1 <?xml version="1.0" encoding="UTF-8"?>
4 <title>Issues in Cell Configuration and Administration</title>
6 <para>This chapter discusses many of the issues to consider when
7 configuring and administering a cell, and directs you to detailed
8 related information available elsewhere in this guide. It is assumed you
9 are already familiar with the material in <link linkend="HDRWQ5">An
10 Overview of OpenAFS Administration</link>.</para>
12 <para>It is best to read this chapter before installing your cell's
13 first file server machine or performing any other administrative
17 <primary>AFS</primary>
19 <secondary>differences from UNIX summarized</secondary>
23 <primary>UNIX</primary>
25 <secondary>differences from AFS summarized</secondary>
29 <primary>differences</primary>
31 <secondary>between AFS and UNIX, summarized</secondary>
35 <title>Differences between AFS and UNIX: A Summary</title>
37 <para>AFS behaves like a standard UNIX file system in most respects,
38 while also making file sharing easy within and between cells. This
39 section describes some differences between AFS and the UNIX file
40 system, referring you to more detailed information as
44 <primary>protection</primary>
46 <secondary>AFS compared to UNIX</secondary>
49 <sect2 id="Header_35">
50 <title>Differences in File and Directory Protection</title>
52 <para>AFS augments the standard UNIX file protection mechanism in
53 two ways: it associates an <emphasis>access control list
54 (ACL)</emphasis> with each directory, and it enables users to define
55 a large number of their own groups, which can be placed on
58 <para>AFS uses ACLs to protect files and directories, rather than
59 relying exclusively on the mode bits. This has several implications,
60 which are discussed further in the indicated sections:
63 <para>AFS ACLs use seven access permissions rather than the
64 three UNIX mode bits. See <link linkend="HDRWQ567">The AFS ACL
65 Permissions</link>.</para>
69 <para>For directories, AFS ignores the UNIX mode bits. For
70 files, AFS uses only the first set of mode bits (the <emphasis
71 role="bold">owner</emphasis> bits), and their meaning
72 interacts with permissions on the directory's ACL. See <link
73 linkend="HDRWQ580">How AFS Interprets the UNIX Mode
78 <para>A directory's ACL protects all of the files in a
79 directory in the same manner. To apply a more restrictive set
80 of AFS permissions to certain file, place it in directory with
81 a different ACL. If a directory must contain files with
82 different permissions, use symbolic links to point to files
83 stored in directories with different ACLs.</para>
87 <para>Moving a file to a different directory changes its
88 protection. See <link linkend="HDRWQ566">Differences Between
89 UFS and AFS Data Protection</link>.</para>
93 <para>An ACL can include about 20 entries granting different
94 combinations of permissions to different users or groups,
95 rather than only the three UNIX entities represented by the
96 three sets of mode bits. See <link
97 linkend="HDRWQ566">Differences Between UFS and AFS Data
98 Protection</link>.</para>
102 <para>You can designate an AFS file as write-only as in the
103 UNIX file system, by setting only the <emphasis
104 role="bold">w</emphasis> (<emphasis
105 role="bold">write</emphasis>) mode bit. You cannot designate
106 an AFS directory as write-only, because AFS ignores the mode
107 bits on a directory. See <link linkend="HDRWQ580">How AFS
108 Interprets the UNIX Mode Bits</link>.</para>
113 <para>AFS enables users to create groups and add other users to
114 those groups. Placing these groups on ACLs extends the same
115 permissions to a number of exactly specified users at the same time,
116 which is much more convenient than placing the individuals on the
117 ACLs directly. See <link linkend="HDRWQ531">Administering the
118 Protection Database</link>.</para>
120 <para>There are also system-defined groups, <emphasis
121 role="bold">system:anyuser</emphasis> and <emphasis
122 role="bold">system:authuser</emphasis>, whose presence on an ACL
123 extends access to a wide range of users at once. See <link
124 linkend="HDRWQ535">The System Groups</link> and <link
125 linkend="HDRWQ571">Using Groups on ACLs</link>.</para>
128 <primary>authentication</primary>
130 <secondary>AFS compared to UNIX</secondary>
134 <primary>password</primary>
136 <secondary>AFS compared to UNIX</secondary>
141 <title>Differences in Authentication</title>
143 <para>Just as the AFS filespace is distinct from each machine's
144 local file system, AFS authentication is separate from local
145 login. This has two practical implications, which will already be
146 familiar to users and system administrators who use Kerberos for
150 <para>To access AFS files, users must log into the local
151 machine as normal, obtain Kerberos tickets, and then obtain
152 AFS tokens. This process can often be automated through the
153 system authentication configuration so that the user logs into
154 the system as normal and obtains Kerberos tickets and AFS
155 tokens transparently. If you cannot or chose not to configure
156 the system this way, your users must login and authenticate in
157 separate steps, as detailed in the <emphasis>OpenAFS User
158 Guide</emphasis>.</para>
162 <para>Passwords may be stored in two separate places: the
163 Kerberos KDC and, optionally, each machine's local user
164 database (<emphasis role="bold">/etc/passwd</emphasis> or
165 equivalent) for the local system. A user's passwords in the
166 two places can differ if desired.</para>
172 <sect2 id="Header_37">
173 <title>Differences in the Semantics of Standard UNIX
176 <para>This section summarizes how AFS modifies the functionality of
180 <term><emphasis role="bold">The chmod
181 command</emphasis></term>
185 <primary>chmod command</primary>
186 <secondary>AFS compared to UNIX</secondary>
190 <primary>commands</primary>
191 <secondary>chmod (AFS compared to UNIX)</secondary>
195 <primary>setuid programs</primary>
196 <secondary>setting mode bits</secondary>
199 <para>Only members of the <emphasis
200 role="bold">system:administrators</emphasis> group can use
201 this command to turn on the setuid, setgid or sticky mode
202 bits on AFS files. For more information, see <link
203 linkend="HDRWQ409">Determining if a Client Can Run Setuid
204 Programs</link>.</para>
209 <term><emphasis role="bold">The chown
210 command</emphasis></term>
214 <primary>chown command</primary>
215 <secondary>AFS compared to UNIX</secondary>
219 <primary>commands</primary>
220 <secondary>chown (AFS compared to UNIX)</secondary>
223 <para>Only members of the <emphasis
224 role="bold">system:administrators</emphasis> group can issue
225 this command on AFS files.</para>
230 <term><emphasis role="bold">The chgrp
231 command</emphasis></term>
235 <primary>chgrp command</primary>
236 <secondary>AFS compared to UNIX</secondary>
240 <primary>commands</primary>
241 <secondary>chgrp (AFS compared to UNIX)</secondary>
244 <para>Only members of the <emphasis
245 role="bold">system:administrators</emphasis> can issue this
246 command on AFS files and directories.</para>
251 <term><emphasis role="bold">The groups and id
252 commands</emphasis></term>
256 <primary>groups command</primary>
257 <secondary>AFS compared to UNIX</secondary>
261 <primary>id command</primary>
262 <secondary>AFS compared to UNIX</secondary>
266 <primary>commands</primary>
267 <secondary>groups (AFS compared to UNIX)</secondary>
271 <primary>commands</primary>
272 <secondary>id (AFS compared to UNIX)</secondary>
275 <para>If the user's AFS tokens are associated with a process
276 authentication group (PAG), the output of these commands may
277 include one or two large numbers. These are artificial
278 groups used by the OpenAFS Cache Manager to track the PAG on
279 some platforms. Other platforms may use other methods, such
280 as native kernel support for a PAG or a similar concept, in
281 which case the large GIDs may not appear. To learn about
282 PAGs, see <link linkend="HDRWQ64">Identifying AFS Tokens by
288 <term><emphasis role="bold">The ln command</emphasis></term>
292 <primary>ln command</primary>
293 <secondary>AFS compared to UNIX</secondary>
297 <primary>commands</primary>
298 <secondary>ln (AFS compared to UNIX)</secondary>
301 <para>This command cannot create hard links between files in
302 different AFS directories. See <link
303 linkend="HDRWQ32">Creating Hard Links</link>.</para>
308 <term><emphasis role="bold">The sshd daemon and ssh
309 command</emphasis></term>
313 <primary>sshd command</primary>
314 <secondary>AFS compared to UNIX</secondary>
318 <primary>commands</primary>
319 <secondary>sshd (AFS compared to UNIX)</secondary>
323 <primary>ssh command</primary>
324 <secondary>AFS compared to UNIX</secondary>
328 <primary>commands</primary>
329 <secondary>ssh (AFS compared to UNIX)</secondary>
332 <para>In order for a user to have access to files stored in
333 AFS, that user needs to have Kerberos tickets and an AFS token
334 on the system from which they're accessing AFS. This has an
335 implication for users who log in remotely via protocols such
336 as Secure Shell (SSH): that log-in process must create local
337 Kerberos tickets and an AFS token on the system, or the user
338 will have to separately authenticate to Kerberos and AFS
339 after logging in.</para>
341 <para>The <ulink url="http://www.openssh.org/">OpenSSH
342 project</ulink> provides an SSH client and server that uses
343 the GSS-API protocol to pass Kerberos tickets between
344 machines. With a suitable SSH client, this allows users to
345 delegate their Kerberos tickets to the remote machine, and
346 that machine to store those tickets and obtain AFS tokens as
347 part of the log-in process.</para>
354 <primary>fsck command</primary>
356 <secondary>AFS compared to UNIX</secondary>
360 <primary>file server machine</primary>
361 <secondary>inode-based</secondary>
365 <primary>file server machine</primary>
366 <secondary>namei-based</secondary>
370 <primary>namei</primary>
371 <secondary>definition</secondary>
375 <primary>commands</primary>
377 <secondary>fsck (AFS compared to UNIX)</secondary>
381 <primary>fsck command</primary>
383 <secondary>AFS version</secondary>
387 <primary>commands</primary>
389 <secondary>fsck (AFS version)</secondary>
393 <primary>directories</primary>
395 <secondary>lost+found</secondary>
399 <primary>lost+found directory</primary>
403 <sect2 id="Header_38">
404 <title>The AFS version of the fsck Command and inode-based
408 <para>The fileserver uses either of two formats for storing data
409 on disk. The inode-based format uses a combination of regular
410 files and extra fields stored in the inode data structures that
411 are normally reserved for use by the operating system. The namei
412 format uses normal file storage and does not use special
413 structures. The choice of storage formats is chosen at compile
414 time and the two formats are incompatible. The inode format is
415 only available on certain platforms. The storage format must be
416 consistent for the fileserver binaries and all vice partitions on
417 a given file server machine.</para>
421 <para>This section on fsck advice only applies to the inode-based
422 fileserver binaries. On servers using namei-based binaries, the
423 vendor-supplied fsck can be used as normal.</para>
426 <para>If you are using AFS fileserver binaries compiled with the
427 inode-based format, never run the standard UNIX <emphasis
428 role="bold">fsck</emphasis> command on an AFS file server
429 machine. It does not understand how the File Server organizes volume
430 data on disk, and so moves all AFS data into the <emphasis
431 role="bold">lost+found</emphasis> directory on the partition.</para>
433 <para>Instead, use the version of the <emphasis
434 role="bold">fsck</emphasis> program that is included in the AFS
435 distribution. The <emphasis>OpenAFS Quick Start Guide</emphasis>
436 explains how to replace the vendor-supplied <emphasis
437 role="bold">fsck</emphasis> program with the AFS version as you
438 install each server machine.</para>
440 <para>The AFS version functions like the standard <emphasis
441 role="bold">fsck</emphasis> program on data stored on both UFS and
442 AFS partitions. The appearance of a banner like the following as the
443 <emphasis role="bold">fsck</emphasis> program initializes confirms
444 that you are running the correct one:</para>
447 --- AFS (R) version fsck---
450 <para>where <emphasis>version</emphasis> is the AFS version. For
451 correct results, it must match the AFS version of the server
452 binaries in use on the machine.</para>
454 <para>If you ever accidentally run the standard version of the
455 program, contact your AFS support provider, contact the OpenAFS
456 mailing lists, or refer to the <ulink
457 url="http://www.openafs.org/support.html">OpenAFS support web
458 page</ulink> for support options. It is sometimes possible to
459 recover volume data from the <emphasis
460 role="bold">lost+found</emphasis> directory. If the data is not
461 recoverabled, then restoring from backup is recommended.</para>
464 <para>Running the fsck binary supplied by the operating system
465 vendor on an fileserver using inode-based file storage will result
466 in data corruption!</para>
471 <title>Creating Hard Links</title>
474 <primary>hard link</primary>
476 <secondary>AFS restrictions on</secondary>
480 <primary>restrictions</primary>
482 <secondary>on hard links in AFS</secondary>
485 <para>AFS does not allow hard links (created with the UNIX <emphasis
486 role="bold">ln</emphasis> command) between files that reside in
487 different directories, because in that case it is unclear which of
488 the directory's ACLs to associate with the link.</para>
490 <para>AFS also does not allow hard links to directories, in order to
491 keep the file system organized as a tree.</para>
493 <para>It is possible to create symbolic links (with the UNIX
494 <emphasis role="bold">ln -s</emphasis> command) between elements in
495 two different AFS directories, or even between an element in AFS and
496 one in a machine's local UNIX file system. Do not create a symbolic
497 link in AFS to a file whose name begins with either a number sign
498 (<emphasis role="bold">#</emphasis>) or a percent sign (<emphasis
499 role="bold">%</emphasis>), however. The Cache Manager interprets
500 such links as a mount point to a regular or read/write volume,
505 <title>AFS Implements Save on Close</title>
508 <primary>fsync system call</primary>
510 <secondary>for files saved on AFS client</secondary>
514 <primary>close system call</primary>
516 <secondary>for files saved on AFS client</secondary>
520 <primary>write</primary>
522 <secondary>system call for files saved on AFS client</secondary>
525 <para>When an application issues the UNIX <emphasis
526 role="bold">close</emphasis> system call on a file, the Cache
527 Manager performs a synchronous write of the data to the File Server
528 that maintains the central copy of the file. It does not return
529 control to the application until the File Server has acknowledged
530 receipt of the data. For the <emphasis role="bold">fsync</emphasis>
531 system call, control does not return to the application until the
532 File Server indicates that it has written the data to non-volatile
533 storage on the file server machine.</para>
535 <para>When an application issues the UNIX <emphasis
536 role="bold">write</emphasis> system call, the Cache Manager writes
537 modifications to the local AFS client cache only. If the local
538 machine crashes or an application program exits without issuing the
539 <emphasis role="bold">close</emphasis> system call, it is possible
540 that the modifications are not recorded in the central copy of the
541 file maintained by the File Server. The Cache Manager does sometimes
542 write this type of modified data from the cache to the File Server
543 without receiving the <emphasis role="bold">close</emphasis> or
544 <emphasis role="bold">fsync</emphasis> system call, such as when it
545 needs to free cache chunks for new data. However, it is not
546 generally possible to predict when the Cache Manager transfers
547 modified data to the File Server in this way.</para>
549 <para>The implication is that if an application's <emphasis
550 role="bold">Save</emphasis> option invokes the <emphasis
551 role="bold">write</emphasis> system call rather than <emphasis
552 role="bold">close</emphasis> or <emphasis
553 role="bold">fsync</emphasis>, the changes are not necessarily stored
554 permanently on the File Server machine. Most application programs
555 issue the <emphasis role="bold">close</emphasis> system call for
556 save operations, as well as when they finish handling a file and
557 when they exit.</para>
560 <sect2 id="Header_41">
561 <title>Setuid Programs</title>
564 <primary>setuid programs</primary>
566 <secondary>restrictions on</secondary>
569 <para>The UNIX setuid bit is ignored by default for programs run
570 from AFS, but can be enabled by the system administrator on a client
571 machine. The <emphasis role="bold">fs setcell</emphasis> command
572 determines whether setuid programs that originate in a particular
573 cell can run on a given client machine. Running setuid binaries from
574 AFS poses a security risk due to weaknesses in the integrity checks
575 of the AFS protocol and should normally not be permitted. See <link
576 linkend="HDRWQ409">Determining if a Client Can Run Setuid
577 Programs</link>.</para>
579 <para>Set the UNIX setuid bit only for files whose owner is UID 0
580 (the local superuser <emphasis role="bold">root</emphasis>). This
581 does not present an automatic security risk: the local superuser has
582 no special privilege in AFS, but only in the local machine's UNIX
583 file system and kernel. Setting the UNIX setuid bit for files owned
584 with a different UID will have unpredictable resuilts, since that
585 UID will be interpreted as possibly different users on each AFS
586 client machine.</para>
588 <para>Any file can be marked with the setuid bit, but only members
589 of the <emphasis role="bold">system:administrators</emphasis> group
590 can issue the <emphasis role="bold">chown</emphasis> system call or
591 the <emphasis role="bold">chown</emphasis> command, or issue the
592 <emphasis role="bold">chmod</emphasis> system call or the <emphasis
593 role="bold">chmod</emphasis> command to set the setuid bit.</para>
598 <title>Choosing a Cell Name</title>
601 <primary>cell</primary>
603 <secondary>name</secondary>
605 <tertiary>choosing</tertiary>
609 <primary>choosing</primary>
611 <secondary>name</secondary>
613 <tertiary>cell</tertiary>
617 <primary>conventions</primary>
619 <secondary>cell name</secondary>
623 <primary>Internet</primary>
625 <secondary>conventions for cell name</secondary>
628 <para>This section explains how to choose a cell name and explains why
629 choosing an appropriate cell name is important.</para>
631 <para>Your cell name must distinguish your cell from all others in the
632 AFS global namespace. By convention, the cell name is the second
633 element in any AFS pathname; therefore, a unique cell name guarantees
634 that every AFS pathname uniquely identifies a file, even if cells use
635 the same directory names at lower levels in their local AFS
636 filespace. For example, both the ABC Corporation cell and the State
637 University cell can have a home directory for the user <emphasis
638 role="bold">pat</emphasis>, because the pathnames are distinct:
639 <emphasis role="bold">/afs/abc.com/usr/pat</emphasis> and <emphasis
640 role="bold">/afs/stateu.edu/usr/pat</emphasis>.</para>
642 <para>By convention, cell names follow the Domain Name System (DNS)
643 conventions for domain names. If you are already an Internet site,
644 then it is simplest and strongly recommended to choose your Internet
645 domain name as the cell name.</para>
647 <para>If you are not an Internet site, it is best to choose a unique
648 DNS-style name, particularly if you plan to connect to the Internet in
649 the future. There are a few constraints on AFS cell names:
652 <para>It can contain as many as 64 characters, but shorter names
653 are better because the cell name frequently is part of machine
654 and file names. If your cell name is long, you can reduce
655 pathname length either by creating a symbolic link to the
656 complete cell name, at the second level in your file tree or by
657 using the <emphasis role="bold">CellAlias</emphasis>
658 configuration file on a client machine. See <link
659 linkend="HDRWQ42">The Second (Cellname) Level</link>.</para>
663 <para>To guarantee it is suitable for different operating system
664 types, the cell name can contain only lowercase characters,
665 numbers, underscores, dashes, and periods. Do not include
666 command shell metacharacters.</para>
670 <para>It can include any number of fields, which are
671 conventionally separated by periods (see the examples
677 <sect2 id="Header_43">
678 <title>How to Set the Cell Name</title>
681 <primary>setting</primary>
682 <secondary>cell name</secondary>
686 <primary>cell</primary>
687 <secondary>name</secondary>
688 <tertiary>setting</tertiary>
692 <primary>server machine</primary>
693 <secondary>setting home cell</secondary>
697 <primary>client machine</primary>
698 <secondary>setting home cell</secondary>
701 <para>The cell name is recorded in two files on the local disk of
702 each file server and client machine. Among other functions, these
703 files define the machine's cell membership and so affect how
704 programs and processes run on the machine; see <link
705 linkend="HDRWQ35">Why Choosing the Appropriate Cell Name is
706 Important</link>. The procedure for setting the cell name is
707 different for the two types of machines.</para>
709 <para>For file server machines, the two files that record the cell
710 name are the <emphasis role="bold">/usr/afs/etc/ThisCell</emphasis>
711 and <emphasis role="bold">/usr/afs/etc/CellServDB</emphasis>
712 files. As described more explicitly in the <emphasis>OpenAFS Quick
713 Start Guide</emphasis>, you set the cell name in both by issuing the
714 <emphasis role="bold">bos setcellname</emphasis> command on the
715 first file server machine you install in your cell. It is not
716 usually necessary to issue the command again. If you use the Update
717 Server, it distributes its copy of the <emphasis
718 role="bold">ThisCell</emphasis> and <emphasis
719 role="bold">CellServDB</emphasis> files to additional server
720 machines that you install. If you do not use the Update Server, the
721 <emphasis>OpenAFS Quick Start Guide</emphasis> explains how to copy
722 the files manually.</para>
724 <para>For client machines, the two files that record the cell name
725 are the <emphasis role="bold">/usr/vice/etc/ThisCell</emphasis> and
726 <emphasis role="bold">/usr/vice/etc/CellServDB</emphasis> files. You
727 create these files on a per-client basis, either with a text editor
728 or by copying them onto the machine from a central source in AFS.
