sure that a valid device mode is set for the driver. We now
explain what that means.
• A valid device mode generated by the driver for the printer (defining
things like paper size, orientation and duplex settings).
If either of these is incomplete, the clients can produce less than optimal out-
put at best. In the worst cases, unreadable garbage or nothing at all comes
from the printer, or it produces a harvest of error messages when attempting
398 Classical Printing Support Chapter 20
to print. Samba stores the named values and all printing-related informa-
tion in its internal TDB database files (ntprinters.tdb, ntdrivers.tdb,
printing.tdb, and ntforms.tdb).
The device mode and the set of printer driver data are basically collections of
settings for all print queue properties, initialized in a sensible way. Device
modes and printer driver data should initially be set on the print server
(the Samba host) to healthy values so the clients can start to use them
immediately. How do we set these initial healthy values? This can be
achieved by accessing the drivers remotely from an NT (or 200x/XP) client,
as discussed in the following paragraphs.
Be aware that a valid device mode can only be initiated by a printer admin
or root (the reason should be obvious). Device modes can be correctly set
only by executing the printer driver program itself. Since Samba cannot
execute this Win32 platform driver code, it sets this field initially to NULL
(which is not a valid setting for clients to use). Fortunately, most drivers
automatically generate the printer driver data that is needed when they are
uploaded to the [print$] share with the help of the APW or rpcclient.
The generation and setting of a first valid device mode, however, requires
some tickling from a client to set it on the Samba server. The easiest means
of doing so is to simply change the page orientation on the server’s printer.
This executes enough of the printer driver program on the client for the
desired effect to happen and feeds back the new device mode to our Samba
server. You can use the native Windows NT/200x/XP printer properties
page from a Window client for this: Procedure to Initialize the Printer
Driver Settings
1. Browse the Network Neighborhood.
2. Find the Samba server.
3. Open the Samba server’s Printers and Faxes folder.
4. Highlight the shared printer in question.
5. Right-click on the printer (you may already be here if you followed the
last section’s description).
6. At the bottom of the context menu select Properties (if the menu still
offers the Connect entry further above, you need to click on that one
first to achieve the driver installation, as shown in the last section).
7. Go to the Advanced tab; click on Printing Defaults.
Section 20.7. Client Driver Installation Procedure 399
10. While you are at it, you may also want to set the desired printing
defaults here, which then apply to all future client driver installations.
This procedure executes the printer driver program on the client platform
and feeds back the correct device mode to Samba, which now stores it in
its TDB files. Once the driver is installed on the client, you can follow
the analogous steps by accessing the local Printers folder, too, if you are a
Samba printer admin user. From now on, printing should work as expected.
Every additional driver may be installed in the same way as just described.
Browse Network Neighborhood, open the Printers folder on Samba server,
right-click on Printer, and choose Connect.... Once this completes (should
be not more than a few seconds, but could also take a minute, depend-
ing on network conditions), you should find the new printer in your client
workstation local Printers and Faxes folder.
You can also open your local Printers and Faxes folder by using this com-
mand on Windows 200x/XP Professional workstations:
You can enter the commands either inside a DOS box window or in the Run
command... field from the Start menu.
After you installed the driver on the Samba server (in its [print$] share),
you should always make sure that your first client installation completes
correctly. Make it a habit for yourself to build the very first connection
from a client as printer admin. This is to make sure that:
• A first valid device mode is really initialized (see above Section 20.7.2)
for more explanation details).
• The default print settings of your printer for all further client instal-
lations are as you want them.
You will be prompted for root’s Samba password; type it, wait a few sec-
onds, click on Printing Defaults, and proceed to set the job options that
should be used as defaults by all clients. Alternatively, instead of root you
can name one other member of the printer admin from the setting.
Now all the other users downloading and installing the driver the same way
(using Point’n’Print) will have the same defaults set for them. If you miss
this step, you’ll get a lot of help desk calls from your users, but maybe you
like to talk to people.
Section 20.8. Other Gotchas 401
The last sentence might be viewed with mixed feelings by some users and
Admins. They have struggled for hours and could not arrive at a point
where their settings seemed to be saved. It is not their fault. The confusing
thing is that in the multitabbed dialog that pops up when you right-click on
the printer name and select Properties, you can arrive at two dialogs that
appear identical, each claiming that they help you to set printer options
in three different ways. Here is the definitive answer to the Samba default
driver setting FAQ:
“I can not set and save default print options for all users on Win-
dows 200x/XP. Why not?”. How are you doing it? I bet the wrong way.
(It is not easy to find out, though.) There are three different ways to bring
you to a dialog that seems to set everything. All three dialogs look the same,
but only one of them does what you intend. You need to be Administrator
or Print Administrator to do this for all users. Here is how I reproduce it
in an XP Professional:
Tip
One issue that has arisen during the recent development phase of Samba
is the need to support driver downloads for hundreds of printers. Using
Windows NT APW for this task is somewhat awkward (to say the least). If
you do not want to acquire RSS pains from the printer installation clicking
orgy alone, you need to think about a non-interactive script.
If more than one printer is using the same driver, the rpcclient setdriver
command can be used to set the driver associated with an installed queue.
If the driver is uploaded to [print$] once and registered with the printing
TDBs, it can be used by multiple print queues. In this case, you just need to
repeat the setprinter subcommand of rpcclient for every queue (without
the need to conduct the adddriver repeatedly). The following is an example
of how this can be accomplished:
[Windows NT x86]
Printer Driver Info 1:
404 Classical Printing Support Chapter 20
[....]
It may not be easy to recognize that the first call to enumprinters showed
the “dm9110” printer with an empty string where the driver should have
been listed (between the two commas in the description field). After the
setdriver command succeeds, all is well.
Tip
• ... contains the setting show add printer wizard = yes (the default).
The APW can do various things:
• Upload a new driver to the Samba [print$] share.
• Associate an uploaded driver with an existing (but still driverless)
print queue.
• Exchange the currently used driver for an existing print queue with
one that has been uploaded before.
• Add an entirely new printer to the Samba host (only in conjunction
with a working add printer command. A corresponding delete printer
command for removing entries from the Printers folder may also be
provided).
The last one (add a new printer) requires more effort than the previous
ones. To use the APW to successfully add a printer to a Samba server,
the add printer command must have a defined value. The program hook
must successfully add the printer to the UNIX print system (i.e., to /etc/
printcap, /etc/cups/printers.conf or other appropriate files) and to
smb.conf if necessary.
When using the APW from a client, if the named printer share does not
exist, smbd will execute the add printer command and reparse to attempt
to locate the new printer share. If the share is still not defined, an error
of ”Access Denied” is returned to the client. The add printer command is
executed under the context of the connected user, not necessarily a root
account. A map to guest = bad user may have connected you unwittingly
under the wrong privilege. You should check it by using the smbstatus
command.
Section 20.8. Other Gotchas 407
Once you are connected with the wrong credentials, there is no means to
reverse the situation other than to close all Explorer windows, and perhaps
reboot.
So you close all connections. You try again. You get the same message. You
check from the Samba side, using smbstatus. Yes, there are more connec-
tions. You kill them all. The client still gives you the same error message.
You watch the smbd.log file on a high debug level and try reconnect. Same
error message, but not a single line in the log. You start to wonder if there
was a connection attempt at all. You run ethereal and tcpdump while you
try to connect. Result: not a single byte goes on the wire. Windows still
gives the error message. You close all Explorer windows and start it again.
You try to connect — and this times it works! Windows seems to cache
connection information somewhere and does not keep it up to date (if you
are unlucky, you might need to reboot to get rid of the error message).
The easiest way to forcefully terminate all connections from your client to a
server is by executing:
This will also disconnect all mapped drives and will allow you create fresh
connection as required.
408 Classical Printing Support Chapter 20
You need to be extremely careful when you take notes about the files be-
longing to a particular driver. Don’t confuse the files for driver version “0”
(for Windows 9x/Me, going into [print$]/WIN/0/), driver version 2 (kernel
mode driver for Windows NT, going into [print$]/W32X86/2/; may be used
on Windows 200x/XP also), and driver version “3” (non-kernel mode driver
going into [print$]/W32X86/3/; cannot be used on Windows NT). Quite
often these different driver versions contain files that have the same name
but actually are very different. If you look at them from the Windows Ex-
plorer (they reside in %WINDOWS%\system32\spool\drivers\W32X86\), you
will probably see names in capital letters, while an enumdrivers command
from Samba would show mixed or lowercase letters, so it is easy to confuse
them. If you install them manually using rpcclient and subcommands, you
may even succeed without an error message. Only later, when you try in-
stall on a client, you will encounter error messages like This server has
no appropriate driver for the printer.
Here is an example. You are invited to look closely at the various files,
compare their names and their spelling, and discover the differences in the
composition of the version 2 and 3 sets. Note: the version 0 set contained
40 Dependentfiles, so I left it out for space reasons:
Dependentfiles: [\\10.160.50.8\print$\W32X86\3\aucplmNT.dll]
Dependentfiles: [\\10.160.50.8\print$\W32X86\3\ucs32p.dll]
Dependentfiles: [\\10.160.50.8\print$\W32X86\3\tnl32.dll]
Dependentfiles: [\\10.160.50.8\print$\W32X86\3\aussdrv.dll]
Dependentfiles: [\\10.160.50.8\print$\W32X86\3\cnspdc.dll]
Dependentfiles: [\\10.160.50.8\print$\W32X86\3\aussapi.dat]
Section 20.8. Other Gotchas 409
Dependentfiles: [\\10.160.50.8\print$\W32X86\3\cns3407.dll]
Dependentfiles: [\\10.160.50.8\print$\W32X86\3\CnS3G.cnt]
Dependentfiles: [\\10.160.50.8\print$\W32X86\3\NBAPI.DLL]
Dependentfiles: [\\10.160.50.8\print$\W32X86\3\NBIPC.DLL]
Dependentfiles: [\\10.160.50.8\print$\W32X86\3\cpcview.exe]
Dependentfiles: [\\10.160.50.8\print$\W32X86\3\cpcdspl.exe]
Dependentfiles: [\\10.160.50.8\print$\W32X86\3\cpcedit.dll]
Dependentfiles: [\\10.160.50.8\print$\W32X86\3\cpcqm.exe]
Dependentfiles: [\\10.160.50.8\print$\W32X86\3\cpcspl.dll]
Dependentfiles: [\\10.160.50.8\print$\W32X86\3\cfine32.dll]
Dependentfiles: [\\10.160.50.8\print$\W32X86\3\cpcr407.dll]
Dependentfiles: [\\10.160.50.8\print$\W32X86\3\Cpcqm407.hlp]
Dependentfiles: [\\10.160.50.8\print$\W32X86\3\cpcqm407.cnt]
Dependentfiles: [\\10.160.50.8\print$\W32X86\3\cns3ggr.dll]
Monitorname: []
Defaultdatatype: []
Dependentfiles: [\\10.160.50.8\print$\W32X86\2\AUCPLMNT.DLL]
Dependentfiles: [\\10.160.50.8\print$\W32X86\2\aussdrv.dll]
Dependentfiles: [\\10.160.50.8\print$\W32X86\2\cnspdc.dll]
Dependentfiles: [\\10.160.50.8\print$\W32X86\2\aussapi.dat]
Dependentfiles: [\\10.160.50.8\print$\W32X86\2\cns3407.dll]
Dependentfiles: [\\10.160.50.8\print$\W32X86\2\CnS3G.cnt]
Dependentfiles: [\\10.160.50.8\print$\W32X86\2\NBAPI.DLL]
Dependentfiles: [\\10.160.50.8\print$\W32X86\2\NBIPC.DLL]
Dependentfiles: [\\10.160.50.8\print$\W32X86\2\cns3gum.dll]
If we write the “version 2” files and the “version 3” files into different text
files and compare the result, we see this picture:
cns3g.dll cns3g.dll
iR8500sg.xpd iR8500sg.xpd
cns3gui.dll cns3gui.dll
cns3g.hlp cns3g.hlp
AUCPLMNT.DLL | aucplmNT.dll
> ucs32p.dll
> tnl32.dll
aussdrv.dll aussdrv.dll
cnspdc.dll cnspdc.dll
aussapi.dat aussapi.dat
cns3407.dll cns3407.dll
CnS3G.cnt CnS3G.cnt
NBAPI.DLL NBAPI.DLL
NBIPC.DLL NBIPC.DLL
cns3gum.dll | cpcview.exe
> cpcdspl.exe
> cpcqm.exe
> cpcspl.dll
> cfine32.dll
> cpcr407.dll
> Cpcqm407.hlp
> cpcqm407.cnt
> cns3ggr.dll
Do not be fooled! Driver files for each version with identical names may be
different in their content, as you can see from this size comparison:
done
In my example were even more differences than shown here. Conclusion: you
must be careful to select the correct driver files for each driver version. Don’t
rely on the names alone, and don’t interchange files belonging to different
driver versions.
Windows NT/2000 print servers associate a port with each printer. These
normally take the form of LPT1:, COM1:, FILE:, and so on. Samba must also
support the concept of ports associated with a printer. By default, only one
printer port, named “Samba Printer Port”, exists on a system. Samba does
not really need such a “port” in order to print; rather it is a requirement
of Windows clients. They insist on being told about an available port when
they request this information; otherwise, they throw an error message at
you. So Samba fakes the port information to keep the Windows clients
happy.
Samba does not support the concept of Printer Pooling internally either.
Printer pooling assigns a logical printer to multiple ports as a form of load
balancing or failover.
If you require multiple ports to be defined for some reason or another (my
users and my boss should not know that they are working with Samba),
configure the enumports command, which can be used to define an external
program that generates a listing of ports on a system.
412 Classical Printing Support Chapter 20
So now the printing works, but there are still problems. Most jobs print
well, some do not print at all. Some jobs have problems with fonts, which
do not look good. Some jobs print fast and some are dead-slow. We cannot
cover it all, but we want to encourage you to read the brief paragraph about
“Avoiding the Wrong PostScript Driver Settings” in Chapter 21, “CUPS
Printing Support”, Section 21.10.16.
The Imprints tool set provides a UNIX equivalent of the Windows NT APW.
For complete information, please refer to the Imprints3 Web site as well as
the documentation included with the Imprints source distribution. This
section provides only a brief introduction to the features of Imprints.
Unfortunately, the Imprints toolset is no longer maintained. As of December
2000, the project is in need of a new maintainer. The most important
skill to have is Perl coding and an interest in MS-RPC-based printing used
in Samba. If you wish to volunteer, please coordinate your efforts on the
Samba technical mailing list. The toolset is still in usable form, but only for
a series of older printer models where there are prepared packages to use.
Packages for more up-to-date print devices are needed if Imprints should
have a future. Information regarding the Imprints toolset can be obtained
from the Imprints4 home page.
The process of creating printer driver packages is beyond the scope of this
document (refer to Imprints.txt, included with the Samba distribution for
more information). In short, an Imprints driver package is a gzipped tarball
containing the driver files, related INF files, and a control file needed by the
installation client.
The Imprints server is really a database server that may be queried via
standard HTTP mechanisms. Each printer entry in the database has an
associated URL for the actual downloading of the package. Each package is
digitally signed via GnuPG, which can be used to verify that the package
downloaded is actually the one referred in the Imprints database. It is
strongly recommended that this security check not be disabled.
The following MS Knowledge Base article may be of some help if you need to
handle Windows 2000 clients: How to Add Printers with No User Interaction
in Windows 2000, (Microsoft KB 1891055 ). It also applies to Windows XP
Professional clients. The ideas sketched out in this section are inspired by
this article, which describes a command-line method that can be applied
to install network and local printers and their drivers. This is most useful
if integrated in Logon Scripts. You can see what options are available by
typing in the command prompt (DOS box):
5
<http://support.microsoft.com/default.aspx?scid=kb;en-us;189105>
Section 20.10. Adding Network Printers without User Interaction 415
rundll32 printui.dll,PrintUIEntry /?
A window pops up that shows you all of the command-line switches available.
An extensive list of examples is also provided. This is only for Windows
200x/XP; it does not work on Windows NT. Windows NT probably has
some other tools in the respective Resource Kit. Here is a suggestion about
what a client logon script might contain, with a short explanation of what
the lines actually do (it works if 200x/XP Windows clients access printers
via Samba, and works for Windows-based print servers too):
/q quiet modus.
/n names a printer.
The basic NT-style printer driver management has not changed considerably
in 3.0 over the 2.2.x releases (apart from many small improvements). Here
migration should be quite easy, especially if you followed previous advice
to stop using deprecated parameters in your setup. For migrations from
an existing 2.0.x setup, or if you continued Windows 9x/Me-style printing
in your Samba 2.2 installations, it is more of an effort. Please read the
appropriate release notes and the HOWTO Collection for Samba-2.2.x. You
can follow several paths. Here are possible scenarios for migration:
• You need to study and apply the new Windows NT printer and driver
support. Previously used parameters printer driver file, printer
driver, and printer driver location are no longer supported.
• You need to install a Windows 9x/Me driver into the [print$] share
for a printer on your Samba host. The driver files will be stored in
the “WIN40/0” subdirectory of [print$], and some other settings and
information go into the printing-related TDBs.
the Imprints6 web site for example. See also the discussion of rpcclient
usage in Chapter 21, “CUPS Printing Support”.
This topic has also been addressed in Chapter 12, “Remote and Local Man-
agement: The Net Command”. If you wish to volunteer your services to
help document this further, please contact John H. Terpstra7 .
Do not confuse the root password, which is valid for the UNIX system (and
in most cases stored in the form of a one-way hash in a file named /etc/
shadow), with the password used to authenticate against Samba. Samba
does not know the UNIX password. Root access to Samba resources requires
that a Samba account for root must first be created. This is done with the
smbpasswd command as follows:
Do not use the existing UNIX print system spool directory for the Samba
spool directory. It may seem convenient and a savings of space, but it only
leads to problems. The two must be separate.
6
<http://imprints.sourceforge.net/>
7
<mail://[email protected]>
Chapter 21
21.1 Introduction
The Common UNIX Print System (CUPS1 ) has become quite popular. All
major Linux distributions now ship it as their default printing system. To
many, it is still a mystical tool. Mostly, it just works. People tend to regard
it as a “black box” that they do not want to look into as long as it works.
But once there is a little problem, they have trouble finding out where to
start debugging it. Refer to Chapter 20, “Classical Printing Support”, which
contains a much information that is also relevant to CUPS.
CUPS sports quite a few unique and powerful features. While its basic
functions may be grasped quite easily, they are also new. Because it is
different from other, more traditional printing systems, it is best not to try
to apply any prior knowledge about printing to this new system. Rather,
try to understand CUPS from the beginning. This documentation will lead
you to a complete understanding of CUPS. Let’s start with the most basic
things first.
21.1.2 Overview
419
420 CUPS Printing Support Chapter 21
Printing with CUPS in the most basic smb.conf setup in Samba-3.0 (as was
true for 2.2.x) requires just two parameters: printing = cups and printcap
= cups. CUPS does not need a printcap file. However, the cupsd.conf
configuration file knows of two related directives that control how such a
file will be automatically created and maintained by CUPS for the conve-
nience of third-party applications (example: Printcap /etc/printcap and
PrintcapFormat BSD). Legacy programs often require the existence of a
printcap file containing printer names or they will refuse to print. Make
sure CUPS is set to generate and maintain a printcap file. For details, see
man cupsd.conf and other CUPS-related documentation, like the wealth
of documents regarding the CUPS server itself available from the CUPS3
web site.
Tip
it is there), and it allows access from only three hosts. To prevent CUPS
from kicking in and taking over the print jobs for that share, we need to set
printing = sysv and printcap = lpstat.
424 CUPS Printing Support Chapter 21
Before we delve into all the configuration options, let us clarify a few points.
Network printing needs to be organized and set up correctly. This frequently
doesn’t happen. Legacy systems or small business LAN environments often
lack design and good housekeeping.
Many small office or home networks, as well as badly organized larger en-
vironments, allow each client a direct access to available network printers.
This is generally a bad idea. It often blocks one client’s access to the printer
when another client’s job is printing. It might freeze the first client’s ap-
plication while it is waiting to get rid of the job. Also, there are frequent
complaints about various jobs being printed with their pages mixed with
each other. A better concept is the use of a print server: it routes all jobs
through one central system, which responds immediately, takes jobs from
multiple concurrent clients, and transfers them to the printer(s) in the cor-
rect order.
the Windows client. You then redirect output to a raw network print queue.
This procedure may be followed to achieve this: Configuration Steps for
Raw CUPS Printing Support
1. Edit /etc/cups/mime.types to uncomment the line near the end of
the file that has:
#application/octet-...
The printer drivers on the Windows clients may be installed in two func-
tionally different ways:
• Manually install the drivers locally on each client, one by one; this
yields the old LanMan style printing and uses a \\sambaserver\printershare
type of connection.
• Deposit and prepare the drivers (for later download) on the print server
(Samba); this enables the clients to use “Point’n’Print” to get drivers
semi-automatically installed the first time they access the printer; with
this method NT/200x/XP clients use the SPOOLSS/MS-RPC type
printing calls.
426 CUPS Printing Support Chapter 21
If you use the first option (drivers are installed on the client side), there
is one setting to take care of: CUPS needs to be told that it should allow
“raw” printing of deliberate (binary) file formats. The CUPS files that need
to be correctly set for raw mode printers to work are:
• /etc/cups/mime.types
• /etc/cups/mime.convs
Both contain entries (at the end of the respective files) that must be uncom-
mented to allow RAW mode operation. In /etc/cups/mime.types, make
sure this line is present:
application/octet-stream
application/octet-stream application/vnd.cups-raw 0 -
If these two files are not set up correctly for raw Windows client printing,
you may encounter the dreaded Unable to convert file 0 in your CUPS
error log file.
Note
of a lot of paper and ink. “Unknown” data are tagged by CUPS as MIME
type: application/octet-stream and not allowed to go to the printer.
By default, you can only send other (known) MIME types “raw”. Sending
data “raw” means that CUPS does not try to convert them and passes
them to the printer untouched (see Chapter 21, “CUPS Printing Support”
for more background explanations).
This is all you need to know to get the CUPS/Samba combo printing “raw”
files prepared by Windows clients, which have vendor drivers locally in-
stalled. If you are not interested in background information about more
advanced CUPS/Samba printing, simply skip the remaining sections of this
chapter.
This section describes three familiar methods, plus one new one, by which
printer drivers may be uploaded.
If you want to use the MS-RPC-type printing, you must upload the drivers
onto the Samba server first ([print$] share). For a discussion on how to de-
posit printer drivers on the Samba host (so the Windows clients can down-
load and use them via “Point’n’Print”), please refer to the Chapter 20,
“Classical Printing Support” of this book. There you will find a description
or reference to three methods of preparing the client drivers on the Samba
server:
These three methods apply to CUPS all the same. The cupsaddsmb utility
is new and more convenient way to load the Windows drivers into Samba is
provided if you use CUPS.
We now know how to set up a “dump” print server, that is, a server that
spools print jobs “raw”, leaving the print data untouched.
You might need to set up CUPS in a smarter way. The reasons could be
manifold:
• Maybe your boss wants to get monthly statistics: Which printer did
how many pages? What was the average data size of a job? What was
the average print run per day? What are the typical hourly peaks in
printing? Which department prints how much?
• Maybe you are asked to set up a print quota system: Users should not
be able to print more jobs once they have surpassed a given limit per
period.
• Maybe your previous network printing setup is a mess and must be
re-organized from a clean beginning.
• Maybe you are experiencing too many “blue screens” originating from
poorly debugged printer drivers running in NT “kernel mode”?
These goals cannot be achieved by a raw print server. To build a server
meeting these requirements, you’ll first need to learn how CUPS works and
how you can enable its features.
What follows is the comparison of some fundamental concepts for Windows
and UNIX printing, then a description of the CUPS filtering system, how it
works, and how you can tweak it.
In Windows OS, the format conversion job is done by the printer drivers.
On MS Windows OS platforms all application programmers have at their
disposal a built-in API, the graphical device interface (GDI), as part and
parcel of the OS itself to base themselves on. This GDI core is used as one
common unified ground for all Windows programs to draw pictures, fonts,
and documents on screen as well as on paper (print). Therefore, printer
driver developers can standardize on a well-defined GDI output for their
own driver input. Achieving WYSIWYG (What You See Is What You Get)
is relatively easy, because the on-screen graphic primitives, as well as the
on-paper drawn objects, come from one common source. This source, the
GDI, often produces a file format called Enhanced MetaFile (EMF). The
EMF is processed by the printer driver and converted to the printer-specific
file format.
Note
In UNIX and Linux, there is no comparable layer built into the OS kernel(s)
or the X (screen display) server. Every application is responsible for itself
to create its print output. Fortunately, most use PostScript and that at
430 CUPS Printing Support Chapter 21
least gives some common ground. Unfortunately, there are many different
levels of quality for this PostScript. And worse, there is a huge difference
(and no common root) in the way the same document is displayed on screen
and how it is presented on paper. WYSIWYG is more difficult to achieve.
This goes back to the time, decades ago, when the predecessors of X.org,
designing the UNIX foundations and protocols for graphical user interfaces,
refused to take responsibility for “paper output”, as some had demanded at
the time, and restricted itself to “on-screen only.” (For some years now, the
“Xprint” project has been under development, attempting to build printing
support into the X framework, including a PostScript and a PCL driver, but
it is not yet ready for prime time.) You can see this unfavorable inheritance
up to the present day by looking into the various “font” directories on your
system; there are separate ones for fonts used for X display and fonts to be
used on paper.
plotters). Of course, you can look at a PostScript file in its textual form and
you will be reading its PostScript code, the language instructions that need
to be interpreted by a rasterizer. Rasterizers produce pixel images, which
may be displayed on screen by a viewer program or on paper by a printer.
So, UNIX is lacking a common ground for printing on paper and display-
ing on screen. Despite this unfavorable legacy for UNIX, basic printing is
fairly easy if you have PostScript printers at your disposal. The reason is
these devices have a built-in PostScript language “interpreter,” also called
a raster image processor (RIP), (which makes them more expensive than
other types of printers; throw PostScript toward them, and they will spit
out your printed pages. The RIP does all the hard work of converting the
PostScript drawing commands into a bitmap picture as you see it on paper,
in a resolution as done by your printer. This is no different than PostScript
printing a file from a Windows origin.
Note
However, there are other types of printers out there. These do not know how
to print PostScript. They use their own PDL, often proprietary. To print
to them is much more demanding. Since your UNIX applications mostly
produce PostScript, and since these devices do not understand PostScript,
you need to convert the print files to a format suitable for your printer on
the host before you can send it away.
432 CUPS Printing Support Chapter 21
Here is where Ghostscript kicks in. Ghostscript is the traditional (and quite
powerful) PostScript interpreter used on UNIX platforms. It is a RIP in soft-
ware, capable of doing a lot of file format conversions for a very broad spec-
trum of hardware devices as well as software file formats. Ghostscript tech-
nology and drivers are what enable PostScript printing to non-PostScript
hardware. This is shown in Figure 21.3.
Tip
Use the “gs -h” command to check for all built-in “de-
vices” on your Ghostscript version. If you specify a pa-
rameter of -sDEVICE=png256 on your Ghostscript com-
mand line, you are asking Ghostscript to convert the input
into a PNG file. Naming a “device” on the command line
is the most important single parameter to tell Ghostscript
exactly how it should render the input. New Ghostscript
versions are released at fairly regular intervals, now by
artofcode LLC. They are initially put under the “AFPL”
license, but re-released under the GNU GPL as soon as
the next AFPL version appears. GNU Ghostscript is
probably the version installed on most Samba systems.
But it has some deficiencies. Therefore, ESP Ghostscript
was developed as an enhancement over GNU Ghostscript,
with lots of bug-fixes, additional devices, and improve-
ments. It is jointly maintained by developers from CUPS,
Gimp-Print, MandrakeSoft, SuSE, Red Hat, and Debian.
It includes the “cups” device (essential to print to non-PS
printers from CUPS).
PPDs contain all the information about general and special features of the
given printer model: Which different resolutions can it handle? Does it have
a duplexing unit? How many paper trays are there? What media types and
sizes does it take? For each item, it also names the special command string
to be sent to the printer (mostly inside the PostScript file) in order to enable
it.
434 CUPS Printing Support Chapter 21
Warning
Tip
Warning
Tip
hope it works for the non-PostScript version too). To understand how these
PPDs work for non-PS printers, we first need to dive deeply into the CUPS
filtering and file format conversion architecture. Stay tuned.
CUPS reads the file /etc/cups/mime.types (and all other files carrying a
*.types suffix in the same directory) upon startup. These files contain the
MIME type recognition rules that are applied when CUPS runs its autotyp-
ing routines. The rule syntax is explained in the man page for mime.types
and in the comments section of the mime.types file itself. A simple rule
reads like this:
This means if a filename has a .pdf suffix or if the magic string %PDF is
right at the beginning of the file itself (offset 0 from the start), then it is a
PDF file (application/pdf). Another rule is this:
If the filename has one of the suffixes .ai, .eps, .ps, or if the file itself
starts with one of the strings %! or <04>%!, it is a generic PostScript file
(application/postscript).