729 See <link linkend="HDRWQ406">Maintaining Knowledge of Database
730 Server Machines</link> for details.</para>
732 <para>Change the cell name in these files only when you want to
733 transfer the machine to a different cell (client machines can only
734 have one default cell at a time and server machines can only belong
735 to one cell at a time). If the machine is a file server, follow the
736 complete set of instructions in the <emphasis>OpenAFS Quick Start
737 Guide</emphasis> for configuring a new cell. If the machine is a
738 client, all you need to do is change the files appropriately and
739 reboot the machine. The next section explains further the negative
740 consequences of changing the name of an existing cell.</para>
742 <para>To set the default cell name used by most AFS commands without
743 changing the local <emphasis
744 role="bold">/usr/vice/etc/ThisCell</emphasis> file, set the AFSCELL
745 environment variable in the command shell. It is worth setting this
746 variable if you need to complete significant administrative work in
747 a foreign cell.</para>
750 <para>The <emphasis role="bold">fs checkservers</emphasis> and
751 <emphasis role="bold">fs mkmount</emphasis> commands do not use
752 the AFSCELL variable. The <emphasis role="bold">fs
753 checkservers</emphasis> command always defaults to the cell named
754 in the <emphasis role="bold">ThisCell</emphasis> file, unless the
755 <emphasis role="bold">-cell</emphasis> argument is used. The
756 <emphasis role="bold">fs mkmount</emphasis> command defaults to
757 the cell in which the parent directory of the new mount point
763 <title>Why Choosing the Appropriate Cell Name is Important</title>
766 <primary>ThisCell file (client)</primary>
767 <secondary>how used by programs</secondary>
770 <para>Take care to select a cell name that is suitable for long-term
771 use. Changing a cell name later is complicated. An appropriate cell
772 name is important because it is the second element in the pathname
773 of all files in a cell's file tree. Because each cell name is
774 unique, its presence in an AFS pathname makes the pathname unique in
775 the AFS global namespace, even if multiple cells use similar
776 filespace organization at lower levels. For instance, it means that
777 every cell can have a home directory called <emphasis
778 role="bold">/afs/</emphasis><replaceable>cellname</replaceable><emphasis
779 role="bold">/usr/pat</emphasis> without causing a conflict. The
780 presence of the cell name in pathnames also means that users in
781 every cell use the same pathname to access a file, whether the file
782 resides in their local cell or in a foreign cell.</para>
784 <para>Another reason to choose the correct cell name early in the
785 process of installing your cell is that the cell membership defined
786 in each machine's <emphasis role="bold">ThisCell</emphasis> file
787 affects the performance of many programs and processes running on
788 the machine. For instance, AFS commands (<emphasis
789 role="bold">fs</emphasis>, <emphasis role="bold">pts</emphasis>, and
790 <emphasis role="bold">vos</emphasis> commands, for example) by
791 default execute in the cell of the machine on which they are
792 issued. The command interpreters check the <emphasis
793 role="bold">ThisCell</emphasis> file on the local disk and then
794 contact the database server machines listed in the <emphasis
795 role="bold">CellServDB</emphasis> file or configured in DNS for the
796 indicated cell. (The <emphasis role="bold">bos</emphasis> commands
797 work differently because the issuer always has to name of the
798 machine on which to run the command.)</para>
800 <para>The <emphasis role="bold">ThisCell</emphasis> file also
801 normally determines the cell for which a user receives an AFS token
802 when he or she logs in to a machine.</para>
804 <para>If you change the cell name, you must change the <emphasis
805 role="bold">ThisCell</emphasis> and <emphasis
806 role="bold">CellServDB</emphasis> files on every server and client
807 machine. Failure to change them all will cause many commands from
808 the AFS suites to not work as expected.</para>
813 <title>Participating in the AFS Global Namespace</title>
816 <primary>participation</primary>
817 <secondary>in AFS global namespace</secondary>
821 <primary>AFS</primary>
822 <secondary>global namespace</secondary>
826 <primary>global namespace</primary>
829 <para>Participating in the AFS global namespace makes your cell's
830 local file tree visible to AFS users in foreign cells and makes other
831 cells' file trees visible to your local users. It makes file sharing
832 across cells just as easy as sharing within a cell. This section
833 outlines the procedures necessary for participating in the global
837 <para>Participation in the global namespace is not
838 mandatory. Some cells use AFS primarily to facilitate file
839 sharing within the cell, and are not interested in providing
840 their users with access to foreign cells.</para>
844 <para>Making your file tree visible does not mean making it
845 vulnerable. You control how foreign users access your cell using
846 the same protection mechanisms that control local users'
847 access. See <link linkend="HDRWQ40">Granting and Denying Foreign
848 Users Access to Your Cell</link>.</para>
852 <para>The two aspects of participation are independent. A cell
853 can make its file tree visible without allowing its users to see
854 foreign cells' file trees, or can enable its users to see other
855 file trees without advertising its own.</para>
859 <para>You make your cell visible to others by advertising your
860 database server machines and allowing users at other sites to
861 access your database server and file server machines. See <link
862 linkend="HDRWQ38">Making Your Cell Visible to
863 Others</link>.</para>
867 <para>You control access to foreign cells on a per-client
868 machine basis. In other words, it is possible to make a foreign
869 cell accessible from one client machine in your cell but not
870 another. See <link linkend="HDRWQ39">Making Other Cells Visible
871 in Your Cell</link>.</para>
877 <title>What the Global Namespace Looks Like</title>
880 <primary>conventions</primary>
881 <secondary>AFS pathnames</secondary>
885 <primary>AFS</primary>
886 <secondary>root directory (/afs)</secondary>
887 <tertiary>on client machine</tertiary>
891 <primary>directories</primary>
892 <secondary>/afs</secondary>
896 <primary>directories</primary>
897 <secondary>/afs/<emphasis>cellname</emphasis></secondary>
901 <primary>cell</primary>
902 <secondary>name</secondary>
903 <tertiary>at second level in file tree</tertiary>
906 <para>The AFS global namespace appears the same to all AFS cells
907 that participate in it, because they all agree to follow a small set
908 of conventions in constructing pathnames.</para>
910 <para>The first convention is that all AFS pathnames begin with the
911 string <emphasis role="bold">/afs</emphasis> to indicate that they
912 belong to the AFS global namespace.</para>
914 <para>The second convention is that the cell name is the second
915 element in an AFS pathname; it indicates where the file resides
916 (that is, the cell in which a file server machine houses the
917 file). As noted, the presence of a cell name in pathnames makes the
918 global namespace possible, because it guarantees that all AFS
919 pathnames are unique even if cells use the same directory names at
920 lower levels in their AFS filespace.</para>
922 <para>What appears at the third and lower levels in an AFS pathname
923 depends on how a cell has chosen to arrange its filespace. There
924 are some suggested conventional directories at the third level; see
925 <link linkend="HDRWQ43">The Third Level</link>.</para>
929 <title>Making Your Cell Visible to Others</title>
932 <primary>cell</primary>
933 <secondary>making local visible to foreign</secondary>
937 <primary>local cell</primary>
938 <secondary>making visible to foreign cells</secondary>
942 <primary>foreign cell</primary>
943 <secondary>making local cell visible</secondary>
946 <para>You make your cell visible to others by advertising your cell
947 name and database server machines. Just like client machines in the
948 local cell, the Cache Manager on machines in foreign cells use the
949 information to reach your cell's Volume Location (VL) Servers when
950 they need volume and file location information. For authenticated
951 access, foreign clients must be configured with the necessary
952 Kerberos version 5 domain-to-realm mappings and Key Distribution
953 Center (KDC) location information for both the local and remote
954 Kerberos version 5 realms.</para>
956 <para>There are two places you can make this information available:
959 <primary>files</primary>
961 <secondary>global CellServDB</secondary>
965 <primary>CellServDB file maintained by the AFS
968 <secondary>as global update source</secondary>
972 <para>In the global <emphasis
973 role="bold">CellServDB</emphasis> file maintained by the AFS
974 Registrar. This file lists the name and database server
975 machines of every cell that has agreed to make this
976 information available to other cells. This file is available
978 url="http://grand.central.org/csdb.html">http://grand.central.org/csdb.html</ulink></para>
980 <para>To add or change your cell's listing in this file,
981 follow the instructions at <ulink
982 url="http://grand.central.org/csdb.html">http://grand.central.org/csdb.html</ulink>.
983 It is a good policy to check the file for changes on a
984 regular schedule. An updated copy of this file is included
985 with new releases of OpenAFS.</para>
988 <primary>files</primary>
990 <secondary>CellServDB.local</secondary>
994 <primary>CellServDB.local file</primary>
999 <para>A file called <emphasis
1000 role="bold">CellServDB.local</emphasis> in the <emphasis
1001 role="bold">/afs/</emphasis><replaceable>cellname</replaceable><emphasis
1002 role="bold">/service/etc</emphasis> directory of your cell's
1003 filespace. List only your cell's database server
1009 <para>Update the files whenever you change the identity of your
1010 cell's database server machines. Also update the copies of the
1011 <emphasis role="bold">CellServDB</emphasis> files on all of your
1012 server machines (in the <emphasis
1013 role="bold">/usr/afs/etc</emphasis> directory) and client machines
1014 (in the <emphasis role="bold">/usr/vice/etc</emphasis>
1015 directory). For instructions, see <link
1016 linkend="HDRWQ118">Maintaining the Server CellServDB File</link> and
1017 <link linkend="HDRWQ406">Maintaining Knowledge of Database Server
1018 Machines</link>.</para>
1020 <para>Once you have advertised your database server machines, it can
1021 be difficult to make your cell invisible again. You can remove the
1022 <emphasis role="bold">CellServDB.local</emphasis> file and ask the
1023 AFS Registrar to remove your entry from the global <emphasis
1024 role="bold">CellServDB</emphasis> file, but other cells probably
1025 have an entry for your cell in their local <emphasis
1026 role="bold">CellServDB</emphasis> files already. To make those
1027 entries invalid, you must change the names or IP addresses of your
1028 database server machines.</para>
1030 <para>Your cell does not have to be invisible to be inaccessible,
1031 however. To make your cell completely inaccessible to foreign users,
1032 remove the <emphasis role="bold">system:anyuser</emphasis> group
1033 from all ACLs at the top three levels of your filespace; see <link
1034 linkend="HDRWQ40">Granting and Denying Foreign Users Access to Your
1038 <primary>cell</primary>
1040 <secondary>making foreign visible to local</secondary>
1044 <primary>local cell</primary>
1046 <secondary>making foreign cells visible in</secondary>
1050 <primary>foreign cell</primary>
1052 <secondary>making visible in local cell</secondary>
1056 <primary>client machine</primary>
1058 <secondary>making foreign cell visible</secondary>
1062 <sect2 id="HDRWQ39">
1063 <title>Making Other Cells Visible in Your Cell</title>
1065 <para>To make a foreign cell's filespace visible on a client machine
1066 in your cell that is not configured for <emphasis
1067 role="bold">Freelance Mode</emphasis> or <emphasis
1068 role="bold">Dynamic Root</emphasis> mode, perform the following
1072 <para>Mount the cell's <emphasis
1073 role="bold">root.cell</emphasis> volume at the second level in
1074 your cell's filespace just below the <emphasis
1075 role="bold">/afs</emphasis> directory. Use the <emphasis
1076 role="bold">fs mkmount</emphasis> command with the <emphasis
1077 role="bold">-cell</emphasis> argument as instructed in <link
1078 linkend="HDRWQ213">To create a cellular mount
1079 point</link>.</para>
1083 <para>Mount AFS at the <emphasis role="bold">/afs</emphasis>
1084 directory on the client machine. The <emphasis
1085 role="bold">afsd</emphasis> program, which initializes the
1086 Cache Manager, performs the mount automatically at the
1087 directory named in the first field of the local <emphasis
1088 role="bold">/usr/vice/etc/cacheinfo</emphasis> file or by the
1089 command's <emphasis role="bold">-mountdir</emphasis>
1090 argument. Mounting AFS at an alternate location makes it
1091 impossible to reach the filespace of any cell that mounts its
1092 <emphasis role="bold">root.afs</emphasis> and <emphasis
1093 role="bold">root.cell</emphasis> volumes at the conventional
1094 locations. See <link linkend="HDRWQ395">Displaying and Setting
1095 the Cache Size and Location</link>.</para>
1099 <para>Create an entry for the cell in the list of database
1100 server machines which the Cache Manager maintains in kernel
1104 role="bold">/usr/vice/etc/CellServDB</emphasis> file on every
1105 client machine's local disk lists the database server machines
1106 for the local and foreign cells. The <emphasis
1107 role="bold">afsd</emphasis> program reads the contents of the
1108 <emphasis role="bold">CellServDB</emphasis> file into kernel
1109 memory as it initializes the Cache Manager. You can also use
1110 the <emphasis role="bold">fs newcell</emphasis> command to add
1111 or alter entries in kernel memory directly between reboots of
1112 the machine. See <link linkend="HDRWQ406">Maintaining
1113 Knowledge of Database Server Machines</link>.</para>
1118 <para>Non-windows client machines may enable <emphasis
1119 role="bold">Dynamic Root Mode</emphasis> by using the <emphasis
1120 role="bold">-dynroot</emphasis> option to <emphasis
1121 role="bold">afsd</emphasis>. When this option is enabled, all cells
1122 listed in the <emphasis role="bold">CellServDB</emphasis> file will
1123 appear in the <emphasis role="bold">/afs</emphasis> directory. The
1124 contents of the <emphasis role="bold">root.afs</emphasis> volume
1125 will be ignored. </para>
1127 <para>Windows client machines may enable <emphasis
1128 role="bold">Freelance Mode</emphasis> during client installation or
1129 by setting the <emphasis role="bold">FreelanceClient</emphasis>
1130 setting under <emphasis role="bold">Service Parameters</emphasis> in
1131 the Windows Registry as mentioned in the <ulink
1132 url="http://docs.openafs.org/ReleaseNotesWindows/">Release
1133 Notes</ulink>. When this option is enabled, the <emphasis
1134 role="bold">root.afs</emphasis> volume is ignored and a mounpoint
1135 for each cell is automatically created in the the <emphasis
1136 role="bold">\\AFS</emphasis> directory when the folder <emphasis
1137 role="bold">\\AFS\<replaceable>cellname</replaceable></emphasis> is
1138 accessed and the foreign Volume Location servers can be reached.
1139 </para> <para>Note that making a foreign cell visible to client
1140 machines does not guarantee that your users can access its
1141 filespace. The ACLs in the foreign cell must also grant them the
1142 necessary permissions.</para>
1145 <primary>cell</primary>
1147 <secondary>granting local access to foreign users</secondary>
1151 <primary>local cell</primary>
1153 <secondary>granting foreign users access to</secondary>
1157 <sect2 id="HDRWQ40">
1158 <title>Granting and Denying Foreign Users Access to Your
1161 <para>Making your cell visible in the AFS global namespace does not
1162 take away your control over the way in which users from foreign
1163 cells access your file tree.</para>
1165 <para>By default, foreign users access your cell as the user
1166 <emphasis role="bold">anonymous</emphasis>, which means they have
1167 only the permissions granted to the <emphasis
1168 role="bold">system:anyuser</emphasis> group on each directory's
1169 ACL. Normally these permissions are limited to the <emphasis
1170 role="bold">l</emphasis> (<emphasis role="bold">lookup</emphasis>)
1171 and <emphasis role="bold">r</emphasis> (<emphasis
1172 role="bold">read</emphasis>) permissions.</para>
1174 <para>There are three ways to grant wider access to foreign users:
1177 <para>Grant additional permissions to the <emphasis
1178 role="bold">system:anyuser</emphasis> group on certain
1179 ACLs. Keep in mind, however, that all users can then access
1180 that directory in the indicated way (not just specific foreign
1181 users you have in mind).</para>
1185 <para>Enable automatic registration for users in the foreign
1186 cell. This may be done by creating a cross-realm trust in the
1187 <emphasis role="bold">Kerberos Database</emphasis>. Then add a
1188 PTS group named <emphasis
1189 role="bold">system:authuser<replaceable>@FOREIGN.REALM</replaceable></emphasis>
1190 and give it a group quota greater than the number of foreign
1191 users expected to be registered. After the cross-realm trust
1192 and the PTS group are created, the <ulink
1193 url="http://docs.openafs.org/Reference/1/aklog.html">aklog</ulink>
1194 command will automatically register foreign users as
1195 needed. Consult the documentation for your <emphasis
1196 role="bold">Kerberos Server</emphasis> for instructions on how
1197 to establish a cross-realm trust. </para>
1201 <para>Create a local authentication account for specific
1202 foreign users, by creating entries in the Protection Database,
1203 the Kerberos Database, and the local password file.</para>
1209 <primary>cell</primary>
1211 <secondary>filespace configuration issues</secondary>
1215 <primary>configuring</primary>
1217 <secondary>filespace, issues</secondary>
1221 <primary>file tree</primary>
1223 <secondary>conventions</secondary>
1225 <tertiary>for configuring</tertiary>
1230 <sect1 id="HDRWQ41">
1231 <title>Configuring Your AFS Filespace</title>
1233 <para>This section summarizes the issues to consider when configuring
1234 your AFS filespace. For a discussion of creating volumes that
1235 correspond most efficiently to the filespace's directory structure,
1236 see <link linkend="HDRWQ44">Creating Volumes to Simplify
1237 Administration</link>.</para>
1240 <para><emphasis role="bold">For Windows users:</emphasis> Windows
1241 uses a backslash (<emphasis role="bold">\</emphasis>) rather than a
1242 forward slash (<emphasis role="bold">/</emphasis>) to separate the
1243 elements in a pathname. The hierarchical organization of the
1244 filespace is however the same as on a UNIX machine.</para>
1247 <para>AFS pathnames must follow a few conventions so the AFS global
1248 namespace looks the same from any AFS client machine. There are
1249 corresponding conventions to follow in building your file tree, not
1250 just because pathnames reflect the structure of a file tree, but also
1251 because the AFS Cache Manager expects a certain configuration.</para>
1254 <primary>AFS</primary>
1256 <secondary>root directory (/afs)</secondary>
1258 <tertiary>in cell filespace</tertiary>
1262 <primary>directories</primary>
1264 <secondary>/afs</secondary>
1267 <sect2 id="Header_51">
1268 <title>The Top /afs Level</title>
1270 <para>The first convention is that the top level in your file tree
1271 be called the <emphasis role="bold">/afs</emphasis> directory. If
1272 you name it something else, then you must use the <emphasis
1273 role="bold">-mountdir</emphasis> argument with the <emphasis
1274 role="bold">afsd</emphasis> program to get Cache Managers to mount
1275 AFS properly. You cannot participate in the AFS global namespace in
1279 <primary>cell</primary>
1281 <secondary>name</secondary>
1283 <tertiary>at second level in file tree</tertiary>
1287 <primary>directories</primary>
1289 <secondary>/afs/<emphasis>cellname</emphasis></secondary>
1293 <primary>symbolic link</primary>
1295 <secondary>at second level of AFS pathname</secondary>
1299 <sect2 id="HDRWQ42">
1300 <title>The Second (Cellname) Level</title>
1302 <para>The second convention is that just below the <emphasis
1303 role="bold">/afs</emphasis> directory you place directories
1304 corresponding to each cell whose file tree is visible and accessible
1305 from the local cell. Minimally, there must be a directory for the
1306 local cell. Each such directory is a mount point to the indicated
1307 cell's <emphasis role="bold">root.cell</emphasis> volume. For
1308 example, in the ABC Corporation cell, <emphasis
1309 role="bold">/afs/abc.com</emphasis> is a mount point for the cell's
1310 own <emphasis role="bold">root.cell</emphasis> volume and <emphasis
1311 role="bold">stateu.edu</emphasis> is a mount point for the State
1312 University cell's <emphasis role="bold">root.cell</emphasis>
1313 volume. The <emphasis role="bold">fs lsmount</emphasis> command
1314 displays the mount points.</para>
1317 % <emphasis role="bold">fs lsmount /afs/abc.com</emphasis>
1318 '/afs/abc.com' is a mount point for volume '#root.cell'
1319 % <emphasis role="bold">fs lsmount /afs/stateu.edu</emphasis>
1320 '/afs/stateu.edu' is a mount point for volume '#stateu.edu:root.cell'
1323 <para>To reduce the amount of typing necessary in pathnames, you can
1324 create a symbolic link with an abbreviated name to the mount point
1325 of each cell your users frequently access (particularly the home
1326 cell). In the ABC Corporation cell, for instance, <emphasis
1327 role="bold">/afs/abc</emphasis> is a symbolic link to the <emphasis
1328 role="bold">/afs/abc.com</emphasis> mount point, as the <emphasis
1329 role="bold">fs lsmount</emphasis> command reveals.</para>
1332 % <emphasis role="bold">fs lsmount /afs/abc</emphasis>
1333 '/afs/abc' is a symbolic link, leading to a mount point for volume
1334 '#root.cell' </programlisting>
1337 <primary>file tree</primary>
1339 <secondary>conventions</secondary>
1341 <tertiary>third level</tertiary>
1345 <primary>directories</primary>
1347 <secondary>conventional under /afs/cellname</secondary>
1351 <sect2 id="HDRWQ43">
1352 <title>The Third Level</title>
1354 <para>You can organize the third level of your cell's file tree any
1355 way you wish. The following list describes directories that appear
1356 at this level in the conventional configuration:
1359 <term><emphasis role="bold">common</emphasis></term>
1362 <para>This directory contains programs and files needed by
1363 users working on machines of all system types, such as text
1364 editors, online documentation files, and so on. Its
1365 <emphasis role="bold">/etc</emphasis> subdirectory is a
1366 logical place to keep the central update sources for files
1367 used on all of your cell's client machines, such as the
1368 <emphasis role="bold">ThisCell</emphasis> and <emphasis
1369 role="bold">CellServDB</emphasis> files.</para>
1374 <term><emphasis role="bold">public</emphasis></term>
1377 <para>A directory accessible to anyone who can access your
1378 filespace, because its ACL grants the <emphasis
1379 role="bold">l</emphasis> (<emphasis
1380 role="bold">lookup</emphasis>) and <emphasis
1381 role="bold">r</emphasis> (<emphasis
1382 role="bold">read</emphasis>) permissions to the <emphasis
1383 role="bold">system:anyuser</emphasis> group. It is useful if
1384 you want to enable your users to make selected information
1385 available to everyone, but do not want to grant foreign
1386 users access to the contents of the <emphasis
1387 role="bold">usr</emphasis> directory which houses user home
1388 directories (and is also at this level). It is conventional
1389 to create a subdirectory for each of your cell's
1395 <term><emphasis role="bold">service</emphasis></term>
1398 <para>This directory contains files and subdirectories that
1399 help cells coordinate resource sharing. For a list of the
1400 proposed standard files and subdirectories to create, call
1401 or write to AFS Product Support.</para>
1403 <para>As an example, files that other cells expect to find
1404 in this directory's <emphasis role="bold">etc</emphasis>
1405 subdirectory can include the following: <itemizedlist>
1408 role="bold">CellServDB.export</emphasis>, a list of
1409 database server machines for many cells</para>
1414 role="bold">CellServDB.local</emphasis>, a list of the
1415 cell's own database server machines</para>
1419 <para><emphasis role="bold">passwd</emphasis>, a copy
1420 of the local password file (<emphasis
1421 role="bold">/etc/passwd</emphasis> or equivalent) kept
1422 on the local disk of the cell's client machines</para>
1426 <para><emphasis role="bold">group</emphasis>, a copy
1427 of the local groups file (<emphasis
1428 role="bold">/etc/group</emphasis> or equivalent) kept
1429 on the local disk of the cell's client machines</para>
1437 <term><emphasis>sys_type</emphasis></term>
1440 <para>A separate directory for storing the server and client
1441 binaries for each system type you use in the cell.
1442 Configuration is simplest if you use the system type names
1443 assigned in the AFS distribution, particularly if you wish
1444 to use the <emphasis role="bold">@sys</emphasis> variable in
1445 pathnames (see <link linkend="HDRWQ56">Using the @sys
1446 Variable in Pathnames</link>). The <emphasis>OpenAFS Release
1447 Notes</emphasis> lists the conventional name for each
1448 supported system type.</para>
1450 <para>Within each such directory, create directories named
1451 <emphasis role="bold">bin</emphasis>, <emphasis
1452 role="bold">etc</emphasis>, <emphasis
1453 role="bold">usr</emphasis>, and so on, to store the programs
1454 normally kept in the <emphasis role="bold">/bin</emphasis>,
1455 <emphasis role="bold">/etc</emphasis> and <emphasis
1456 role="bold">/usr</emphasis> directories on a local
1457 disk. Then create symbolic links from the local directories
1458 on client machines into AFS; see <link
1459 linkend="HDRWQ55">Configuring the Local Disk</link>. Even if
1460 you do not choose to use symbolic links in this way, it can
1461 be convenient to have central copies of system binaries in
1462 AFS. If binaries are accidentally removed from a machine,
1463 you can recopy them onto the local disk from AFS rather than
1464 having to recover them from tape</para>
1469 <term><emphasis role="bold">usr</emphasis></term>
1472 <para>This directory contains home directories for your
1473 local users. As discussed in the previous entry for the
1474 <emphasis role="bold">public</emphasis> directory, it is
1475 often practical to protect this directory so that only
1476 locally authenticated users can access it. This keeps the
1477 contents of your user's home directories as secure as
1480 <para>If your cell is quite large, directory lookup can be
1481 slowed if you put all home directories in a single <emphasis
1482 role="bold">usr</emphasis> directory. For suggestions on
1483 distributing user home directories among multiple grouping
1484 directories, see <link linkend="HDRWQ59">Grouping Home
1485 Directories</link>.</para>
1486 </listitem> </varlistentry>
1489 <term><emphasis role="bold">wsadmin</emphasis></term>
1492 <para>This directory contains prototype, configuration and
1493 library files for use with the <emphasis
1494 role="bold">package</emphasis> program. See <link
1495 linkend="HDRWQ419">Configuring Client Machines with the
1496 package Program</link>.</para>
1503 <primary>volume name</primary>
1505 <secondary>conventions for</secondary>
1509 <primary>conventions</primary>
1511 <secondary>volume names</secondary>
1515 <primary>volume</primary>
1517 <secondary>separate for each top level directory</secondary>
1521 <primary>file tree</primary>
1523 <secondary>creating volumes to match top level
1524 directories</secondary>
1529 <sect1 id="HDRWQ44">
1530 <title>Creating Volumes to Simplify Administration</title>
1532 <para>This section discusses how to create volumes in ways that make
1533 administering your system easier.</para>
1535 <para>At the top levels of your file tree (at least through the third
1536 level), each directory generally corresponds to a separate
1537 volume. Some cells also configure the subdirectories of some third
1538 level directories as separate volumes. Common examples are the
1540 role="bold">/afs/</emphasis><replaceable>cellname</replaceable><emphasis
1541 role="bold">/common</emphasis> and <emphasis
1542 role="bold">/afs/</emphasis><replaceable>cellname</replaceable><emphasis
1543 role="bold">/usr</emphasis> directories.</para>
1545 <para>You do not have to create a separate volume for every directory
1546 level in a tree, but the advantage is that each volume tends to be
1547 smaller and easier to move for load balancing. The overhead for a
1548 mount point is no greater than for a standard directory, nor does the
1549 volume structure itself require much disk space. Most cells find that
1550 below the fourth level in the tree, using a separate volume for each
1551 directory is no longer efficient. For instance, while each user's home
1552 directory (at the fourth level in the tree) corresponds to a separate
1553 volume, all of the subdirectories in the home directory normally
1554 reside in the same volume.</para>
1556 <para>Keep in mind that only one volume can be mounted at a given
1557 directory location in the tree. In contrast, a volume can be mounted
1558 at several locations, though this is not recommended because it
1559 distorts the hierarchical nature of the file tree, potentially causing
1563 <primary>volume name</primary>
1565 <secondary>restrictions</secondary>
1569 <primary>restrictions</primary>
1571 <secondary>on volume names</secondary>
1575 <primary>volume name</primary>
1577 <secondary>two required</secondary>
1581 <primary>volume</primary>
1583 <secondary>root (root.afs and root.cell)</secondary>
1587 <primary>root volumes (root.afs and root.cell)</primary>
1590 <sect2 id="Header_55">
1591 <title>Assigning Volume Names</title>
1593 <para>You can name your volumes anything you choose, subject to a
1597 <para>Read/write volume names can be up to 22 characters in
1598 length. The maximum length for volume names is 31 characters,
1599 and there must be room to add the <emphasis
1600 role="bold">.readonly</emphasis> extension on read-only
1605 <para>Do not add the <emphasis
1606 role="bold">.readonly</emphasis> and <emphasis
1607 role="bold">.backup</emphasis> extensions to volume names
1608 yourself, even if they are appropriate. The Volume Server adds
1609 them automatically as it creates a read-only or backup version
1614 <para>There must be volumes named <emphasis
1615 role="bold">root.afs</emphasis> and <emphasis
1616 role="bold">root.cell</emphasis>, mounted respectively at the
1617 top (<emphasis role="bold">/afs</emphasis>) level in the
1618 filespace and just below that level, at the cell's name (for
1619 example, at <emphasis role="bold">/afs/abc.com</emphasis> in
1620 the ABC Corporation cell).</para>
1622 <para>Deviating from these names only creates confusion and
1623 extra work. Changing the name of the <emphasis
1624 role="bold">root.afs</emphasis> volume, for instance, means
1625 that you must use the <emphasis
1626 role="bold">-rootvol</emphasis> argument to the <emphasis
1627 role="bold">afsd</emphasis> program on every client machine,
1628 to name the alternate volume.</para>