Warning
5
<http://www.linuxprinting.org/>
438 CUPS Printing Support Chapter 21
Note
CUPS can handle ASCII text, HP-GL, PDF, PostScript, DVI, and many
image formats (GIF, PNG, TIFF, JPEG, Photo-CD, SUN-Raster, PNM,
PBM, SGI-RGB, and more) and their associated MIME types with its filters.
CUPS reads the file /etc/cups/mime.convs (and all other files named with
a *.convs suffix in the same directory) upon startup. These files contain
lines naming an input MIME type, an output MIME type, a format con-
version filter that can produce the output from the input type, and virtual
costs associated with this conversion. One example line reads like this:
This means that the pdftops filter will take application/pdf as input
and produce application/postscript as output; the virtual cost of this
Section 21.5. The CUPS Filtering Architecture 439
application/octet-stream
The last two examples name the texttops filter to work on text/plain as
well as on application/x-shell. (Hint: This differentiation is needed for
the syntax highlighting feature of texttops).
The “CUPS requirements” for filters are simple. Take filenames or stdin
as input and write to stdout. They should take these arguments:
printer The name of the printer queue (normally this is the name of the
filter being run).
440 CUPS Printing Support Chapter 21
filename (optionally) The print request file (if missing, filters expected data
fed through stdin). In most cases, it is easy to write a simple wrapper
script around existing filters to make them work with CUPS.
21.5.4 Prefilters
21.5.5 pstops
21.5.6 pstoraster
for the first stage of the rasterization process. Its input is of MIME type
application/vnd.cups-postscript; its output is application/vnd.cups-raster.
This output format is not yet meant to be printable. Its aim is to serve as
a general-purpose input format for more specialized raster drivers that are
able to generate device-specific printer data. This is shown in Figure 21.6.
CUPS versions before version 1.1.15 shipped a binary (or source code) stan-
dalone filter, named pstoraster. pstoraster, which was derived from
GNU Ghostscript 5.50 and could be installed besides and in addition to any
GNU or AFPL Ghostscript package without conflicting.
From version 1.1.15, this feature has changed. The functions for this filter
have been integrated back into Ghostscript (now based on GNU Ghostscript
version 7.05). The pstoraster filter is now a simple shell script calling gs
with the -sDEVICE=cups parameter. If your Ghostscript fails when this
command is executed: gs -h |grep cups, you might not be able to print,
update your Ghostscript.
Section 21.5. The CUPS Filtering Architecture 443
CUPS ships with quite a variety of raster drivers for processing CUPS raster.
On my system I find in /usr/lib/cups/filter/ the following: rastertoalps,
rastertobj, rastertoepson, rastertoescp, rastertopcl, rastertotur-
boprint, rastertoapdk, rastertodymo, rastertoescp, rastertohp, and
rastertoprinter. Don’t worry if you have fewer drivers than this; some
of these are installed by commercial add-ons to CUPS (like rastertotur-
boprint), and others (like rastertoprinter) by third-party driver devel-
opment projects (such as Gimp-Print) wanting to cooperate as closely as
possible with CUPS. See Figure 21.9.
444 CUPS Printing Support Chapter 21
The last part of any CUPS filtering chain is a backend. Backends are special
programs that send the print-ready file to the final device. There is a sep-
arate backend program for any transfer protocol for sending print jobs over
the network, and one for every local interface. Every CUPS print queue
needs to have a CUPS “device-URI” associated with it. The device URI
is the way to encode the backend used to send the job to its destination.
Network device-URIs use two slashes in their syntax, local device URIs only
one, as you can see from the following list. Keep in mind that local interface
names may vary greatly from my examples, if your OS is not Linux:
serial This backend sends print files to serially connected printers. An ex-
ample for the CUPS device-URI to use is serial:/dev/ttyS0?baud=11500.
parallel This backend sends print files to printers connected to the parallel
port. An example for the CUPS device-URI to use is parallel:/dev/
lp0.
SCSI This backend sends print files to printers attached to the SCSI in-
terface. An example for the CUPS device-URI to use is scsi:/dev/
sr1.
ipp This backend sends print files to IPP-connected network printers (or
to other CUPS servers). Examples for CUPS device-URIs to use are
ipp:://192.193.194.195/ipp (for many HP printers) and ipp://
remote cups server/printers/remote printer name.
smb This backend sends print files to printers shared by a Windows host.
Examples of CUPS device-URIs that may be used includes:
smb://workgroup/server/printersharename
smb://server/printersharename
smb://username:password@workgroup/server/printersharename
smb://username:password@server/printersharename
It is easy to write your own backends as shell or Perl scripts if you need
any modification or extension to the CUPS print system. One reason could
be that you want to create “special” printers that send the print jobs as
email (through a “mailto:/” backend), convert them to PDF (through a
“pdfgen:/” backend) or dump them to “/dev/null”. (In fact, I have the
systemwide default printer set up to be connected to a devnull:/ backend:
there are just too many people sending jobs without specifying a printer, and
scripts and programs that do not name a printer. The systemwide default
deletes the job and sends a polite email back to the $USER asking him or
her to always specify the correct printer name.)
Not all of the mentioned backends may be present on your system or usable
(depending on your hardware configuration). One test for all available CUPS
backends is provided by the lpinfo utility. Used with the -v parameter, it
lists all available backends:
Section 21.5. The CUPS Filtering Architecture 447
$ lpinfo -v
cupsomatic filters may be the most widely used on CUPS installations. You
must be clear that these were not developed by the CUPS people. They are a
third-party add-on to CUPS. They utilize the traditional Ghostscript devices
to render jobs for CUPS. When troubleshooting, you should know about
the difference. Here the whole rendering process is done in one stage, inside
Ghostscript, using an appropriate device for the target printer. cupsomatic
uses PPDs that are generated from the Foomatic Printer & Driver Database
at Linuxprinting.org.
You can recognize these PPDs from the line calling the cupsomatic filter:
You may find this line among the first 40 or so lines of the PPD file. If you
have such a PPD installed, the printer shows up in the CUPS Web interface
with a foomatic namepart for the driver description. cupsomatic is a Perl
script that runs Ghostscript with all the complicated command-line options
autoconstructed from the selected PPD and command line options give to
the print job.
However, cupsomatic is now deprecated. Its PPDs (especially the first
generation of them, still in heavy use out there) are not meeting the Adobe
specifications. You might also suffer difficulties when you try to download
them with “Point’n’Print” to Windows clients. A better and more powerful
successor is now in a stable beta-version: it is called foomatic-rip. To use
foomatic-rip as a filter with CUPS, you need the new type of PPDs, which
have a similar but different line:
specify different quality levels (hi-res photo, normal color, grayscale, and
draft) with a single click, whereas before you could have required five or
more different selections (media type, resolution, inktype, and dithering al-
gorithm). There is support for custom-size media built in. There is support
to switch print options from page to page in the middle of a job. And the
best thing is that the new foomatic-rip works seamlessly with all legacy
spoolers too (like LPRng, BSD-LPD, PDQ, PPR, and so on), providing for
them access to use PPDs for their printing.
If you want to see an overview of all the filters and how they relate to each
other, the complete picture of the puzzle is at the end of this chapter.
21.5.12 mime.convs
CUPS autoconstructs all possible filtering chain paths for any given MIME
type and every printer installed. But how does it decide in favor of or against
a specific alternative? (There may be cases where there is a choice of two or
more possible filtering chains for the same target printer.) Simple. You may
have noticed the figures in the third column of the mime.convs file. They
represent virtual costs assigned to this filter. Every possible filtering chain
will sum up to a total “filter cost.” CUPS decides for the most “inexpensive”
route.
Tip
You can tell CUPS to print (nearly) any file “raw”. “Raw” means it will not
be filtered. CUPS will send the file to the printer “as is” without bothering if
the printer is able to digest it. Users need to take care themselves that they
send sensible data formats only. Raw printing can happen on any queue if
the “-o raw” option is specified on the command line. You can also set up
raw-only queues by simply not associating any PPD with it. This command:
application/octet-stream
This line (with no specific autotyping rule set) makes all files not otherwise
auto-typed a member of application/octet-stream. In /etc/cups/mime.
convs, have this line:
application/octet-stream application/vnd.cups-raw 0 -
This line tells CUPS to use the Null Filter (denoted as “-”, doing nothing at
all) on application/octet-stream, and tag the result as application/vnd.
cups-raw. This last one is always a green light to the CUPS scheduler to
now hand the file over to the backend connecting to the printer and sending
it over.
Note
Originally PPDs were meant to be used for PostScript printers only. Here,
they help to send device-specific commands and settings to the RIP, which
processes the job file. CUPS has extended this scope for PPDs to cover non-
PostScript printers too. This was not difficult, because it is a standardized
file format. In a way it was logical too: CUPS handles PostScript and uses
a PostScript RIP (Ghostscript) to process the job files. The only difference
is that a PostScript printer has the RIP built-in, for other types of printers
the Ghostscript RIP runs on the host computer.
PPDs for a non-PostScript printer have a few lines that are unique to CUPS.
The most important one looks similar to this:
It is the last piece in the CUPS filtering puzzle. This line tells the CUPS dae-
mon to use as a last filter rastertoprinter. This filter should be served
as input an application/vnd.cups-raster MIME type file. Therefore,
CUPS should autoconstruct a filtering chain, which delivers as its last out-
put the specified MIME type. This is then taken as input to the specified
rastertoprinter filter. After the last filter has done its work (rasterto-
printer is a Gimp-Print filter), the file should go to the backend, which
sends it to the output device.
CUPS by default ships only a few generic PPDs, but they are good for several
hundred printer models. You may not be able to control different paper
trays, or you may get larger margins than your specific model supports. See
Table 21.1Table 21.1 for summary information.
Some are free software, some are shareware, and some are proprietary.
Often this produces better quality (and has several more advantages) than
other methods. This is shown in Figure 21.10.
One other method is the cupsomatic/foomatic-rip way. Note that cup-
somatic is not made by the CUPS developers. It is an independent contri-
bution to printing development, made by people from Linuxprinting.org.6
cupsomatic is no longer developed, maintained, or supported. It now been
replaced by foomatic-rip. foomatic-rip is a complete rewrite of the old
cupsomatic idea, but very much improved and generalized to other (non-
CUPS) spoolers. An upgrade to foomatic-rip is strongly advised, espe-
cially if you are upgrading to a recent version of CUPS, too.
Like the old cupsomatic method, the foomatic-rip (new) method from
Linuxprinting.org uses the traditional Ghostscript print file processing, do-
ing everything in a single step. It therefore relies on all the other devices
built into Ghostscript. The quality is as good (or bad) as Ghostscript render-
ing is in other spoolers. The advantage is that this method supports many
printer models not supported (yet) by the more modern CUPS method.
Of course, you can use both methods side by side on one system (and even
for one printer, if you set up different queues) and find out which works best
6
See also <http://www.cups.org/cups-help.html>
Section 21.5. The CUPS Filtering Architecture 453
for you.
cupsomatic kidnaps the print file after the application/vnd.cups-postscript
stage and deviates it through the CUPS-external, systemwide Ghostscript
installation. Therefore, the print file bypasses the pstoraster filter (and
also bypasses the CUPS raster drivers rastertosomething). After Ghostscript
finished its rasterization, cupsomatic hands the rendered file directly to the
CUPS backend. Figure 21.10, Figure 21.10, illustrates the difference be-
tween native CUPS rendering and the Foomatic/cupsomatic method.
printer, but you want to print pages 3-5, 7, and 11-13 only, and you want
to print them “two-up” and “duplex”:
• Your print options (page selection as required, two-up, duplex) are
passed to CUPS on the command line.
• The (complete) PDF file is sent to CUPS and autotyped as applica-
tion/pdf.
• The file therefore must first pass the pdftops prefilter, which pro-
duces PostScript MIME type application/postscript (a preview
here would still show all pages of the original PDF).
• The file then passes the pstops filter that applies the command-line
options: it selects pages 2-5, 7, and 11-13, creates the imposed layout
“two pages on one sheet”, and inserts the correct “duplex” command
(as defined in the printer’s PPD) into the new PostScript file; the file is
now of PostScript MIME type application/vnd.cups-postscript.
• The file goes to the socket backend, which transfers the job to the
printers.
The resulting filter chain, therefore, is as shown in Figure 21.11.
Assume you want to print the same filter to an USB-connected Epson Stylus
Photo printer installed with the CUPS stphoto2.ppd. The first few filtering
stages are nearly the same:
• Your print options (page selection as required, two-up, duplex) are
passed to CUPS on the command line.
• The (complete) PDF file is sent to CUPS and autotyped as applica-
tion/pdf.
• The file must first pass the pdftops prefilter, which produces PostScript
MIME type application/postscript (a preview here would still
show all pages of the original PDF).
• The file then passes the “pstops” filter that applies the command-
line options: it selects the pages 2-5, 7, and 11-13, creates the imposed
Section 21.5. The CUPS Filtering Architecture 455
layout “two pages on one sheet,” and inserts the correct “duplex” com-
mand (oops — this printer and PPD do not support duplex printing at
all, so this option will be ignored) into the new PostScript file; the file is
now of PostScript MIME type application/vnd.cups-postscript.
• The file then passes the pstoraster stage and becomes MIME type
application/cups-raster.
• Finally, the rastertoepson filter does its work (as indicated in the
printer’s PPD), creating the printer-specific raster data and embedding
any user-selected print options into the print data stream.
• The file goes to the usb backend, which transfers the job to the print-
ers.
The resulting filter chain therefore is as shown in Figure 21.12.
usb rastertoepson
On the Internet you can now find many thousands of CUPS-PPD files (with
their companion filters), in many national languages supporting more than
1,000 non-PostScript models.
• ESP PrintPro7 (commercial, non-free) is packaged with more than
3,000 PPDs, ready for successful use “out of the box” on Linux, Mac
OS X, IBM-AIX, HP-UX, Sun-Solaris, SGI-IRIX, Compaq Tru64, Dig-
ital UNIX, and other commercial Unices (it is written by the CUPS
developers themselves and its sales help finance the further develop-
ment of CUPS, as they feed their creators).
• The Gimp-Print Project8 (GPL, free software) provides around 140
7
<http://wwwl.easysw.com/printpro/>
8
<http://gimp-print.sourceforge.net/>
456 CUPS Printing Support Chapter 21
CUPS also supports the use of “interface scripts” as known from System
V AT&T printing systems. These are often used for PCL printers, from
applications that generate PCL print jobs. Interface scripts are specific to
printer models. They have a role similar to PPDs for PostScript printers.
Interface scripts may inject the Escape sequences as required into the print
data stream if the user, for example, selects a certain paper tray, or changes
paper orientation, or uses A3 paper. Interface scripts are practically un-
known in the Linux realm. On HP-UX platforms they are more often used.
You can use any working interface script on CUPS too. Just install the
printer with the -i option:
9
<http://www.turboprint.com/>
10
<http://www-124.ibm.com/developerworks/oss/linux/projects/omni/>
11
<http://hpinkjet.sourceforge.net/>
12
<http://www.linuxprinting.org/>
Section 21.6. Network Printing (Purely Windows) 457
• Execute the driver locally and render the GDI output (EMF) into the
printer-specific format on their own.
• Send the GDI output (EMF) to the server, where the driver is executed
to render the printer-specific output.
Both print paths are shown in the flowcharts in Figure 21.13, and Fig-
ure 21.14.
In the first case the print server must spool the file as raw, meaning it
shouldn’t touch the job file and try to convert it in any way. This is what a
traditional UNIX-based print server can do too, and at a better performance
and more reliably than an NT print server. This is what most Samba ad-
ministrators probably are familiar with. One advantage of this setup is that
this “spooling-only” print server may be used even if no driver(s) for UNIX
are available. It is sufficient to have the Windows client drivers available
and installed on the clients. This is illustrated in Figure 21.13.
13
<http://playground.sun.com/printing/documentation/interface.html>
458 CUPS Printing Support Chapter 21
The other path executes the printer driver on the server. The client transfers
print files in EMF format to the server. The server uses the PostScript, PCL,
ESC/P, or other driver to convert the EMF file into the printer-specific
language. It is not possible for UNIX to do the same. Currently, there
is no program or method to convert a Windows client’s GDI output on a
UNIX server into something a printer could understand. This is illustrated
in Figure 21.14.
Since UNIX print servers cannot execute the Win32 program code on their
platform, the picture is somewhat different. However, this does not limit
your options all that much. On the contrary, you may have a way here to
implement printing features that are not possible otherwise.
Here is a simple recipe showing how you can take advantage of CUPS’s
powerful features for the benefit of your Windows network printing clients:
• Let the CUPS server render the PostScript into device-specific raster
format.
This requires the clients to use a PostScript driver (even if the printer is a
non-PostScript model. It also requires that you have a driver on the CUPS
server.
When these parameters are specified, all manually set print directives (like
print command or lppause command ) in smb.conf (as well as in Samba
itself) will be ignored. Instead, Samba will directly interface with CUPS
through its application program interface (API), as long as Samba has been
compiled with CUPS library (libcups) support. If Samba has not been
compiled with CUPS support, and if no other print commands are set up,
then printing will use the System V AT&T command set, with the -oraw
option automatically passing through (if you want your own defined print
commands to work with a Samba server that has CUPS support compiled
in, simply use classicalprinting = sysv). This is illustrated in Figure 21.15.
460 CUPS Printing Support Chapter 21
Samba must use its own spool directory (it is set by a line similar to path =
/var/spool/samba, in the [printers] or [printername] section of smb.conf).
Samba receives the job in its own spool space and passes it into the spool
directory of CUPS (the CUPS spool directory is set by the RequestRoot
directive in a line that defaults to RequestRoot /var/spool/cups). CUPS
checks the access rights of its spool directory and resets it to healthy values
with every restart. We have seen quite a few people who used a common
spooling space for Samba and CUPS, and struggled for weeks with this
“problem.”
This section discusses the use of CUPS filters on the server — configuration
where clients make use of a PostScript driver with CUPS-PPDs.
PPDs can control all print device options. They are usually provided by
the manufacturer — if you own a PostScript printer, that is. PPD files
are always a component of PostScript printer drivers on MS Windows or
Apple Mac OS systems. They are ASCII files containing user-selectable
print options, mapped to appropriate PostScript, PCL, or PJL commands
for the target printer. Printer driver GUI dialogs translate these options
“on the fly” into buttons and drop-down lists for the user to select.
CUPS can load, without any conversions, the PPD file from any Windows
(NT is recommended) PostScript driver and handle the options. There is
a Web browser interface to the print options (select <http://localhost:
631/printers/> and click on one Configure Printer button to see it) or a
command-line interface (see man lpoptions or see if you have lphelp on
your system). There are also some different GUI front-ends on Linux/UNIX,
which can present PPD options to users. PPD options are normally meant
to be evaluated by the PostScript RIP on the real PostScript printer.
CUPS does not limit itself to “real” PostScript printers in its use of PPDs.
The CUPS developers have extended the scope of the PPD concept to
also describe available device and driver options for non-PostScript printers
through CUPS-PPDs.
because until now there have been only two different PostScript drivers: the
one from Adobe and the one from Microsoft. Both are well tested and are
as stable as you can imagine on Windows. The CUPS driver is derived from
the Microsoft one.
More recent printer drivers on W200x and XP no longer run in kernel mode
(unlike Windows NT). However, both operating systems can still use the
NT drivers, running in kernel mode (you can roughly tell which is which
as the drivers in subdirectory “2” of “W32X86” are “old” ones). As was
said before, the Adobe as well as the Microsoft PostScript drivers are not
known to cause any stability problems. The CUPS driver is derived from
the Microsoft one. There is a simple reason for this: the MS DDK (Device
464 CUPS Printing Support Chapter 21
The cupsaddsmb utility (shipped with all current CUPS versions) is an al-
ternative method to transfer printer drivers into the Samba [print$] share.
Remember, this share is where clients expect drivers deposited and set up
for download and installation. It makes the sharing of any (or all) installed
CUPS printers quite easy. cupsaddsmb can use the Adobe PostScript
driver as well as the newly developed CUPS PostScript driver for Win-
dows NT/200x/XP. cupsaddsmb does not work with arbitrary vendor printer
drivers, but only with the exact driver files that are named in its man page.
The CUPS printer driver is available from the CUPS download site. Its
package name is cups-samba-[version].tar.gz. It is preferred over the
Adobe drivers because it has a number of advantages:
• It supports a much more accurate page accounting.
• It supports banner pages and page labels on all printers.
• It supports the setting of a number of job IPP attributes (such as job
priority, page label, and job billing).
However, currently only Windows NT, 2000, and XP are supported by the
CUPS drivers. You will also need to get the respective part of the Adobe
Section 21.10. Configuring CUPS for Driver Download 465
CUPS users may get the exact same package from <http://www.cups.org/
software.html>. It is a separate package from the CUPS-based software
files, tagged as CUPS 1.1.x Windows NT/200x/XP Printer Driver for Samba
(tar.gz, 192k). The filename to download is cups-samba-1.1.x.tar.gz.
Upon untar and unzipping, it will reveal these files:
466 CUPS Printing Support Chapter 21
These have been packaged with the ESP meta-packager software EPM. The
*.install and *.remove files are simple shell scripts, which untar the *.ss
(the *.ss is nothing else but a tar archive, which can be untarred by “tar”
too). Then it puts the content into /usr/share/cups/drivers/. This
content includes three files:
root# ./cups-samba.install
[....]
Installing software...
Updating file permissions...
Running post-install commands...
Installation is complete.
The script should automatically put the driver files into the /usr/share/
cups/drivers/ directory:
Warning
This new CUPS PostScript driver is currently binary only, but free of charge.
No complete source code is provided (yet). The reason is that it has been
developed with the help of the Microsoft DDK and compiled with Microsoft
Visual Studio 6. Driver developers are not allowed to distribute the whole
of the source code as free software. However, CUPS developers released the
“diff” in source code under the GPL, so anybody with a license for Visual
Studio and a DDK will be able to compile for himself or herself.
The CUPS drivers do not support the older Windows 95/98/Me, but only
the Windows NT/2000/XP client.
Windows NT, 2000, and XP are supported by:
• cups.hlp
• cupsdrvr.dll
• cupsui.dll
Adobe drivers are available for the older Windows 95/98/Me as well as for
Windows NT/2000/XP clients. The set of files is different from the different
platforms.
Windows 95, 98, and ME are supported by:
• ADFONTS.MFM
• ADOBEPS4.DRV
• ADOBEPS4.HLP
• DEFPRTR2.PPD
468 CUPS Printing Support Chapter 21
• ICONLIB.DLL
• PSMON.DLL
Windows NT, 2000, and XP are supported by:
• ADOBEPS5.DLL
• ADOBEPSU.DLL
• ADOBEPSU.HLP
Note
If both the Adobe driver files and the CUPS driver files
for the support of Windows NT/200x/XP are present in-
stalled on the server, the Adobe files will be ignored and
the CUPS files will be used. If you prefer — for what-
ever reason — to use Adobe-only drivers, move away the
three CUPS driver files. The Windows 9x/Me clients use
the Adobe drivers in any case.
Acquiring the Adobe driver files seems to be unexpectedly difficult for many
users. They are not available on the Adobe Web site as single files, and
the self-extracting and/or self-installing Windows-.exe is not easy to locate
either. You probably need to use the included native installer and run the
installation process on one client once. This will install the drivers (and one
generic PostScript printer) locally on the client. When they are installed,
share the generic PostScript printer. After this, the client’s [print$] share
holds the Adobe files, which you can get with smbclient from the CUPS
host.
Users of the ESP Print Pro software are able to install the ESP print drivers
package as an alternative to the Adobe postscript drivers. To do so, retrieve
Section 21.10. Configuring CUPS for Driver Download 469
the driver files from the normal download area of the ESP Print Pro software
at Easy Software14 web site. You need to locate the link labeled “SAMBA”
among the Download Printer Drivers for ESP Print Pro 4.x area and
download the package. Once installed, you can prepare any driver by simply
highlighting the printer in the Printer Manager GUI and selecting Export
Driver... from the menu. Of course, you need to have prepared Samba
beforehand to handle the driver files; that is, set up the [print$] share, and
so on. The ESP Print Pro package includes the CUPS driver files as well as
a (licensed) set of Adobe drivers for the Windows 95/98/Me client family.
Once you have run the install script (and possibly manually moved the cups.
hlp file to /usr/share/cups/drivers/), the driver is ready to be put into
Samba’s [print$] share (which often maps to /etc/samba/drivers/ and
contains a subdirectory tree with WIN40 and W32X86 branches). You do
this by running cupsaddsmb (see also man cupsaddsmb for CUPS since
release 1.1.16).
Tip
Once the driver files are in the [print$] share and are initialized, they are
ready to be downloaded and installed by the Windows NT/200x/XP clients.
14
<http://www.easysw.com/software.html>
470 CUPS Printing Support Chapter 21
Note
Note
Note
Note
Are you interested in a comparison between the CUPS and the Adobe
PostScript drivers? For our purposes these are the most important items
that weigh in favor of CUPS:
• No hassle with the Adobe EULA.
• No hassle with the question, “Where do I get the ADOBE*.* driver
files?”
• The Adobe drivers (on request of the printer PPD associated with
them) often put a PJL header in front of the main PostScript part
of the print file. Thus, the print file starts with <1B >%-12345X or
<escape>%-12345X instead of %!PS. This leads to the CUPS dae-
mon autotyping the incoming file as a print-ready file, not initiating a
pass through the pstops filter (to speak more technically, it is not re-
garded as the generic MIME-type application/postscript, but as
the more special MIME type application/cups.vnd-postscript),
which therefore also leads to the page accounting in /var/log/cups/-
page log not receiving the exact number of pages; instead the dummy
page number of “1” is logged in a standard setup).
• The Adobe driver has more options to misconfigure the PostScript
generated by it (like setting it inadvertently to Optimize for Speed
instead of Optimize for Portability, which could lead to CUPS being
unable to process it).
• The CUPS PostScript driver output sent by Windows clients to the
CUPS server is guaranteed to autotype as the generic MIME type
application/postscript, thus passing through the CUPS pstops
filter and logging the correct number of pages in the page log for
accounting and quota purposes.
• The CUPS PostScript driver supports the sending of additional stan-
dard (IPP) print options by Windows NT/200x/XP clients. Such ad-
ditional print options are naming the CUPS standard banner pages
(or the custom ones, should they be installed at the time of driver
download), using the CUPS page-label option, setting a job priority,
and setting the scheduled time of printing (with the option to support
additional useful IPP job attributes in the future).
• The CUPS PostScript driver supports the inclusion of the new *cup-
Section 21.10. Configuring CUPS for Driver Download 473
The cupsaddsmb command copies the needed files into your [print$] share.
Additionally, the PPD associated with this printer is copied from /etc/
cups/ppd/ to [print$]. There the files wait for convenient Windows client
installations via Point’n’Print. Before we can run the command successfully,
we need to be sure that we can authenticate toward Samba. If you have a
small network, you are probably using user-level security (security = user).
Here is an example of a successfully run cupsaddsmb command:
To share all printers and drivers, use the -a parameter instead of a printer
name. Since cupsaddsmb “exports” the printer drivers to Samba, it should
be obvious that it only works for queues with a CUPS driver associated.
Probably you want to see what’s going on. Use the -v parameter to get a
more verbose output. The output below was edited for better readability:
all “\” at the end of a line indicate that I inserted an artificial line break
plus some indentation here:
Warning
You will see the root password for the Samba account
printed on screen.
If you look closely, you’ll discover your root password was transferred un-
encrypted over the wire, so beware! Also, if you look further, you may
discover error messages like NT STATUS OBJECT NAME COLLISION in
the output. This will occur when the directories WIN40 and W32X86 al-
ready existed in the [print$] driver download share (from a previous driver
installation). These are harmless warning messages.
What has happened? What did cupsaddsmb do? There are five stages of
the procedure:
1. Call the CUPS server via IPP and request the driver files and the PPD
file for the named printer.
476 CUPS Printing Support Chapter 21
2. Store the files temporarily in the local TEMPDIR (as defined in cupsd.
conf).
3. Connect via smbclient to the Samba server’s [print$] share and put
the files into the share’s WIN40 (for Windows 9x/Me) and W32X86
(for Windows NT/200x/XP) subdirectories.
4. Connect via rpcclient to the Samba server and execute the adddriver
command with the correct parameters.
5. Connect via rpcclient to the Samba server a second time and execute
the setdriver command.
Note
You must always check if the utility completed successfully in all fields. You
need at minimum these three messages among the output:
1. Printer Driver infotec 2105 successfully installed. # (for the W32X86
== Windows NT/200x/XP architecture).
2. Printer Driver infotec 2105 successfully installed. # (for the WIN40
== Windows 9x/Me architecture).
3. Successfully set [printerXPZ] to driver [printerXYZ].
Section 21.10. Configuring CUPS for Driver Download 477
These messages are probably not easily recognized in the general output.
If you run cupsaddsmb with the -a parameter (which tries to prepare all
active CUPS printer drivers for download), you might miss if individual
printer drivers had problems installing properly. A redirection of the output
will help you analyze the results in retrospective.
If you get:
it means that you might have set use client driver = yes for this printer.
Setting it to “no” will solve the problem. Refer to the smb.conf man page
for explanantion of the use client driver.
Note
(Note the two backslashes: the first one is required to “escape” the second
one).
478 CUPS Printing Support Chapter 21
Figure 21.16 shows a chart about the procedures, command flows, and data
flows of the cupaddsmb command. Note again: cupsaddsmb is not in-
tended to, and does not work with, raw print queues!
should you desire to use the CUPS networked PostScript RIP functions.
(Note that user “ntadmin” needs to be a valid Samba user with the required
privileges to access the printershare.) This sets up the printer connection in
the traditional LanMan way (not using MS-RPC).
Printing works, but there are still problems. Most jobs print well, some do
not print at all. Some jobs have problems with fonts, which do not look
very good. Some jobs print fast and some are dead-slow. Many of these
problems can be greatly reduced or even completely eliminated if you follow
a few guidelines. Remember, if your print device is not PostScript-enabled,
you are treating your Ghostscript installation on your CUPS host with the
output your client driver settings produce. Treat it well:
• Avoid the PostScript Output Option: Optimize for Speed setting. Use
the Optimize for Portability instead (Adobe PostScript driver).
480 CUPS Printing Support Chapter 21
• Choose True Type Font: Download as Softfont into Printer over the
default Replace by Device Font (for exotic fonts, you may need to
change it back to get a printout at all; Adobe).
Of course, you can run all the commands that are embedded into the cup-
saddsmb convenience utility yourself, one by one, and upload and prepare
the driver files for future client downloads.
1. Prepare Samba (a CUPS print queue with the name of the printer
should be there. We are providing the driver now).
We are going to do this now. First, read the man page on rpcclient to
get a first idea. Look at all the printing-related subcommands: enumprint-
ers, enumdrivers, enumports, adddriver, and setdriver are among the
most interesting ones. rpcclient implements an important part of the MS-
RPC protocol. You can use it to query (and command) a Windows NT
(or 200x/XP) PC, too. MS-RPC is used by Windows clients, among other
things, to benefit from the Point’n’Print features. Samba can now mimic
this as well.
Section 21.11. Installing PostScript Driver Files Manually Using rpcclient 481
First let’s check the rpcclient man page. Here are two relevant passages:
adddriver <arch> <config> Execute an AddPrinterDriver() RPC to
install the printer driver information on the server. The driver files should
already exist in the directory returned by getdriverdir. Possible values for
arch are the same as those for the getdriverdir command. The config
parameter is defined as follows:
The exact format isn’t made too clear by the man page, since you have to
deal with some parameters containing spaces. Here is a better description
for it. We have line-broken the command and indicated the breaks with “\”.
Usually you would type the command in one line without the line breaks:
482 CUPS Printing Support Chapter 21
adddriver "Architecture" \
"LongPrinterName:DriverFile:DataFile:ConfigFile:HelpFile:\
LanguageMonitorFile:DataType:ListOfFiles,Comma-separated"
What the man pages denote as a simple <config> keyword in reality con-
sists of eight colon-separated fields. The last field may take multiple (in
some very insane cases, even 20 different additional) files. This might sound
confusing at first. What the man pages call the “LongPrinterName” in re-
ality should be called the “Driver Name”. You can name it anything you
want, as long as you use this name later in the rpcclient ... setdriver
command. For practical reasons, many name the driver the same as the
printer.
It isn’t simple at all. I hear you asking: “How do I know which files are Driver
File”, “Data File”, “Config File”, “Help File” and “Language Monitor File in
each case?” For an answer, you may want to have a look at how a Windows
NT box with a shared printer presents the files to us. Remember that this
whole procedure has to be developed by the Samba Team by listening to the
traffic caused by Windows computers on the wire. We may as well turn to a
Windows box now and access it from a UNIX workstation. We will query it
with rpcclient to see what it tells us and try to understand the man page
more clearly.
From the result it should become clear which is which. Here is an example
from my installation:
[Windows NT x86]
Printer Driver Info 3:
Version: [2]
Driver Name: [DANKA InfoStream]
Architecture: [Windows NT x86]
Driver Path: [C:\WINNT\System32\spool\DRIVERS\W32X86\2\PSCRIPT.DLL]
Datafile: [C:\WINNT\System32\spool\DRIVERS\W32X86\2\INFOSTRM.PPD]
Configfile: [C:\WINNT\System32\spool\DRIVERS\W32X86\2\PSCRPTUI.DLL]
Helpfile: [C:\WINNT\System32\spool\DRIVERS\W32X86\2\PSCRIPT.HLP]
Dependentfiles: []
Dependentfiles: []
Dependentfiles: []
Dependentfiles: []
Dependentfiles: []
Dependentfiles: []
Dependentfiles: []
Monitorname: []
Defaultdatatype: []
Some printer drivers list additional files under the label Dependentfiles,
and these would go into the last field ListOfFiles,Comma-separated. For
the CUPS PostScript drivers, we do not need any (nor would we for the
Adobe PostScript driver); therefore, the field will get a “NULL” entry.
From the man page (and from the quoted output of cupsaddsmb above)
it becomes clear that you need to have certain conditions in order to make
the manual uploading and initializing of the driver files succeed. The two
rpcclient subcommands (adddriver and setdriver) need to encounter the
following preconditions to complete successfully:
• You are connected as printer admin or root (this is not the “Printer
Operators” group in NT, but the printer admin group as defined in
the [global] section of smb.conf).
• Copy all required driver files to \\SAMBA\print$\w32x86 and \\SAMBA\print$\win40
484 CUPS Printing Support Chapter 21
We are going to install a printer driver now by manually executing all re-
quired commands. Because this may seem a rather complicated process
at first, we go through the procedure step by step, explaining every single
action item as it comes up. Manual Driver Installation
1. Install the printer on CUPS.
This installs a printer with the name mysmbtstprn to the CUPS sys-
tem. The printer is accessed via a socket (a.k.a. JetDirect or Direct
TCP/IP) connection. You need to be root for this step.
2. (Optional.) Check if the printer is recognized by Samba.
This should show the printer in the list. If not, stop and restart the
Samba daemon (smbd) or send a HUP signal:
drivername:[]
servername:[\\kde-bitshop]
printername:[\\kde-bitshop\mysmbtstprn]
sharename:[mysmbtstprn]
portname:[Samba Printer Port]
drivername:[]
comment:[mysmbtstprn]
location:[]
sepfile:[]
printprocessor:[winprint]
486 CUPS Printing Support Chapter 21
None of the three commands shown above should show a driver. This
step was done for the purpose of demonstrating this condition. An
attempt to connect to the printer at this stage will prompt a message
along the lines of, “The server does not have the required printer driver
installed.”
(This command should be entered in one long single line. Line breaks
and the line ends indicated by “\” have been inserted for readability
reasons.) This step is required for the next one to succeed. It makes the
driver files physically present in the [print$] share. However, clients
would still not be able to install them, because Samba does not yet
treat them as driver files. A client asking for the driver would still be
presented with a “not installed here” message.
root# ls -l /etc/samba/drivers/W32X86/
total 669
drwxr-sr-x 2 root ntadmin 532 May 25 23:08 2
drwxr-sr-x 2 root ntadmin 670 May 16 03:15 3
-rwxr--r-- 1 root ntadmin 14234 May 25 23:21 cups.hlp
-rwxr--r-- 1 root ntadmin 278380 May 25 23:21 cupsdrvr.dll
-rwxr--r-- 1 root ntadmin 215848 May 25 23:21 cupsui.dll
-rwxr--r-- 1 root ntadmin 169458 May 25 23:21 mysmbtstprn.PPD
Section 21.11. Installing PostScript Driver Files Manually Using rpcclient 487
The driver files now are in the W32X86 architecture “root” of [print$].
You cannot repeat this step if it fails. It could fail even as a result of
a simple typo. It will most likely have moved a part of the driver files
into the “2” subdirectory. If this step fails, you need to go back to the
fourth step and repeat it before you can try this one again. In this
step, you need to choose a name for your driver. It is normally a good
idea to use the same name as is used for the printer name; however, in
big installations you may use this driver for a number of printers that
obviously have different names, so the name of the driver is not fixed.
root# ls -l /etc/samba/drivers/W32X86/
total 1
drwxr-sr-x 2 root ntadmin 532 May 25 23:22 2
drwxr-sr-x 2 root ntadmin 670 May 16 03:15 3
root# ls -l /etc/samba/drivers/W32X86/2
total 5039
[....]
-rwxr--r-- 1 root ntadmin 14234 May 25 23:21 cups.hlp
-rwxr--r-- 1 root ntadmin 278380 May 13 13:53 cupsdrvr.dll
-rwxr--r-- 1 root ntadmin 215848 May 13 13:53 cupsui.dll
-rwxr--r-- 1 root ntadmin 169458 May 25 23:21 mysmbtstprn.PPD
Notice how step 6 also moved the driver files to the appropriate sub-
directory. Compare this with the situation after step 5.
Remember, this command greps for the name you chose for the driver
in step 6. This command must succeed before you can proceed.
9. Tell Samba which printer should use these driver files (setdriver).
Since you can bind any printer name (print queue) to any driver, this is
a convenient way to set up many queues that use the same driver. You
do not need to repeat all the previous steps for the setdriver command
to succeed. The only preconditions are that enumdrivers must find
the driver and enumprinters must find the printer.
10. (Optional) Verify if Samba has recognized this association.
drivername:[mydrivername]
comment:[mysmbtstprn]
location:[]
sepfile:[]
printprocessor:[winprint]
Compare these results with the ones from steps 2 and 3. Every one of
these commands show the driver is installed. Even the enumprinters
command now lists the driver on the “description” line.
11. (Optional.) Tickle the driver into a correct device mode. You
certainly know how to install the driver on the client. In case you are
not particularly familiar with Windows, here is a short recipe: Browse
the Network Neighborhood, go to the Samba server, and look for the
shares. You should see all shared Samba printers. Double-click on
the one in question. The driver should get installed and the network
connection set up. Another way is to open the Printers (and Faxes)
folder, right-click on the printer in question, and select Connect or
490 CUPS Printing Support Chapter 21
Then hit [TAB] five times, [ENTER] twice, [TAB] once, and [ENTER]
again, and march to the printer.
14. (Recommended.) Study the test page. Hmmm. Just kidding!
By now you know everything about printer installations and you do not
need to read a word. Just put it in a frame and bolt it to the wall with
the heading ”MY FIRST RPCCLIENT-INSTALLED PRINTER” —
why not just throw it away!
15. (Obligatory.) Enjoy. Jump. Celebrate your success.
The setdriver command will fail if in Samba’s mind the queue is not already
there. A successful installation displys the promising message that the:
following the adddriver parts of the procedure. But you may also see a
disappointing message like this one: result was NT STATUS UNSUCCESSFUL
It is not good enough that you can see the queue in CUPS, using the lpstat
-p ir85wm command. A bug in most recent versions of Samba prevents the
proper update of the queue list. The recognition of newly installed CUPS
printers fails unless you restart Samba or send a HUP to all smbd processes.
To verify if this is the reason why Samba does not execute the setdriver
command successfully, check if Samba “sees” the printer:
By the way, you can use these commands, plus a few more, of course, to
install drivers on remote Windows NT print servers too!
Some mystery is associated with the series of files with a tdb suffix appear-
ing in every Samba installation. They are connections.tdb, printing.tdb,
492 CUPS Printing Support Chapter 21
*.tdb files are not human readable. They are written in a binary format.
“Why not ASCII?”, you may ask. “After all, ASCII configuration files are
a good and proven tradition on UNIX.” The reason for this design decision
by the Samba Team is mainly performance. Samba needs to be fast; it runs
a separate smbd process for each client connection, in some environments
many thousands of them. Some of these smbds might need to write-access
the same *.tdb file at the same time. The file format of Samba’s *.tdb files
allows for this provision. Many smbd processes may write to the same *.
tdb file at the same time. This wouldn’t be possible with pure ASCII files.
It is very important that all *.tdb files remain consistent over all write and
read accesses. However, it may happen that these files do get corrupted.
(A kill -9 ‘pidof smbd’ while a write access is in progress could do the
damage, as could a power interruption, etc.). In cases of trouble, a deletion
of the old printing-related *.tdb files may be the only option. After that,
you need to re-create all print-related setups unless you have made a backup
of the *.tdb files in time.
Section 21.13. CUPS Print Drivers from Linuxprinting.org 493
Samba ships with a little utility that helps the root user of your system to
backup your *.tdb files. If you run it with no argument, it prints a usage
message:
root# tdbbackup
Usage: tdbbackup [options] <fname...>
Version:3.0a
-h this help message
-s suffix set the backup suffix
-v verify mode (restore if corrupt)
root# ls
. browse.dat locking.tdb ntdrivers.tdb printing.tdb
.. share_info.tdb connections.tdb messages.tdb ntforms.tdb
printing.tdbkp unexpected.tdb brlock.tdb gmon.out namelist.debug
ntprinters.tdb sessionid.tdb
root# ls -l printing.tdb*
-rw------- 1 root root 40960 May 2 03:44 printing.tdb
-rw------- 1 root root 40960 May 2 03:44 printing.tdb.bak
CUPS ships with good support for HP LaserJet-type printers. You can
install the generic driver as follows:
The -m switch will retrieve the laserjet.ppd from the standard repository
for not-yet-installed PPDs, which CUPS typically stores in /usr/share/
cups/model. Alternatively, you may use -P /path/to/your.ppd.
The generic laserjet.ppd, however, does not support every special option
for every LaserJet-compatible model. It constitutes a sort of “least common
denominator” of all the models. If for some reason you must pay for the
commercially available ESP Print Pro drivers, your first move should be to
consult the database on the Linuxprinting15 Web site. Linuxprinting.org has
excellent recommendations about which driver is best used for each printer.
Its database is kept current by the tireless work of Till Kamppeter from
Mandrakesoft, who is also the principal author of the foomatic-rip utility.
Note
a
<http://www.linuxprinting.org/driver_list.cgi>
Nowadays, most Linux distributions rely on the utilities from the Linux-
printing.org16 to create their printing-related software (which, by the way,
15
<http://www.linuxprinting.org/printer_list.cgi>
16
<http://www.linuxprinting.org/>
Section 21.13. CUPS Print Drivers from Linuxprinting.org 495
works on all UNIXes and on Mac OS X and Darwin, too). The utilities
from this sire have a very end-user-friendly interface that allows for an easy
update of drivers and PPDs for all supported models, all spoolers, all oper-
ating systems, and all package formats (because there is none). Its history
goes back a few years.
Recently, Foomatic has achieved the astonishing milestone of 1,000 listed17
printer models. Linuxprinting.org keeps all the important facts about printer
drivers, supported models, and which options are available for the various
driver/printer combinations in its Foomatic18 database. Currently there are
245 drivers19 in the database. Many drivers support various models, and
many models may be driven by different drivers — its your choice!
At present, there are 690 devices dubbed as working perfectly: 181 are
mostly perfect, 96 are partially perfect, and 46 are paperweights. Keeping
in mind that most of these are non-PostScript models (PostScript print-
ers are automatically supported by CUPS to perfection by using their own
manufacturer-provided Windows PPD), and that a multifunctional device
never qualifies as working perfectly if it does not also scan and copy and fax
under GNU/Linux — then this is a truly astonishing achievement! Three
years ago the number was not more than 500, and Linux or UNIX printing
at the time wasn’t anywhere near the quality it is today.
A few years ago Grant Taylor20 started it all. The roots of today’s Linux-
printing.org are in the first Linux Printing HOWTO21 that he authored. As
a side-project to this document, which served many Linux users and admins
to guide their first steps in this complicated and delicate setup (to a scientist,
printing is “applying a structured deposition of distinct patterns of ink or
toner particles on paper substrates”), he started to build in a little Postgres
database with information about the hardware and driver zoo that made
up Linux printing of the time. This database became the core component
17
<http://www.linuxprinting.org/printer_list.cgi?make=Anyone>
18
<http://www.linuxprinting.org/foomatic.html>
19
<http://www.linuxprinting.org/driver_list.cgi>
20
<http://www2.picante.com:81/~gtaylor/>
21
<http://www.linuxprinting.org/foomatic2.9/howto/>
496 CUPS Printing Support Chapter 21
“Why the funny name?” you ask. When it really took off, around spring
2000, CUPS was far less popular than today, and most systems used LPD,
LPRng, or even PDQ to print. CUPS shipped with a few generic drivers
(good for a few hundred different printer models). These didn’t support
many device-specific options. CUPS also shipped with its own built-in ras-
terization filter (pstoraster, derived from Ghostscript). On the other hand,
CUPS provided brilliant support for controlling all printer options through
standardized and well-defined PPD files. Plus, CUPS was designed to be
easily extensible.
Taylor already had in his database a respectable compilation of facts about
many more printers and the Ghostscript “drivers” they run with. His idea,
to generate PPDs from the database information and use them to make
standard Ghostscript filters work within CUPS, proved to work very well.
It also killed several birds with one stone:
• It made all current and future Ghostscript filter developments available
for CUPS.
• It made available a lot of additional printer models to CUPS users
(because often the traditional Ghostscript way of printing was the
only one available).
• It gave all the advanced CUPS options (Web interface, GUI driver
configurations) to users wanting (or needing) to use Ghostscript filters.
and the respective printer/filter combo, another Perl script named CUPS-O-
Matic did the PPD generation. After that was working, Taylor implemented
within a few days a similar thing for two other spoolers. Names chosen for
the config-generator scripts were PDQ-O-Matic23 (for PDQ) and LPD-O-
Matic24 (for — you guessed it — LPD); the configuration here didn’t use
PPDs but other spooler-specific files.
From late summer of that year, Till Kamppeter25 started to put work into
the database. Kamppeter had been newly employed by Mandrakesoft26 to
convert its printing system over to CUPS, after they had seen his FLTK27 -
based XPP28 (a GUI front-end to the CUPS lp-command). He added a huge
amount of new information and new printers. He also developed the support
for other spoolers, like PPR29 (via ppromatic), GNUlpr30 , and LPRng31
(both via an extended lpdomatic) and spooler-less printing (directomatic32 ).
So, to answer your question, “Foomatic” is the general name for all the
overlapping code and data behind the “*omatic” scripts. Foomatic, up to
versions 2.0.x, required (ugly) Perl data structures attached to Linuxprint-
ing.org PPDs for CUPS. It had a different “*omatic” script for every spooler,
as well as different printer configuration files.
This has all changed in Foomatic versions 2.9 (beta) and released as “stable”
3.0. It has now achieved the convergence of all *omatic scripts and is called
the foomatic-rip33 . This single script is the unification of the previously
different spooler-specific *omatic scripts. foomatic-rip is used by all the
different spoolers alike, and because it can read PPDs (both the original
PostScript printer PPDs and the Linuxprinting.org-generated ones), all of a
23
<http://www.linuxprinting.org/download.cgi?filename=lpdomatic&show=0>
24
<http://www.linuxprinting.org/download.cgi?filename=lpdomatic&show=0>
25
<http://www.linuxprinting.org/till/>
26
<http://www.mandrakesoft.com/>
27
<http://www.fltk.org/>
28
<http://cups.sourceforge.net/xpp/>
29
<http://ppr.sourceforge.net/>
30
<http://sourceforge.net/projects/lpr/>
31
<http://www.lprng.org/>
32
<http://www.linuxprinting.org/download.cgi?filename=directomatic&show=
0>
33
<http://www.linuxprinting.org/foomatic2.9/download.cgi?filename=
foomatic-rip&show=0>
498 CUPS Printing Support Chapter 21
sudden all supported spoolers can have the power of PPDs at their disposal.
Users only need to plug foomatic-rip into their system. For users there is
improved media type and source support — paper sizes and trays are easier
to configure.
Also, the new generation of Linuxprinting.org PPDs no longer contains Perl
data structures. If you are a distro maintainer and have used the previous
version of Foomatic, you may want to give the new one a spin, but remember
to generate a new-version set of PPDs via the new foomatic-db-engine!34 .
Individual users just need to generate a single new PPD specific to their
model by following the steps35 outlined in the Foomatic tutorial or in this
chapter. This new development is truly amazing.
foomatic-rip is a very clever wrapper around the need to run Ghostscript
with a different syntax, options, device selections, and/or filters for each
different printer or spooler. At the same time, it can read the PPD associated
with a print queue and modify the print job according to the user selections.
Together with this comes the 100% compliance of the new Foomatic PPDs
with the Adobe spec. Some innovative features of the Foomatic concept
may surprise users. It will support custom paper sizes for many printers
and will support printing on media drawn from different paper trays within
the same job (in both cases, even where there is no support for this from
Windows-based vendor printer drivers).
37
<http://hpinkjet.sf.net/>
38
<http://gimp-print.sf.net/>
39
<http://www.linuxprinting.org//kpfeifle/LinuxKongress2002/Tutorial/>
40
<http://www.linuxprinting.org/newsportal/>
41
<http://www.linuxprinting.org/macosx/>
42
<http://www.linuxprinting.org/newsportal/thread.php3?name=linuxprinting.
macosx.general>
500 CUPS Printing Support Chapter 21
Note
43
<http://printing.kde.org/overview/kprinter.phtml>
44
<http://gtklp.sourceforge.net/>
Section 21.13. CUPS Print Drivers from Linuxprinting.org 501
or “spoolerless” printing53 .
• You can view the PPD in your browser through this link: <http://
www.linuxprinting.org/ppd-o-matic.cgi?driver=ljet4&printer=
HP-LaserJet_4_Plus&show=1>
• Most importantly, you can also generate and download the PPD54 .
• The PPD contains all the information needed to use our model and
the driver; once installed, this works transparently for the user. Later
you’ll only need to choose resolution, paper size, and so on, from the
Web-based menu, or from the print dialog GUI, or from the command
line.
• If you ended up on the drivers page55 , you can choose to use the
“PPD-O-Matic” online PPD generator program.
• Select the exact model and check either Download or Display PPD
file and click Generate PPD file.
• If you save the PPD file from the browser view, please do not use
cut and paste (since it could possibly damage line endings and tabs,
which makes the PPD likely to fail its duty), but use Save as... in
your browser’s menu. (It is best to use the Download option directly
from the Web page.)
• Another interesting part on each driver page is the Show execution
details button. If you select your printer model and click on that but-
ton, a complete Ghostscript command line will be displayed, enumer-
ating all options available for that combination of driver and printer
model. This is a great way to “learn Ghostscript by doing”. It is also
an excellent cheat sheet for all experienced users who need to recon-
struct a good command line for that darned printing script, but can’t
remember the exact syntax.
• Sometime during your visit to Linuxprinting.org, save the PPD to a
suitable place on your hard disk, say /path/to/my-printer.ppd (if
you prefer to install your printers with the help of the CUPS Web
interface, save the PPD to the /usr/share/cups/model/ path and
restart cupsd).
53
<http://www.linuxprinting.org/direct-doc.html>
54
<http://www.linuxprinting.org/ppd-o-matic.cgi?driver=ljet4&printer=
HP-LaserJet_4_Plus&show=0>
55
<http://www.linuxprinting.org/show_driver.cgi?driver=ljet4>
Section 21.13. CUPS Print Drivers from Linuxprinting.org 503
• Then install the printer with a suitable command line, like this:
CUPS will discover this new available filter at startup after restarting
cupsd.
Once you print to a print queue set up with the Foomatic PPD, CUPS will in-
sert the appropriate commands and comments into the resulting PostScript
job file. foomatic-rip is able to read and act upon these and uses some
specially encoded Foomatic comments embedded in the job file. These in
turn are used to construct (transparently for you, the user) the complicated
Ghostscript command line telling the printer driver exactly how the result-
ing raster data should look and which printer commands to embed into the
data stream. You need:
• A “foomatic+something” PPD — but this is not enough to print with
CUPS (it is only one important component).
• The foomatic-rip filter script (Perl) in /usr/lib/cups/filters/.
56
<http://www.linuxprinting.org/foomatic2.9/download.cgi?filename=
foomatic-rip&show=1>
504 CUPS Printing Support Chapter 21
Often there are questions regarding print quotas where Samba users (that
is, Windows clients) should not be able to print beyond a certain number of
pages or data volume per day, week, or month. This feature is dependent
on the real print subsystem you’re using. Samba’s part is always to receive
the job files from the clients (filtered or unfiltered) and hand them over to
this printing subsystem.
Of course one could hack things with one’s own scripts. But then there is
CUPS. CUPS supports quotas that can be based on the size of jobs or on
the number of pages or both, and can span any time period you want.
This is an example command of how root would set a print quota in CUPS,
assuming an existing printer named “quotaprinter”:
This would limit every single user to print no more than 100 pages or 1024
KB of data (whichever comes first) within the last 604,800 seconds ( = 1
week).
Section 21.14. Page Accounting with CUPS 505
For CUPS to count correctly, the printfile needs to pass the CUPS pstops
filter; otherwise it uses a dummy count of “one”. Some print files do not pass
it (e.g., image files), but then those are mostly one-page jobs anyway. This
also means that proprietary drivers for the target printer running on the
client computers and CUPS/Samba, which then spool these files as “raw”
(i.e., leaving them untouched, not filtering them), will be counted as one-
pagers too!
You need to send PostScript from the clients (i.e., run a PostScript driver
there) to have the chance to get accounting done. If the printer is a non-
PostScript model, you need to let CUPS do the job to convert the file to
a print-ready format for the target printer. This is currently working for
about a thousand different printer models. Linuxprinting.org has a driver
list57 .
Before CUPS 1.1.16, your only option was to use the Adobe PostScript
driver on the Windows clients. The output of this driver was not always
passed through the pstops filter on the CUPS/Samba side, and therefore
was not counted correctly (the reason is that it often, depending on the
PPD being used, wrote a PJL-header in front of the real PostScript, which
caused CUPS to skip pstops and go directly to the pstoraster stage).
From CUPS 1.1.16 onward, you can use the CUPS PostScript driver for
Windows NT/200x/XP clients (which is tagged in the download area of
http://www.cups.org/ as the cups-samba-1.1.16.tar.gz package). It
does not work for Windows 9x/Me clients, but it guarantees:
• To not write a PJL-header.
• To still read and support all PJL-options named in the driver PPD
with its own means.
• That the file will pass through the pstops filter on the CUPS/Samba
server.
• To page-count correctly the print file.
57
<http://www.linuxprinting.org/printer_list.cgi>
506 CUPS Printing Support Chapter 21
You can read more about the setup of this combination in the man page for
cupsaddsmb (which is only present with CUPS installed, and only current
from CUPS 1.1.16).
These are the items CUPS logs in the page log for every page of a job:
• Printer name
• User name
• Job ID
• Time of printing
• Page number
• Number of copies
• A billing information string (optional)
• The host that sent the job (included since version 1.1.19)
Here is an extract of my CUPS server’s page log file to illustrate the format
and included items:
This was job ID 401, printed on tec IS2027 by user kurt, a 64-page job
printed in three copies, billed to #marketing, and sent from IP address
10.160.50.13. The next job had ID 402, was sent by user boss from IP
address 10.160.51.33, printed from one page 440 copies, and is set to be
billed to finance-dep.
• The ones named above (wrongly logged job in case of printer hardware
failure, and so on).
• In reality, CUPS counts the job pages that are being processed in
software (that is, going through the RIP) rather than the physical
sheets successfully leaving the printing device. Thus, if there is a jam
while printing the fifth sheet out of 1,000 and the job is aborted by
the printer, the page count will still show the figure of 1,000 for that
job.
• All quotas are the same for all users (no flexibility to give the boss a
higher quota than the clerk) and no support for groups.
• A user being denied a job because of a filled-up quota does not get
a meaningful error message from CUPS other than “client-error-not-
possible”.
This is the best system currently available, and there are huge improvements
under development for CUPS 1.2:
• Page counting will go into the backends (these talk directly to the
printer and will increase the count in sync with the actual printing
process; thus, a jam at the fifth sheet will lead to a stop in the count-
ing).
• Probably there will be support for users to inquire about their accounts
in advance.
• Probably there will be support for some other tools around this topic.
508 CUPS Printing Support Chapter 21
cupsomatic/Foomatic are not native CUPS drivers and they do not ship
with CUPS. They are a third-party add-on developed at Linuxprinting.org.
As such, they are a brilliant hack to make all models (driven by Ghostscript
drivers/filters in traditional spoolers) also work via CUPS, with the same
(good or bad!) quality as in these other spoolers. cupsomatic is only a
vehicle to execute a Ghostscript command line at that stage in the CUPS
filtering chain where normally the native CUPS pstoraster filter would kick
in. cupsomatic bypasses pstoraster, kidnaps the print file from CUPS,
and redirects it to go through Ghostscript. CUPS accepts this because the
associated cupsomatic/foomatic-PPD specifies:
This line persuades CUPS to hand the file to cupsomatic once it has success-
fully converted it to the MIME type application/vnd.cups-postscript.
This conversion will not happen for jobs arriving from Windows that are au-
totyped application/octet-stream, with the according changes in /etc/
cups/mime.types in place.