1630 <para>Similarly, changing the <emphasis
1631 role="bold">root.cell</emphasis> volume name prevents users in
1632 foreign cells from accessing your filespace, if the mount
1633 point for your cell in their filespace refers to the
1634 conventional <emphasis role="bold">root.cell</emphasis>
1635 name. Of course, this is one way to make your cell invisible
1636 to other cells.</para>
1641 <para>It is best to assign volume names that indicate the type of
1642 data they contain, and to use similar names for volumes with similar
1643 contents. It is also helpful if the volume name is similar to (or at
1644 least has elements in common with) the name of the directory at
1645 which it is mounted. Understanding the pattern then enables you
1646 accurately to guess what a volume contains and where it is
1649 <para>Many cells find that the most effective volume naming scheme
1650 puts a common prefix on the names of all related volumes. <link
1651 linkend="TBLVOL-PREFIX">Table 1</link> describes the recommended
1652 prefixing scheme.</para>
1654 <table id="TBLVOL-PREFIX" label="1">
1655 <title>Suggested volume prefixes</title>
1658 <colspec colwidth="14*" />
1660 <colspec colwidth="28*" />
1662 <colspec colwidth="22*" />
1664 <colspec colwidth="36*" />
1668 <entry><emphasis role="bold">Prefix</emphasis></entry>
1670 <entry><emphasis role="bold">Contents</emphasis></entry>
1672 <entry><emphasis role="bold">Example Name</emphasis></entry>
1674 <entry><emphasis role="bold">Example Mount
1675 Point</emphasis></entry>
1681 <entry><emphasis role="bold">common.</emphasis></entry>
1683 <entry>popular programs and files</entry>
1685 <entry><emphasis role="bold">common.etc</emphasis></entry>
1688 role="bold">/afs/</emphasis><replaceable>cellname</replaceable><emphasis
1689 role="bold">/common/etc</emphasis></entry>
1693 <entry><emphasis role="bold">src.</emphasis></entry>
1695 <entry>source code</entry>
1697 <entry><emphasis role="bold">src.afs</emphasis></entry>
1700 role="bold">/afs/</emphasis><replaceable>cellname</replaceable><emphasis
1701 role="bold">/src/afs</emphasis></entry>
1705 <entry><emphasis role="bold">proj.</emphasis></entry>
1707 <entry>project data</entry>
1709 <entry><emphasis role="bold">proj.portafs</emphasis></entry>
1712 role="bold">/afs/</emphasis><replaceable>cellname</replaceable><emphasis
1713 role="bold">/proj/portafs</emphasis></entry>
1717 <entry><emphasis role="bold">test.</emphasis></entry>
1719 <entry>testing or other temporary data</entry>
1721 <entry><emphasis role="bold">test.smith</emphasis></entry>
1724 role="bold">/afs/</emphasis><replaceable>cellname</replaceable><emphasis
1725 role="bold">/usr/smith/test</emphasis></entry>
1729 <entry><emphasis role="bold">user.</emphasis></entry>
1731 <entry>user home directory data</entry>
1733 <entry><emphasis role="bold">user.terry</emphasis></entry>
1736 role="bold">/afs/</emphasis><replaceable>cellname</replaceable><emphasis
1737 role="bold">/usr/terry</emphasis></entry>
1741 <entry>sys_type<emphasis role="bold">.</emphasis></entry>
1743 <entry>programs compiled for an operating system
1746 <entry><emphasis role="bold">rs_aix42.bin</emphasis></entry>
1749 role="bold">/afs/</emphasis><replaceable>cellname</replaceable><emphasis
1750 role="bold">/rs_aix42/bin</emphasis></entry>
1756 <para><link linkend="TBLPREFIX-EXAMPLE">Table 2</link> is a more
1757 specific example for a cell's <emphasis
1758 role="bold">rs_aix42</emphasis> system volumes and
1761 <table id="TBLPREFIX-EXAMPLE" label="2">
1762 <title>Example volume-prefixing scheme</title>
1765 <colspec colwidth="14*" />
1767 <colspec colwidth="28*" />
1769 <colspec colwidth="22*" />
1771 <colspec colwidth="36*" />
1775 <entry><emphasis role="bold">Example Name</emphasis></entry>
1777 <entry><emphasis role="bold">Example Mount
1778 Point</emphasis></entry>
1784 <entry><emphasis role="bold">rs_aix42.bin</emphasis></entry>
1787 role="bold">/afs/</emphasis><replaceable>cellname</replaceable><emphasis
1788 role="bold">/rs_aix42/bin</emphasis>, <emphasis
1789 role="bold">/afs/</emphasis><replaceable>cellname</replaceable><emphasis
1790 role="bold">/rs_aix42/bin</emphasis></entry>
1794 <entry><emphasis role="bold">rs_aix42.etc</emphasis></entry>
1797 role="bold">/afs/</emphasis><replaceable>cellname</replaceable><emphasis
1798 role="bold">/rs_aix42/etc</emphasis></entry>
1802 <entry><emphasis role="bold">rs_aix42.usr</emphasis></entry>
1805 role="bold">/afs/</emphasis><replaceable>cellname</replaceable><emphasis
1806 role="bold">/rs_aix42/usr</emphasis></entry>
1811 role="bold">rs_aix42.usr.afsws</emphasis></entry>
1814 role="bold">/afs/</emphasis><replaceable>cellname</replaceable><emphasis
1815 role="bold">/rs_aix42/usr/afsws</emphasis></entry>
1820 role="bold">rs_aix42.usr.lib</emphasis></entry>
1823 role="bold">/afs/</emphasis><replaceable>cellname</replaceable><emphasis
1824 role="bold">/rs_aix42/usr/lib</emphasis></entry>
1829 role="bold">rs_aix42.usr.bin</emphasis></entry>
1832 role="bold">/afs/</emphasis><replaceable>cellname</replaceable><emphasis
1833 role="bold">/rs_aix42/usr/bin</emphasis></entry>
1838 role="bold">rs_aix42.usr.etc</emphasis></entry>
1841 role="bold">/afs/</emphasis><replaceable>cellname</replaceable><emphasis
1842 role="bold">/rs_aix42/usr/etc</emphasis></entry>
1847 role="bold">rs_aix42.usr.inc</emphasis></entry>
1850 role="bold">/afs/</emphasis><replaceable>cellname</replaceable><emphasis
1851 role="bold">/rs_aix42/usr/inc</emphasis></entry>
1856 role="bold">rs_aix42.usr.man</emphasis></entry>
1859 role="bold">/afs/</emphasis><replaceable>cellname</replaceable><emphasis
1860 role="bold">/rs_aix42/usr/man</emphasis></entry>
1865 role="bold">rs_aix42.usr.sys</emphasis></entry>
1868 role="bold">/afs/</emphasis><replaceable>cellname</replaceable><emphasis
1869 role="bold">/rs_aix42/usr/sys</emphasis></entry>
1874 role="bold">rs_aix42.usr.local</emphasis></entry>
1877 role="bold">/afs/</emphasis><replaceable>cellname</replaceable><emphasis
1878 role="bold">/rs_aix42/usr/local</emphasis></entry>
1884 <para>There are several advantages to this scheme:
1887 <para>The volume name is similar to the mount point name in
1888 the filespace. In all of the entries in <link
1889 linkend="TBLPREFIX-EXAMPLE">Table 2</link>, for example, the
1890 only difference between the volume and mount point name is
1891 that the former uses periods as separators and the latter uses
1892 slashes. Another advantage is that the volume name indicates
1893 the contents, or at least suggests the directory on which to
1894 issue the <emphasis role="bold">ls</emphasis> command to learn
1895 the contents.</para>
1899 <para>It makes it easy to manipulate groups of related volumes
1900 at one time. In particular, the <emphasis role="bold">vos
1901 backupsys</emphasis> command's <emphasis
1902 role="bold">-prefix</emphasis> argument enables you to create
1903 a backup version of every volume whose name starts with the
1904 same string of characters. Making a backup version of each
1905 volume is one of the first steps in backing up a volume with
1906 the AFS Backup System, and doing it for many volumes with one
1907 command saves you a good deal of typing. For instructions for
1908 creating backup volumes, see <link linkend="HDRWQ201">Creating
1909 Backup Volumes</link>, For information on the AFS Backup
1910 System, see <link linkend="HDRWQ248">Configuring the AFS
1911 Backup System</link> and <link linkend="HDRWQ283">Backing Up
1912 and Restoring AFS Data</link>.</para>
1916 <para>It makes it easy to group related volumes together on a
1917 partition. Grouping related volumes together has several
1918 advantages of its own, discussed in <link
1919 linkend="HDRWQ49">Grouping Related Volumes on a
1920 Partition</link>.</para>
1926 <primary>volume</primary>
1928 <secondary>grouping related on same partition</secondary>
1932 <primary>disk partition</primary>
1934 <secondary>grouping related volumes on</secondary>
1938 <sect2 id="HDRWQ49">
1939 <title>Grouping Related Volumes on a Partition</title>
1941 <para>If your cell is large enough to make it practical, consider
1942 grouping related volumes together on a partition. In general, you
1943 need at least three file server machines for volume grouping to be
1944 effective. Grouping has several advantages, which are most obvious
1945 when the file server machine becomes inaccessible:
1948 <para>If you keep a hardcopy record of the volumes on a
1949 partition, you know which volumes are unavailable. You can
1950 keep such a record without grouping related volumes, but a
1951 list composed of unrelated volumes is much harder to maintain.
1952 Note that the record must be on paper, because the outage can
1953 prevent you from accessing an online copy or from issuing the
1954 <emphasis role="bold">vos listvol</emphasis> command, which
1955 gives you the same information.</para>
1959 <para>The effect of an outage is more localized. For example,
1960 if all of the binaries for a given system type are on one
1961 partition, then only users of that system type are
1962 affected. If a partition houses binary volumes from several
1963 system types, then an outage can affect more people,
1964 particularly if the binaries that remain available are
1965 interdependent with those that are not available.</para>
1970 <para>The advantages of grouping related volumes on a partition do
1971 not necessarily extend to the grouping of all related volumes on one
1972 file server machine. For instance, it is probably unwise in a cell
1973 with two file server machines to put all system volumes on one
1974 machine and all user volumes on the other. An outage of either
1975 machine probably affects everyone.</para>
1977 <para>Admittedly, the need to move volumes for load balancing
1978 purposes can limit the practicality of grouping related volumes.
1979 You need to weigh the complementary advantages case by case.</para>
1982 <primary>replication</primary>
1984 <secondary>appropriate volumes</secondary>
1988 <primary>volume</primary>
1990 <secondary>type to replicate</secondary>
1994 <primary>volume</primary>
1996 <secondary>where to place replicated</secondary>
2000 <primary>read-only volume</primary>
2002 <secondary>selecting site</secondary>
2006 <sect2 id="HDRWQ50">
2007 <title>When to Replicate Volumes</title>
2009 <para>As discussed in <link linkend="HDRWQ15">Replication</link>,
2010 replication refers to making a copy, or clone, of a read/write
2011 source volume and then placing the copy on one or more additional
2012 file server machines. Replicating a volume can increase the
2013 availability of the contents. If one file server machine housing the
2014 volume becomes inaccessible, users can still access the copy of the
2015 volume stored on a different machine. No one machine is likely to
2016 become overburdened with requests for a popular file, either,
2017 because the file is available from several machines.</para>
2019 <para>However, replication is not appropriate for all cells. If a
2020 cell does not have much disk space, replication can be unduly
2021 expensive, because each clone not on the same partition as the
2022 read/write source takes up as much disk space as its source volume
2023 did at the time the clone was made. Also, if you have only one file
2024 server machine, replication uses up disk space without increasing
2025 availability.</para>
2027 <para>Replication is also not appropriate for volumes that change
2028 frequently. You must issue the <emphasis role="bold">vos
2029 release</emphasis> command every time you need to update a read-only
2030 volume to reflect changes in its read/write source.</para>
2032 <para>For both of these reasons, replication is appropriate only for
2033 popular volumes whose contents do not change very often, such as
2034 system binaries and other volumes mounted at the upper levels of
2035 your filespace. User volumes usually exist only in a read/write
2036 version since they change so often.</para>
2038 <para>If you are replicating any volumes, you must replicate the
2039 <emphasis role="bold">root.afs</emphasis> and <emphasis
2040 role="bold">root.cell</emphasis> volumes, preferably at two or three
2041 sites each (even if your cell only has two or three file server
2042 machines). The Cache Manager needs to pass through the directories
2043 corresponding to the <emphasis role="bold">root.afs</emphasis> and
2044 <emphasis role="bold">root.cell</emphasis> volumes as it interprets
2045 any pathname. The unavailability of these volumes makes all other
2046 volumes unavailable too, even if the file server machines storing
2047 the other volumes are still functioning.</para>
2049 <para>Another reason to replicate the <emphasis
2050 role="bold">root.afs</emphasis> volume is that it can lessen the
2051 load on the File Server machine. The Cache Manager has a bias to
2052 access a read-only version of the <emphasis
2053 role="bold">root.afs</emphasis> volume if it is replicate, which
2054 puts the Cache Manager onto the <emphasis>read-only path</emphasis>
2055 through the AFS filespace. While on the read-only path, the Cache
2056 Manager attempts to access a read-only copy of replicated
2057 volumes. The File Server needs to track only one callback per Cache
2058 Manager for all of the data in a read-only volume, rather than the
2059 one callback per file it must track for read/write volumes. Fewer
2060 callbacks translate into a smaller load on the File Server.</para>
2062 <para>If the <emphasis role="bold">root.afs</emphasis> volume is not
2063 replicated, the Cache Manager follows a read/write path through the
2064 filespace, accessing the read/write version of each volume. The File
2065 Server distributes and tracks a separate callback for each file in a
2066 read/write volume, imposing a greater load on it.</para>
2068 <para>For more on read/write and read-only paths, see <link
2069 linkend="HDRWQ209">The Rules of Mount Point Traversal</link>.</para>
2071 <para>It also makes sense to replicate system binary volumes in many
2072 cases, as well as the volume corresponding to the <emphasis
2073 role="bold">/afs/</emphasis><replaceable>cellname</replaceable><emphasis
2074 role="bold">/usr</emphasis> directory and the volumes corresponding
2076 role="bold">/afs/</emphasis><replaceable>cellname</replaceable><emphasis
2077 role="bold">/common</emphasis> directory and its
2078 subdirectories.</para>
2080 <para>It is a good idea to place a replica on the same partition as
2081 the read/write source. In this case, the read-only volume is a clone
2082 (like a backup volume): it is a copy of the source volume's vnode
2083 index, rather than a full copy of the volume contents. Only if the
2084 read/write volume moves to another partition or changes
2085 substantially does the read-only volume consume significant disk
2086 space. Read-only volumes kept on other partitions always consume the
2087 full amount of disk space that the read/write source consumed when
2088 the read-only volume was created.</para>
2091 <sect2 id="Header_58">
2092 <title>The Default Quota and ACL on a New Volume</title>
2094 <para>Every AFS volume has associated with it a quota that limits
2095 the amount of disk space the volume is allowed to use. To set and
2096 change quota, use the commands described in <link
2097 linkend="HDRWQ234">Setting and Displaying Volume Quota and Current
2100 <para>By default, every new volume is assigned a space quota of 5000
2101 KB blocks unless you include the <emphasis
2102 role="bold">-maxquota</emphasis> argument to the <emphasis
2103 role="bold">vos create</emphasis> command. Also by default, the ACL
2104 on the root directory of every new volume grants all permissions to
2105 the members of the <emphasis
2106 role="bold">system:administrators</emphasis> group. To learn how to
2107 change these values when creating an account with individual
2108 commands, see <link linkend="HDRWQ503">To create one user account
2109 with individual commands</link>. When using <emphasis
2110 role="bold">uss</emphasis> commands to create accounts, you can
2111 specify alternate ACL and quota values in the template file's
2112 <emphasis role="bold">V</emphasis> instruction; see <link