CUPS is widely configurable and flexible, even regarding its filtering mech-
anism. Another workaround in some situations would be to have in /etc/
cups/mime.types entries as follows:
Section 21.15. Additional Material 509
application/postscript application/vnd.cups-raw 0 -
application/vnd.cups-postscript application/vnd.cups-raw 0 -
This would prevent all PostScript files from being filtered (rather, they will
through the virtual nullfilter denoted with “-”). This could only be useful for
PostScript printers. If you want to print PostScript code on non-PostScript
printers (provided they support ASCII text printing), an entry as follows
could be useful:
*/* application/vnd.cups-raw 0 -
and would effectively send all files to the backend without further processing.
application/vnd.cups-postscript application/vnd.cups-raw 0 \
my_PJL_stripping_filter
You will need to write a my PJL stripping filter (which could be a shell
script) that parses the PostScript and removes the unwanted PJL. This needs
to conform to CUPS filter design (mainly, receive and pass the parameters
printername, job-id, username, jobtitle, copies, print options, and possi-
bly the filename). It is installed as world executable into /usr/lib/cups/
filters/ and is called by CUPS if it encounters a MIME type application/vnd.
cups-postscript.
Samba print files pass through two spool directories. One is the incoming
directory managed by Samba (set in the path = /var/spool/samba directive
in the [printers] section of smb.conf). The other is the spool directory of
your UNIX print subsystem. For CUPS it is normally /var/spool/cups/,
as set by the cupsd.conf directive RequestRoot /var/spool/cups.
PreserveJobFiles Yes This keeps the job files themselves in cupsd’s mind
(it keeps the d12345, d12346, etc., files in the CUPS spool directory).
This is set to “No” as the CUPS default.
“MaxJobs 500” This directive controls the maximum number of jobs that
are kept in memory. Once the number of jobs reaches the limit, the
oldest completed job is automatically purged from the system to make
room for the new one. If all of the known jobs are still pending or
active, then the new job will be rejected. Setting the maximum to 0
disables this functionality. The default setting is 0.
(There are also additional settings for MaxJobsPerUser and MaxJobsPer-
Printer.)
21.16.2 Preconditions
Note
From time to time the question arises, how can you print to a Windows-
attached printer from Samba? Normally the local connection from Windows
host to printer would be done by USB or parallel cable, but this does not
matter to Samba. From here only an SMB connection needs to be opened to
the Windows host. Of course, this printer must be shared first. As you have
learned by now, CUPS uses backends to talk to printers and other servers.
To talk to Windows shared printers, you need to use the smb (surprise,
surprise!) backend. Check if this is in the CUPS backend directory. This
usually resides in /usr/lib/cups/backend/. You need to find an smb file
there. It should be a symlink to smbspool, and the file must exist and be
executable:
root# ls -l /usr/lib/cups/backend/
total 253
drwxr-xr-x 3 root root 720 Apr 30 19:04 .
512 CUPS Printing Support Chapter 21
smbspool was written by Mike Sweet from the CUPS folks. It is included
and ships with Samba. It may also be used with print subsystems other
than CUPS, to spool jobs to Windows printer shares. To set up printer
winprinter on CUPS, you need to have a driver for it. Essentially this
means to convert the print data on the CUPS/Samba host to a format that
the printer can digest (the Windows host is unable to convert any files you
may send). This also means you should be able to print to the printer if
it were hooked directly at your Samba/CUPS host. For troubleshooting
purposes, this is what you should do to determine if that part of the process
chain is in order. Then proceed to fix the network connection/authentication
to the Windows host, and so on.
To install a printer with the smb backend on CUPS, use this command:
The PPD must be able to direct CUPS to generate the print data for the
target model. For PostScript printers, just use the PPD that would be used
with the Windows NT PostScript driver. But what can you do if the printer
is only accessible with a password? Or if the printer’s host is part of another
workgroup? This is provided for: You can include the required parameters
as part of the smb:// device-URI like this:
• smb://WORKGROUP/WINDOWSNETBIOSNAME/printersharename
• smb://username:password@WORKGROUP/WINDOWSNETBIOSNAME/printersharename
• smb://username:password@WINDOWSNETBIOSNAME/printersharename
Note that the device URI will be visible in the process list of the Samba
server (e.g., when someone uses the ps -aux command on Linux), even if
the username and passwords are sanitized before they get written into the
log files. This is an inherently insecure option; however, it is the only one.
Don’t use it if you want to protect your passwords. Better share the printer
in a way that does not require a password! Printing will only work if you
have a working NetBIOS name resolution up and running. Note that this is
a feature of CUPS and you do not necessarily need to have smbd running.
The diagrams in Figure 21.17 and Figure 21.18 show how CUPS handles
print jobs.
For Windows 9x/Me, clients require the printer names to be eight characters
(or “8 plus 3 chars suffix”) max; otherwise, the driver files will not get
transferred when you want to download them from Samba.
Have you set security = user? Have you used smbpasswd to give root
a Samba account? You can do two things: open another terminal and
514 CUPS Printing Support Chapter 21
CUPS in and of itself has this (general) filter chain (italic letters
are file-formats or MIME types, other are filters (this is
true for pre-1.1.15 of pre-4.3 versions of CUPS and ESP PrintPro):
something-fileformat
something-device-specific backend
somethingtops
application/vnd.cups-raster
pstops
as shipped with CUPS,
independent from any
GhostScript installation on the
application/vnd.cups-postscript pstoraster
system
(= "postscript interpreter")
ESP PrintPro has some enhanced "rastertosomething" filters as compared to
CUPS, and also a somewhat improved "pstoraster" filter.
The use of “cupsaddsmb” gives “No PPD file for printer...” message while
PPD file is present. What might the problem be?
Have you enabled printer sharing on CUPS? This means, do you have a <Lo-
cation /printers>....</Location> section in CUPS server’s cupsd.
conf that does not deny access to the host you run “cupsaddsmb” from?
It could be an issue if you use cupsaddsmb remotely, or if you use it with
a -h parameter: cupsaddsmb -H sambaserver -h cupsserver -v print-
ername.
Use smbstatus to check which user you are from Samba’s point of view.
Do you have the privileges to write into the [print$] share?
Once you are connected as the wrong user (for example, as nobody, which
often occurs if you have map to guest = bad user), Windows Explorer will
not accept an attempt to connect again as a different user. There will
not be any bytes transferred on the wire to Samba, but still you’ll see a
stupid error message that makes you think Samba has denied access. Use
smbstatus to check for active connections. Kill the PIDs. You still can’t
re-connect, and you get the dreaded You can’t connect with a second
account from the same machine message as soon as you try. And you do
not see a single byte arriving at Samba (see logs; use “ethereal”) indicating a
renewed connection attempt. Shut all Explorer Windows. This makes Win-
dows forget what it has cached in its memory as established connections.
Then reconnect as the right user. The best method is to use a DOS terminal
window and first do net use z: \\GANDALF\print$ /user:root. Check
with smbstatus that you are connected under a different account. Now
open the Printers folder (on the Samba server in the Network Neighbor-
hood), right-click on the printer in question, and select Connect.....
516 CUPS Printing Support Chapter 21
You see per smbstatus that you are connected as user nobody, but you
want to be root or printer admin. This is probably due to map to guest =
bad user, which silently connected you under the guest account when you
gave (maybe by accident) an incorrect username. Remove map to guest if
you want to prevent this.
This information came from a mailing list posting regarding problems expe-
rienced when upgrading from Adobe drivers to CUPS drivers on Microsoft
Windows NT/200x/XP clients.
First delete all old Adobe-using printers. Then delete all old Adobe drivers.
(On Windows 200x/XP, right-click in the background of Printers folder,
select Server Properties..., select tab Drivers, and delete here).
Do you use the “naked” root user name? Try to do it this way: cupsaddsmb
-U DOMAINNAME\\root -v printername> (note the two backslashes: the
first one is required to “escape” the second one).
Deleting a printer on the client will not delete the driver too (to verify,
right-click on the white background of the Printers folder, select Server
Properties and click on the Drivers tab). These same old drivers will be
re-used when you try to install a printer with the same name. If you want
to update to a new driver, delete the old ones first. Deletion is only possible
if no other printer uses the same driver.
Local security policies may not allow the installation of unsigned drivers —
“local security policies” may not allow the installation of printer drivers at
Section 21.19. Common Errors 517
all.
For print change, notify functions on NT++ clients. These need to run the
Server service first (renamed to File & Print Sharing for MS Networks
in XP).
Win XP-SP1 introduced a Point and Print Restriction Policy (this restric-
tion does not apply to “Administrator” or “Power User” groups of users).
In Group Policy Object Editor, go to User Configuration -> Administra-
tive Templates -> Control Panel -> Printers. The policy is automatically
set to Enabled and the Users can only Point and Print to machines
in their Forest . You probably need to change it to Disabled or Users
can only Point and Print to these servers to make driver downloads
from Samba possible.
How are you doing it? I bet the wrong way (it is not easy to find out,
though). There are three different ways to bring you to a dialog that seems
to set everything. All three dialogs look the same, yet only one of them does
what you intend. You need to be Administrator or Print Administrator to
do this for all users. Here is how I do in on XP:
A The first wrong way: .
518 CUPS Printing Support Chapter 21
Don’t use Optimize for Speed, but use Optimize for Portability in-
stead (Adobe PS Driver). Don’t use Page Independence: No. Always
Section 21.19. Common Errors 519
Have you ever by accident set the CUPS spool directory to the same location
(RequestRoot /var/spool/samba/ in cupsd.conf or the other way round:
/var/spool/cups/ is set as path> in the [printers] section)? These must be
different. Set RequestRoot /var/spool/cups/ in cupsd.conf and path =
/var/spool/samba in the [printers] section of smb.conf. Otherwise, cupsd
will sanitize permissions to its spool directory with each restart and printing
will not work reliably.
In this case a print queue called “lp” intermittently swallows jobs and spits
out completely different ones from what was sent.
It is a bad idea to name any printer “lp”. This is the traditional UNIX name
for the default printer. CUPS may be set up to do an automatic creation of
Implicit Classes. This means, to group all printers with the same name to a
pool of devices and load-balance the jobs across them in a round-robin fash-
ion. Chances are high that someone else has a printer named “lp” too. You
may receive that person’s jobs and send your own to his or her device unwit-
tingly. To have tight control over the printer names, set BrowseShortNames
520 CUPS Printing Support Chapter 21
something-fileformat
somethingtops
application/postscript
(constructs complicated
pstops
Ghostscript commandline
to let the file be processed by a
"-sDEVICE-s.th." call...)
application/vnd.cups-postscript cupsomatic
pstoraster
(= "postscript interpreter")
application/vnd.cups-raster
rastertosomething
(= "raster driver")
Ghostscript at work....
backend
Stackable VFS (Virtual File System) modules support was new to Samba-
3 and has proven quite popular. Samba passes each request to access the
UNIX file system through the loaded VFS modules. This chapter covers the
modules that come with the Samba source and provides references to some
external modules.
22.2 Discussion
If not supplied with your platform distribution binary Samba package you
may have problems compiling these modules, as shared libraries are compiled
and linked in different ways on different systems. They currently have been
tested against GNU/Linux and IRIX.
To use the VFS modules, create a share similar to the one below. The
important parameter is the vfs objects parameter where you can list one or
more VFS modules by name. For example, to log all access to files and put
deleted files in a recycle bin, see Example 22.2.1:
The modules are used in the order in which they are specified. Let’s say that
you want to both have a virus scanner module and a recycle bin module.
It is wise to put the virus scanner module as the first one so that it is the
first that get run an may detect a virus immediately, before any action is
performed on that file. vfs objects = vscan-clamav recycle
Samba will attempt to load modules from the /lib directory in the root
directory of the Samba installation (usually /usr/lib/samba/vfs or /usr/
523
524 Stackable VFS modules Chapter 22
local/samba/lib/vfs).
Some modules can be used twice for the same share. This can be done using
a configuration similar to the one shown in Example 22.2.2.
22.3.1 audit
A simple module to audit file access to the syslog facility. The following
operations are logged:
• share
• connect/disconnect
Section 22.3. Included Modules 525
• directory opens/create/remove
• file open/close/rename/unlink/chmod
This module is identical with the audit module above except that it sends
audit logs to both syslog as well as the smbd log files. The log level for this
module is set in the smb.conf file.
Valid settings and the information that will be recorded are shown in Ta-
ble 22.1.
This auditing tool is more felxible than most people readily will recognize.
There are a number of ways by which useful logging information can be
recorded.
• Syslog can be used to record all transaction. This can be disabled by
setting in the smb.conf file syslog = 0.
• Logging can take place to the default log file (log.smbd) for all loaded
VFS modules just by setting in the smb.conf file log level = 0
vfs:x, where x is the log level. This will disable general logging while
activating all logging of VFS module activity at the log level specified.
• Detailed logging can be obtained per user, per client machine, etc.
This requires the above together with the creative use of the log file
settings.
An example of detailed per-user and per-machine logging can be ob-
tained by setting log file = /var/log/samba/%U.%m.log.
526 Stackable VFS modules Chapter 22
Auditing information often must be preserved for a long time. So that the
log files do not get rotated it is essential that the max log size = 0 be set in
the smb.conf file.
This module was created to allow Roaming Profile files and directories to
be set (on the Samba server under UNIX) as read only. This module will, if
installed on the Profiles share, report to the client that the Profile files and
directories are writeable. This satisfies the client even though the files will
never be overwritten as the client logs out or shuts down.
22.3.4 recycle
A Recycle Bin-like module. Where used, unlink calls will be intercepted and
files moved to the recycle directory instead of being deleted. This gives the
same effect as the Recycle Bin on Windows computers.
The Recycle Bin will not appear in Windows Explorer views of the network
file system (share) nor on any mapped drive. Instead, a directory called .
recycle will be automatically created when the first file is deleted. Users
can recover files from the .recycle directory. If the recycle:keeptree has
been specified, deleted files will be found in a path identical with that from
which the file was deleted.
Supported options for the recycle module are as follow:
recycle:versions If this option is set, two files with the same name that
are deleted will both be kept in the recycle bin. Newer deleted versions
of a file will be called “Copy #x of filename”.
Section 22.3. Included Modules 527
recycle:maxsize Files that are larger than the number of bytes specified
by this parameter will not be put into the recycle bin.
recycle:exclude List of files that should not be put into the recycle bin
when deleted, but deleted in the regular way.
22.3.5 netatalk
A netatalk module will ease co-existence of Samba and netatalk file sharing
services.
• Does not care about creating .AppleDouble forks, just keeps them in
sync.
Warning
The shadow copy module allows you to setup functionality that is similar
to MS shadow copy services. When setup properly, this module allows
Microsoft shadow copy clients to browse ”shadow copies” on samba shares.
You will need to install the shadow copy client. You can get the MS shadow
copy client here.1 . Note the additional requirements for pre-Windows XP
clients. I did not test this functionality with any pre-Windows XP clients.
You should be able to get more information about MS Shadow Copy from
the Microsoft’s site2 .
The shadow copy VFS module requires some underlying file system setup
with some sort of Logical Volume Manager (LVM) such as LVM1, LVM2, or
EVMS. Setting up LVM is beyond the scope of this document; however, we
will outline the steps we took to test this functionality for example purposes
only. You need to make sure the LVM implementation you choose to deploy
is ready for production. Make sure you do plenty of tests.
• See Learning Linux LVM, Part 16 and Learning Linux LWM, Part 27
for Daniel Robbins’ well written a two part tutorial on Linux and LVM
using LVM source code and reiserfs.
At the time of this writing, not much testing has been done. I tested the
shadow copy VFS module with a specific scenario which was not deployed
in a production environment, but more as a proof of concept. The scenario
involved a Samba 3 file server on Debian Sarge with an XFS file system and
LVM1. I do NOT recommend you use this as a solution without doing your
own due diligence with regard to all the components presented here. That
said, following is an basic outline of how I got things going.
1. Installed Operating System. In my tests, I used Debian Sarge8
(i.e. testing) on an XFS file system. Setting up the OS is a bit beyond
the scope of this document. It is assumed that you have a working OS
capable of running Samba.
2. Install & Configure Samba. See the Part I of this HOWTO for
more detail on this. It doesn’t matter if it is a Domain Controller or
Member File Server, but it is assumed that you have a working Samba
3.0.3 or newer server running.
3. Install & Configure LVM. Before you can make shadow copies
available to the client, you have to create the shadow copies. This
is done by taking some sort of file system snapshot. Snapshots are a
typical feature of Logical Volume Managers such as LVM, so we first
need to have that setup.
The following is provided as an example and will be most helpful for
Debian users. Again, this was tested using the ”testing” or ”Sarge”
distribution.
• Install lvm10 and devfsd packages if you have not done so already.
On Debian systems, you are warned of the interaction of devfs
and lvm1 which requires the use of devfs filenames. Running
apt-get update && apt-get install lvm10 devfsd xfsprogs
should do the trick for this example.
6
<http://www-106.ibm.com/developerworks/linux/library/l-lvm/>
7
<http://www-106.ibm.com/developerworks/library/l-lvm2.html>
8
<http://www.debian.org/devel/debian-installer/>
530 Stackable VFS modules Chapter 22
#!/bin/bash
# This is a test, this is only a test
SNAPNAME=‘date +%Y.%m.%d-%H.%M.%S‘
xfs_freeze -f /data/shadow_share/
lvcreate -L10M -s -n $SNAPNAME /dev/shadowvol/sh_test
xfs_freeze -u /data/shadow_share/
mkdir /data/shadow_share/@GMT-$SNAPNAME
mount /dev/shadowvol/$SNAPNAME /data/shadow_share/@GMT-$SNAPNAME -onouuid,ro
532 Stackable VFS modules Chapter 22
Note that the script does not handle other things like remounting
snapshots on reboot.
6. Test From Client. To test, you will need to install the shadow copy
client which you can obtain from the Microsoft web site.9 I only tested
this with an XP client so your results may vary with other pre-XP
clients. Once installed, with your XP client you can right-click on
specific files or in the empty space of the shadow share and view the
”properties”. If anything has changed, then you will see it on the
”Previous Versions” tab of the properties window.
This section contains a listing of various other VFS modules that have been
posted but do not currently reside in the Samba CVS tree for one reason or
another (e.g., it is easy for the maintainer to have his or her own CVS tree).
No statements about the stability or functionality of any module should be
implied due to its presence here.
22.4.1 DatabaseFS
22.4.2 vscan
12
<http://www.openantivirus.org/projects.php#samba-vscan>
Chapter 23
534
Section 23.2. Introduction 535
Note
23.2 Introduction
It is well known that UNIX and Microsoft Windows NT have different mod-
els for representing user and group information and use different technologies
for implementing them. This fact has made it difficult to integrate the two
systems in a satisfactory manner.
One common solution in use today has been to create identically named user
accounts on both the UNIX and Windows systems and use the Samba suite
of programs to provide file and print services between the two. This solution
is far from perfect, however, because adding and deleting users on both sets
of machines becomes a chore, and two sets of passwords are required —
both of which can lead to synchronization problems between the UNIX and
Windows systems and confusion for users.
We divide the unified logon problem for UNIX machines into three smaller
problems:
• Obtaining Windows NT user and group information.
536 Winbind: Use of Domain Accounts Chapter 23
The term foreign SID is often met with the reaction that it is not relevant
to a particular environment. The following documents an interchange that
538 Winbind: Use of Domain Accounts Chapter 23
took place on the Samba mailing list. It is a good example of the confusion
often expressed regarding the use of winbind.
Fact: Winbind is needed to handle users who use workstations that are NOT
part of the local domain.
Response: “Why? I’ve used Samba with workstations that are not part of
my domains lots of times without using winbind. I though winbind was
for using Samba as a member server in a domain controlled by another
Samba/Windows PDC.”
If the Samba server will be accessed from a domain other than the local
Samba domain, or if there will be access from machines that are not local
domain members, winbind will permit the allocation of UIDs and GIDs from
the assigned pool that will keep the identity of the foreign user separate from
users that are members of the Samba domain.
This means that winbind is eminently useful in cases where a single Samba
PDC on a local network is combined with both domain member and non-
domain member workstations. If winbind is not used, the user george on a
Windows workstation that is not a domain member will be able to access
the files of a user called george in the account database of the Samba server
that is acting as a PDC. When winbind is used, the default condition is
that the local user george will be treated as the account DOMAIN\george
and the foreign (non-member of the domain) account will be treated as
MACHINE\george because each has a different SID.
Over the last few years, efforts have been underway by various Samba Team
members to decode various aspects of the Microsoft Remote Procedure Call
(MSRPC) system. This system is used for most network-related operations
Section 23.4. How Winbind Works 539
Since late 2001, Samba has gained the ability to interact with Microsoft
Windows 2000 using its “native mode” protocols rather than the NT4 RPC
services. Using LDAP and Kerberos, a domain member running Winbind
can enumerate users and groups in exactly the same way as a Windows 200x
client would, and in so doing provide a much more efficient and effective
Winbind implementation.
then the C library will first load a module called /lib/libnss files.so
followed by the module /lib/libnss example.so. The C library will dy-
namically load each of these modules in turn and call resolver functions
within the modules to try to resolve the request. Once the request is re-
solved, the C library returns the result to the application.
This NSS interface provides an easy way for Winbind to hook into the op-
erating system. All that needs to be done is to put libnss winbind.so
in /lib/ then add “winbind” into /etc/nsswitch.conf at the appropriate
place. The C library will then call Winbind to resolve user and group names.
check and in what order. This interface makes adding a new authentication
service for Winbind very easy: simply copy the pam winbind.so module to /
lib/security/, and the PAM control files for relevant services are updated
to allow authentication via Winbind. See the PAM documentation in Chap-
ter 27, “PAM-Based Distributed Authentication”, for more information.
An active system can generate a lot of user and group name lookups. To
reduce the network cost of these lookups, Winbind uses a caching scheme
based on the SAM sequence number supplied by NT domain controllers.
User or group information returned by a PDC is cached by Winbind along
with a sequence number also returned by the PDC. This sequence number
is incremented by Windows NT whenever any user or group information is
modified. If a cached entry has expired, the sequence number is requested
from the PDC and compared against the sequence number of the cached
entry. If the sequence numbers do not match, then the cached information
is discarded and up-to-date information is requested directly from the PDC.
542 Winbind: Use of Domain Accounts Chapter 23
23.5.1 Introduction
This section describes the procedures used to get Winbind up and run-
ning. Winbind is capable of providing access and authentication control for
Windows Domain users through an NT or Windows 200x PDC for regular
services, such as telnet and ftp, as well for Samba services.
• Why should I do this?
This allows the Samba administrator to rely on the authentication
mechanisms on the Windows NT/200x PDC for the authentication of
domain members. Windows NT/200x users no longer need to have
separate accounts on the Samba server.
• Who should be reading this document?
This document is designed for system administrators. If you are im-
plementing Samba on a file server and wish to (fairly easily) integrate
existing Windows NT/200x users from your PDC onto the Samba
server, this document is for you.
23.5.2 Requirements
If you have a Samba configuration file that you are currently using, BACK
IT UP! If your system already uses PAM, back up the /etc/pam.d directory
contents! If you haven’t already made a boot disk, MAKE ONE NOW!
Messing with the PAM configuration files can make it nearly impossible
to log in to your machine. That’s why you want to be able to boot back
into your machine in single-user mode and restore your /etc/pam.d to the
original state it was in if you get frustrated with the way things are going.
The latest version of Samba-3 includes a functioning winbindd daemon.
Please refer to the main Samba Web page1 , or better yet, your closest Samba
mirror site for instructions on downloading the source code.
To allow domain users the ability to access Samba shares and files, as well as
potentially other services provided by your Samba machine, PAM must be
set up properly on your machine. In order to compile the Winbind modules,
1
<http://samba.org/>
Section 23.5. Installation and Configuration 543
you should have at least the PAM development libraries installed on your
system. Please refer the PAM Web site <http://www.kernel.org/pub/
linux/libs/pam/>.
Before starting, it is probably best to kill off all the Samba-related daemons
running on your server. Kill off all smbd, nmbd, and winbindd processes that
may be running. To use PAM, make sure that you have the standard PAM
package that supplies the /etc/pam.d directory structure, including the
PAM modules that are used by PAM-aware services, several PAM libraries,
and the /usr/doc and /usr/man entries for PAM. Winbind is built better
in Samba if the pam-devel package is also installed. This package includes
the header files needed to compile PAM-aware applications.
Now, as root, you need to edit /etc/nsswitch.conf to allow user and group
entries to be visible from the winbindd daemon. My /etc/nsswitch.conf
file looked like this after editing:
WINBIND:
program = /usr/lib/security/WINBIND
options = authonly
Several parameters are needed in the smb.conf file to control the behavior of
winbindd. These are described in more detail in the winbindd(8) man page.
My smb.conf file, as shown in Example 23.5.1, was modified to include the
necessary entries in the [global] section.
2
<http://publibn.boulder.ibm.com/doc_link/en_US/a_doc_lib/aixprggd/
kernextc/sec_load_mod.htm>
3
<http://publibn.boulder.ibm.com/doc_link/en_US/a_doc_lib/aixbman/
baseadmn/iandaadmin.htm>
546 Winbind: Use of Domain Accounts Chapter 23
Note
Eventually, you will want to modify your Samba startup script to automati-
cally invoke the winbindd daemon when the other parts of Samba start, but
it is possible to test out just the Winbind portion first. To start up Winbind
services, enter the following command as root:
root#/usr/local/samba/sbin/winbindd
Section 23.5. Installation and Configuration 547
Note
Winbindd can now also run in “dual daemon mode”. This will make it run
as two processes. The first will answer all requests from the cache, thus
making responses to clients faster. The other will update the cache for the
query to which the first has just responded. The advantage of this is that
responses stay accurate and are faster. You can enable dual daemon mode
by adding -B to the command line:
root#/usr/local/samba/sbin/winbindd -B
I’m always paranoid and like to make sure the daemon is really running.
root#ps -ae | grep winbindd
This command should produce output like the following if the daemon is
running.
Now, for the real test, try to get some information about the users on your
PDC:
root#/usr/local/samba/bin/wbinfo -u
This should echo back a list of users on your Windows users on your PDC.
For example, I get the following response:
CEO\Administrator
CEO\burdell
CEO\Guest
CEO\jt-ad
CEO\krbtgt
548 Winbind: Use of Domain Accounts Chapter 23
CEO\TsInternetUser
You can do the same sort of thing to get group information from the PDC:
root# /usr/local/samba/bin/wbinfo -g
CEO\Domain Admins
CEO\Domain Users
CEO\Domain Guests
CEO\Domain Computers
CEO\Domain Controllers
CEO\Cert Publishers
CEO\Schema Admins
CEO\Enterprise Admins
CEO\Group Policy Creator Owners
The function getent can now be used to get unified lists of both local and
PDC users and groups. Try the following command:
root#getent passwd
You should get a list that looks like your /etc/passwd list followed by the
domain users with their new UIDs, GIDs, home directories, and default
shells.
The same thing can be done for groups with the command:
root#getent group
Linux The winbindd daemon needs to start up after the smbd and nmbd
daemons are running. To accomplish this task, you need to modify the
startup scripts of your system. They are located at /etc/init.d/smb in
Red Hat Linux and in /etc/init.d/samba in Debian Linux. Edit your
script to add commands to invoke this daemon in the proper sequence. My
startup script starts up smbd, nmbd, and winbindd from the /usr/local/
Section 23.5. Installation and Configuration 549
samba/bin directory directly. The start function in the script looks like
this:
start() {
KIND="SMB"
echo -n $"Starting $KIND services: "
daemon /usr/local/samba/bin/smbd $SMBDOPTIONS
RETVAL=$?
echo
KIND="NMB"
echo -n $"Starting $KIND services: "
daemon /usr/local/samba/bin/nmbd $NMBDOPTIONS
RETVAL2=$?
echo
KIND="Winbind"
echo -n $"Starting $KIND services: "
daemon /usr/local/samba/sbin/winbindd
RETVAL3=$?
echo
[ $RETVAL -eq 0 -a $RETVAL2 -eq 0 -a $RETVAL3 -eq 0 ] && \
touch /var/lock/subsys/smb || RETVAL=1
return $RETVAL
}
If you would like to run winbindd in dual daemon mode, replace the line:
daemon /usr/local/samba/sbin/winbindd
daemon /usr/local/samba/sbin/winbindd -B
The stop function has a corresponding entry to shut down the services and
looks like this:
550 Winbind: Use of Domain Accounts Chapter 23
stop() {
KIND="SMB"
echo -n $"Shutting down $KIND services: "
killproc smbd
RETVAL=$?
echo
KIND="NMB"
echo -n $"Shutting down $KIND services: "
killproc nmbd
RETVAL2=$?
echo
KIND="Winbind"
echo -n $"Shutting down $KIND services: "
killproc winbindd
RETVAL3=$?
[ $RETVAL -eq 0 -a $RETVAL2 -eq 0 -a $RETVAL3 -eq 0 ] && \
rm -f /var/lock/subsys/smb
echo ""
return $RETVAL
}
Solaris Winbind does not work on Solaris 9; see Section 40.6.2 for details.