2113 linkend="HDRWQ473">Creating a Volume with the V
2114 Instruction</link>.</para>
2117 <primary>server machine</primary>
2119 <secondary>configuration issues</secondary>
2123 <primary>configuring</primary>
2125 <secondary>file server machine, issues</secondary>
2129 <primary>roles for server machine</primary>
2131 <secondary>summary</secondary>
2135 <primary>server machine</primary>
2137 <secondary>roles for</secondary>
2139 <tertiary>summary</tertiary>
2143 <primary>server machine</primary>
2145 <secondary>first installed</secondary>
2150 <sect1 id="HDRWQ51">
2151 <title>Configuring Server Machines</title>
2153 <para>This section discusses some issues to consider when configuring
2154 server machines, which store AFS data, transfer it to client machines
2155 on request, and house the AFS administrative databases. To learn about
2156 client machines, see <link linkend="HDRWQ54">Configuring Client
2157 Machines</link>.</para>
2159 <para>If your cell has more than one AFS server machine, you can
2160 configure them to perform specialized functions. A machine can assume
2161 one or more of the roles described in the following list. For more
2162 details, see <link linkend="HDRWQ90">The Four Roles for File Server
2166 <para>A <emphasis>simple file server</emphasis> machine runs
2167 only the processes that store and deliver AFS files to client
2168 machines. You can run as many simple file server machines as you
2169 need to satisfy your cell's performance and disk space
2170 requirements.</para>
2174 <para>A <emphasis>database server machine</emphasis> runs the
2175 four database server processes that maintain AFS's replicated
2176 administrative databases: the Authentication, Backup,
2177 Protection, and Volume Location (VL) Server processes.</para>
2181 <para>A <emphasis>binary distribution machine</emphasis>
2182 distributes the AFS server binaries for its system type to all
2183 other server machines of that system type.</para>
2187 <para>The single <emphasis>system control machine</emphasis>
2188 distributes common server configuration files to all other
2189 server machines in the cell, in a cell that runs the United
2190 States edition of AFS (cells that use the international edition
2191 of AFS must not use the system control machine for this
2192 purpose). The machine conventionally also serves as the time
2193 synchronization source for the cell, adjusting its clock
2194 according to a time source outside the cell.</para>
2199 <para>The <emphasis>OpenAFS Quick Beginnings</emphasis> explains how
2200 to configure your cell's first file server machine to assume all four
2201 roles. The <emphasis>OpenAFS Quick Beginnings</emphasis> chapter on
2202 installing additional server machines also explains how to configure
2203 them to perform one or more roles.</para>
2206 <primary>database server machine</primary>
2208 <secondary>reason to run three</secondary>
2212 <primary>distribution</primary>
2214 <secondary>of databases</secondary>
2218 <primary>databases, distributed</primary>
2222 <primary>distributed databases</primary>
2225 <sect2 id="HDRWQ52">
2226 <title>Replicating the OpenAFS Administrative Databases</title>
2228 <para>The AFS administrative databases are housed on database server
2229 machines and store information that is crucial for correct cell
2230 functioning. Both server processes and Cache Managers access the
2231 information frequently:
2234 <para>Every time a Cache Manager fetches a file from a
2235 directory that it has not previously accessed, it must look up
2236 the file's location in the Volume Location Database
2241 <para>Every time a user obtains an AFS token from the
2242 Authentication Server, the server looks up the user's password
2243 in the Authentication Database.</para>
2247 <para>The first time that a user accesses a volume housed on a
2248 specific file server machine, the File Server contacts the
2249 Protection Server for a list of the user's group memberships
2250 as recorded in the Protection Database.</para>
2254 <para>Every time you back up a volume using the AFS Backup
2255 System, the Backup Server creates records for it in the Backup
2261 <para>Maintaining your cell is simplest if the first machine has the
2262 lowest IP address of any machine you plan to use as a database
2263 server machine. If you later decide to use a machine with a lower IP
2264 address as a database server machine, you must update the <emphasis
2265 role="bold">CellServDB</emphasis> file on all clients before
2266 introducing the new machine.</para>
2268 <para>If your cell has more than one server machine, it is best to
2269 run more than one as a database server machine (but more than three
2270 are rarely necessary). Replicating the administrative databases in
2271 this way yields the same benefits as replicating volumes: increased
2272 availability and reliability. If one database server machine or
2273 process stops functioning, the information in the database is still
2274 available from others. The load of requests for database information
2275 is spread across multiple machines, preventing any one from becoming
2278 <para>Unlike replicated volumes, however, replicated databases do
2279 change frequently. Consistent system performance demands that all
2280 copies of the database always be identical, so it is not acceptable
2281 to record changes in only some of them. To synchronize the copies of
2282 a database, the database server processes use AFS's distributed
2283 database technology, Ubik. See <link linkend="HDRWQ102">Replicating
2284 the OpenAFS Administrative Databases</link>.</para>
2286 <para>If your cell has only one file server machine, it must also
2287 serve as a database server machine. If you cell has two file server
2288 machines, it is not always advantageous to run both as database
2289 server machines. If a server, process, or network failure interrupts
2290 communications between the database server processes on the two
2291 machines, it can become impossible to update the information in the
2292 database because neither of them can alone elect itself as the
2293 synchronization site.</para>
2296 <primary>server machine</primary>
2298 <secondary>protecting directories on local disk</secondary>
2302 <primary>local disk</primary>
2304 <secondary>protecting on file server machine</secondary>
2308 <sect2 id="HDRWQ53">
2309 <title>AFS Files on the Local Disk</title>
2311 <para>It is generally simplest to store the binaries for all AFS
2312 server processes in the <emphasis
2313 role="bold">/usr/afs/bin</emphasis> directory on every file server
2314 machine, even if some processes do not actively run on the
2315 machine. This makes it easier to reconfigure a machine to fill a new
2318 <para>For security reasons, the <emphasis
2319 role="bold">/usr/afs</emphasis> directory on a file server machine
2320 and all of its subdirectories and files must be owned by the local
2321 superuser <emphasis role="bold">root</emphasis> and have only the
2322 first <emphasis role="bold">w</emphasis> (<emphasis
2323 role="bold">write</emphasis>) mode bit turned on. Some files even
2324 have only the first <emphasis role="bold">r</emphasis> (<emphasis
2325 role="bold">read</emphasis>) mode bit turned on (for example, the
2326 <emphasis role="bold">/usr/afs/etc/KeyFile</emphasis> file, which
2327 lists the AFS server encryption keys). Each time the BOS Server
2328 starts, it checks that the mode bits on certain files and
2329 directories match the expected values. For a list, see the
2330 <emphasis>OpenAFS Quick Beginnings</emphasis> section about
2331 protecting sensitive AFS directories, or the discussion of the
2332 output from the <emphasis role="bold">bos status</emphasis> command
2333 in <link linkend="HDRWQ159">To display the status of server
2334 processes and their BosConfig entries</link>.</para>
2336 <para>For a description of the contents of all AFS directories on a
2337 file server machine's local disk, see <link
2338 linkend="HDRWQ80">Administering Server Machines</link>.</para>
2341 <sect2 id="Header_62">
2342 <title>Configuring Partitions to Store AFS Data</title>
2344 <para>The partitions that house AFS volumes on a file server machine
2345 must be mounted at directories named</para>
2348 role="bold">/vicep</emphasis><emphasis>index</emphasis></para>
2350 <para>where <emphasis>index</emphasis> is one or two lowercase
2351 letters. By convention, the first AFS partition created is mounted
2352 at the <emphasis role="bold">/vicepa</emphasis> directory, the
2353 second at the <emphasis role="bold">/vicepb</emphasis> directory,
2354 and so on through the <emphasis role="bold">/vicepz</emphasis>
2355 directory. The names then continue with <emphasis
2356 role="bold">/vicepaa</emphasis> through <emphasis
2357 role="bold">/vicepaz</emphasis>, <emphasis
2358 role="bold">/vicepba</emphasis> through <emphasis
2359 role="bold">/vicepbz</emphasis>, and so on, up to the maximum
2360 supported number of server partitions, which is specified in the
2361 OpenAFS Release Notes.</para>
2363 <para>Each <emphasis role="bold">/vicep</emphasis>x directory must
2364 correspond to an entire partition or logical volume, and must be a
2365 subdirectory of the root directory (/). It is not acceptable to
2366 configure part of (for example) the <emphasis
2367 role="bold">/usr</emphasis> partition as an AFS server partition and
2368 mount it on a directory called <emphasis
2369 role="bold">/usr/vicepa</emphasis>.</para>
2371 <para>Also, do not store non-AFS files on AFS server partitions. The
2372 File Server and Volume Server expect to have available all of the
2373 space on the partition. Sharing space also creates competition
2374 between AFS and the local UNIX file system for access to the
2375 partition, particularly if the UNIX files are frequently
2379 <primary>server machine</primary>
2381 <secondary>monitoring</secondary>
2385 <primary>file server machine</primary>
2387 <secondary>rebooting, about</secondary>
2391 <primary>rebooting</primary>
2393 <secondary>file server machine, limiting</secondary>
2397 <primary>weekly restart of BOS Server (automatic)</primary>
2399 <secondary>about</secondary>
2403 <primary>restart times for BOS Server</primary>
2405 <secondary>about</secondary>
2409 <sect2 id="Header_63">
2410 <title>Monitoring, Rebooting and Automatic Process Restarts</title>
2412 <para>AFS provides several tools for monitoring the File Server,
2413 including the <emphasis role="bold">scout</emphasis> and <emphasis
2414 role="bold">afsmonitor</emphasis> programs. You can configure them
2415 to alert you when certain threshold values are exceeded, for example
2416 when a server partition is more than 95% full. See <link
2417 linkend="HDRWQ323">Monitoring and Auditing AFS
2418 Performance</link>.</para>
2420 <para>Rebooting a file server machine requires shutting down the AFS
2421 processes and so inevitably causes a service outage. Reboot file
2422 server machines as infrequently as possible. For instructions, see
2423 <link linkend="HDRWQ139">Rebooting a Server Machine</link>.</para>
2425 <para>The BOS Server checks each morning at 5:00 a.m. for any newly
2426 installed binary files in the <emphasis
2427 role="bold">/usr/afs/bin</emphasis> directory. It compares the
2428 timestamp on each binary file to the time at which the corresponding
2429 process last restarted. If the timestamp on the binary is later, the
2430 BOS Server restarts the corresponding process to start using
2433 <para>The BOS server also supports performing a weekly restart of
2434 all AFS server processes, including itself. This functionality is
2435 disabled on new installs, but historically it was set to 4:00am on
2436 Sunday. Administrators may find that installations predating OpenAFS
2437 1.6.0 have weekly restarts enabled.</para>
2439 <para>The default times are in the early morning hours when the
2440 outage that results from restarting a process is likely to disturb
2441 the fewest number of people. You can display the restart times for
2442 each machine with the <emphasis role="bold">bos
2443 getrestart</emphasis> command, and set them with the <emphasis
2444 role="bold">bos setrestart</emphasis> command. The latter command
2445 enables you to disable automatic restarts entirely, by setting the
2446 time to <emphasis role="bold">never</emphasis>. See <link
2447 linkend="HDRWQ171">Setting the BOS Server's Restart
2448 Times</link>.</para>
2451 <primary>client machine</primary>
2453 <secondary>configuration issues</secondary>
2457 <primary>configuring</primary>
2459 <secondary>client machine, issues</secondary>
2464 <sect1 id="HDRWQ54">
2465 <title>Configuring Client Machines</title>
2467 <para>This section summarizes issues to consider as you install and
2468 configure client machines in your cell.</para>
2471 <primary>client machine</primary>
2473 <secondary>files required on local disk</secondary>
2477 <primary>local disk</primary>
2479 <secondary>files required on client machine</secondary>
2483 <primary>file</primary>
2485 <secondary>required on client machine local disk</secondary>
2488 <sect2 id="HDRWQ55">
2489 <title>Configuring the Local Disk</title>
2491 <para>You can often free up significant amounts of local disk space
2492 on AFS client machines by storing standard UNIX files in AFS and
2493 creating symbolic links to them from the local disk. The <emphasis
2494 role="bold">@sys</emphasis> pathname variable can be useful in links
2495 to system-specific files; see <link linkend="HDRWQ56">Using the @sys
2496 Variable in Pathnames</link>.</para>
2498 <para>There are two types of files that must actually reside on the
2499 local disk: boot sequence files needed before the <emphasis
2500 role="bold">afsd</emphasis> program is invoked, and files that can
2501 be helpful during file server machine outages.</para>
2503 <para>During a reboot, AFS is inaccessible until the <emphasis
2504 role="bold">afsd</emphasis> program executes and initializes the
2505 Cache Manager. (In the conventional configuration, the AFS
2506 initialization file is included in the machine's initialization
2507 sequence and invokes the <emphasis role="bold">afsd</emphasis>
2508 program.) Files needed during reboot prior to that point must reside
2509 on the local disk. They include the following, but this list is not
2510 necessarily exhaustive.
2513 <para>Standard UNIX utilities including the following or their
2517 <para>Machine initialization files (stored in the
2518 <emphasis role="bold">/etc</emphasis> or <emphasis
2519 role="bold">/sbin</emphasis> directory on many system
2524 <para>The <emphasis role="bold">fstab</emphasis>
2529 <para>The <emphasis role="bold">mount</emphasis> command
2534 <para>The <emphasis role="bold">umount</emphasis>
2535 command binary</para>
2542 <para>All subdirectories and files in the <emphasis
2543 role="bold">/usr/vice</emphasis> directory, including the
2548 role="bold">/usr/vice/cache</emphasis> directory</para>
2553 role="bold">/usr/vice/etc/afsd</emphasis> command
2559 role="bold">/usr/vice/etc/cacheinfo</emphasis> file</para>
2564 role="bold">/usr/vice/etc/CellServDB</emphasis>
2570 role="bold">/usr/vice/etc/ThisCell</emphasis> file</para>
2575 <para>For more information on these files, see <link
2576 linkend="HDRWQ391">Configuration and Cache-Related Files on
2577 the Local Disk</link>.</para>
2582 <para>The other type of files and programs to retain on the local
2583 disk are those you need when diagnosing and fixing problems caused
2584 by a file server outage, because the outage can make inaccessible
2585 the copies stored in AFS. Examples include the binaries for a text
2586 editor (such as <emphasis role="bold">ed</emphasis> or <emphasis
2587 role="bold">vi</emphasis>) and for the <emphasis
2588 role="bold">fs</emphasis> and <emphasis role="bold">bos</emphasis>
2589 commands. Store copies of AFS command binaries in the <emphasis
2590 role="bold">/usr/vice/etc</emphasis> directory as well as including
2591 them in the <emphasis role="bold">/usr/afsws</emphasis> directory,
2592 which is normally a link into AFS. Then place the <emphasis
2593 role="bold">/usr/afsws</emphasis> directory before the <emphasis
2594 role="bold">/usr/vice/etc</emphasis> directory in users'
2595 <envar>PATH</envar> environment variable definition. When AFS is
2596 functioning normally, users access the copy in the <emphasis
2597 role="bold">/usr/afsws</emphasis> directory, which is more likely to
2598 be current than a local copy.</para>
2600 <para>You can automate the configuration of client machine local
2601 disks by using the <emphasis role="bold">package</emphasis> program,
2602 which updates the contents of the local disk to match a
2603 configuration file. See <link linkend="HDRWQ419">Configuring Client
2604 Machines with the package Program</link>.</para>
2607 <sect2 id="Header_66">
2608 <title>Enabling Access to Foreign Cells</title>
2611 <primary>client machine</primary>
2613 <secondary>enabling access to foreign cell</secondary>
2616 <para>As detailed in <link linkend="HDRWQ39">Making Other Cells
2617 Visible in Your Cell</link>, you enable the Cache Manager to access
2618 a cell's AFS filespace by storing a list of the cell's database
2619 server machines in the local <emphasis
2620 role="bold">/usr/vice/etc/CellServDB</emphasis> file. The Cache
2621 Manager reads the list into kernel memory at reboot for faster
2622 retrieval. You can change the list in kernel memory between reboots
2623 by using the <emphasis role="bold">fs newcell</emphasis> command. It
2624 is often practical to store a central version of the <emphasis
2625 role="bold">CellServDB</emphasis> file in AFS and use the <emphasis
2626 role="bold">package</emphasis> program periodically to update each
2627 client's version with the source copy. See <link
2628 linkend="HDRWQ406">Maintaining Knowledge of Database Server
2629 Machines</link>.</para>
2631 <para>Because each client machine maintains its own copy of the
2632 <emphasis role="bold">CellServDB</emphasis> file, you can in theory
2633 enable access to different foreign cells on different client
2634 machines. This is not usually practical, however, especially if
2635 users do not always work on the same machine.</para>
2638 <primary>at-sys (@sys) variable in pathnames</primary>
2642 <primary>sys (@sys) variable in pathnames</primary>
2646 <primary>variables</primary>
2648 <secondary>@sys in pathnames</secondary>
2652 <sect2 id="HDRWQ56">
2653 <title>Using the @sys Variable in Pathnames</title>
2655 <para>When creating symbolic links into AFS on the local disk, it is
2656 often practical to use the @sys variable in pathnames. The Cache
2657 Manager automatically substitutes the local machine's AFS system
2658 name (CPU/operating system type) for the @sys variable. This means
2659 you can place the same links on machines of various system types and
2660 still have each machine access the binaries for its system type. For
2661 example, the Cache Manager on a machine running AIX 4.2 converts
2662 <emphasis role="bold">/afs/abc.com/@sys</emphasis> to <emphasis
2663 role="bold">/afs/abc.com/rs_aix42</emphasis>, whereas a machine
2664 running Solaris 7 converts it to <emphasis
2665 role="bold">/afs/abc.com/sun4x_57</emphasis>.</para>
2667 <para>If you want to use the @sys variable, it is simplest to use
2668 the conventional AFS system type names as specified in the OpenAFS
2669 Release Notes. The Cache Manager records the local machine's system
2670 type name in kernel memory during initialization. If you do not use
2671 the conventional names, you must use the <emphasis role="bold">fs
2672 sysname</emphasis> command to change the value in kernel memory from
2673 its default just after Cache Manager initialization, on every client
2674 machine of the relevant system type. The <emphasis role="bold">fs
2675 sysname</emphasis> command also displays the current value; see
2676 <link linkend="HDRWQ417">Displaying and Setting the System Type
2679 <para>In pathnames in the AFS filespace itself, use the @sys
2680 variable carefully and sparingly, because it can lead to unexpected
2681 results. It is generally best to restrict its use to only one level
2682 in the filespace. The third level is a common choice, because that
2683 is where many cells store the binaries for different machine
2686 <para>Multiple instances of the @sys variable in a pathname are
2687 especially dangerous to people who must explicitly change
2688 directories (with the <emphasis role="bold">cd</emphasis> command,
2689 for example) into directories that store binaries for system types
2690 other than the machine on which they are working, such as
2691 administrators or developers who maintain those directories. After
2692 changing directories, it is recommended that such people verify they
2693 are in the desired directory.</para>
2696 <sect2 id="Header_68">
2697 <title>Setting Server Preferences</title>
2699 <para>The Cache Manager stores a table of preferences for file
2700 server machines in kernel memory. A preference rank pairs a file
2701 server machine interface's IP address with an integer in the range
2702 from 1 to 65,534. When it needs to access a file, the Cache Manager
2703 compares the ranks for the interfaces of all machines that house the
2704 file, and first attempts to access the file via the interface with
2705 the best rank. As it initializes, the Cache Manager sets default
2706 ranks that bias it to access files via interfaces that are close to
2707 it in terms of network topology. You can adjust the preference ranks
2708 to improve performance if you wish.</para>
2710 <para>The Cache Manager also uses similar preferences for Volume
2711 Location (VL) Server machines. Use the <emphasis role="bold">fs
2712 getserverprefs</emphasis> command to display preference ranks and
2713 the <emphasis role="bold">fs setserverprefs</emphasis> command to
2714 set them. See <link linkend="HDRWQ414">Maintaining Server Preference
2715 Ranks</link>.</para>
2718 <primary>user account</primary>
2720 <secondary>configuration issues</secondary>
2724 <sect1 id="HDRWQ57">
2725 <title>Configuring AFS User Accounts</title>
2727 <para>This section discusses some of the issues to consider when
2728 configuring AFS user accounts. Because AFS is separate from the UNIX
2729 file system, a user's AFS account is separate from her UNIX
2732 <para>The preferred method for creating a user account is with the
2733 <emphasis role="bold">uss</emphasis> suite of commands. With a single
2734 command, you can create all the components of one or many accounts,
2735 after you have prepared a template file that guides the account
2736 creation. See <link linkend="HDRWQ449">Creating and Deleting User
2737 Accounts with the uss Command Suite</link>.</para>
2739 <para>Alternatively, you can issue the individual commands that create
2740 each component of an account. For instructions, along with
2741 instructions for removing user accounts and changing user passwords,
2742 user volume quotas and usernames, see <link
2743 linkend="HDRWQ491">Administering User Accounts</link>.</para>
2745 <para>When users leave your system, it is often good policy to remove
2746 their accounts. Instructions appear in <link
2747 linkend="HDRWQ486">Deleting Individual Accounts with the uss delete
2748 Command</link> and <link linkend="HDRWQ524">Removing a User
2749 Account</link>.</para>
2751 <para>An AFS user account consists of the following components, which
2752 are described in greater detail in <link linkend="HDRWQ494">The
2753 Components of an AFS User Account</link>.
2756 <para>A Protection Database entry</para>
2760 <para>An Authentication Database entry</para>
2764 <para>A volume</para>
2768 <para>A home directory at which the volume is mounted</para>
2772 <para>Ownership of the home directory and full permissions on
2777 <para>An entry in the local password file (<emphasis
2778 role="bold">/etc/passwd</emphasis> or equivalent) of each
2779 machine the user needs to log into</para>
2783 <para>Optionally, standard files and subdirectories that make
2784 the account more useful</para>
2789 <para>By creating some components but not others, you can create
2790 accounts at different levels of functionality, using either <emphasis
2791 role="bold">uss</emphasis> commands as described in <link
2792 linkend="HDRWQ449">Creating and Deleting User Accounts with the uss
2793 Command Suite</link> or individual commands as described in <link
2794 linkend="HDRWQ491">Administering User Accounts</link>. The levels of
2795 functionality include the following
2798 <para>An authentication-only account enables the user to obtain
2799 AFS tokens and so to access protected AFS data and to issue
2800 privileged commands. It consists only of entries in the
2801 Authentication and Protection Database. This type of account is
2802 suitable for administrative accounts and for users from foreign
2803 cells who need to access protected data. Local users generally
2804 also need a volume and home directory.</para>
2808 <para>A basic user account includes a volume for the user, in
2809 addition to Authentication and Protection Database entries. The
2810 volume is mounted in the AFS filespace as the user's home
2811 directory, and provides a repository for the user's personal
2816 <para>A full account adds configuration files for basic
2817 functions such as logging in, printing, and mail delivery to a
2818 basic account, making it more convenient and useful. For a
2819 discussion of some useful types of configuration files, see
2820 <link linkend="HDRWQ60">Creating Standard Files in New AFS
2821 Accounts</link>.</para>
2826 <para>If your users have UNIX user accounts that predate the
2827 introduction of AFS in the cell, you possibly want to convert them
2828 into AFS accounts. There are three main issues to consider:
2831 <para>Making UNIX and AFS UIDs match</para> </listitem>
2834 <para>Setting the password field in the local password file
2835 appropriately</para>
2839 <para>Moving files from the UNIX file system into AFS</para>
2844 <para>For further discussion, see <link linkend="HDRWQ459">Converting
2845 Existing UNIX Accounts with uss</link> or <link
2846 linkend="HDRWQ498">Converting Existing UNIX Accounts</link>.</para>
2849 <primary>username</primary>
2851 <secondary>choosing</secondary>
2855 <primary>user</primary>
2857 <secondary>name</secondary>
2863 <primary>choosing</primary>
2865 <secondary>name</secondary>
2867 <tertiary>user</tertiary>
2871 <primary>anonymous user</primary>
2873 <secondary>AFS UID reserved</secondary>
2877 <primary>AFS UID</primary>
2879 <secondary>reserved</secondary>
2881 <tertiary>anonymous user</tertiary>
2884 <sect2 id="HDRWQ58">
2885 <title>Choosing Usernames and Naming Other Account
2888 <para>This section suggests schemes for choosing usernames, AFS
2889 UIDs, user volume names and mount point names, and also outlines
2890 some restrictions on your choices.</para>
2893 <title>Usernames</title>
2895 <para>AFS imposes very few restrictions on the form of
2896 usernames. It is best to keep usernames short, both because many
2897 utilities and applications can handle usernames of no more than
2898 eight characters and because by convention many components of and
2899 AFS account incorporate the name. These include the entries in the
2900 Protection and Authentication Databases, the volume, and the mount
2901 point. Depending on your electronic mail delivery system, the
2902 username can become part of the user's mailing address. The
2903 username is also the string that the user types when logging in to
2904 a client machine.</para>
2907 <para>Some common choices for usernames are last names, first names,
2908 initials, or a combination, with numbers sometimes added. It is
2909 also best to avoid using the following characters, many of which
2910 have special meanings to the command shell.
2913 <para>The comma (<emphasis role="bold">,</emphasis>)</para>
2917 <para>The colon (<emphasis role="bold">:</emphasis>), because
2918 AFS reserves it as a field separator in protection group
2919 names; see <link linkend="HDRWQ62">The Two Types of
2920 User-Defined Groups</link></para>
2924 <para>The semicolon (<emphasis
2925 role="bold">;</emphasis>)</para>
2929 <para>The "at-sign" (<emphasis role="bold">@</emphasis>); this
2930 character is reserved for Internet mailing addresses</para>
2938 <para>The newline character</para>
2942 <para>The period (<emphasis role="bold">.</emphasis>); it is
2943 conventional to use this character only in the special
2944 username that an administrator adopts while performing
2945 privileged tasks, such as <emphasis
2946 role="bold">pat.admin</emphasis></para>
2952 <title>AFS UIDs and UNIX UIDs</title>
2954 <para>AFS associates a unique identification number, the AFS UID,
2955 with every username, recording the mapping in the user's
2956 Protection Database entry. The AFS UID functions within AFS much
2957 as the UNIX UID does in the local file system: the AFS server
2958 processes and the Cache Manager use it internally to identify a
2959 user, rather than the username.</para>
2962 <para>Every AFS user also must have a UNIX UID recorded in the local
2963 password file (<emphasis role="bold">/etc/passwd</emphasis> or
2964 equivalent) of each client machine they log onto. Both
2965 administration and a user's AFS access are simplest if the AFS UID
2966 and UNIX UID match. One important consequence of matching UIDs is
2967 that the owner reported by the <emphasis role="bold">ls
2968 -l</emphasis> command matches the AFS username.</para>
2970 <para>It is usually best to allow the Protection Server to allocate
2971 the AFS UID as it creates the Protection Database entry. However,
2972 both the <emphasis role="bold">pts createuser</emphasis> command and
2973 the <emphasis role="bold">uss</emphasis> commands that create user
2974 accounts enable you to assign AFS UIDs explicitly. This is
2975 appropriate in two cases:
2978 <para>You wish to group together the AFS UIDs of related
2983 <para>You are converting an existing UNIX account into an AFS
2984 account and want to make the AFS UID match the existing UNIX
2990 <para>After the Protection Server initializes for the first time on
2991 a cell's first file server machine, it starts assigning AFS UIDs at
2992 a default value. To change the default before creating any user
2993 accounts, or at any time, use the <emphasis role="bold">pts
2994 setmax</emphasis> command to reset the <computeroutput>max user id
2995 counter</computeroutput>. To display the counter, use the <emphasis
2996 role="bold">pts listmax</emphasis> command. See <link
2997 linkend="HDRWQ560">Displaying and Setting the AFS UID and GID
2998 Counters</link>.</para>
3000 <para>AFS reserves one AFS UID, 32766, for the user <emphasis
3001 role="bold">anonymous</emphasis>. The AFS server processes assign
3002 this identity and AFS UID to any user who does not possess a token
3003 for the local cell. Do not assign this AFS UID to any other user or
3004 hardcode its current value into any programs or a file's owner
3005 field, because it is subject to change in future releases.</para>
3008 <primary>username</primary>
3010 <secondary>part of volume name</secondary>
3014 <primary>choosing</primary>
3016 <secondary>name</secondary>
3018 <tertiary>user volume</tertiary>
3022 <title>User Volume Names</title>
3024 <para>Like any volume name, a user volume's base (read/write) name
3025 cannot exceed 22 characters in length or include the <emphasis
3026 role="bold">.readonly</emphasis> or <emphasis
3027 role="bold">.backup</emphasis> extension. See <link
3028 linkend="HDRWQ44">Creating Volumes to Simplify
3029 Administration</link>. By convention, user volume names have the
3030 format <emphasis role="bold">user.</emphasis>username. Using the
3031 <emphasis role="bold">user.</emphasis> prefix not only makes it
3032 easy to identify the volume's contents, but also to create a
3033 backup version of all user volumes by issuing a single <emphasis
3034 role="bold">vos backupsys</emphasis> command.</para>
3038 <primary>mount point</primary>
3040 <secondary>choosing name for user volume</secondary>
3044 <primary>choosing</primary>
3046 <secondary>name</secondary>
3048 <tertiary>user volume mount point</tertiary>
3052 <title>Mount Point Names</title>
3054 <para>By convention, the mount point for a user's volume is named
3055 after the username. Many cells follow the convention of mounting
3056 user volumes in the <emphasis
3057 role="bold">/afs/</emphasis><replaceable>cellname</replaceable><emphasis
3058 role="bold">/usr</emphasis> directory, as discussed in <link
3059 linkend="HDRWQ43">The Third Level</link>. Very large cells
3060 sometimes find that mounting all user volumes in the same
3061 directory slows directory lookup, however; for suggested
3062 alternatives, see the following section.</para>
3066 <primary>directories</primary>
3068 <secondary>for grouping user home directories</secondary>
3072 <primary>user account</primary>
3074 <secondary>suggestions for grouping home directories</secondary>
3078 <sect2 id="HDRWQ59">
3079 <title>Grouping Home Directories</title>
3081 <para>Mounting user volumes in the <emphasis
3082 role="bold">/afs/</emphasis><replaceable>cellname</replaceable><emphasis
3083 role="bold">/usr</emphasis> directory is an AFS-appropriate
3084 variation on the standard UNIX practice of putting user home
3085 directories under the <emphasis role="bold">/usr</emphasis>
3086 subdirectory. However, cells with more than a few hundred users
3087 sometimes find that mounting all user volumes in a single directory
3088 results in slow directory lookup. The solution is to distribute user
3089 volume mount points into several directories; there are a number of
3090 alternative methods to accomplish this.