On Solaris, you need to modify the /etc/init.d/samba.server startup
script. It usually only starts smbd and nmbd but should now start winbindd,
too. If you have Samba installed in /usr/local/samba/bin, the file could
contains something like this:
##
## samba.server
##
if [ ! -d /usr/bin ]
then # /usr not mounted
exit
fi
pid=‘/usr/bin/ps -e |
/usr/bin/grep -w $1 |
/usr/bin/sed -e ’s/^ *//’ -e ’s/ .*//’‘
[ "$pid" != "" ] && kill $pid
}
case "$1" in
’start’)
#
# Edit these lines to suit your installation (paths, workgroup, host)
#
echo Starting SMBD
/usr/local/samba/bin/smbd -D -s \
/usr/local/samba/smb.conf
’stop’)
killproc nmbd
killproc smbd
killproc winbindd
;;
*)
echo "Usage: /etc/init.d/samba.server { start | stop }"
;;
esac
Again, if you would like to run Samba in dual daemon mode, replace:
552 Winbind: Use of Domain Accounts Chapter 23
/usr/local/samba/sbin/winbindd
/usr/local/samba/sbin/winbindd -B
Restarting If you restart the smbd, nmbd, and winbindd daemons at this
point, you should be able to connect to the Samba server as a domain
member just as if you were a local user.
If you have made it this far, you know that winbindd and Samba are
working together. If you want to use Winbind to provide authentication for
other services, keep reading. The PAM configuration files need to be altered
in this step. (Did you remember to make backups of your original /etc/
pam.d files? If not, do it now.)
You will need a PAM module to use winbindd with these other services. This
module will be compiled in the ../source/nsswitch directory by invoking
the command:
from the ../source directory. The pam winbind.so file should be copied
to the location of your other PAM security modules. On my Red Hat sys-
tem, this was the /lib/security directory. On Solaris, the PAM security
modules reside in /usr/lib/security.
The other services that I modified to allow the use of Winbind as an authen-
tication service were the normal login on the console (or a terminal session),
telnet logins, and ftp service. In order to enable these services, you may
first need to change the entries in /etc/xinetd.d (or /etc/inetd.conf).
Red Hat Linux 7.1 and later uses the new xinetd.d structure, in this case
you need to change the lines in /etc/xinetd.d/telnet and /etc/xinetd.
d/wu-ftp from
enable = no
to
enable = yes
For ftp services to work properly, you will also need to either have individual
directories for the domain users already present on the server or change the
home directory template to a general directory for all domain users. These
can be easily set using the smb.conf global entry template homedir.
Note
The /etc/pam.d/login file can be changed in nearly the same way. It now
looks like this:
#
Section 23.5. Installation and Configuration 555
I also added a try first pass line after the winbind.so line to get rid of
annoying double prompts for passwords.
Now restart your Samba and try connecting through your application that
you configured in the pam.conf.
23.6 Conclusion
The Winbind system, through the use of the NSS, PAMs, and appropriate
Microsoft RPC calls, have allowed us to provide seamless integration of
Microsoft Windows NT domain users on a UNIX system. The result is a
great reduction in the administrative cost of running a mixed UNIX and NT
network.
• Winbind is currently only available for the Linux, Solaris, AIX, and
IRIX operating systems, although ports to other operating systems
are certainly possible. For such ports to be feasible, we require the C
library of the target operating system to support the NSS and PAM
systems. This is becoming more common as NSS and PAM gain sup-
port among UNIX vendors.
• The mappings of Windows NT RIDs to UNIX IDs is not made al-
gorithmically and depends on the order in which unmapped users or
groups are seen by Winbind. It may be difficult to recover the map-
pings of RID to UNIX ID if the file containing this information is
corrupted or destroyed.
• Currently the Winbind PAM module does not take into account pos-
sible workstation and logon time restrictions that may be set for Win-
dows NT users; this is instead up to the PDC to enforce.
Warning
root# wbinfo -u
MIDEARTH\maryo
558 Winbind: Use of Domain Accounts Chapter 23
MIDEARTH\jackb
MIDEARTH\ameds
...
MIDEARTH\root
root# wbinfo -g
MIDEARTH\Domain Users
MIDEARTH\Domain Admins
MIDEARTH\Domain Guests
...
MIDEARTH\Accounts
ADVANCED NETWORK
MANAGEMENT
This section documents peripheral issues that are of great importance to net-
work administrators who want to improve network resource access control,
to automate the user environment, and to make their lives a little easier.
This chapter presents information on each of these areas. They are placed
here, and not in other chapters, for ease of reference.
Since I do not need to buy an NT4 server, how do I get the User Manager
for Domains and the Server Manager?
559
560 Advanced Network Management Chapter 24
• Server Manager
• Event Viewer
The Windows NT 4.0 version of the User Manager for Domains and Server
Manager are available from Microsoft via ftp2 .
The following information was posted to the Samba mailing list at Apr 3
23:33:50 GMT 2003. It is presented in slightly edited form (with author
details omitted for privacy reasons). The entire answer is reproduced below
with some comments removed.
Answer provided: Check out the new offer of “NX” software from NoMa-
chine3 .
It implements an easy-to-use interface to the Remote X protocol as well as
incorporating VNC/RFB and rdesktop/RDP into it, but at a speed perfor-
mance much better than anything you may have ever seen.
Remote X is not new at all, but what they did achieve successfully is a
new way of compression and caching technologies that makes the thing fast
enough to run even over slow modem/ISDN connections.
I test drove their (public) Red Hat machine in Italy, over a loaded Inter-
net connection, with enabled thumbnail previews in KDE konqueror, which
popped up immediately on “mouse-over”. From inside that (remote X) ses-
sion I started a rdesktop session on another, a Windows XP machine. To
test the performance, I played Pinball. I am proud to announce that my
score was 631,750 points at first try.
NX performs better on my local LAN than any of the other “pure” connec-
tion methods I use from time to time: TightVNC, rdesktop or Remote X.
It is even faster than a direct crosslink connection between two nodes.
I even got sound playing from the Remote X app to my local boxes, and had
a working “copy’n’paste” from an NX window (running a KDE session in
Italy) to my Mozilla mailing agent. These guys are certainly doing something
right!
I recommend test driving NX to anybody with a only a passing interest in
remote computing the NX4 utility.
Just download the free-of-charge client software (available for Red Hat,
SuSE, Debian and Windows) and be up and running within 5 minutes (they
need to send you your account data, though, because you are assigned a real
UNIX account on their testdrive.nomachine.com box).
They plan to get to the point were you can have NX application servers
running as a cluster of nodes, and users simply start an NX session locally
and can select applications to run transparently (apps may even run on
another NX node, but pretend to be on the same as used for initial login,
because it displays in the same window. You also can run it full-screen, and
after a short time you forget that it is a remote session at all).
3
<http://www.nomachine.com/>
4
<http://www.nomachine.com/testdrive.php>
562 Advanced Network Management Chapter 24
Now the best thing for last: All the core compression and caching technolo-
gies are released under the GPL and available as source code to anybody
who wants to build on it! These technologies are working, albeit started
from the command line only (and very inconvenient to use in order to get a
fully running remote X session up and running).
To answer your questions:
• You do not need to install a terminal server; XP has RDP support
built in.
• NX is much cheaper than Citrix — and comparable in performance,
probably faster.
• You do not need to hack XP — it just works.
• You log into the XP box from remote transparently (and I think there
is no need to change anything to get a connection, even if authentica-
tion is against a domain).
• The NX core technologies are all Open Source and released under the
GPL — you can now use a (very inconvenient) command line at no
cost, but you can buy a comfortable (proprietary) NX GUI front end
for money.
• NoMachine is encouraging and offering help to OSS/Free Software
implementations for such a front-end too, even if it means competition
to them (they have written to this effect even to the LTSP, KDE, and
GNOME developer mailing lists).
There are several opportunities for creating a custom network startup con-
figuration environment.
• No Logon Script.
• Simple universal Logon Script that applies to all users.
• Use of a conditional Logon Script that applies per-user or per-group
attributes.
• Use of Samba’s preexec and postexec functions on access to the NET-
LOGON share to create a custom logon script and then execute it.
Section 24.4. Network Logon Script Magic 563
#!/usr/bin/perl
#
# genlogon.pl
#
# Perl script to generate user logon scripts on the fly, when users
# connect from a Windows client. This script should be called from
# smb.conf with the %U, %G and %L parameters. I.e:
#
# root preexec = genlogon.pl %U %G %L
#
# The script generated will perform
# the following:
#
# 1. Log the user connection to /var/log/samba/netlogon.log
# 2. Set the PC’s time to the Linux server time (which is maintained
# daily to the National Institute of Standards Atomic clock on the
# internet.
# 3. Connect the user’s home drive to H: (H for Home).
# 4. Connect common drives that everyone uses.
# 5. Connect group-specific drives for certain user groups.
# 6. Connect user-specific drives for certain users.
# 7. Connect network printers.
[myshare]
...
5
<http://support.microsoft.com/default.asp?scid=kb;en-us;189105>
566 Advanced Network Management Chapter 24
#!/bin/bash
IFS="-"
RESULT=$(smbstatus -S -u $1 2> /dev/null | awk ’NF > 6 {print $1}’ | sort | uniq
if [ "X${RESULT}" == X ]; then
exit 0
else
exit 1
fi
Chapter 25
This chapter summarizes the current state of knowledge derived from per-
sonal practice and knowledge from Samba mailing list subscribers. Before
reproduction of posted information, every effort has been made to validate
the information given. Where additional information was uncovered through
this validation, it is provided also.
When MS Windows NT 3.5 was introduced, the hot new topic was the
ability to implement Group Policies for users and groups. Then along came
MS Windows NT4 and a few sites started to adopt this capability. How do
we know that? By the number of “boo-boos” (or mistakes) administrators
made and then requested help to resolve.
By the time that MS Windows 2000 and Active Directory was released,
administrators got the message: Group Policies are a good thing! They
can help reduce administrative costs and actually make happier users. But
adoption of the true potential of MS Windows 200x Active Directory and
Group Policy Objects (GPOs) for users and machines were picked up on
rather slowly. This was obvious from the Samba mailing list back in 2000
and 2001 when there were few postings regarding GPOs and how to replicate
them in a Samba environment.
Judging by the traffic volume since mid 2002, GPOs have become a standard
part of the deployment in many sites. This chapter reviews techniques and
567
568 System and Account Policies Chapter 25
There are a large number of documents in addition to this old one that
should also be read and understood. Try searching on the Microsoft Web
site for “Group Policies”.
What follows is a brief discussion with some helpful notes. The information
provided here is incomplete — you are warned.
You need the Windows 98 Group Policy Editor to set up Group Profiles un-
der Windows 9x/Me. It can be found on the original full-product Windows
98 installation CD-ROM under tools/reskit/netadmin/poledit. Install
this using the Add/Remove Programs facility, and then click on Have Disk.
Use the Group Policy Editor to create a policy file that specifies the location
of user profiles and/or My Documents, and so on. Then save these settings
in a file called Config.POL that needs to be placed in the root of the [NET-
LOGON] share. If Windows 98 is configured to log onto the Samba domain,
it will automatically read this file and update the Windows 9x/Me registry
of the machine as it logs on.
Further details are covered in the Windows 98 Resource Kit documentation.
If you do not take the correct steps, then every so often Windows 9x/Me will
check the integrity of the registry and restore its settings from the backup
copy of the registry it stores on each Windows 9x/Me machine. So, you will
occasionally notice things changing back to the original settings.
Install the Group Policy handler for Windows 9x/Me to pick up Group Poli-
cies. Look on the Windows 98 CD-ROM in \tools\reskit\netadmin\poledit.
Install Group Policies on a Windows 9x/Me client by double-clicking on
grouppol.inf. Log off and on again a couple of times and see if Windows
98 picks up Group Policies. Unfortunately, this needs to be done on every
Windows 9x/Me machine that uses Group Policies.
To create or edit ntconfig.pol, you must use the NT Server Policy Editor,
poledit.exe, which is included with NT4 Server but not with NT worksta-
tion. There is a Policy Editor on an NT4 Workstation but it is not suitable
for creating domain policies. Furthermore, although the Windows 95 Policy
570 System and Account Policies Chapter 25
Windows NT4 system policies allow the setting of registry parameters spe-
cific to users, groups, and computers (client workstations) that are members
of the NT4-style domain. Such policy files will work with MS Windows
200x/XP clients also.
New to MS Windows 2000, Microsoft recently introduced a style of Group
Policy that confers a superset of capabilities compared with NT4-style poli-
cies. Obviously, the tool used to create them is different, and the mechanism
for implementing them is much improved.
The older NT4-style registry-based policies are known as Administrative
Templates in MS Windows 2000/XP GPOs. The latter includes the abil-
Section 25.2. Creating and Managing System Policies 571
Remember, NT4 policy files are named NTConfig.POL and are stored in the
root of the NETLOGON share on the domain controllers. A Windows NT4
user enters a username and password and selects the domain name to which
the logon will attempt to take place. During the logon process, the client
machine reads the NTConfig.POL file from the NETLOGON share on the
authenticating server and modifies the local registry values according to the
settings in this file.
Windows 200x GPOs are feature-rich. They are not stored in the NET-
LOGON share, but rather part of a Windows 200x policy file is stored in
the Active Directory itself and the other part is stored in a shared (and
replicated) volume called the SYSVOL folder. This folder is present on all
Active Directory domain controllers. The part that is stored in the Active
Directory itself is called the Group Policy Container (GPC), and the part
that is stored in the replicated share called SYSVOL is known as the Group
Policy Template (GPT).
With NT4 clients, the policy file is read and executed only as each user
logs onto the network. MS Windows 200x policies are much more complex
— GPOs are processed and applied at client machine startup (machine
specific part), and when the user logs onto the network, the user-specific part
is applied. In MS Windows 200x-style policy management, each machine
and/or user may be subject to any number of concurrently applicable (and
applied) policy sets (GPOs). Active Directory allows the administrator to
also set filters over the policy settings. No such equivalent capability exists
with NT4-style policy files.
Instead of using the tool called the System Policy Editor, commonly called
Poledit (from the executable name poledit.exe), GPOs are created and
managed using a Microsoft Management Console (MMC) snap-in as follows:
572 System and Account Policies Chapter 25
Note
Policies can define a specific user’s settings or the settings for a group of
users. The resulting policy file contains the registry settings for all users,
groups, and computers that will be using the policy file. Separate policy
files for each user, group, or computer are not necessary.
Section 25.3. Managing Account/User Policies 573
A new tool called editreg is under development. This tool can be used
to edit registry files (called NTUser.DAT) that are stored in user and group
profiles. NTConfig.POL files have the same structure as the NTUser.DAT file
and can be edited using this tool. editreg is being built with the intent
to enable NTConfig.POL files to be saved in text format and to permit the
building of new NTConfig.POL files with extended capabilities. It is proving
difficult to realize this capability, so do not be surprised if this feature does
not materialize. Formal capabilities will be announced at the time that this
tool is released for production use.
The tools that may be used to configure these types of controls from the MS
Windows environment are the NT4 User Manager for Domains, the NT4
System and Group Policy Editor, and the Registry Editor (regedt32.exe).
Under MS Windows 200x/XP, this is done using the MMC with appropriate
“snap-ins,” the registry editor, and potentially also the NT4 System and
Group Policy Editor.
With a Samba domain controller, the new tools for managing user account
and policy information include: smbpasswd, pdbedit, net, and rpc-
client. The administrator should read the man pages for these tools and
become familiar with their use.
The following attempts to document the order of processing the system and
user policies following a system reboot and as part of the user logon:
Section 25.5. System Startup and Logon Processing Overview 575
“We have created the Config.POL file and put it in the NETLOGON share.
It has made no difference to our Win XP Pro machines, they just do not
see it. It worked fine with Win 98 but does not work any longer since we
upgraded to Win XP Pro. Any hints?”
Policy files are not portable between Windows 9x/Me and MS Windows
NT4/200x/XP-based platforms. You need to use the NT4 Group Policy
Editor to create a file called NTConfig.POL so it is in the correct format for
your MS Windows XP Pro clients.
Chapter 26
DESKTOP PROFILE
MANAGEMENT
Roaming profiles are feared by some, hated by a few, loved by many, and a
godsend for some administrators.
Roaming profiles allow an administrator to make available a consistent user
desktop as the user moves from one machine to another. This chapter pro-
vides much information regarding how to configure and manage roaming
profiles.
While roaming profiles might sound like nirvana to some, they are a real and
tangible problem to others. In particular, users of mobile computing tools,
where often there may not be a sustained network connection, are often
better served by purely local profiles. This chapter provides information to
help the Samba administrator deal with those situations.
Warning
577
578 Desktop Profile Management Chapter 26
For example, to support Windows NT4/200x clients, set the following in the
[global] section of the smb.conf file:
l o g o n path = \\ p r o f i l e s e r v e r \ p r o f i l e s h a r e \ ←-
p r o f i l e p a t h \%U\ m o r e p r o f i l e p a t h
where “%L” translates to the name of the Samba server and “%U” translates
to the username.
Note
To support Windows 9x/Me clients, you must use the logon home parameter.
Samba has been fixed so net use /home now works as well and it, too, relies
on the logon home parameter.
By using the logon home parameter, you are restricted to putting Windows
9x/Me profiles in the user’s home directory. But wait! There is a trick you
can use. If you set the following in the [global] section of your smb.conf file:
l o g o n home = \\%L\%U\ . p r o f i l e s
then your Windows 9x/Me clients will dutifully put their clients in a subdi-
rectory of your home directory called .profiles (making them hidden).
Not only that, but net use /home will also work because of a feature in
Windows 9x/Me. It removes any directory stuff off the end of the home
directory area and only uses the server and share portion. That is, it looks
like you specified \\%L\%U for logon home.
You can support profiles for Windows 9x and Windows NT clients by setting
both the logon home and logon path parameters. For example,
l o g o n home = \\%L\%u \ . p r o f i l e s
l o g o n path = \\%L\ p r o f i l e s \%u
Windows 9x/Me and NT4 and later profiles should not be stored in the
same location because Windows NT4 and later will experience problems
580 Desktop Profile Management Chapter 26
The question often asked is, “How may I enforce use of local profiles?” or
“How do I disable roaming profiles?”
There are three ways of doing this:
In smb.conf Affect the following settings and ALL clients will be forced to
use a local profile: logon home = and logon path =
The arguments to these parameters must be left blank. It is necessary
to include the = sign to specifically assign the empty value.
Change of Profile Type: From the start menu right-click on the My Com-
puter icon, select Properties, click on the User Profiles tab, select the
profile you wish to change from Roaming type to Local, and click on
Change Type.
Consult the MS Windows registry guide for your particular MS Windows
version for more information about which registry keys to change to enforce
use of only local user profiles.
Section 26.2. Roaming Profiles 581
Note
When a user first logs in on Windows 9x, the file user.DAT is created, as
are folders Start Menu, Desktop, Programs, and Nethood. These direc-
tories and their contents will be merged with the local versions stored in
c:\windows\profiles\username on subsequent logins, taking the most re-
cent from each. You will need to use the [global] options preserve case =
yes, short preserve case = yes, and case sensitive = no in order to maintain
capital letters in shortcuts in any of the profile folders.
The user.DAT file contains all the user’s preferences. If you wish to enforce
a set of preferences, rename their user.DAT file to user.MAN, and deny them
write access to this file.
Under Windows 9x/Me, profiles are downloaded from the Primary Logon.
If you have the Primary Logon as “Client for Novell Networks”, then the
profiles and logon script will be downloaded from your Novell server. If
you have the Primary Logon as “Windows Logon”, then the profiles will
582 Desktop Profile Management Chapter 26
be loaded from the local machine — a bit against the concept of roaming
profiles, it would seem!
You will now find that the Microsoft Networks Login box contains [user,
password, domain] instead of just [user, password]. Type in the Samba
server’s domain name (or any other domain known to exist, but bear in
mind that the user will be authenticated against this domain and profiles
downloaded from it if that domain logon server supports it), user name and
user’s password.
Once the user has been successfully validated, the Windows 9x/Me machine
informs you that The user has not logged on before and asks Do you
wish to save the user’s preferences? Select Yes.
Once the Windows 9x/Me client comes up with the desktop, you should be
able to examine the contents of the directory specified in the logon path on
the Samba server and verify that the Desktop, Start Menu, Programs, and
Nethood folders have been created.
These folders will be cached locally on the client and updated when the user
logs off (if you haven’t made them read-only by then). You will find that if
the user creates further folders or shortcuts, the client will merge the profile
contents downloaded with the contents of the profile directory already on
the local client, taking the newest folders and shortcut from each set.
If you have made the folders/files read-only on the Samba server, then you
will get errors from the Windows 9x/Me machine on logon and logout as
it attempts to merge the local and remote profile. Basically, if you have
any errors reported by the Windows 9x/Me machine, check the UNIX file
permissions and ownership rights on the profile directory contents, on the
Samba server.
If you have problems creating user profiles, you can reset the user’s local
desktop cache, as shown below. When this user next logs in, the user will
be told that he/she is logging in “for the first time”.
1. Instead of logging in under the [user, password, domain] dialog, press
escape.
2. Run the regedit.exe program, and look in:
HKEY LOCAL MACHINE\Windows\CurrentVersion\ProfileList
You will find an entry for each user of ProfilePath. Note the contents
of this key (likely to be c:\windows\profiles\username), then delete
Section 26.2. Roaming Profiles 583
6. Check the contents of the profile path (see logon path described above)
and delete the user.DAT or user.MAN file for the user, making a backup
if required.
Warning
If all else fails, increase Samba’s debug log levels to between 3 and 10, and/or
run a packet sniffer program such as ethereal or netmon.exe, and look for
error messages.
If you have access to an Windows NT4/200x server, then first set up roam-
ing profiles and/or netlogons on the Windows NT4/200x server. Make a
packet trace, or examine the example packet traces provided with Windows
NT4/200x server, and see what the differences are with the equivalent Samba
trace.
584 Desktop Profile Management Chapter 26
You must first convert the profile from a local profile to a domain profile on
the MS Windows workstation as follows:
1. Log on as the local workstation administrator.
2. Right-click on the My Computer icon, and select Properties.
3. Click on the User Profiles tab.
4. Select the profile you wish to convert (click it once).
Section 26.2. Roaming Profiles 585
Note
Note
Does the new version of Samba have an Active Directory analogue? If so,
then you may be able to set the policy through this.
If you cannot set group policies in Samba, then you may be able to set the
policy locally on each machine. If you want to try this, then do the following:
1. On the XP workstation, log in with an administrative account.
2. Click on Start -> Run.
3. Type mmc.
4. Click on OK.
5. A Microsoft Management Console should appear.
6. Click on File -> Add/Remove Snap-in -> Add.
7. Double-click on Group Policy.
8. Click on Finish -> Close.
9. Click on OK.
10. In the “Console Root” window expand Local Computer Policy ->
Computer Configuration -> Administrative Templates -> System
-> User Profiles.
11. Double-click on Do not check for user ownership of Roaming Profile
Folders.
12. Select Enabled.
13. Click on OK.
14. Close the whole console. You do not need to save the settings (this
refers to the console settings rather than the policies you have changed).
15. Reboot.
There are certain situations that cause a cached local copy of roaming profile
not to be deleted on exit, even if the policy to force such deletion is set. To
deal with that situation, a special service was created. The application
UPHClean (User Profile Hive Cleanup) can be installed as a service on
Windows NT4/2000/XP Professional and Windows 2003.
Section 26.2. Roaming Profiles 587
The UPHClean software package can be downloaded from User Profile Hive
Cleanup Service.1
There is nothing to stop you from specifying any path that you like for the
location of users’ profiles. Therefore, you could specify that the profile be
stored on a Samba server or any other SMB server, as long as that SMB
server supports encrypted passwords.
Note
4. In the box labeled Copy Profile to add your new path, such as,
c:\temp\foobar
6. Click on the group “Everyone”, click on OK. This closes the “choose
user” box.
You should obtain the SID of your NT4 domain. You can use the net rpc
info to do this. See Chapter 12, “Remote and Local Management: The Net
Command”, Section 12.13 for more information.
26.2.5.3 moveuser.exe
You can identify the SID by using GetSID.exe from the Windows NT
Server 4.0 Resource Kit.
Windows NT 4.0 stores the local profile information in the registry un-
der the following key: HKEY LOCAL MACHINE\SOFTWARE\Microsoft\Windows
NT\CurrentVersion\ProfileList
Under the ProfileList key, there will be subkeys named with the SIDs of the
users who have logged on to this computer. (To find the profile information
for the user whose locally cached profile you want to move, find the SID
for the user with the GetSID.exe utility.) Inside the appropriate user’s
subkey, you will see a string value named ProfileImagePath.
A mandatory profile is a profile that the user does not have the ability
to overwrite. During the user’s session, it may be possible to change the
desktop environment; however, as the user logs out, all changes made will
be lost. If it is desired to not allow the user any ability to change the
desktop environment, then this must be done through policy settings. See
Chapter 25, “System and Account Policies”.
Note
into a mandatory profile, simply locate the NTUser.DAT file in the copied
profile and rename it to NTUser.MAN.
Note
MS Windows 9x/Me and NT4/200x/XP will use a default profile for any
user for whom a profile does not already exist. Armed with a knowledge
of where the default profile is located on the Windows workstation, and
knowing which registry keys affect the path from which the default profile
is created, it is possible to modify the default profile to one that has been
optimized for the site. This has significant administrative advantages.
Section 26.5. Default Profile for Windows Users 591
To enable default per-use profiles in Windows 9x/Me, you can either use the
Windows 98 System Policy Editor or change the registry directly.
To enable default per-user profiles in Windows 9x/Me, launch the System
Policy Editor, then select File -> Open Registry. Next click on the Local
Computer icon, click on Windows 98 System, select User Profiles, and
click on the enable box. Remember to save the registry changes.
To modify the registry directly, launch the Registry Editor (regedit.exe)
and select the hive HKEY LOCAL MACHINE\Network\Logon. Now add a DWORD
type key with the name “User Profiles.” To enable user profiles to set the
value to 1; to disable user profiles set it to 0.
When a user logs on to a Windows 9x/Me machine, the local profile path,
HKEY LOCAL MACHINE\Software\Microsoft\Windows\CurrentVersion\ProfileList,
is checked for an existing entry for that user.
If the user has an entry in this registry location, Windows 9x/Me checks
for a locally cached version of the user profile. Windows 9x/Me also checks
the user’s home directory (or other specified directory if the location has
been modified) on the server for the user profile. If a profile exists in both
locations, the newer of the two is used. If the user profile exists on the
server but does not exist on the local machine, the profile on the server is
downloaded and used. If the user profile only exists on the local machine,
that copy is used.
If a user profile is not found in either location, the default user profile from
the Windows 9x/Me machine is used and copied to a newly created folder for
the logged on user. At log off, any changes that the user made are written
to the user’s local profile. If the user has a roaming profile, the changes are
written to the user’s profile on the server.
On MS Windows NT4, the default user profile is obtained from the location
%SystemRoot%\Profiles, which in a default installation will translate to
592 Desktop Profile Management Chapter 26
will also remain stored in the same way, unless the following registry key is
created:
HKEY_LOCAL_MACHINE\SYSTEM\Software\Microsoft\Windows NT\CurrentVersion\
winlogon\"DeleteRoamingCache"=dword:0000000
The Registry Hive key that affects the behavior of folders that are part of
the default user profile are controlled by entries on Windows NT4 is:
HKEY_CURRENT_USER
\Software
\Microsoft
\Windows
\CurrentVersion
\Explorer
\User Shell Folders
The above hive key contains a list of automatically managed folders. The
default entries are shown in Table 26.1.
The registry key that contains the location of the default profile settings is:
HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Windows\CurrentVersion\Explorer\
User Shell Folders
Note
When a new user first logs onto an MS Windows 200x/XP machine, the de-
fault profile is obtained from C:\Documents and Settings\Default User.
The administrator can modify or change the contents of this location, and
MS Windows 200x/XP will gladly use it. This is far from the optimum
arrangement, since it will involve copying a new default profile to every MS
Windows 200x/XP client workstation.
When MS Windows 200x/XP participates in a domain security context, and
if the default user profile is not found, then the client will search for a default
profile in the NETLOGON share of the authenticating server. In MS Win-
dows parlance, it is
%LOGONSERVER%\NETLOGON\Default User, and if one exists there, it will
copy this to the workstation in the C:\Documents and Settings\ under
the Windows login name of the use.
Note
If a default profile does not exist in this location, then MS Windows 200x/XP
will use the local default profile.
On logging out, the user’s desktop profile is stored to the location specified
in the registry settings that pertain to the user. If no specific policies have
been created or passed to the client during the login process (as Samba does
automatically), then the user’s profile is written to the local machine only
under the path C:\Documents and Settings\%USERNAME%.
Those wishing to modify the default behavior can do so through these three
methods:
• Modify the registry keys on the local machine manually and place
the new default profile in the NETLOGON share root. This is not
recommended because it is maintenance intensive.
• Create an NT4-style NTConfig.POL file that specifies this behavior
and locate this file in the root of the NETLOGON share along with
596 Desktop Profile Management Chapter 26
There is also an entry called “Default” that has no value set. The default
entry is of type REG SZ; all the others are of type REG EXPAND SZ.
It makes a huge difference to the speed of handling roaming user profiles if
all the folders are stored on a dedicated location on a network server. This
Section 26.6. Common Errors 597
means that it will not be necessary to write the Outlook PST file over the
network for every login and logout.
To set this to a network location, you could use the following examples:
%LOGONSERVER%\%USERNAME%\Default Folders
This stores the folders in the user’s home directory under a directory called
Default Folders. You could also use:
\\SambaServer\FolderShare\%USERNAME%
in which case the default folders are stored in the server named SambaServer
in the share called FolderShare under a directory that has the name of the
MS Windows user as seen by the Linux/UNIX file system.
Please note that once you have created a default profile share, you must
migrate a user’s profile (default or custom) to it.
HKEY_LOCAL_MACHINE\SYSTEM\Software\Microsoft\Windows NT\CurrentVersion\
winlogon\"DeleteRoamingCache"=dword:00000001
The following are some typical errors, problems, and questions that have
been asked on the Samba mailing lists.
With Samba-2.2.x, the choice you have is to enable or disable roaming pro-
files support. It is a global-only setting. The default is to have roaming
profiles, and the default path will locate them in the user’s home directory.
598 Desktop Profile Management Chapter 26
If disabled globally, then no one will have roaming profile ability. If enabled
and you want it to apply only to certain machines, then on those machines
on which roaming profile support is not wanted, it is necessary to disable
roaming profile handling in the registry of each such machine.
With Samba-3, you can have a global profile setting in smb.conf, and you
can override this by per-user settings using the Domain User Manager (as
with MS Windows NT4/200x).
In any case, you can configure only one profile per user. That profile can be
either:
Roaming profiles As a user logs onto the network, a centrally stored pro-
file is copied to the workstation to form a local profile. This local
profile will persist (remain on the workstation disk) unless a registry
key is changed that will cause this profile to be automatically deleted
on logout.
The point of this discussion is to show that roaming profiles and good con-
trols of how they can be changed as well as good discipline make for a
problem-free site.
• If each machine is used by many users, then much local disk storage
is needed for local profiles.
• Every workstation the user logs into has its own profile; these can be
very different from machine to machine.
“When the client logs onto the domain controller, it searches for a profile to
download. Where do I put this default profile?”
First, the Samba server needs to be configured as a domain controller. This
can be done by setting in smb.conf:
security = user
o s l e v e l = 32 ( o r more )
domain l o g o n s = Yes
Note
PAM-BASED DISTRIBUTED
AUTHENTICATION
Note
601
602 PAM-Based Distributed Authentication Chapter 27
/etc/passwd There are several PAM modules that interact with this stan-
dard UNIX user database. The most common are called pam unix.
so, pam unix2.so, pam pwdb.so and pam userdb.so.
Kerberos The pam krb5.so module allows the use of any Kerberos-compliant
server. This tool is used to access MIT Kerberos, Heimdal Kerberos,
and potentially Microsoft Active Directory (if enabled).
LDAP The pam ldap.so module allows the use of any LDAP v2- or v3-
compatible backend server. Commonly used LDAP backend servers
include OpenLDAP v2.0 and v2.1, Sun ONE iDentity server, Novell
eDirectory server, and Microsoft Active Directory.
SMB Password This module, called pam smbpass.so, allows user authen-
tication of the passdb backend that is configured in the Samba smb.
conf file.
Section 27.2. Technical Discussion 603
SMB Server The pam smb auth.so module is the original MS Windows
networking authentication tool. This module has been somewhat out-
dated by the Winbind module.
In this section we discuss the correct syntax of and generic options respected
by entries to these files. PAM-specific tokens in the configuration file are
case insensitive. The module paths, however, are case sensitive, since they
indicate a file’s name and reflect the case dependence of typical file systems.
604 PAM-Based Distributed Authentication Chapter 27
The case sensitivity of the arguments to any given module is defined for each
module in turn.
In addition to the lines described below, there are two special characters
provided for the convenience of the system administrator: comments are
preceded by a “#” and extend to the next end-of-line; also, module specifi-
cation lines may be extended with a “\”-escaped newline.
If the PAM authentication module (loadable link library file) is located in
the default location, then it is not necessary to specify the path. In the case
of Linux, the default location is /lib/security. If the module is located
outside the default, then the path must be specified as:
The remaining information in this subsection was taken from the documen-
tation of the Linux-PAM project. For more information on PAM, see the
Official Linux-PAM home page1 .
A general configuration line of the /etc/pam.conf file has the following
form:
We explain the meaning of each of these tokens. The second (and more
recently adopted) way of configuring Linux-PAM is via the contents of the
/etc/pam.d/ directory. Once we have explained the meaning of the tokens,
we describe this method.
service-name The name of the service associated with this entry. Fre-
quently, the service-name is the conventional name of the given appli-
cation — for example, ftpd, rlogind and su, and so on.
There is a special service-name reserved for defining a default authen-
tication mechanism. It has the name OTHER and may be specified in
1
<http://ftp.kernel.org/pub/linux/libs/pam/>
Section 27.2. Technical Discussion 605
module-type One of (currently) four types of module. The four types are
as follows:
control-flag The control-flag is used to indicate how the PAM library will
react to the success or failure of the module it is associated with. Since
modules can be stacked (modules of the same type execute in series,
one after another), the control-flags determine the relative importance
of each module. The application is not made aware of the individual
success or failure of modules listed in the /etc/pam.conf file. Instead,
it receives a summary success or fail response from the Linux-PAM
library. The order of execution of these modules is that of the entries
in the /etc/pam.conf file; earlier entries are executed before later
606 PAM-Based Distributed Authentication Chapter 27
The more elaborate (newer) syntax is much more specific and gives
the administrator a great deal of control over how the user is authen-
ticated. This form of the control-flag is delimited with square brackets
and consists of a series of value=action tokens:
The last of these (default) can be used to set the action for those
return values that are not explicitly defined.
The action1 can be a positive integer or one of the following tokens:
ignore; ok; done; bad; die; and reset. A positive integer, J, when
specified as the action, can be used to indicate that the next J modules
of the current module-type will be skipped. In this way, the adminis-
trator can develop a moderately sophisticated stack of modules with
a number of different paths of execution. Which path is taken can be
determined by the reactions of individual modules.
• ignore: When used with a stack of modules, the module’s re-
turn status will not contribute to the return code the application
obtains.
• bad: This action indicates that the return code should be thought
of as indicative of the module failing. If this module is the first
in the stack to fail, its status value will be used for that of the
whole stack.
• die: Equivalent to bad with the side effect of terminating the
module stack and PAM immediately returning to the application.
608 PAM-Based Distributed Authentication Chapter 27
• ok: This tells PAM that the administrator thinks this return code
should contribute directly to the return code of the full stack of
modules. In other words, if the former state of the stack would
lead to a return of PAM SUCCESS, the module’s return code will
override this value. Note, if the former state of the stack holds
some value that is indicative of a module’s failure, this ok value
will not be used to override that value.
• reset: Clears all memory of the state of the module stack and
starts again with the next stacked module.
Just to get a feel for the power of this new syntax, here is a taste of
what you can do with it. With Linux-PAM-0.63, the notion of client
plug-in agents was introduced. This makes it possible for PAM to sup-
port machine-machine authentication using the transport protocol in-
herent to the client/server application. With the [ ... value=action
... ] control syntax, it is possible for an application to be configured
to support binary prompts with compliant clients, but to gracefully fail
over into an alternative authentication mode for legacy applications.
The arguments are a list of tokens that are passed to the module when
it is invoked, much like arguments to a typical Linux shell command.
Generally, valid arguments are optional and are specific to any given
module. Invalid arguments are ignored by a module; however, when
encountering an invalid argument, the module is required to write an
error to syslog(3). For a list of generic options, see the next section.
When using this convention, you can include “[” characters inside the
string, and if you wish to have a “]” character inside the string that
will survive the argument parsing, you should use “\[”. In other words,
Any line in one of the configuration files that is not formatted cor-
rectly will generally tend (erring on the side of caution) to make the
authentication process fail. A corresponding error is written to the
system log files with a call to syslog(3).
#%PAM-1.0
# The PAM configuration file for the login service
#
auth required pam_securetty.so
auth required pam_nologin.so
# auth required pam_dialup.so
# auth optional pam_mail.so
auth required pam_pwdb.so shadow md5
# account requisite pam_time.so
account required pam_pwdb.so
session required pam_pwdb.so
# session optional pam_lastlog.so
# password required pam_cracklib.so retry=3
password required pam_pwdb.so shadow md5
The following example for the login program replaces the use of the pam
pwdb.so module that uses the system password database (/etc/passwd, /
etc/shadow, /etc/group) with the module pam smbpass.so, which uses the
Samba database containing the Microsoft MD4 encrypted password hashes.
This database is stored either in /usr/local/samba/private/smbpasswd,
/etc/samba/smbpasswd or in /etc/samba.d/smbpasswd, depending on the
Samba implementation for your UNIX/Linux system. The pam smbpass.
so module is provided by Samba version 2.2.1 or later. It can be compiled
by specifying the --with-pam smbpass options when running Samba’s con-
figure script. For more information on the pam smbpass module, see the
documentation in the source/pam smbpass directory of the Samba source
distribution.
#%PAM-1.0
# The PAM configuration file for the login service
#
auth required pam_smbpass.so nodelay
account required pam_smbpass.so nodelay
session required pam_smbpass.so nodelay
password required pam_smbpass.so nodelay
The following is the PAM configuration file for a particular Linux system.
The default condition uses pam pwdb.so.
#%PAM-1.0
# The PAM configuration file for the samba service
#
auth required pam_pwdb.so nullok nodelay shadow audit
account required pam_pwdb.so audit nodelay
session required pam_pwdb.so nodelay
password required pam_pwdb.so shadow md5
In the following example, the decision has been made to use the smbpasswd
database even for basic Samba authentication. Such a decision could also
be made for the passwd program and would thus allow the smbpasswd
passwords to be changed using the passwd program:
612 PAM-Based Distributed Authentication Chapter 27
#%PAM-1.0
# The PAM configuration file for the samba service
#
auth required pam_smbpass.so nodelay
account required pam_pwdb.so audit nodelay
session required pam_pwdb.so nodelay
password required pam_smbpass.so nodelay smbconf=/etc/samba.d/smb.conf
Note
yes. The reason is that PAM modules cannot support the chal-
lenge/response authentication mechanism needed in the presence
of SMB password encryption.
As winbind users and groups are resolved from a server, user and group
IDs are allocated from a specified range. This is done on a first come, first
served basis, although all existing users and groups will be mapped as soon
as a client performs a user or group enumeration command. The allocated
UNIX IDs are stored in a database file under the Samba lock directory and
will be remembered.
The astute administrator will realize from this that the combination of pam
smbpass.so, winbindd, and a distributed passdb backend such as ldap will
allow the establishment of a centrally managed, distributed user/password
database that can also be used by all PAM-aware (e.g., Linux) programs
and applications. This arrangement can have particularly potent advantages
compared with the use of Microsoft Active Directory Service (ADS) insofar
as the reduction of wide-area network authentication traffic.
614 PAM-Based Distributed Authentication Chapter 27
Warning
The following is a sample PAM configuration that shows the use of pam smbpass
to make sure private/smbpasswd is kept in sync when /etc/passwd (/
etc/shadow) is changed. It is useful when an expired password might be
changed by an application (such as ssh).
#%PAM-1.0
# password-sync
#
auth requisite pam_nologin.so
Section 27.2. Technical Discussion 615
The following PAM configuration shows the use of pam smbpass to migrate
from plaintext to encrypted passwords for Samba. Unlike other methods,
this can be used for users who have never connected to Samba shares: pass-
word migration takes place when users ftp in, login using ssh, pop their
mail, and so on.
616 PAM-Based Distributed Authentication Chapter 27
#%PAM-1.0
# password-migration
#
auth requisite pam_nologin.so
# pam_smbpass is called IF pam_unix succeeds.
auth requisite pam_unix.so
auth optional pam_smbpass.so migrate
account required pam_unix.so
password requisite pam_cracklib.so retry=3
password requisite pam_unix.so shadow md5 use_authtok try_first_pass
password optional pam_smbpass.so nullok use_authtok try_first_pass
session required pam_unix.so
#%PAM-1.0
# password-mature
#
auth requisite pam_nologin.so
auth required pam_unix.so
account required pam_unix.so
password requisite pam_cracklib.so retry=3
password requisite pam_unix.so shadow md5 use_authtok try_first_pass
password required pam_smbpass.so use_authtok use_first_pass
session required pam_unix.so
The following is a sample PAM configuration that shows pam smbpass used
together with pam krb5. This could be useful on a Samba PDC that is also
a member of a Kerberos realm.
Section 27.3. Common Errors 617
#%PAM-1.0
# kdc-pdc
#
auth requisite pam_nologin.so
auth requisite pam_krb5.so
auth optional pam_smbpass.so migrate
account required pam_krb5.so
password requisite pam_cracklib.so retry=3
password optional pam_smbpass.so nullok use_authtok try_first_pass
password required pam_krb5.so use_authtok try_first_pass
session required pam_krb5.so
When I open a new console with [ctrl][alt][F1], I can’t log in with my user
“pitie.” I have tried with user “scienceu\pitie” also.
The problem may lie with the inclusion of pam stack.so service=system-auth.
That file often contains a lot of stuff that may duplicate what you are already
doing. Try commenting out the pam stack lines for auth and account and
618 PAM-Based Distributed Authentication Chapter 27
root# wbinfo -u
MIDEARTH\maryo
MIDEARTH\jackb
MIDEARTH\ameds
...
MIDEARTH\root
root# wbinfo -g
MIDEARTH\Domain Users
MIDEARTH\Domain Admins
MIDEARTH\Domain Guests
...
MIDEARTH\Accounts
Your system is likely running nscd, the name service caching daemon. Shut
it down, do not restart it! You will find your problem resolved.
Chapter 28
INTEGRATING MS
WINDOWS NETWORKS
WITH SAMBA
This chapter deals with NetBIOS over TCP/IP name to IP address resolu-
tion. If your MS Windows clients are not configured to use NetBIOS over
TCP/IP, then this section does not apply to your installation. If your in-
stallation involves the use of NetBIOS over TCP/IP, then this chapter may
help you to resolve networking problems.
Note
620
Section 28.2. Background Information 621
system. Likewise, many UNIX and Linux administrators have not been
exposed to the intricacies of MS Windows TCP/IP-based networking (and
may have no desire to be, either).
This chapter gives a short introduction to the basics of how a name can be
resolved to its IP address for each operating system environment.
Note
When NetBIOS over TCP/IP is disabled, the use of DNS is essential. Most
installations that disable NetBIOS over TCP/IP today use MS Active Di-
rectory Service (ADS). ADS requires dynamic DNS with Service Resource
Records (SRV RR) and with Incremental Zone Transfers (IXFR). Use of
DHCP with ADS is recommended as a further means of maintaining central
control over the client workstation network configuration.
• /etc/resolv.conf
• /etc/host.conf
• /etc/nsswitch.conf
28.3.1 /etc/hosts
Network packets that are sent over the physical network transport layer
communicate not via IP addresses but rather using the Media Access Control
address, or MAC address. IP addresses are currently 32 bits in length and
are typically presented as four decimal numbers that are separated by a dot
(or period) — for example, 168.192.1.1.
MAC addresses use 48 bits (or 6 bytes) and are typically represented as
two-digit hexadecimal numbers separated by colons: 40:8e:0a:12:34:56.
Every network interface must have a MAC address. Associated with a MAC
address may be one or more IP addresses. There is no relationship between
an IP address and a MAC address; all such assignments are arbitrary or dis-
cretionary in nature. At the most basic level, all network communications
take place using MAC addressing. Since MAC addresses must be globally
unique and generally remain fixed for any particular interface, the assign-
ment of an IP address makes sense from a network management perspective.
More than one IP address can be assigned per MAC address. One address
must be the primary IP address — this is the address that will be returned
in the Address Resolution Protocol (ARP) reply.
28.3.2 /etc/resolv.conf
28.3.3 /etc/host.conf
order hosts,bind
multi on
Both addresses should be returned. Please refer to the man page for host.
conf for further details.
624 Integrating MS Windows Networks with Samba Chapter 28
28.3.4 /etc/nsswitch.conf
This file controls the actual name resolution targets. The file typically has
resolver object specifications as follows:
# /etc/nsswitch.conf
#
# Name Service Switch configuration file.
#
passwd: compat
# Alternative entries for password authentication are:
# passwd: compat files nis ldap winbind
shadow: compat
group: compat
Table 28.1 and Table 28.2 tables list typical NetBIOS name/service type
registrations.
It should be noted that all NetBIOS machines register their own names as
per Table 28.1 and Table 28.2. This is in vast contrast to TCP/IP instal-
lations where the system administrator traditionally determines in the /
etc/hosts or in the DNS database what names are associated with each IP
address.
One further point of clarification should be noted. The /etc/hosts file and
the DNS records do not provide the NetBIOS name information that MS
Windows clients depend on to locate the type of service that may be needed.
An example of this is what happens when an MS Windows client wants to
locate a domain logon server. It finds this service and the IP address of a
server that provides it by performing a lookup (via a NetBIOS broadcast)
for enumeration of all machines that have registered the name type *<1c>.
A logon request is then sent to each IP address that is returned in the
enumerated list of IP addresses. Whichever machine first replies, it then
ends up providing the logon services.
The name “workgroup” or “domain” really can be confusing, since these
have the added significance of indicating what is the security architecture
of the MS Windows network. The term “workgroup” indicates that the
primary nature of the network environment is that of a peer-to-peer design.
In a workgroup, all machines are responsible for their own security, and
generally such security is limited to the use of just a password (known as
share-level security). In most situations with peer-to-peer networking, the
users who control their own machines will simply opt to have no security at
all. It is possible to have user-level security in a workgroup environment,
thus requiring the use of a username and a matching password.
MS Windows networking is thus predetermined to use machine names for
all local and remote machine message passing. The protocol used is called
Server Message Block (SMB), and this is implemented using the NetBIOS
protocol (Network Basic Input/Output System). NetBIOS can be encap-
sulated using LLC (Logical Link Control) protocol — in which case the
resulting protocol is called NetBEUI (Network Basic Extended User Inter-
face). NetBIOS can also be run over IPX (Internetworking Packet Exchange)
protocol as used by Novell NetWare, and it can be run over TCP/IP pro-
tocols — in which case the resulting protocol is called NBT or NetBT, the
NetBIOS over TCP/IP.
MS Windows machines use a complex array of name resolution mechanisms.
Since we are primarily concerned with TCP/IP, this demonstration is limited
to this area.
Section 28.4. Name Resolution as Used within MS Windows Networking 627
# This file is compatible with Microsoft LAN Manager 2.x TCP/IP lmhosts
# files and offers the following extensions:
#
# #PRE
# #DOM:<domain>
# #INCLUDE <filename>
# #BEGIN_ALTERNATE
# #END_ALTERNATE
# \0xnn (non-printing character support)
#
# Following any entry in the file with the characters "#PRE" will cause
# the entry to be preloaded into the name cache. By default, entries are
# not preloaded, but are parsed only after dynamic name resolution fails.
#
# Following an entry with the "#DOM:<domain>" tag will associate the
# entry with the domain specified by <domain>. This effects how the
# browser and logon services behave in TCP/IP environments. To preload
# the host name associated with #DOM entry, it is necessary to also add a
# #PRE to the line. The <domain> is always pre-loaded although it will not
# be shown when the name cache is viewed.
#
# Specifying "#INCLUDE <filename>" will force the RFC NetBIOS (NBT)
# software to seek the specified <filename> and parse it as if it were
# local. <filename> is generally a UNC-based name, allowing a
# centralized lmhosts file to be maintained on a server.
# It is ALWAYS necessary to provide a mapping for the IP address of the
# server prior to the #INCLUDE. This mapping must use the #PRE directive.
# In addition the share "public" in the example below must be in the
# LanMan Server list of "NullSessionShares" in order for client machines to
# be able to read the lmhosts file successfully. This key is under
# \machine\system\currentcontrolset\services\lanmanserver\
# parameters\nullsessionshares
# in the registry. Simply add "public" to the list found there.
#
# The #BEGIN_ and #END_ALTERNATE keywords allow multiple #INCLUDE
# statements to be grouped together. Any single successful include
# will cause the group to succeed.
#
# Finally, non-printing characters can be embedded in mappings by
# first surrounding the NetBIOS name in quotations, then using the
Section 28.4. Name Resolution as Used within MS Windows Networking 629
This capability is configured in the TCP/IP setup area in the network con-
figuration facility. If enabled, an elaborate name resolution sequence is fol-
630 Integrating MS Windows Networks with Samba Chapter 28
lowed, the precise nature of which is dependent on how the NetBIOS Node
Type parameter is configured. A Node Type of 0 means that NetBIOS
broadcast (over UDP broadcast) is used if the name that is the subject of
a name lookup is not found in the NetBIOS name cache. If that fails, then
DNS, HOSTS, and LMHOSTS are checked. If set to Node Type 8, then a
NetBIOS Unicast (over UDP Unicast) is sent to the WINS server to obtain
a lookup before DNS, HOSTS, LMHOSTS, or broadcast lookup is used.
“I can ping my Samba server from Windows, but I cannot ping my Windows
machine from the Samba server.”
The Windows machine was at IP address 192.168.1.2 with netmask 255.255.255.0,
the Samba server (Linux) was at IP address 192.168.1.130 with netmask
255.255.255.128. The machines were on a local network with no external
connections.
Due to inconsistent netmasks, the Windows machine was on network 192.168.1.0/24,
while the Samba server was on network 192.168.1.128/25 — logically a dif-
ferent network.
“The name of the Samba server was changed, Samba was restarted, and now
the Samba server cannot be pinged by its new name from an MS Windows
NT4 workstation, but it does still respond to pinging using the old name.
Why?”
From this description, three things are obvious:
• WINS is not in use; only broadcast-based name resolution is used.
• The Samba server was renamed and restarted within the last 10 or 15
minutes.
632 Integrating MS Windows Networks with Samba Chapter 28
• The old Samba server name is still in the NetBIOS name cache on the
MS Windows NT4 workstation.
To find what names are present in the NetBIOS name cache on the MS
Windows NT4 machine, open a cmd shell and then:
C:\> nbtstat -n
C:\> nbtstat -c
C:\>
UNICODE/CHARSETS
633
634 Unicode/Charsets Chapter 29
As of Samba-3, Samba can (and will) talk Unicode over the wire. Internally,
Samba knows of three kinds of character sets:
unix charset This is the charset used internally by your operating system.
The default is UTF-8, which is fine for most systems and covers all
characters in all languages. The default in previous Samba releases
was to save filenames in the encoding of the clients — for example,
cp850 for Western European countries.
display charset This is the charset Samba uses to print messages on your
screen. It should generally be the same as the unix charset.
dos charset This is the charset Samba uses when communicating with
DOS and Windows 9x/Me clients. It will talk Unicode to all newer
2
<http://www.unicode.org/>
Section 29.4. Conversion from Old Names 635
Because previous Samba versions did not do any charset conversion, char-
acters in filenames are usually not correct in the UNIX charset but only for
the local charset used by the DOS/Windows clients.
Bjoern Jacke has written a utility named convmv3 that can convert whole
directory structures to different charsets with one single command.
The dos charset and display charset should be set to the locale compatible
with the character set and encoding method used on Windows. This is
usually CP932 but sometimes has a different name.
The unix charset can be either Shift JIS series, EUC-JP series, or UTF-8.
UTF-8 is always available, but the availability of other locales and the name
itself depends on the system.
Additionally, you can consider using the Shift JIS series as the value of the
unix charset parameter by using the vfs cap module, which does the same
thing as setting “coding system = CAP” in the Samba 2.2 series.
Where to set unix charset to is a difficult question. Here is a list of details,
advantages, and disadvantages of using a certain value.
Shift JIS series Shift JIS series means a locale that is equivalent to Shift JIS,
used as a standard on Japanese Windows. In the case of Shift JIS,
for example, if a Japanese filename consists of 0x8ba4 and 0x974c (a
4-bytes Japanese character string meaning “share”) and “.txt” is writ-
ten from Windows on Samba, the filename on UNIX becomes 0x8ba4,
0x974c, “.txt” (an 8-byte BINARY string), same as Windows.
Since Shift JIS series is usually used on some commercial-based UNIXes;
hp-ux and AIX as the Japanese locale (however, it is also possible to
use the EUC-JP locale series). To use Shift JIS series on these plat-
forms, Japanese filenames created from Windows can be referred to
also on UNIX.
If your UNIX is already working with Shift JIS and there is a user who
needs to use Japanese filenames written from Windows, the Shift JIS
Section 29.5. Japanese Charsets 637
Shift JIS series + vfs cap (CAP encoding) CAP encoding means a spec-
ification used in CAP and NetAtalk, file server software for Macintosh.
In the case of CAP encoding, for example, if a Japanese filename con-
sists of 0x8ba4 and 0x974c, and “.txt” is written from Windows on
Samba, the filename on UNIX becomes “:8b:a4:97L.txt” (a 14 bytes
ASCII string).
For CAP encoding, a byte that cannot be expressed as an ASCII char-
acter (0x80 or above) is encoded in an “:xx” form. You need to take
care of containing a “\(0x5c)” in a filename, but filenames are not
Section 29.5. Japanese Charsets 639
You should set CP932 if using GNU libiconv for unix charset. With
this setting, filenames in the “cap-share” share are written with CAP
encoding.
GNU libiconv To handle Japanese correctly, you should apply the patch
640 Unicode/Charsets Chapter 29
libiconv-1.8-cp932-patch.diff.gz4 to libiconv-1.8.
Other Japanese locales (for example, Shift JIS and EUC-JP) should
not be used because of the lack of the compatibility with Windows.
Other Japanese locales (for example, Shift JIS and EUC-JP) should
not be used because of the lack of the compatibility with Windows.
Prior to Samba-2.2 series, the “coding system” parameter was used. The
default codepage in Samba 2.x was code page 850. In the Samba-3 series
this has been replaced with the unix charset parameter. Table 29.1 shows
the mapping table when migrating from the Samba-2.2 series to Samba-3.
4
<http://www2d.biglobe.ne.jp/~msyk/software/libiconv-patch.html>
5
<http://www2d.biglobe.ne.jp/~msyk/software/glibc/>
Section 29.6. Common Errors 641
BACKUP TECHNIQUES
The Samba project is over 10 years old. During the early history of Samba,
UNIX administrators were its key implementors. UNIX administrators use
UNIX system tools to backup UNIX system files. Over the past 4 years,
an increasing number of Microsoft network administrators have taken an
interest in Samba. This is reflected in the questions about backup in general
on the Samba mailing lists.
Please note that all information here is provided as is and without recom-
mendation of fitness or suitability. The network administrator is strongly en-
couraged to perform due diligence research before implementing any backup
solution, whether free software or commercial.
642
Section 30.2. Discussion of Backup Solutions 643
A useful Web site I recently stumbled across that you might like to refer to
is located at www.allmerchants.com1 .
The following three free software projects might also merit consideration.
30.2.1 BackupPC
30.2.2 Rsync
30.2.3 Amanda
3
<http://www.amanda.org/>
4
<http://bobs.sourceforge.net/>
Chapter 31
HIGH AVAILABILITY
Network administrators are often concerned about the availability of file and
print services. Network users are inclined toward intolerance of the services
they depend on to perform vital task responsibilities.
All humans fail, in both great and small ways we fail continually.
Machines fail too. Computers are machines that are managed
by humans, the fallout from failure can be spectacular. Your
responsibility is to deal with failure, to anticipate it and to elim-
inate it as far as is humanly and economically wise to achieve.
Are your actions part of the problem or part of the solution?
If we are to deal with failure in a planned and productive manner, then first
we must understand the problem. That is the purpose of this chapter.
645
646 High Availability Chapter 31
SMB requests are sent by vuid to their associated server. No code exists
today to effect this solution. This problem is conceptually similar to the
problem of correctly handling requests from multiple requests from Windows
2000 Terminal Server in Samba.
There exists many distributed file systems for UNIX and Linux.
• NFS
• AFS
• OpenGFS
• Lustre
The server pool (cluster) can use any distributed file system backend if all
SMB semantics are performed within this pool.
Where a clustered server provides purely SMB services, oplock handling may
be done within the server pool without imposing a need for this to be passed
to the backend file system pool.
On the other hand, where the server pool also provides NFS or other file
services, it will be essential that the implementation be oplock-aware so it
can interoperate with SMB services. This is a significant challenge today.
A failure to provide this interoperability will result in a significant loss of
performance that will be sorely noted by users of Microsoft Windows clients.
Last, all state information must be shared across the server pool.
Section 31.2. Technical Discussion 649
Most backend file systems support POSIX file semantics. This makes it
difficult to push SMB semantics back into the file system. POSIX locks
have different properties and semantics from SMB locks.
All smbd processes in the server pool must of necessity communicate very
quickly. For this, the current tdb file structure that Samba uses is not
suitable for use across a network. Clustered smbds must use something
else.
If only one server is active in a pair, the need for high-speed server intercon-
nect is avoided. This allows the use of existing high-availability solutions,
instead of inventing a new one. This simpler solution comes at a price —
the cost of which is the need to manage a more complex file name space.
Since there is now not a single file system, administrators must remember
where all services are located — a complexity not easily dealt with.
Many failover solutions (like Red Hat Cluster Manager and Microsoft Wolf-
pack) can use a shared SCSI of Fiber Channel disk storage array for failover
communication. Information regarding Red Hat high availability solutions
for Samba may be obtained from www.redhat.com1 .