3093 <para>Distribute user home directories into multiple
3094 directories that reflect organizational divisions, such as
3095 academic or corporate departments. For example, a company can
3096 create group directories called <emphasis
3097 role="bold">usr/marketing</emphasis>, <emphasis
3098 role="bold">usr/research</emphasis>, <emphasis
3099 role="bold">usr/finance</emphasis>. A good feature of this
3100 scheme is that knowing a user's department is enough to find
3101 the user's home directory. Also, it makes it easy to set the
3102 ACL to limit access to members of the department only. A
3103 potential drawback arises if departments are of sufficiently
3104 unequal size that users in large departments experience slower
3105 lookup than users in small departments. This scheme is also
3106 not appropriate in cells where users frequently change between
3111 <para>Distribute home directories into alphabetic
3112 subdirectories of the <emphasis role="bold">usr</emphasis>
3113 directory (the <emphasis role="bold">usr/a</emphasis>
3114 subdirectory, the <emphasis role="bold">usr/b</emphasis>
3115 subdirectory, and so on), based on the first letter of the
3116 username. If the cell is very large, create subdirectories
3117 under each letter that correspond to the second letter in the
3118 user name. This scheme has the same advantages and
3119 disadvantages of a department-based scheme. Anyone who knows
3120 the user's username can find the user's home directory, but
3121 users with names that begin with popular letters sometimes
3122 experience slower lookup.</para>
3126 <para>Distribute home directories randomly but evenly into
3127 more than one grouping directory. One cell that uses this
3128 scheme has over twenty such directories called the <emphasis
3129 role="bold">usr1</emphasis> directory, the <emphasis
3130 role="bold">usr2</emphasis> directory, and so on. This scheme
3131 is especially appropriate in cells where the other two schemes
3132 do not seem feasible. It eliminates the potential problem of
3133 differences in lookup speed, because all directories are about
3134 the same size. Its disadvantage is that there is no way to
3135 guess which directory a given user's volume is mounted in, but
3136 a solution is to create a symbolic link in the regular
3137 <emphasis role="bold">usr</emphasis> directory that references
3138 the actual mount point. For example, if user <emphasis
3139 role="bold">smith</emphasis>'s volume is mounted at the
3140 <emphasis role="bold">/afs/bigcell.com/usr17/smith</emphasis>
3141 directory, then the <emphasis
3142 role="bold">/afs/bigcell.com/usr/smith</emphasis> directory is
3143 a symbolic link to the <emphasis
3144 role="bold">../usr17/smith</emphasis> directory. This way, if
3145 someone does not know which directory the user <emphasis
3146 role="bold">smith</emphasis> is in, he or she can access it
3147 through the link called <emphasis
3148 role="bold">usr/smith</emphasis>; people who do know the
3149 appropriate directory save lookup time by specifying
3155 <para>For instructions on how to implement the various schemes when
3156 using the <emphasis role="bold">uss</emphasis> program to create
3157 user accounts, see <link linkend="HDRWQ472">Evenly Distributing User
3158 Home Directories with the G Instruction</link> and <link
3159 linkend="HDRWQ473">Creating a Volume with the V
3160 Instruction</link>.</para>
3163 <sect2 id="Header_72">
3164 <title>Making a Backup Version of User Volumes Available</title>
3166 <para>Mounting the backup version of a user's volume is a simple way
3167 to enable users themselves to restore data they have accidentally
3168 removed or deleted. It is conventional to mount the backup version
3169 at a subdirectory of the user's home directory (called perhaps the
3170 <emphasis role="bold">OldFiles</emphasis> subdirectory), but other
3171 schemes are possible. Once per day you create a new backup version
3172 to capture the changes made that day, overwriting the previous day's
3173 backup version with the new one. Users can always retrieve the
3174 previous day's copy of a file without your assistance, freeing you
3175 to deal with more pressing tasks.</para>
3177 <para>Users sometimes want to delete the mount point to their backup
3178 volume, because they erroneously believe that the backup volume's
3179 contents count against their quota. Remind them that the backup
3180 volume is separate, so the only space it uses in the user volume is
3181 the amount needed for the mount point.</para>
3183 <para>For further discussion of backup volumes, see <link
3184 linkend="HDRWQ77">Backing Up AFS Data</link> and <link
3185 linkend="HDRWQ201">Creating Backup Volumes</link>.</para>
3188 <primary>file</primary>
3190 <secondary>creating standard ones in new user account</secondary>
3194 <primary>user account</primary>
3196 <secondary>creating</secondary>
3198 <tertiary>standard files in</tertiary>
3202 <primary>creating</primary>
3204 <secondary>standard files in new user account</secondary>
3208 <sect2 id="HDRWQ60">
3209 <title>Creating Standard Files in New AFS Accounts</title>
3211 <para>From your experience as a UNIX administrator, you are probably
3212 familiar with the use of login and shell initialization files (such
3213 as the <emphasis role="bold">.login</emphasis> and <emphasis
3214 role="bold">.cshrc</emphasis> files) to make an account easier to
3217 <para>It is often practical to add some AFS-specific directories to
3218 the definition of the user's <envar>PATH</envar> environment
3219 variable, including the following:
3222 <para>The path to a <emphasis role="bold">bin</emphasis>
3223 subdirectory in the user's home directory for binaries the
3224 user has created (that is, <emphasis
3225 role="bold">/afs/</emphasis><replaceable>cellname</replaceable><emphasis
3226 role="bold">/usr/</emphasis><replaceable>username</replaceable><emphasis
3227 role="bold">/bin</emphasis>)</para>
3231 <para>The <emphasis role="bold">/usr/afsws/bin</emphasis>
3232 path, which conventionally includes programs like <emphasis
3233 role="bold">fs</emphasis>, <emphasis
3234 role="bold">klog</emphasis>, <emphasis
3235 role="bold">kpasswd</emphasis>, <emphasis
3236 role="bold">pts</emphasis>, <emphasis
3237 role="bold">tokens</emphasis>, and <emphasis
3238 role="bold">unlog</emphasis></para>
3242 <para>The <emphasis role="bold">/usr/afsws/etc</emphasis>
3243 path, if the user is an administrator; it usually houses the
3244 AFS command suites that require privilege (the <emphasis
3245 role="bold">backup</emphasis>, <emphasis
3246 role="bold">butc</emphasis>, <emphasis
3247 role="bold">kas</emphasis>, <emphasis
3248 role="bold">uss</emphasis>, <emphasis
3249 role="bold">vos</emphasis> commands), the <emphasis
3250 role="bold">package</emphasis> program, and others</para>
3255 <para>If you are not using an AFS-modified login utility, it can be
3256 helpful to users to invoke the <emphasis role="bold">klog</emphasis>
3257 command in their <emphasis role="bold">.login</emphasis> file so
3258 that they obtain AFS tokens as part of logging in. In the following
3259 example command sequence, the first line echoes the string
3260 <computeroutput>klog</computeroutput> to the standard output stream,
3261 so that the user understands the purpose of the
3262 <computeroutput>Password:</computeroutput> prompt that appears when
3263 the second line is executed. The <emphasis
3264 role="bold">-setpag</emphasis> flag associates the new tokens with a
3265 process authentication group (PAG), which is discussed further in
3266 <link linkend="HDRWQ64">Identifying AFS Tokens by PAG</link>.</para>
3269 echo -n "klog " klog -setpag
3272 <para>The following sequence of commands has a similar effect,
3273 except that the <emphasis role="bold">pagsh</emphasis> command forks
3274 a new shell with which the PAG and tokens are associated.</para>
3277 pagsh echo -n "klog " klog
3280 <para>If you use an AFS-modified login utility, this sequence is not
3281 necessary, because such utilities both log a user in locally and
3282 obtain AFS tokens.</para>
3285 <primary>group</primary>
3287 <secondary>AFS GID</secondary>
3291 <primary>group</primary>
3293 <secondary>restrictions</secondary>
3297 <primary>group</primary>
3299 <secondary>privacy flags</secondary>
3303 <primary>privacy flags on Protection Database entry</primary>
3308 <sect1 id="HDRWQ61">
3309 <title>Using AFS Protection Groups</title>
3311 <para>AFS enables users to define their own groups of other users or
3312 machines. The groups are placed on ACLs to grant the same permissions
3313 to many users without listing each user individually. For group
3314 creation instructions, see <link linkend="HDRWQ531">Administering the
3315 Protection Database</link>.</para>
3317 <para>Groups have AFS ID numbers, just as users do, but an AFS group
3318 ID (GID) is a negative integer whereas a user's AFS UID is a positive
3319 integer. By default, the Protection Server allocates a new group's AFS
3320 GID automatically, but members of the <emphasis
3321 role="bold">system:administrators</emphasis> group can assign a GID
3322 when issuing the <emphasis role="bold">pts creategroup</emphasis>
3323 command. Before explicitly assigning a GID, it is best to verify that
3324 it is not already in use.</para>
3326 <para>A group cannot belong to another group, but it can own another
3327 group or even itself as long as it (the owning group) has at least one
3328 member. The current owner of a group can transfer ownership of the
3329 group to another user or group, even without the new owner's
3330 permission. At that point the former owner loses administrative
3331 control over the group.</para>
3333 <para>By default, each user can create 20 groups. A system
3334 administrator can increase or decrease this group creation quota with
3335 the <emphasis role="bold">pts setfields</emphasis> command.</para>
3337 <para>Each Protection Database entry (group or user) is protected by a
3338 set of five privacy flagswhich limit who can administer the entry and
3339 what they can do. The default privacy flags are fairly restrictive,
3340 especially for user entries. See <link linkend="HDRWQ559">Setting the
3341 Privacy Flags on Database Entries</link>.</para>
3344 <primary>system:administrators group</primary>
3346 <secondary>about</secondary>
3350 <primary>system:anyuser group</primary>
3352 <secondary>about</secondary>
3356 <primary>system:authuser group</primary>
3358 <secondary>about</secondary>
3362 <primary>group</primary>
3364 <secondary>system-defined</secondary>
3367 <sect2 id="Header_75">
3368 <title>The Three System Groups</title>
3370 <para>As the Protection Server initializes for the first time on a
3371 cell's first database server machine, it automatically creates three
3372 group entries: the <emphasis role="bold">system:anyuser</emphasis>,
3373 <emphasis role="bold">system:authuser</emphasis>, and <emphasis
3374 role="bold">system:administrators</emphasis> groups.</para>
3377 <primary>AFS UID</primary>
3379 <secondary>reserved</secondary>
3381 <tertiary>system-defined groups</tertiary>
3384 <para>The first two system groups are unlike any other groups in the
3385 Protection Database in that they do not have a stable membership:
3388 <para>The <emphasis role="bold">system:anyuser</emphasis>
3389 group includes everyone who can access a cell's AFS filespace:
3390 users who have tokens for the local cell, users who have
3391 logged in on a local AFS client machine but not obtained
3392 tokens (such as the local superuser <emphasis
3393 role="bold">root</emphasis>), and users who have connected to
3394 a local machine from outside the cell. Placing the <emphasis
3395 role="bold">system:anyuser</emphasis> group on an ACL grants
3396 access to the widest possible range of users. It is the only
3397 way to extend access to users from foreign AFS cells that do
3398 not have local accounts.</para>
3402 <para>The <emphasis role="bold">system:authuser</emphasis>
3403 group includes everyone who has a valid token obtained from
3404 the cell's AFS authentication service.</para>
3409 <para>Because the groups do not have a stable membership, the
3410 <emphasis role="bold">pts membership</emphasis> command produces no
3411 output for them. Similarly, they do not appear in the list of groups
3412 to which a user belongs.</para>
3414 <para>The <emphasis role="bold">system:administrators</emphasis>
3415 group does have a stable membership, consisting of the cell's
3416 privileged administrators. Members of this group can issue any
3417 <emphasis role="bold">pts</emphasis> command, and are the only ones
3418 who can issue several other restricted commands (such as the
3419 <emphasis role="bold">chown</emphasis> command on AFS files). By
3420 default, they also implicitly have the <emphasis
3421 role="bold">a</emphasis> (<emphasis
3422 role="bold">administer</emphasis>) and <emphasis
3423 role="bold">l</emphasis> (<emphasis role="bold">lookup</emphasis>)
3424 permissions on every ACL in the filespace. For information about
3425 changing this default, see <link linkend="HDRWQ586">Administering
3426 the system:administrators Group</link>.</para>
3428 <para>For a discussion of how to use system groups effectively on
3429 ACLs, see <link linkend="HDRWQ571">Using Groups on
3433 <sect2 id="HDRWQ62">
3434 <title>The Two Types of User-Defined Groups</title>
3436 <para>All users can create regular groups. A regular group name has
3437 two fields separated by a colon, the first of which must indicate
3438 the group's ownership. The Protection Server refuses to create or
3439 change the name of a group if the result does not accurately
3440 indicate the ownership.</para>
3442 <para>Members of the <emphasis
3443 role="bold">system:administrators</emphasis> group can create
3444 prefix-less groups whose names do not have the first field that
3445 indicates ownership. For suggestions on using the two types of
3446 groups effectively, see <link linkend="HDRWQ545">Using Groups
3447 Effectively</link>.</para>
3450 <primary>authentication</primary>
3452 <secondary>AFS separate from UNIX</secondary>
3456 <primary>AFS</primary>
3458 <secondary>authentication separate from UNIX</secondary>
3463 <sect1 id="HDRWQ63">
3464 <title>Login and Authentication in AFS</title>
3466 <para>As explained in <link linkend="HDRWQ31">Differences in
3467 Authentication</link>, AFS authentication is separate from UNIX
3468 authentication because the two file systems are separate. The
3469 separation has two practical implications:
3472 <para>To access AFS files, users must both log into the local
3473 file system and authenticate with the AFS authentication
3474 service. (Logging into the local file system is necessary
3475 because the only way to access the AFS filespace is through a
3476 Cache Manager, which resides in the local machine's
3481 <para>Passwords are stored in two separate places: in the
3482 Kerberos Database for AFS and in the each machine's local
3483 password file (the <emphasis role="bold">/etc/passwd</emphasis>
3484 file or equivalent) for the local file system.</para>
3489 <para>When a user successfully authenticates, the AFS authentication
3490 service passes a token to the user's Cache Manager. The token is a
3491 small collection of data that certifies that the user has correctly
3492 provided the password associated with a particular AFS identity. The
3493 Cache Manager presents the token to AFS server processes along with
3494 service requests, as proof that the user is genuine. To learn about
3495 the mutual authentication procedure they use to establish identity,
3496 see <link linkend="HDRWQ75">A More Detailed Look at Mutual
3497 Authentication</link>.</para>
3499 <para>The Cache Manager stores tokens in the user's credential
3500 structure in kernel memory. To distinguish one user's credential
3501 structure from another's, the Cache Manager identifies each one either
3502 by the user's UNIX UID or by a process authentication group (PAG),
3503 which is an identification number guaranteed to be unique in the
3504 cell. For further discussion, see <link linkend="HDRWQ64">Identifying
3505 AFS Tokens by PAG</link>.</para>
3508 <primary>tokens</primary>
3510 <secondary>one-per-cell rule</secondary>
3513 <para>A user can have only one token per cell in each separately
3514 identified credential structure. To obtain a second token for the same
3515 cell, the user must either log into a different machine or obtain
3516 another credential structure with a different identifier than any
3517 existing credential structure, which is most easily accomplished by
3518 issuing the <emphasis role="bold">pagsh</emphasis> command (see <link
3519 linkend="HDRWQ64">Identifying AFS Tokens by PAG</link>). In a single
3520 credential structure, a user can have one token for each of many cells
3521 at the same time. As this implies, authentication status on one
3522 machine or PAG is independent of authentication status on another
3523 machine or PAG, which can be very useful to a user or system
3524 administrator.</para>
3526 <para>The AFS distribution includes library files that enable each
3527 system type's login utility to authenticate users with AFS and log
3528 them into the local file system in one step. If you do not configure
3529 an AFS-modified login utility on a client machine, its users must
3530 issue the <emphasis role="bold">klog</emphasis> command to
3531 authenticate with AFS after logging in.</para>
3534 <para>The AFS-modified libraries do not necessarily support all
3535 features available in an operating system's proprietary login
3536 utility. In some cases, it is not possible to support a utility at
3537 all. For more information about the supported utilities in each AFS
3538 version, see the OpenAFS Release Notes.</para>
3542 <primary>commands</primary>
3544 <secondary>pagsh</secondary>
3548 <primary>pagsh command</primary>
3552 <primary>commands</primary>
3554 <secondary>klog with -setpag flag</secondary>
3558 <primary>klog command</primary>
3560 <secondary>with -setpag flag</secondary>
3564 <primary>PAG</primary>
3566 <secondary>creating with klog or pagsh command</secondary>
3570 <primary>creating</primary>
3572 <secondary>PAG with klog or pagsh command</secondary>
3576 <primary>process authentication group</primary>
3578 <secondary></secondary>
3583 <sect2 id="HDRWQ64">
3584 <title>Identifying AFS Tokens by PAG</title>
3586 <para>As noted, the Cache Manager identifies user credential
3587 structures either by UNIX UID or by PAG. Using a PAG is preferable
3588 because it guaranteed to be unique: the Cache Manager allocates it
3589 based on a counter that increments with each use. In contrast,
3590 multiple users on a machine can share or assume the same UNIX UID,
3591 which creates potential security problems. The following are two
3592 common such situations:
3595 <para>The local superuser <emphasis
3596 role="bold">root</emphasis> can always assume any other user's
3597 UNIX UID simply by issuing the <emphasis
3598 role="bold">su</emphasis> command, without providing the
3599 user's password. If the credential structure is associated
3600 with the user's UNIX UID, then assuming the UID means
3601 inheriting the AFS tokens.</para>
3605 <para>Two users working on different NFS client machines can
3606 have the same UNIX UID in their respective local file
3607 systems. If they both access the same NFS/AFS Translator
3608 machine, and the Cache Manager there identifies them by their
3609 UNIX UID, they become indistinguishable. To eliminate this
3610 problem, the Cache Manager on a translator machine
3611 automatically generates a PAG for each user and uses it,
3612 rather than the UNIX UID, to tell users apart.</para>
3617 <para>Yet another advantage of PAGs over UIDs is that processes
3618 spawned by the user inherit the PAG and so share the token; thus
3619 they gain access to AFS as the authenticated user. In many
3620 environments, for example, printer and other daemons run under
3621 identities (such as the local superuser <emphasis
3622 role="bold">root</emphasis>) that the AFS server processes recognize
3623 only as the <emphasis role="bold">anonymous</emphasis> user. Unless
3624 PAGs are used, such daemons cannot access files for which the
3625 <emphasis role="bold">system:anyuser</emphasis> group does not have
3626 the necessary ACL permissions.</para>
3628 <para>Once a user has a PAG, any new tokens the user obtains are
3629 associated with the PAG. The PAG expires two hours after any
3630 associated tokens expire or are discarded. If the user issues the
3631 <emphasis role="bold">klog</emphasis> command before the PAG
3632 expires, the new token is associated with the existing PAG (the PAG
3633 is said to be recycled in this case).</para>
3635 <para>AFS-modified login utilities automatically generate a PAG, as
3636 described in the following section. If you use a standard login
3637 utility, your users must issue the <emphasis
3638 role="bold">pagsh</emphasis> command before the <emphasis
3639 role="bold">klog</emphasis> command, or include the latter command's
3640 <emphasis role="bold">-setpag</emphasis> flag. For instructions, see
3641 <link linkend="HDRWQ69">Using Two-Step Login and
3642 Authentication</link>.</para>
3644 <para>Users can also use either command at any time to create a new
3645 PAG. The difference between the two commands is that the <emphasis
3646 role="bold">klog</emphasis> command replaces the PAG associated with
3647 the current command shell and tokens. The <emphasis
3648 role="bold">pagsh</emphasis> command initializes a new command shell
3649 before creating a new PAG. If the user already had a PAG, any
3650 running processes or jobs continue to use the tokens associated with
3651 the old PAG whereas any new jobs or processes use the new PAG and
3652 its associated tokens. When you exit the new shell (by pressing
3653 <<emphasis role="bold">Ctrl-d</emphasis>>, for example), you
3654 return to the original PAG and shell. By default, the <emphasis
3655 role="bold">pagsh</emphasis> command initializes a Bourne shell, but
3656 you can include the <emphasis role="bold">-c</emphasis> argument to
3657 initialize a C shell (the <emphasis role="bold">/bin/csh</emphasis>
3658 program on many system types) or Korn shell (the <emphasis
3659 role="bold">/bin/ksh</emphasis> program) instead.</para>
3662 <primary>login utility</primary>
3664 <secondary>AFS version</secondary>
3668 <sect2 id="HDRWQ65">
3669 <title>Using an AFS-modified login Utility</title>
3671 <para>As previously mentioned, an AFS-modified login utility
3672 simultaneously obtains an AFS token and logs the user into the local
3673 file system. This section outlines the login and authentication
3674 process and its interaction with the value in the password field of
3675 the local password file.</para>
3677 <para>An AFS-modified login utility performs a sequence of steps
3678 similar to the following; details can vary for different operating
3682 <para>It checks the user's entry in the local password file
3683 (the <emphasis role="bold">/etc/passwd</emphasis> file or
3688 <para>If no entry exists, or if an asterisk
3689 (<computeroutput>*</computeroutput>) appears in the entry's
3690 password field, the login attempt fails. If the entry exists,
3691 the attempt proceeds to the next step.</para>
3695 <para><anchor id="LIWQ66" />The utility obtains a PAG.</para>
3699 <para><anchor id="LIWQ67" />The utility converts the password
3700 provided by the user into an encryption key and encrypts a
3701 packet of data with the key. It sends the packet to the AFS
3702 authentication service (the AFS Authentication Server in the
3703 conventional configuration).</para>
3707 <para>The authentication service decrypts the packet and,
3708 depending on the success of the decryption, judges the
3709 password to be correct or incorrect. (For more details, see
3710 <link linkend="HDRWQ75">A More Detailed Look at Mutual
3711 Authentication</link>.)
3714 <para>If the authentication service judges the password
3715 incorrect, the user does not receive an AFS token. The
3716 PAG is retained, ready to be associated with any tokens
3717 obtained later. The attempt proceeds to Step <link
3718 linkend="LIWQ68">6</link>.</para>
3722 <para>If the authentication service judges the password
3723 correct, it issues a token to the user as proof of AFS
3724 authentication. The login utility logs the user into the
3725 local UNIX file system. Some login utilities echo the
3726 following banner to the screen to alert the user to
3727 authentication with AFS. Step <link
3728 linkend="LIWQ68">6</link> is skipped.