1
<http://www.redhat.com/docs/manuals/enterprise/RHEL-AS-2.1-Manual/
cluster-manager/s1-service-samba.html>
2
<http://www.linux-ha.org/>
Section 31.2. Technical Discussion 651
31.2.6 Conclusions
HANDLING LARGE
DIRECTORIES
The key was fixing the directory handling to read only the current list re-
quested instead of the old (up to samba-3.0.11) behavior of reading the entire
directory into memory before doling out names. Normally this would have
broken OS/2 applications, which have very strange delete semantics, but
by stealing logic from Samba4 (thanks, Tridge), the current code in 3.0.12
handles this correctly.
First, you need to canonicalize all the files in the directory to have one
case, upper or lower — take your pick (I chose upper because all my files
were already uppercase names). Then set up a new custom share for the
application as follows:
[bigshare]
path = /home/jeremy/tmp/manyfilesdir
read only = no
case sensitive = True
default case = upper
preserve case = no
short preserve case = no
652
653
Of course, use your own path and settings, but set the case options to match
the case of all the files in your directory. The path should point at the large
directory needed for the application — any new files created in there and in
any paths under it will be forced by smbd into uppercase, but smbd will no
longer have to scan the directory for names: it knows that if a file does not
exist in uppercase, then it doesn’t exist at all.
The secret to this is really in the case sensitive = True line. This tells smbd
never to scan for case-insensitive versions of names. So if an application
asks for a file called FOO, and it cannot be found by a simple stat call, then
smbd will return file not found immediately without scanning the containing
directory for a version of a different case. The other xxx case xxx lines
make this work by forcing a consistent case on all files created by smbd.
Remember, all files and directories under the path directory must be in
uppercase with this smb.conf stanza because smbd will not be able to find
lowercase filenames with these settings. Also note that this is done on a per-
share basis, allowing this parameter to be set only for a share servicing an
application with this problematic behavior (using large numbers of entries
in a directory) — the rest of your smbd shares don’t need to be affected.
This makes smbd much faster when dealing with large directories. My test
case has over 100,000 files, and smbd now deals with this very efficiently.
Part IV
UPGRADING FROM
SAMBA-2.X TO
SAMBA-3.0.20
This chapter deals exclusively with the differences between Samba-3.0.20 and
Samba-2.2.8a. It points out where configuration parameters have changed,
and provides a simple guide for the move from 2.2.x to 3.0.20.
655
656 Upgrading from Samba-2.x to Samba-3.0.20 Chapter 33
the old LDAP database to the new one through use of the pdbedit. See
Section 10.3.2.
In alphabetical order, these are the parameters eliminated for Samba 3.0.20.
• admin log
• alternate permissions
• character set
• client codepage
• code page directory
• coding system
• domain admin group
• domain guest group
• force unknown acl user
• nt smb support
• post script
• printer driver
• printer driver file
• printer driver location
• status
658 Upgrading from Samba-2.x to Samba-3.0.20 Chapter 33
• strip dot
• total print jobs
• use rhosts
• valid chars
• vfs options
• client signing
• client use spnego
• disable netbios
• ntlm auth
• paranoid server security
• server schannel
• server signing
• smb ports
• use spnego
File Service
• get quota command
• hide special files
• hide unwriteable files
• hostname lookups
• kernel change notify
• mangle prefix
• map acl inherit
• msdfs proxy
• set quota command
• use sendfile
• vfs objects
Printing
• max reported print jobs
Unicode and Character Sets
• display charset
• dos charset
• unicode
660 Upgrading from Samba-2.x to Samba-3.0.20 Chapter 33
• UNIX charset
SID to UID/GID Mappings
• idmap backend
• idmap gid
• idmap uid
• winbind enable local accounts
• winbind trusted domains only
• template primary group
• enable rid algorithm
LDAP
• ldap delete dn
• ldap group suffix
• ldap idmap suffix
• ldap machine suffix
• ldap passwd sync
• ldap user suffix
General Configuration
• preload modules
• privatedir
33.4.1 Databases
The following issues are known changes in behavior between Samba-2.2 and
Samba-3 that may affect certain installations of Samba.
1. When operating as a member of a Windows domain, Samba-2.2 would
map any users authenticated by the remote DC to the “guest account”
if a UID could not be obtained via the getpwnam() call. Samba-3 re-
jects the connection as
NT STATUS LOGON FAILURE. There is no current workaround to
re-establish the Samba-2.2 behavior.
2. When adding machines to a Samba-2.2 controlled domain, the “add
user script” was used to create the UNIX identity of the machine trust
account. Samba-3 introduces a new “add machine script” that must
be specified for this purpose. Samba-3 will not fall back to using the
“add user script” in the absence of an “add machine script”.
There have been a few new changes that Samba administrators should be
aware of when moving to Samba-3.
1. Encrypted passwords have been enabled by default in order to inter-
operate better with out-of-the-box Windows client installations. This
does mean that either (a) a Samba account must be created for each
user, or (b) “encrypt passwords = no” must be explicitly defined in
smb.conf.
2. Inclusion of new security = ads option for integration with an Active
Directory domain using the native Windows Kerberos 5 and LDAP
protocols.
Samba-3 also includes the possibility of setting up chains of authentication
methods (auth methods) and account storage backends (passdb backend ).
Please refer to the smb.conf man page and Chapter 10, Chapter 10, “Ac-
count Information Databases”, for details. While both parameters assume
sane default values, it is likely that you will need to understand what the
values actually mean in order to ensure Samba operates correctly.
Certain functions of the smbpasswd tool have been split between the new
smbpasswd utility, the net tool, and the new pdbedit utility. See the
Section 33.4. New Functionality 663
33.4.4 LDAP
$ smbpasswd -S <DOMAINNAME>
The old sambaAccount schema may still be used by specifying the ldap-
sam compat passdb backend. However, the sambaAccount and associated
attributes have been moved to the historical section of the schema file and
must be uncommented before use if needed. The Samba-2.2 object class dec-
laration for a sambaAccount has not changed in the Samba-3 samba.schema
file.
Other new object classes and their uses include:
664 Upgrading from Samba-2.x to Samba-3.0.20 Chapter 33
The following new smb.conf parameters have been added to aid in directing
certain LDAP queries when passdb backend = ldapsam://... has been
specified.
Samba-3 supports an LDAP backend for the idmap subsystem. The fol-
lowing options inform Samba that the idmap table should be stored on the
directory server onterose in the ou=idmap,dc=quenya,dc=org partition.
[ global ]
... idmap backend = l d a p : l d a p : / / o n t e r o s e /
l d a p idmap s u f f i x = ou=idmap , dc=quenya , dc= ←-
org
idmap u i d = 40000 −50000
idmap g i d = 40000 −50000
This is a rough guide to assist those wishing to migrate from NT4 domain
control to Samba-3-based domain control.
In the IT world there is often a saying that all problems are encountered
because of poor planning. The corollary to this saying is that not all prob-
lems can be anticipated and planned for. Then again, good planning will
anticipate most show-stopper-type situations.
34.1.1 Objectives
The key objective for most organizations is to make the migration from MS
Windows NT4 to Samba-3 domain control as painless as possible. One of
the challenges you may experience in your migration process may well be
convincing management that the new environment should remain in place.
Many who have introduced open source technologies have experienced pres-
sure to return to a Microsoft-based platform solution at the first sign of
trouble.
666
Section 34.1. Planning and Getting Started 667
and the PDC. If it’s long (more than 100 ms), locate a BDC on the remote
segment to serve as the local authentication and access control server.
There are cardinal rules to effective network design that cannot be broken
with impunity. The most important rule: Simplicity is king in every well-
controlled network. Every part of the infrastructure must be managed; the
more complex it is, the greater will be the demand of keeping systems secure
and functional.
Keep in mind the nature of how data must be shared. Physical disk space
layout should be considered carefully. Some data must be backed up. The
simpler the disk layout, the easier it will be to keep track of backup needs.
Identify what backup media will meet your needs; consider backup to tape,
CD-ROM or DVD-ROM, or other offline storage medium. Plan and imple-
ment for minimum maintenance. Leave nothing to chance in your design;
above all, do not leave backups to chance: backup, test, and validate every
backup; create a disaster recovery plan and prove that it works.
Users should be grouped according to data access control needs. File and
directory access is best controlled via group permissions, and the use of
the “sticky bit” on group-controlled directories may substantially avoid file
access complaints from Samba share users.
Logon scripts can help to ensure that all users gain the share and printer
connections they need.
670 Migration from NT4 PDC to Samba-3 PDC Chapter 34
Logon scripts can be created on the fly so all commands executed are specific
to the rights and privileges granted to the user. The preferred controls
should be effected through group membership so group information can be
used to create a custom logon script using the root preexec parameters to
the NETLOGON share.
Some sites prefer to use a tool such as kixstart to establish a controlled
user environment. In any case, you may wish to do a Google search for
logon script process controls. In particular, you may wish to explore the use
of the Microsoft Knowledge Base article KB189105 that deals with how to
add printers without user intervention via the logon script process.
User and group profiles may be migrated using the tools described in the
section titled Desktop Profile Management.
Profiles may also be managed using the Samba-3 tool profiles. This tool
allows the MS Windows NT-style security identifiers (SIDs) that are stored
inside the profile NTuser.DAT file to be changed to the SID of the Samba-3
domain.
1. Create a BDC account in the old NT4 domain for the Samba server
using NT Server Manager. Samba must not be running.
5. Now assign each of the UNIX groups to NT groups: (It may be useful
to copy this text to a script called initGroups.sh)
#!/bin/bash
#### Keep this as a shell script for future re-use
There are three basic choices for sites that intend to migrate from MS Win-
dows NT4 to Samba-3:
• Simple conversion (total replacement).
• Upgraded conversion (could be one of integration).
• Complete redesign (completely new solution).
Minimize downstream problems by:
• Taking sufficient time.
• Avoiding panic.
• Testing all assumptions.
• Testing the full roll-out program, including workstation deployment.
Table 34.2 lists the conversion choices given the type of migration being
contemplated.
Policies (migrate or create new ones) Exercise great caution when mak-
ing registry changes; use the right tool and be aware that changes made
674 Migration from NT4 PDC to Samba-3 PDC Chapter 34
There are many and varied opinions regarding the usefulness of SWAT. No
matter how hard one tries to produce the perfect configuration tool, it re-
mains an object of personal taste. SWAT is a tool that allows Web-based
configuration of Samba. It has a wizard that may help to get Samba con-
figured quickly, it has context-sensitive help on each smb.conf parameter,
it provides for monitoring of current state of connection information, and it
allows networkwide MS Windows network password management.
SWAT is a facility that is part of the Samba suite. The main executable
is called swat and is invoked by the internetworking super daemon. See
Section 35.2.2 for details.
SWAT uses integral Samba components to locate parameters supported by
the particular version of Samba. Unlike tools and utilities that are external
to Samba, SWAT is always up to date as known Samba parameters change.
SWAT provides context-sensitive help for each configuration parameter, di-
rectly from man page entries.
Some network administrators believe that it is a good idea to write systems
documentation inside configuration files, and for them SWAT will always be
a nasty tool. SWAT does not store the configuration file in any intermediate
form; rather, it stores only the parameter settings, so when SWAT writes
the smb.conf file to disk, it writes only those parameters that are at other
676
Section 35.2. Guidelines and Technical Tips 677
than the default settings. The result is that all comments, as well as pa-
rameters that are no longer supported, will be lost from the smb.conf file.
Additionally, the parameters will be written back in internal ordering.
Note
This section aims to unlock the dark secrets behind how SWAT may be
made to work, how it can be made more secure, and how to solve interna-
tionalization support problems.
The very first step that should be taken before attempting to configure a
host system for SWAT operation is to check that it is installed. This may
seem a trivial point to some, but several Linux distributions do not install
SWAT by default, even though they do ship an installable binary support
package containing SWAT on the distribution media.
Finally, when you are sure that SWAT has been fully installed, please check
that SWAT is enabled in the control file for the internetworking super-
daemon (inetd or xinetd) that is used on your operating system platform.
678 SWAT: The Samba Web Administration Tool Chapter 35
To validate that SWAT is installed, first locate the swat binary file on the
system. It may be found under the following directories:
/usr/local/samba/bin — the default Samba location
/usr/sbin — the default location on most Linux systems
/opt/samba/bin
The actual location is much dependent on the choice of the operating system
vendor or as determined by the administrator who compiled and installed
Samba.
There are a number of methods that may be used to locate the swat binary
file. The following methods may be helpful.
frodo:~ # swat -?
Usage: swat [OPTION...]
-a, --disable-authentication Disable authentication (demo mode)
Help options:
-?, --help Show this help message
--usage Display brief usage message
Now that you have found that swat is in the search path, it is easy to
identify where the file is located. Here is another simple way this may be
done:
Section 35.2. Guidelines and Technical Tips 679
If the above measures fail to locate the swat binary, another approach is
needed. The following may be used:
This list shows that there is a control file for xinetd, the internetwork super-
daemon that is installed on this server. The location of the SWAT binary
file is /usr/sbin/swat, and the support files for it are located under the
directory /usr/share/samba/swat.
We must now check where swat expects to find its support files. This can
be done as follows:
/usr/share/samba/swat/lang/ja/help/welcome.html
/usr/share/samba/swat/lang/ja/images
/usr/share/samba/swat/lang/ja/images/home.gif
...
/usr/share/samba/swat/lang/ja/include
/usr/share/samba/swat/lang/ja/include/header.nocss.html
...
/usr/share/samba/swat/lang/tr
/usr/share/samba/swat/lang/tr/help
/usr/share/samba/swat/lang/tr/help/welcome.html
/usr/share/samba/swat/lang/tr/images
/usr/share/samba/swat/lang/tr/images/home.gif
...
/usr/share/samba/swat/lang/tr/include
/usr/share/samba/swat/lang/tr/include/header.html
/usr/share/samba/swat/using_samba
...
/usr/share/samba/swat/images
/usr/share/samba/swat/images/home.gif
...
/usr/share/samba/swat/include
/usr/share/samba/swat/include/footer.html
/usr/share/samba/swat/include/header.html
jht@frodo:/>
If the files needed are not available, it is necessary to obtain and install them
before SWAT can be used.
# default: off
# description: SWAT is the Samba Web Admin Tool. Use swat \
# to configure your Samba server. To use SWAT, \
# connect to port 901 with your favorite web browser.
service swat
{
port = 901
socket_type = stream
wait = no
only_from = localhost
user = root
server = /usr/sbin/swat
log_on_failure += USERID
disable = no
}
In the above, the default setting for disable is yes. This means that SWAT
is disabled. To enable use of SWAT, set this parameter to no as shown.
Both of the previous examples assume that the swat binary has been located
in the /usr/sbin directory. In addition to the above, SWAT will use a
directory access point from which it will load its Help files as well as other
control information. The default location for this on most Linux systems
is in the directory /usr/share/samba/swat. The default location using
Samba defaults will be /usr/local/samba/swat.
Access to SWAT will prompt for a logon. If you log onto SWAT as any
non-root user, the only permission allowed is to view certain aspects of
configuration as well as access to the password change facility. The buttons
that will be exposed to the non-root user are HOME, STATUS, VIEW, and
PASSWORD. The only page that allows change capability in this case is
PASSWORD.
As long as you log onto SWAT as the user root, you should obtain full change
682 SWAT: The Samba Web Administration Tool Chapter 35
and commit ability. The buttons that will be exposed include HOME,
GLOBALS, SHARES, PRINTERS, WIZARD, STATUS, VIEW, and PASS-
WORD.
Many people have asked about how to set up SWAT with SSL to allow
for secure remote administration of Samba. Here is a method that works,
courtesy of Markus Krieger.
Modifications to the SWAT setup are as follows:
1. Install OpenSSL.
2. Generate certificate and private key.
SWAT can be configured to display its messages to match the settings of the
language configurations of your Web browser. It will be passed to SWAT in
the Accept-Language header of the HTTP request.
To enable this feature:
• Install the proper msg files from the Samba source/po directory into
$LIBDIR.
Section 35.3. Overview and Quick Tour 683
The name of the msg file is the same as the language ID sent by the browser.
For example, en means English, ja means Japanese, fr means French.
If you do not like some of messages, or there are no msg files for your locale,
you can create them simply by copying the en.msg files to the directory for
“your language ID.msg” and filling in proper strings to each “msgstr”. For
example, in it.msg, the msg file for the Italian locale, just set:
and so on. If you find a mistake or create a new msg file, please email it to
us so we will consider it in the next release of Samba. The msg file should
be encoded in UTF-8.
Note that if you enable this feature and the display charset is not matched
to your browser’s setting, the SWAT display may be corrupted. In a future
version of Samba, SWAT will always display messages with UTF-8 encoding.
You will then not need to set this smb.conf file parameter.
SWAT is a tool that may be used to configure Samba or just to obtain useful
links to important reference materials such as the contents of this book as
well as other documents that have been found useful for solving Windows
networking problems.
The SWAT title page provides access to the latest Samba documentation.
The manual page for each Samba component is accessible from this page, as
are the Samba HOWTO-Collection (this document) as well as the O’Reilly
book “Using Samba.”
are available from the SWAT home page also. One diagnostic tool that is
not mentioned on this page but that is particularly useful is ethereal1 .
Warning
The GLOBALS button exposes a page that allows configuration of the global
parameters in smb.conf. There are two levels of exposure of the parameters:
To switch to other than Basic editing ability, click on Advanced. You may
also do this by clicking on the radio button, then click on the Commit
Changes button.
After making any changes to configuration parameters, make sure that you
click on the Commit Changes button before moving to another area; oth-
erwise, your changes will be lost.
1
<http://www.ethereal.com/>
Section 35.3. Overview and Quick Tour 685
Note
The Edit button permits the editing (setting) of the minimal set of options
that may be necessary to create a working Samba server.
Finally, there are a limited set of options that determine what type of server
Samba will be configured for, whether it will be a WINS server, participate
as a WINS client, or operate with no WINS support. By clicking one button,
you can elect to expose (or not) user home directories.
The status page serves a limited purpose. First, it allows control of the
Samba daemons. The key daemons that create the Samba server environ-
ment are smbd, nmbd, and winbindd.
Finally, the status page may be used to terminate specific smbd client con-
nections in order to free files that may be locked.
The view page allows you to view the optimized smb.conf file and, if you
are particularly masochistic, permits you also to see all possible global con-
figuration parameters and their settings.
The password change page is a popular tool that allows the creation, dele-
tion, deactivation, and reactivation of MS Windows networking users on the
local machine. You can also use this tool to change a local password for a
user account.
When logged in as a non-root account, the user must provide the old pass-
word as well as the new password (twice). When logged in as root, only the
new password is required.
Section 35.3. Overview and Quick Tour 687
One popular use for this tool is to change user passwords across a range of
remote MS Windows servers.
Part V
Troubleshooting
Chapter 36
36.1 Introduction
This file contains a list of tests you can perform to validate your Samba
server. It also tells you what the likely cause of the problem is if it fails any
one of these steps. If it passes all these tests, then it is probably working
fine.
You should do all the tests in the order shown. We have tried to carefully
choose them so later tests only use capabilities verified in the earlier tests.
However, do not stop at the first error: there have been some instances when
continuing with the tests has helped to solve a problem.
If you send one of the Samba mailing lists an email saying, “It does not
work,” and you have not followed this test procedure, you should not be
surprised if your email is ignored.
36.2 Assumptions
In all of the tests, it is assumed you have a Samba server called BIGSERVER
and a PC called ACLIENT, both in workgroup TESTGROUP.
The procedure is similar for other types of clients.
It is also assumed you know the name of an available share in your smb.
conf. I for our examples this share is called tmp. You can add a tmp share
like this by adding the lines shown in Example 36.2.1.
689
690 The Samba Checklist Chapter 36
Note
Please pay attention to the error messages you receive. If any error message
reports that your server is being unfriendly, you should first check that your
IP name resolution is correctly set up. Make sure your /etc/resolv.conf
file points to name servers that really do exist.
Also, if you do not have DNS server access for name resolution, please check
that the settings for your smb.conf file results in dns proxy = no. The
best way to check this is with testparm smb.conf.
It is helpful to monitor the log files during testing by using the tail -F
log file name in a separate terminal console (use ctrl-alt-F1 through F6 or
multiple terminals in X). Relevant log files can be found (for default instal-
lations) in /usr/local/samba/var. Also, connection logs from machines
can be found here or possibly in /var/log/samba, depending on how or if
you specified logging in your smb.conf file.
If you make changes to your smb.conf file while going through these test,
remember to restart smbd and nmbd.
Note
2. Run the command ping BIGSERVER from the PC and ping ACLIENT
from the UNIX box. If you do not get a valid response, then your
TCP/IP software is not correctly installed. You will need to start a
“DOS prompt” window on the PC to run ping. If you get a message
saying “host not found” or a similar message, then your DNS software
or /etc/hosts file is not correctly set up. It is possible to run Samba
without DNS entries for the server and client, but it is assumed you do
have correct entries for the remainder of these tests. Another reason
why ping might fail is if your host is running firewall software. You will
need to relax the rules to let in the workstation in question, perhaps
by allowing access from another subnet (on Linux this is done via the
appropriate firewall maintenance commands ipchains or iptables).
Note
frodo:~ # iptables -L -v
Chain INPUT (policy DROP 98496 packets, 12M bytes)
pkts bytes target prot opt in out source destination
692 The Samba Checklist Chapter 36
Note
If you get a message saying “session request failed,” the server refused
the connection. If it says “Your server software is being unfriendly,”
then it’s probably because you have invalid command line parameters
to smbd, or a similar fatal problem with the initial startup of smbd.
Also check your config file (smb.conf) for syntax errors with testparm
and that the various directories where Samba keeps its log and lock
files exist. There are a number of reasons for which smbd may refuse
or decline a session request. The most common of these involve one
or more of the smb.conf file entries as shown in Example 36.3.1.
In Example 36.3.1, no allowance has been made for any session re-
quests that will automatically translate to the loopback adapter ad-
dress 127.0.0.1. To solve this problem, change these lines as shown in
Example 36.3.2.
you will need to use the -B option to set the broadcast address to that
of the PC’s subnet. This test will probably fail if your subnet mask
and broadcast address are not correct. (Refer to test 3 notes above).
7. Run the command smbclient //BIGSERVER/TMP. You should
then be prompted for a password. You should use the password of
the account with which you are logged into the UNIX box. If you
want to test with another account, then add the -U accountname
option to the end of the command line — for example, smbclient
//bigserver/tmp -Ujohndoe.
Note
Once you enter the password, you should get the smb> prompt. If you
do not, then look at the error message. If it says “invalid network
name,” then the service tmp is not correctly set up in your smb.conf.
If it says “bad password,” then the likely causes are:
(a) You have shadow passwords (or some other password system) but
didn’t compile in support for them in smbd.
(b) Your valid users configuration is incorrect.
(c) You have a mixed-case password and you haven’t enabled the
password level option at a high enough level.
(d) The path line in smb.conf is incorrect. Check it with testparm.
(e) You enabled password encryption but didn’t map UNIX to Samba
users. Run smbpasswd -a username
Once connected, you should be able to use the commands dir, get,
put, and so on. Type help command for instructions. You should
especially check that the amount of free disk space shown is correct
when you type dir.
8. On the PC, type the command net view \\BIGSERVER. You will
need to do this from within a DOS prompt window. You should get
696 The Samba Checklist Chapter 36
(c) Enable Windows name resolution via DNS in the advanced sec-
tion of the TCP/IP setup.
There are many sources of information available in the form of mailing lists,
RFCs, and documentation. The documentation that comes with the Samba
distribution contains good explanations of general SMB topics such as brows-
ing.
With SMB networking, it is often not immediately clear what the cause is
of a certain problem. Samba itself provides rather useful information, but
in some cases you might have to fall back to using a sniffer. A sniffer is a
program that listens on your LAN, analyzes the data sent on it, and displays
it on the screen.
One of the best diagnostic tools for debugging problems is Samba itself. You
can use the -d option for both smbd and nmbd to specify the debug level
at which to run. See the man pages for smbd, nmbd, and smb.conf for
more information regarding debugging options. The debug level can range
from 1 (the default) to 10 (100 for debugging passwords).
Another helpful method of debugging is to compile Samba using the gcc
-g flag. This will include debug information in the binaries and allow you
to attach gdb to the running smbd/nmbd process. To attach gdb to an
smbd process for an NT workstation, first get the workstation to make
698
Section 37.1. Diagnostics Tools 699
the connection. Pressing ctrl-alt-delete and going down to the domain box
is sufficient (at least, the first time you join the domain) to generate a
LsaEnumTrustedDomains. Thereafter, the workstation maintains an open
connection and there will be an smbd process running (assuming that you
haven’t set a really short smbd idle timeout). So, in between pressing ctrl-
alt-delete and actually typing in your password, you can attach gdb and
continue.
Some useful Samba commands worth investigating are:
$ testparm | more
$ smbclient -L //{netbios name of server}
37.1.2 Tcpdump
Tcpdump1 was the first UNIX sniffer with SMB support. It is a command-
line utility and now, its SMB support is somewhat lagging that of ethereal
and tethereal.
37.1.3 Ethereal
Ethereal2 is a graphical sniffer, available for both UNIX (Gtk) and Windows.
Ethereal’s SMB support is quite good. For details on the use of ethereal,
read the well-written Ethereal User Guide.
Listen for data on ports 137, 138, 139, and 445. For example, use the fil-
ter port 137, port 138, port 139, or port 445 as seen in Figure 37.1
snapshot.
A console version of ethereal is available as well and is called tethereal.
For tracing things on Microsoft Windows NT, Network Monitor (aka Net-
mon) is available on Microsoft Developer Network CDs, the Windows NT
Server install CD, and the SMS CDs. The version of Netmon that ships with
1
<http://www.tcpdump.org/>
2
<http://www.ethereal.com/>
700 Analyzing and Solving Samba Problems Chapter 37
SMS allows for dumping packets between any two computers (i.e., placing
the network interface in promiscuous mode). The version on the NT Server
install CD will only allow monitoring of network traffic directed to the local
NT box and broadcasts on the local subnet. Be aware that Ethereal can
read and write Netmon formatted files.
.
• See how Scott Merrill simulates a BDC behavior at http://www.skippy.net/linux/smb-
howto.html3 .
• FTP site for older SMB specs, ftp://ftp.microsoft.com/developr/drg/CIFS/4
• Always remember that the developers are volunteers; they are not
paid and they never guarantee to produce a particular feature at a
particular time. Any timelines are “best guess,” and nothing more.
• Always mention what version of Samba you are using and what oper-
ating system it’s running under. You should list the relevant sections
of your smb.conf file, at least the options in [global] that affect PDC
support.
• Try to make your questions clear and brief. Lots of long, convoluted
questions get deleted before they are completely read! Do not post
HTML-encoded messages. Most people on mailing lists simply delete
them.
• If you run one of those nifty “I’m on holiday” things when you are
away, make sure its configured to not answer mailing list traffic. Au-
toresponses to mailing lists really irritate the thousands of people who
end up having to deal with such bad netiquet bahavior.
• Don’t cross post. Work out which is the best list to post to and
see what happens. Do not post to both samba-ntdom and samba-
technical. Many people active on the lists subscribe to more than one
list and get annoyed to see the same message two or more times. Often
someone who thinks a message would be better dealt with on another
list will forward it on for you.
• You might include partial log files written at a debug level set to as
much as 20. Please do not send the entire log but just enough to give
the context of the error messages.
• If you have a complete Netmon trace (from the opening of the pipe to
the error), you can send the *.CAP file as well.
To have your name removed from a Samba mailing list, go to the same place
where you went to subscribe to it, go to http://lists.samba.org5 , click on
your nearest mirror, click on Support, and then click on Samba-related
mailing lists.
Please do not post messages to the list asking to be removed. You will only
be referred to the above address (unless that process failed in some way).
5
<http://lists.samba.org/>
Chapter 38
REPORTING BUGS
38.1 Introduction
Please report bugs using Samba’s Bugzilla1 facilities and take the time to
read this file before you submit a bug report. Also, check to see if it has
changed between releases, as we may be changing the bug reporting mech-
anism at some point.
Please do as much as you can yourself to help track down the bug. Samba
is maintained by a dedicated group of people who volunteer their time,
skills, and efforts. We receive far more mail than we can possibly answer,
so you have a much higher chance of a response and a fix if you send us a
“developer-friendly” bug report that lets us fix it fast.
If you post the bug to the comp.protocols.smb newsgroup or the mailing
list, do not assume that we will read it. If you suspect that your problem is
not a bug but a configuration problem, it is better to send it to the Samba
mailing list, as there are thousands of other users on that list who may be
able to help you.
You may also like to look though the recent mailing list archives, which are
conveniently accessible on the Samba Web pages at <http://samba.org/
samba/>.
Before submitting a bug report, check your config for silly errors. Look in
your log files for obvious messages that tell you’ve misconfigured something.
1
<https://bugzilla.samba.org/>
705
706 Reporting Bugs Chapter 38
Have you looked through Chapter 36, “The Samba Checklist”? This is
extremely important.
If you include part of a log file with your bug report, then be sure to annotate
it with exactly what you were doing on the client at the time and exactly
what the results were.
To set the debug level, use the log level in your smb.conf. You may also find
it useful to set the log level higher for just one machine and keep separate
logs for each machine. To do this, add the following lines to your main smb.
conf file:
l o g l e v e l = 10
l o g f i l e = / u s r / l o c a l /samba/ l i b / l o g .%m
i n c l u d e = / u s r / l o c a l /samba/ l i b /smb . c o n f .%m
The smb.conf entry log level is synonymous with the parameter debuglevel
that has been used in older versions of Samba and is being retained for
backward compatibility of smb.conf files.
As the log level value is increased, you will record a significantly greater
level of debugging information. For most debugging operations, you may
not need a setting higher than 3. Nearly all bugs can be tracked at a setting
of 10, but be prepared for a large volume of log data.
Section 38.4. Internal Errors 707
This will cause the level of detail to be expanded to the debug class (log
level) passed to each functional area per the value shown above. The first
value passed to the log level of 0 means turn off all unnecessary debugging
except the debug classes set for the functional areas as specified. The table
shown in Table 38.1 may be used to attain very precise analysis of each SMB
operation Samba is conducting.
If you get the message “INTERNAL ERROR” in your log files, it means
that Samba got an unexpected signal while running. It is probably a seg-
mentation fault and almost certainly means a bug in Samba (unless you
have faulty hardware or system software).
If the message came from smbd, it will probably be accompanied by a mes-
sage that details the last SMB message received by smbd. This information
is often useful in tracking down the problem, so please include it in your bug
report.
708 Reporting Bugs Chapter 38
You should also detail how to reproduce the problem, if possible. Please
make this reasonably detailed.
You may also find that a core file appeared in a corefiles subdirectory
of the directory where you keep your Samba log files. This file is the most
useful tool for tracking down the bug. To use it, you do this:
adding appropriate paths to smbd and core so gdb can find them. If you do
not have gdb, try dbx. Then within the debugger, use the command where
to give a stack trace of where the problem occurred. Include this in your
report.
If you know any assembly language, do a disass of the routine where the
problem occurred (if it’s in a library routine, then disassemble the routine
that called it) and try to work out exactly where the problem is by looking
at the surrounding code. Even if you do not know assembly, including this
information in the bug report can be useful.
Unfortunately, some UNIXes (in particular some recent Linux kernels) refuse
to dump a core file if the task has changed UID (which smbd does often).
To debug with this sort of system, you could try to attach to the running
process using gdb smbd PID, where you get PID from smbstatus. Then use
c to continue and try to cause the core dump using the client. The debugger
should catch the fault and tell you where it occurred.
Compile with -g to ensure you have symbols in place. Add the following
line to the smb.conf file global section:
to catch any panics. If smbd seems to be frozen, look for any sleep processes.
If it is not, and appears to be spinning, find the PID of the spinning process
and type:
gdb /usr/local/samba/sbin/smbd
then “attach ‘pid’” (of the spinning process), then type “bt” to get a back-
trace to see where the smbd is in the call path.
38.6 Patches
The best sort of bug report is one that includes a fix! If you send us patches,
please use diff -u format if your version of diff supports it; otherwise, use
diff -c4. Make sure you do the diff against a clean version of the source
and let me know exactly what version you used.
Part VI
Reference Section
Chapter 39
You can obtain the Samba source file from the Samba Web site1 . To obtain
a development version, you can download Samba from Subversion or using
rsync.
39.1.1 Introduction
711
712 How to Compile Samba Chapter 39
You can access the source code via your favorite WWW browser. This allows
you to access the contents of individual files in the repository and also to
look at the revision history and commit logs of individual files. You can also
ask for a diff listing between any two versions on the repository.
You can also access the source code via a normal Subversion client. This
gives you much more control over what you can do with the repository and
allows you to check out whole source trees and keep them up to date via
normal Subversion commands. This is the preferred method of access if you
are a developer and not just a casual browser.
In order to be able to download the Samba sources off Subversion, you need a
Subversion client. Your distribution might include one, or you can download
the sources from <http://subversion.tigris.org/>.
To gain access via anonymous Subversion, use the following steps. Retriev-
ing Samba using Subversion
3. Whenever you want to merge in the latest code changes, use the fol-
lowing command from within the Samba directory: svn update
Section 39.2. Accessing the Samba Sources via rsync and ftp 713
The disadvantage of the unpacked trees is that they do not support auto-
matic merging of local changes as Subversion does. rsync access is most
convenient for an initial install.
It is strongly recommended that you verify the PGP signature for any source
file before installing it. Even if you’re not downloading from a mirror site,
verifying PGP signatures should be a standard reflex. Many people today
use the GNU GPG tool set in place of PGP. GPG can substitute for PGP.
$ wget http://us1.samba.org/samba/ftp/samba-3.0.20.tar.asc
$ wget http://us1.samba.org/samba/ftp/samba-pubkey.asc
The first file is the PGP signature for the Samba source file; the other is the
Samba public PGP key itself. Import the public PGP key with:
$ gzip -d samba-3.0.20.tar.gz
$ gpg --verify samba-3.0.20.tar.asc
3
<ftp://pserver.samba.org/pub/unpacked>
4
<rsync://pserver.samba.org/ftp/unpacked/>
5
<http://rsync.samba.org/>
714 How to Compile Samba Chapter 39
After the source tarball has been unpacked, the next step involves configu-
ration to match Samba to your operating system platform. If your source
directory does not contain the configure script, it is necessary to build it
before you can continue. Building of the configure script requires the correct
version of the autoconf tool kit. Where the necessary version of autoconf is
present, the configure script can be generated by executing the following:
root# cd samba-3.0.20
root# ./autogen.sh
To build the binaries, run the program ./configure in the source directory.
This should automatically configure Samba for your operating system. If you
have unusual needs, then you may wish to first run:
This will help you to see what special options can be enabled. Now execute
./configure with any arguments it might need:
root# make
Section 39.4. Building the Binaries 715
Once it is successfully compiled, you can execute the command shown here
to install the binaries and manual pages:
Some people prefer to install binary files and man pages separately. If this
is your wish, the binary files can be installed by executing:
Note that if you are upgrading from a previous version of Samba the old
versions of the binaries will be renamed with an “.old” extension. You can
go back to the previous version by executing:
As you can see from this, building and installing Samba does not need to
result in disaster!
In order to compile Samba with ADS support, you need to have installed
on your system:
After you run configure, make sure that the include/config.h it generates
contain lines like this:
#define HAVE_KRB5 1
#define HAVE_LDAP 1
If it does not, configure did not find your KRB5 libraries or your LDAP
libraries. Look in config.log to figure out why and fix it.
SuSE Linux installs Heimdal packages that may be required to allow you
to build binary packages. You should verify that the development libraries
have been installed on your system.
SuSE Linux Samba RPMs support Kerberos. Please refer to the documen-
tation for your SuSE Linux system for information regarding SuSE Linux
Section 39.5. Starting the smbd nmbd and winbindd 717
You must choose to start smbd, winbindd and nmbd either as daemons or
from inetd. Don’t try to do both! Either you can put them in inetd.conf
and have them started on demand by inetd or xinetd, or you can start them
as daemons either from the command-line or in /etc/rc.local. See the
man pages for details on the command line options. Take particular care
to read the bit about what user you need to have to start Samba. In many
cases, you must be root.
The main advantage of starting smbd and nmbd using the recommended
daemon method is that they will respond slightly more quickly to an initial
connection request.
Note
Some distributions use xinetd instead of inetd. Consult the xinetd manual
for configuration information.
Note
Note
Warning
To start the server as a daemon, you should create a script something like
this one, perhaps calling it startsmb.
#!/bin/sh
/usr/local/samba/bin/smbd -D
/usr/local/samba/bin/winbindd
/usr/local/samba/bin/nmbd -D
To kill it, send a kill signal to the processes nmbd and smbd.
Note
If you use the SVR4-style init system, you may like to look
at the examples/svr4-startup script to make Samba
fit into that system.
Red Hat Linux has not always included all Samba components in the stan-
dard installation. So versions of Red Hat Linux do not install the winbind
utility, even though it is present on the installation CDROM media. Check
to see if the winbindd is present on the system:
root# ls /usr/sbin/winbindd
720 How to Compile Samba Chapter 39
/usr/sbin/winbindd
This means that the appropriate RPM package was installed. The following
response means that it is not installed:
Novell SUSE Linux products automatically install all essential Samba com-
ponents in a default installation. Configure your smb.conf file, then execute
the following to start Samba:
PORTABILITY
Samba works on a wide range of platforms, but the interface all the platforms
provide is not always compatible. This chapter contains platform-specific
information about compiling and using Samba.
40.1 HPUX
If you encounter this problem, make sure the programs that are failing to
initgroups() are run as users, not in any groups with GIDs outside the
allowed range.
On HP-UX you must use gcc or the HP ANSI compiler. The free compiler
that comes with HP-UX is not ANSI compliant and cannot compile Samba.
722
Section 40.2. SCO UNIX 723
If you run an old version of SCO UNIX, you may need to get important
TCP/IP patches for Samba to work correctly. Without the patch, you may
encounter corrupt data transfers using Samba.
The patch you need is UOD385 Connection Drivers SLS. It is available from
SCO ftp.sco.com1 , directory SLS, files uod385a.Z and uod385a.ltr.Z).
The information provided here refers to an old version of SCO UNIX. If
you require binaries for more recent SCO UNIX products, please contact
SCO to obtain packages that are ready to install. You should also verify
with SCO that your platform is up to date for the binary packages you
will install. This is important if you wish to avoid data corruption problems
with your installation. To build Samba for SCO UNIX products may require
significant patching of Samba source code. It is much easier to obtain binary
packages directly from SCO.
40.3 DNIX
DNIX has a problem with seteuid() and setegid(). These routines are needed
for Samba to work correctly, but they were left out of the DNIX C library
for some reason.
For this reason Samba by default defines the macro NO EID in the DNIX
section of includes.h. This works around the problem in a limited way, but
it is far from ideal, and some things still will not work right.
To fix the problem properly, you need to assemble the following two functions
and then either add them to your C library or link them into Samba. Put
the following in the file setegid.s:
.globl _setegid
_setegid:
moveq #47,d0
movl #100,a0
moveq #1,d1
movl 4(sp),a1
trap #9
1
<ftp://ftp.sco.com/>
724 Portability Chapter 40
bccs 1$
jmp cerror
1$:
clrl d0
rts
.globl _seteuid
_seteuid:
moveq #47,d0
movl #100,a0
moveq #0,d1
movl 4(sp),a1
trap #9
bccs 1$
jmp cerror
1$:
clrl d0
rts
$ as seteuid.s
$ as setegid.s
#define NO_EID
Section 40.4. Red Hat Linux 725
By default during installation, some versions of Red Hat Linux add an entry
to /etc/hosts as follows:
This causes Samba to loop back onto the loopback interface. The result
is that Samba fails to communicate correctly with the world and therefore
may fail to correctly negotiate who is the master browse list holder and who
is the master browser.
Corrective action: Delete the entry after the word ”loopback” in the line
starting 127.0.0.1.
40.6 Solaris
Sun released patches for Solaris 2.6, 8, and 9. The patch for Solaris 7 has
not been released yet.
The patch revision for 2.6 is 105181-34, for 8 is 108528-19, and for 9 is
112233-04.
After the installation of these patches, it is recommended to reconfigure and
rebuild Samba.
Thanks to Joe Meslovich for reporting this.
Nsswitch on Solaris 9 refuses to use the Winbind NSS module. This behavior
is fixed by Sun in patch 112960-142 .
2
<http://sunsolve.sun.com/search/advsearch.do?collection=PATCH&type=
collections&max=50&language=en&queryKey5=112960;rev=14&toDocument=yes>
Chapter 41
Yes. Thursby1 has a CIFS client/server called DAVE2 . They test it against
Windows 95, Windows NT/200x/XP, and Samba for compatibility issues.
At the time of this writing, DAVE was at version 4.1. Please refer to
Thursby’s Web site for more information regarding this product.
1
<http://www.thursby.com/>
2
<http://www.thursby.com/products/dave.html>
3
<http://www.umich.edu/~rsug/netatalk/>
4
<http://www.cs.mu.oz.au/appletalk/atalk.html>
5
<http://www.eats.com/linux_mac_win.html>
727
728 Samba and Other CIFS Clients Chapter 41
This sections deals with configuring OS/2 Warp 3 (not Connect), OS/2 1.2,
1.3 or 2.x.
You can use the free Microsoft LAN Manager 2.2c Client for OS/2 that
is available from ftp://ftp.microsoft.com/BusSys/Clients/LANMAN.OS2/6 .
In a nutshell, edit the file \OS2VER in the root directory of the OS/2 boot
partition and add the lines:
6
<ftp://ftp.microsoft.com/BusSys/Clients/LANMAN.OS2/>
Section 41.3. Windows for Workgroups 729
20=setup.exe
20=netwksta.sys
20=netvdd.sys
before you install the client. Also, do not use the included NE2000 driver be-
cause it is buggy. Try the NE2000 or NS2000 driver from ftp://ftp.cdrom.com/pub/os2/network/ndis/7
instead.
Use the latest TCP/IP stack from Microsoft if you use Windows for Work-
groups. The early TCP/IP stacks had lots of bugs.
7
<ftp://ftp.cdrom.com/pub/os2/network/ndis/>
730 Samba and Other CIFS Clients Chapter 41
Windows for Workgroups does a lousy job with passwords. When you change
passwords on either the UNIX box or the PC, the safest thing to do is delete
the .pwl files in the Windows directory. The PC will complain about not
finding the files, but will soon get over it, allowing you to enter the new
password.
If you do not do this, you may find that Windows for Workgroups remembers
and uses the old password, even if you told it a new one.
Often Windows for Workgroups will totally ignore a password you give it in
a dialog box.
There is a program call admincfg.exe on the last disk (disk 8) of the WFW
3.11 disk set. To install it, type EXPAND A:\ADMINCFG.EX C:\WINDOWS\ADMINCFG.
EXE. Then add an icon for it via the Program Manager New menu. This
program allows you to control how WFW handles passwords, Disable Pass-
word Caching and so on, for use with security = user.
To support print queue reporting, you may find that you have to use TCP/IP
as the default protocol under Windows for Workgroups. For some reason, if
you leave NetBEUI as the default, it may break the print queue reporting
on some systems. It is presumably a Windows for Workgroups bug.
Note that some people have found that setting DefaultRcvWindow in the
[MSTCP] section of the SYSTEM.INI file under Windows for Workgroups to
3072 gives a big improvement.
When using Windows 95 OEM SR2, the following updates are recommended
where Samba is being used. Please note that the above change (John, specify
the change???????) will affect you once these updates have been installed.
There are more updates than the ones mentioned here. Refer to the Mi-
crosoft Web site for all currently available updates to your specific version
of Windows 95.
Kernel Update: KRNLUPD.EXE
Ping Fix: PINGUPD.EXE
RPC Update: RPCRTUPD.EXE
TCP/IP Update: VIPUPD.EXE
Redirector Update: VRDRUPD.EXE
There are several annoyances with Windows 2000 SP2, one of which only
appears when using a Samba server to host user profiles to Windows 2000
SP2 clients in a Windows domain. This assumes that Samba is a member
of the domain, but the problem will most likely occur if it is not.
In order to serve profiles successfully to Windows 2000 SP2 clients (when not
operating as a PDC), Samba must have nt acl support = no added to the file
share that houses the roaming profiles. If this is not done, then the Windows
2000 SP2 client will complain about not being able to access the profile
(Access Denied) and create multiple copies of it on disk (DOMAIN.user.001,
DOMAIN.user.002, and so on). See the smb.conf man page for more details
on this option. Also note that the nt acl support parameter was formally a
global parameter in releases prior to Samba 2.2.2.
The reason for this bug is that the Windows 200x SP2 client copies the
security descriptor for the profile that contains the Samba server’s SID, and
not the domain SID. The client compares the SID for SAMBA\user and
realizes it is different from the one assigned to DOMAIN\user; hence, access
denied message.
Section 41.6. Windows NT 3.1 733
When the nt acl support parameter is disabled, Samba will send the Win-
dows 200x client a response to the QuerySecurityDescriptor trans2 call,
which causes the client to set a default ACL for the profile. This default
ACL includes:
DOMAIN\user “Full Control”>
Note
9
<http://support.microsoft.com/default.aspx?scid=kb;Q103765>
Chapter 42
SAMBA PERFORMANCE
TUNING
42.1 Comparisons
The Samba server uses TCP to talk to the client, so if you are trying to see
if it performs well, you should really compare it to programs that use the
same protocol. The most readily available programs for file transfer that
use TCP are ftp or another TCP-based SMB server.
If you want to test against something like an NT or Windows for Workgroups
server, then you will have to disable all but TCP on either the client or
server. Otherwise, you may well be using a totally different protocol (such
as NetBEUI) and comparisons may not be valid.
Generally, you should find that Samba performs similarly to ftp at raw
transfer speed. It should perform quite a bit faster than NFS, although this
depends on your system.
Several people have done comparisons between Samba and Novell, NFS, or
Windows NT. In some cases Samba performed the best, in others the worst.
I suspect the biggest factor is not Samba versus some other system, but the
hardware and drivers used on the various systems. Given similar hardware,
Samba should certainly be competitive in speed with other systems.
There are a number of socket options that can greatly affect the performance
of a TCP-based server like Samba.
734
Section 42.3. Read Size 735
The socket options that Samba uses are settable both on the command line
with the -O option and in the smb.conf file.
The socket options section of the smb.conf manual page describes how to
set these and gives recommendations.
Getting the socket options correct can make a big difference to your perfor-
mance, but getting them wrong can degrade it by just as much. The correct
settings are very dependent on your local network.
The socket option TCP NODELAY is the one that seems to make the
biggest single difference for most networks. Many people report that adding
socket options = TCP NODELAY doubles the read performance of a Samba
drive. The best explanation I have seen for this is that the Microsoft TCP/IP
stack is slow in sending TCP ACKs.
The option read size affects the overlap of disk reads/writes with network
reads/writes. If the amount of data being transferred in several of the SMB
commands (currently SMBwrite, SMBwriteX, and SMBreadbraw) is larger
than this value, then the server begins writing the data before it has received
the whole packet from the network, or in the case of SMBreadbraw, it begins
writing to the network before all the data has been read from disk.
This overlapping works best when the speeds of disk and network access are
similar, having little effect when the speed of one is much greater than the
other.
The default value is 16384, but little experimentation has been done as yet
to determine the optimal value, and it is likely that the best value will vary
greatly between systems anyway. A value over 65536 is pointless and will
cause you to allocate memory unnecessarily.
736 Samba Performance Tuning Chapter 42
At startup the client and server negotiate a maximum transmit size, which
limits the size of nearly all SMB commands. You can set the maximum size
that Samba will negotiate using the max xmit option in smb.conf. Note
that this is the maximum size of SMB requests that Samba will accept,
but not the maximum size that the client will accept. The client maximum
receive size is sent to Samba by the client, and Samba honors this limit.
It defaults to 65536 bytes (the maximum), but it is possible that some clients
may perform better with a smaller transmit unit. Trying values of less than
2048 is likely to cause severe problems. In most cases the default is the best
option.
If you set the log level (also known as debug level ) higher than 2, then you
may suffer a large drop in performance. This is because the server flushes
the log file after each operation, which can be quite expensive.
Some machines may find write raw slower than normal write, in which case
you may wish to change this option.
Slow logins are almost always due to the password checking time. Using the
lowest practical password level will improve things.
Often a speed problem can be traced to the client. The client (for example
Windows for Workgroups) can often be tuned for better TCP performance.
Check the sections on the various clients in Chapter 41, “Samba and Other
CIFS Clients”.
Grab the mii-tool and check the duplex settings on the NIC. My guess is
that it is a link layer issue, not an application layer problem. Also run
ifconfig and verify that the framing error, collisions, and so on, look normal
for ethernet. (John, should this be set off as an extract???????)
738 Samba Performance Tuning Chapter 42
Our Samba PDC server has been hosting three TB of data to our 500+ users
[Windows NT/XP] for the last three years using Samba without a problem.
Today all shares went very slow. Also, the main smbd kept spawning new
processes, so we had 1600+ running SMDB’s (normally we average 250). It
crashed the SUN E3500 cluster twice. After a lot of searching, I decided to
rm /var/locks/*.tdb. Happy again.
Question: Is there any method of keeping the *.tdb files in top condition,
or how can I detect early corruption?
Answer: Yes, run tdbbackup each time after stopping nmbd and before
starting nmbd.
Question: What I also would like to mention is that the service latency
seems a lot lower than before the locks cleanup. Any ideas on keeping it top
notch?
Answer: Yes. Same answer as for previous question!
A.1 Preamble
The licenses for most software are designed to take away your freedom to
share and change it. By contrast, the GNU General Public License is in-
tended to guarantee your freedom to share and change free software - to
make sure the software is free for all its users. This General Public Li-
cense applies to most of the Free Software Foundation’s software and to any
other program whose authors commit to using it. (Some other Free Software
Foundation software is covered by the GNU Library General Public License
instead.) You can apply it to your programs, too.
When we speak of free software, we are referring to freedom, not price.
Our General Public Licenses are designed to make sure that you have the
freedom to distribute copies of free software (and charge for this service if
you wish), that you receive source code or can get it if you want it, that you
can change the software or use pieces of it in new free programs; and that
you know you can do these things.
To protect your rights, we need to make restrictions that forbid anyone to
deny you these rights or to ask you to surrender the rights. These restrictions
translate to certain responsibilities for you if you distribute copies of the
software, or if you modify it.
For example, if you distribute copies of such a program, whether gratis or
for a fee, you must give the recipients all the rights that you have. You must
make sure that they, too, receive or can get the source code. And you must
show them these terms so they know their rights.
739
740 GNU General Public License Appendix A
A.2.1 Section 0
This License applies to any program or other work which contains a notice
placed by the copyright holder saying it may be distributed under the terms
of this General Public License. The ”Program”, below, refers to any such
program or work, and a “work based on the Program” means either the
Program or any derivative work under copyright law: that is to say, a work
containing the Program or a portion of it, either verbatim or with modifi-
cations and/or translated into another language. (Hereinafter, translation
is included without limitation in the term “modification”.) Each licensee is
addressed as “you”.
Activities other than copying, distribution and modification are not covered
by this License; they are outside its scope. The act of running the Program
is not restricted, and the output from the Program is covered only if its
contents constitute a work based on the Program (independent of having
Section A.2. TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND
MODIFICATION 741
A.2.2 Section 1
You may copy and distribute verbatim copies of the Program’s source code
as you receive it, in any medium, provided that you conspicuously and appro-
priately publish on each copy an appropriate copyright notice and disclaimer
of warranty; keep intact all the notices that refer to this License and to the
absence of any warranty; and give any other recipients of the Program a
copy of this License along with the Program.
You may charge a fee for the physical act of transferring a copy, and you
may at your option offer warranty protection in exchange for a fee.
A.2.3 Section 2
You may modify your copy or copies of the Program or any portion of it,
thus forming a work based on the Program, and copy and distribute such
modifications or work under the terms of Section A.2.2 above, provided that
you also meet all of these conditions:
1. You must cause the modified files to carry prominent notices stating
that you changed the files and the date of any change.
2. You must cause any work that you distribute or publish, that in whole
or in part contains or is derived from the Program or any part thereof,
to be licensed as a whole at no charge to all third parties under the
terms of this License.
Exception:
A.2.4 Section 3
You may copy and distribute the Program (or a work based on it, under
Section A.2.3 in object code or executable form under the terms of Sec-
tion A.2.2 and Section A.2.3 above provided that you also do one of the
following:
1. Accompany it with the complete corresponding machine-readable source
code, which must be distributed under the terms of Sections 1 and 2
above on a medium customarily used for software interchange; or,
2. Accompany it with a written offer, valid for at least three years, to
give any third party, for a charge no more than your cost of physically
Section A.2. TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND
MODIFICATION 743
A.2.5 Section 4
You may not copy, modify, sublicense, or distribute the Program except
as expressly provided under this License. Any attempt otherwise to copy,
modify, sublicense or distribute the Program is void, and will automatically
terminate your rights under this License. However, parties who have received
copies, or rights, from you under this License will not have their licenses
terminated so long as such parties remain in full compliance.
A.2.6 Section 5
You are not required to accept this License, since you have not signed it.
However, nothing else grants you permission to modify or distribute the
744 GNU General Public License Appendix A
A.2.7 Section 6
Each time you redistribute the Program (or any work based on the Pro-
gram), the recipient automatically receives a license from the original licen-
sor to copy, distribute or modify the Program subject to these terms and
conditions. You may not impose any further restrictions on the recipients’
exercise of the rights granted herein. You are not responsible for enforcing
compliance by third parties to this License.
A.2.8 Section 7
distribute software through any other system and a licensee cannot impose
that choice.
A.2.9 Section 8
A.2.10 Section 9
The Free Software Foundation may publish revised and/or new versions of
the General Public License from time to time. Such new versions will be
similar in spirit to the present version, but may differ in detail to address
new problems or concerns.
A.2.11 Section 10
If you wish to incorporate parts of the Program into other free programs
whose distribution conditions are different, write to the author to ask for
permission. For software which is copyrighted by the Free Software Founda-
tion, write to the Free Software Foundation; we sometimes make exceptions
for this. Our decision will be guided by the two goals of preserving the free
status of all derivatives of our free software and of promoting the sharing
and reuse of software generally.
746 GNU General Public License Appendix A
A.2.13 Section 12
If you develop a new program, and you want it to be of the greatest possible
use to the public, the best way to achieve this is to make it free software
which everyone can redistribute and change under these terms.
To do so, attach the following notices to the program. It is safest to attach
them to the start of each source file to most effectively convey the exclusion
Section A.3. How to Apply These Terms to Your New Programs 747
of warranty; and each file should have at least the ”copyright” line and a
pointer to where the full notice is found.
<one line to give the program’s name and a brief idea of what it does.>
Copyright (C) <year> <name of author>
This program is free software; you can redistribute it and/or modify it under
the terms of the GNU General Public License as published by the Free
Software Foundation; either version 2 of the License, or (at your option)
any later version.
This program is distributed in the hope that it will be useful, but WITH-
OUT ANY WARRANTY; without even the implied warranty of MER-
CHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc., 51
Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
Also add information on how to contact you by electronic and paper mail.
If the program is interactive, make it output a short notice like this when it
starts in an interactive mode:
Gnomovision version 69, Copyright (C) year name of author Gnomovision
comes with ABSOLUTELY NO WARRANTY; for details type ‘show w’.
This is free software, and you are welcome to redistribute it under certain
conditions; type ‘show c’ for details.
The hypothetical commands ‘show w’ and ‘show c’ should show the appro-
priate parts of the General Public License. Of course, the commands you
use may be called something other than ‘show w’ and ‘show c’; they could
even be mouse-clicks or menu items–whatever suits your program.
You should also get your employer (if you work as a programmer) or your
school, if any, to sign a ”copyright disclaimer” for the program, if necessary.
Here is a sample; alter the names:
Yoyodyne, Inc., hereby disclaims all copyright interest in the program ‘Gnomo-
vision’ (which makes passes at compilers) written by James Hacker.
<signature of Ty Coon>, 1 April 1989 Ty Coon, President of Vice
This General Public License does not permit incorporating your program
into proprietary programs. If your program is a subroutine library, you may
748 GNU General Public License Appendix A
749
750 GLOSSARY
753
754 Subject Index
configuration wizard, 128 CUPS, 359, 360, 367, 369, 370, 419,
configure, 714 420, 434
configuring a firewall, 342 Page Accounting, 504
confirm address, 341 quotas, 504
confirm the password, 352 CUPS API, 374
confirm the trust, 351 CUPS backends, 446
connection resources, 105 CUPS filtering, 435, 436
connections, 4 CUPS filtering chain, 444
connections.tdb, see also TDB 491 CUPS libarary API, 119
container, 109 CUPS PostScript, 471
Control Panel, 128 CUPS PostScript driver, 472
controls, 340 CUPS print filters, 119
convert CUPS raster, 436, 442
domain member server, 63 CUPS-PPD, 496
converted, 178 cups.hlp, 469
core files, 708 cupsaddsmb, 427, 464, 469, 473,
core graphic engine, 429 475, 477, 478, 483
corrupted file, 267 cupsd.conf, 374, 420, 448, 510
cupsomatic, 435, 436, 447, 451–
cosine.schema, 195
453, 494, 496
cracker, 341
custom scripts, 191
create, 295
customized print commands, 375
Create a Computer Account, 100
create a domain machine account, daemon, 3, 346, 719
101 data caching, 323
create domain member, 100 data corruption, 173, 325
create machine trust account, 104 data interchange, 291
create mask, 302, 309 data stream, 361
create partition, 530 database, 96
Create the Computer Account, 109 database backend, 177
create user accounts, 115 database backends, 178
create volume, 530 DatabaseFS, 532
credentials, 43, 59, 85, 108, 129, DDK, 463, 467
241, 283 DDNS, 148
credentials validation, 87 Debian, 529
critical aspects of configuration, 120 Debian Sarge, 529
cron, 89 debug, 708
cross-segment browsing, 146 debug level, 698, 736
cross-subnet browsing, 152, 161, 168 debuglevel, 706
csc policy, 318 dedicated print server, 359
Subject Index 759