3730 AFS(R) version Login
3739 <para><anchor id="LIWQ68" />If no AFS token was granted in
3740 Step <link linkend="LIWQ67">4</link>, the login utility
3741 attempts to log the user into the local file system, by
3742 comparing the password provided to the local password file.
3745 <para>If the password is incorrect or any value other
3746 than an encrypted 13-character string appears in the
3747 password field, the login attempt fails.</para>
3751 <para>If the password is correct, the user is logged
3752 into the local file system only.</para> </listitem>
3760 <primary>local password file</primary>
3762 <secondary>when using AFS--modified login utility</secondary>
3766 <primary>login utility</primary>
3768 <secondary>AFS version's interaction with local password
3773 <primary>password</primary>
3775 <secondary>local password file</secondary>
3778 <para>As indicated, when you use an AFS-modified login utility, the
3779 password field in the local password file is no longer the primary
3780 gate for access to your system. If the user provides the correct AFS
3781 password, then the program never consults the local password
3782 file. However, you can still use the password field to control
3783 access, in the following way:
3786 <para>To prevent both local login and AFS authentication,
3787 place an asterisk (<emphasis role="bold">*</emphasis>) in the
3788 field. This is useful mainly in emergencies, when you want to
3789 prevent a certain user from logging into the machine.</para>
3793 <para>To prevent login to the local file system if the user
3794 does not provide the correct AFS password, place a character
3795 string of any length other than the standard thirteen
3796 characters in the field. This is appropriate if you want to
3797 permit only people with local AFS accounts to login on your
3798 machines. A single <emphasis role="bold">X</emphasis> or other
3799 character is the most easily recognizable way to do
3804 <para>To enable a user to log into the local file system even
3805 after providing an incorrect AFS password, record a standard
3806 UNIX encrypted password in the field by issuing the standard
3807 UNIX password-setting command (<emphasis
3808 role="bold">passwd</emphasis> or equivalent).</para>
3813 <para>Systems that use a Pluggable Authentication Module (PAM) for
3814 login and AFS authentication do not necessarily consult the local
3815 password file at all, in which case they do not use the password
3816 field to control authentication and login attempts. Instead,
3817 instructions in the PAM configuration file (on many system types,
3818 <emphasis role="bold">/etc/pam.conf</emphasis>) fill the same
3819 function. See the instructions in the OpenAFS Quick Beginnings for
3820 installing AFS-modified login utilities.</para>
3823 <primary>local password file</primary>
3825 <secondary>when not using AFS-modified login utility</secondary>
3829 <sect2 id="HDRWQ69">
3830 <title>Using Two-Step Login and Authentication</title>
3832 <para>In cells that do not use an AFS-modified login utility, users
3833 must issue separate commands to login and authenticate, as detailed
3834 in the OpenAFS User Guide:
3837 <para>They use the standard <emphasis
3838 role="bold">login</emphasis> program to login to the local
3839 file system, providing the password listed in the local
3840 password file (the <emphasis
3841 role="bold">/etc/passwd</emphasis> file or equivalent).</para>
3845 <para>They must issue the <emphasis
3846 role="bold">klog</emphasis> command to authenticate with the
3847 AFS authentication service, including its <emphasis
3848 role="bold">-setpag</emphasis> flag to associate the new
3849 tokens with a process authentication group (PAG).</para>
3854 <para>As mentioned in <link linkend="HDRWQ60">Creating Standard
3855 Files in New AFS Accounts</link>, you can invoke the <emphasis
3856 role="bold">klog -setpag</emphasis> command in a user's <emphasis
3857 role="bold">.login</emphasis> file (or equivalent) so that the user
3858 does not have to remember to issue the command after logging in. The
3859 user still must type a password twice, once at the prompt generated
3860 by the login utility and once at the <emphasis
3861 role="bold">klog</emphasis> command's prompt. This implies that the
3862 two passwords can differ, but it is less confusing if they do
3865 <para>Another effect of not using an AFS-modified login utility is
3866 that the AFS servers recognize the standard <emphasis
3867 role="bold">login</emphasis> program as the <emphasis
3868 role="bold">anonymous</emphasis> user. If the <emphasis
3869 role="bold">login</emphasis> program needs to access any AFS files
3870 (such as the <emphasis role="bold">.login</emphasis> file in a
3871 user's home directory), then the ACL that protects the file must
3872 include an entry granting the <emphasis role="bold">l</emphasis>
3873 (<emphasis role="bold">lookup</emphasis>) and <emphasis
3874 role="bold">r</emphasis> (<emphasis role="bold">read</emphasis>)
3875 permissions to the <emphasis role="bold">system:anyuser</emphasis>
3878 <para>When you do not use an AFS-modified login utility, an actual
3879 (scrambled) password must appear in the local password file for each
3880 user. Use the <emphasis role="bold">/bin/passwd</emphasis> file to
3881 insert or change these passwords. It is simpler if the password in
3882 the local password file matches the AFS password, but it is not
3886 <primary>tokens</primary>
3888 <secondary>displaying for user</secondary>
3892 <primary>tokens</primary>
3894 <secondary>command</secondary>
3898 <primary>commands</primary>
3900 <secondary>tokens</secondary>
3904 <primary>listing</primary>
3906 <secondary>tokens held by issuer</secondary>
3910 <primary>commands</primary>
3912 <secondary>klog</secondary>
3916 <primary>klog command</primary>
3920 <primary>server process</primary>
3922 <secondary>creating ticket (tokens) for</secondary>
3926 <primary>tickets</primary>
3928 <secondary></secondary>
3934 <primary>tokens</primary>
3936 <secondary>creating for server process</secondary>
3940 <primary>authenticated identity</primary>
3942 <secondary>acquiring with klog command</secondary>
3946 <primary>unlog command</primary>
3950 <primary>commands</primary>
3952 <secondary>unlog</secondary>
3956 <primary>discarding</primary>
3958 <secondary>tokens</secondary>
3962 <primary>tokens</primary>
3964 <secondary>discarding with unlog command</secondary>
3968 <sect2 id="Header_81">
3969 <title>Obtaining, Displaying, and Discarding Tokens</title>
3971 <para>Once logged in, a user can obtain a token at any time with the
3972 <emphasis role="bold">klog</emphasis> command. If a valid token
3973 already exists, the new one overwrites it. If a PAG already exists,
3974 the new token is associated with it.</para>
3976 <para>By default, the <emphasis role="bold">klog</emphasis> command
3977 authenticates the issuer using the identity currently logged in to
3978 the local file system. To authenticate as a different identity, use
3979 the <emphasis role="bold">-principal</emphasis> argument. To obtain
3980 a token for a foreign cell, use the <emphasis
3981 role="bold">-cell</emphasis> argument (it can be combined with the
3982 <emphasis role="bold">-principal</emphasis> argument). See the
3983 OpenAFS User Guide and the entry for the <emphasis
3984 role="bold">klog</emphasis> command in the OpenAFS Administration
3987 <para>To discard either all tokens or the token for a particular
3988 cell, issue the <emphasis role="bold">unlog</emphasis> command. The
3989 command affects only the tokens associated with the current command
3990 shell. See the OpenAFS User Guideand the entry for the <emphasis
3991 role="bold">unlog</emphasis> command in the OpenAFS Administration
3994 <para>To display the tokens associated with the current command
3995 shell, issue the <emphasis role="bold">tokens</emphasis>
3996 command. The following examples illustrate its output in various
3999 <para>If the issuer is not authenticated in any cell:</para>
4002 % <emphasis role="bold">tokens</emphasis>
4003 Tokens held by the Cache Manager:
4007 <para>The following shows the output for a user with AFS UID 1000 in
4008 the ABC Corporation cell:</para>
4011 % <emphasis role="bold">tokens</emphasis>
4012 Tokens held by the Cache Manager:
4013 User's (AFS ID 1000) tokens for afs@abc.com [Expires Jun 2 10:00]
4017 <para>The following shows the output for a user who is authenticated
4018 in ABC Corporation cell, the State University cell and the DEF
4019 Company cell. The user has different AFS UIDs in the three
4020 cells. Tokens for the last cell are expired:</para>
4023 % <emphasis role="bold">tokens</emphasis>
4024 Tokens held by the Cache Manager:
4025 User's (AFS ID 1000) tokens for afs@abc.com [Expires Jun 2 10:00]
4026 User's (AFS ID 4286) tokens for afs@stateu.edu [Expires Jun 3 1:34]
4027 User's (AFS ID 22) tokens for afs@def.com [>>Expired<<]
4031 <para>The Kerberos version of the <emphasis
4032 role="bold">tokens</emphasis> command (the <emphasis
4033 role="bold">tokens.krb</emphasis> command), also reports information
4034 on the ticket-granting ticket, including the ticket's owner, the
4035 ticket-granting service, and the expiration date, as in the
4036 following example. Also see <link linkend="HDRWQ70">Support for
4037 Kerberos Authentication</link>.</para>
4040 % <emphasis role="bold">tokens.krb</emphasis>
4041 Tokens held by the Cache Manager:
4042 User's (AFS ID 1000) tokens for afs@abc.com [Expires Jun 2 10:00]
4043 User smith's tokens for krbtgt.ABC.COM@abc.com [Expires Jun 2 10:00]
4048 <sect2 id="Header_82">
4049 <title>Setting Default Token Lifetimes for Users</title>
4052 <primary>tokens</primary>
4054 <secondary>setting default lifetimes for users</secondary>
4057 <para>The maximum lifetime of a user token is the smallest of the
4058 ticket lifetimes recorded in the following three Authentication
4059 Database entries. The <emphasis role="bold">kas examine</emphasis>
4060 command reports the lifetime as <computeroutput>Max ticket
4061 lifetime</computeroutput>. Administrators who have the
4062 <computeroutput>ADMIN</computeroutput> flag on their Authentication
4063 Database entry can use the <emphasis
4064 role="bold">-lifetime</emphasis> argument to the <emphasis
4065 role="bold">kas setfields</emphasis> command to set an entry's
4069 <para>The <emphasis role="bold">afs</emphasis> entry, which
4070 corresponds to the AFS server processes. The default is 100
4075 <para>The <emphasis role="bold">krbtgt</emphasis>.cellname
4076 entry, which corresponds to the ticket-granting ticket used
4077 internally in generating the token. The default is 720 hours
4082 <para>The entry for the user of the AFS-modified login utility
4083 or issuer of the <emphasis role="bold">klog</emphasis>
4084 command. The default is 25 hours for user entries created
4085 using the AFS 3.1 or later version of the Authentication
4086 Server, and 100 hours for user entries created using the AFS
4087 3.0 version of the Authentication Server. A user can use the
4088 <emphasis role="bold">kas examine</emphasis> command to
4089 display his or her own Authentication Database entry.</para>
4095 <para>An AFS-modified login utility always grants a token with a
4096 lifetime calculated from the previously described three
4097 values. When issuing the <emphasis role="bold">klog</emphasis>
4098 command, a user can request a lifetime shorter than the default by
4099 using the <emphasis role="bold">-lifetime</emphasis> argument. For
4100 further information, see the OpenAFS User Guide and the <emphasis
4101 role="bold">klog</emphasis> reference page in the OpenAFS
4102 Administration Reference.</para>
4106 <sect2 id="Header_83">
4107 <title>Changing Passwords</title>
4110 <primary>password</primary>
4112 <secondary>changing in AFS</secondary>
4116 <primary>kpasswd command</primary>
4120 <primary>commands</primary>
4122 <secondary>kpasswd</secondary>
4126 <primary>kas commands</primary>
4128 <secondary>setpassword</secondary>
4132 <primary>commands</primary>
4134 <secondary>kas setpassword</secondary>
4137 <para>Regular AFS users can change their own passwords by using
4138 either the <emphasis role="bold">kpasswd</emphasis> or <emphasis
4139 role="bold">kas setpassword</emphasis> command. The commands prompt
4140 for the current password and then twice for the new password, to
4141 screen out typing errors.</para>
4143 <para>Administrators who have the
4144 <computeroutput>ADMIN</computeroutput> flag on their Authentication
4145 Database entries can change any user's password, either by using the
4146 <emphasis role="bold">kpasswd</emphasis> command (which requires
4147 knowing the current password) or the <emphasis role="bold">kas
4148 setpassword</emphasis> command.</para>
4150 <para>If your cell does not use an AFS-modified login utility,
4151 remember also to change the local password, using the operating
4152 system's password-changing command. For more instructions on
4153 changing passwords, see <link linkend="HDRWQ516">Changing AFS
4154 Passwords</link>.</para>
4157 <sect2 id="Header_84">
4158 <title>Imposing Restrictions on Passwords and Authentication
4161 <para>You can help to make your cell more secure by imposing
4162 restrictions on user passwords and authentication attempts. To
4163 impose the restrictions as you create an account, use the <emphasis
4164 role="bold">A</emphasis> instruction in the <emphasis
4165 role="bold">uss</emphasis> template file as described in <link
4166 linkend="HDRWQ478">Increasing Account Security with the A
4167 Instruction</link>. To set or change the values on an existing
4168 account, use the <emphasis role="bold">kas setfields</emphasis>
4169 command as described in <link linkend="HDRWQ515">Improving Password
4170 and Authentication Security</link>.</para>
4173 <primary>password</primary>
4175 <secondary>expiration</secondary>
4179 <primary>password</primary>
4181 <secondary>lifetime</secondary>
4185 <primary>kas commands</primary>
4187 <secondary>setfields</secondary>
4191 <primary>commands</primary>
4193 <secondary>kas setfields</secondary>
4197 <primary>Authentication Database</primary>
4199 <secondary>password lifetime, setting</secondary>
4203 <primary>password</primary>
4205 <secondary>restricting reuse</secondary>
4208 <para>By default, AFS passwords never expire. Limiting password
4209 lifetime can help improve security by decreasing the time the
4210 password is subject to cracking attempts. You can choose an lifetime
4211 from 1 to 254 days after the password was last changed. It
4212 automatically applies to each new password as it is set. When the
4213 user changes passwords, you can also insist that the new password is
4214 not similar to any of the 20 passwords previously used.</para>
4217 <primary>password</primary>
4219 <secondary>consequences of multiple failed authentication
4220 attempts</secondary>
4224 <primary>kas commands</primary>
4226 <secondary>setfields</secondary>
4230 <primary>commands</primary>
4232 <secondary>kas setfields</secondary>
4236 <primary>authentication</primary>
4238 <secondary>consequences of multiple failures</secondary>
4241 <para>Unscrupulous users can try to gain access to your AFS cell by
4242 guessing an authorized user's password. To protect against this type
4243 of attack, you can limit the number of times that a user can
4244 consecutively fail to provide the correct password. When the limit
4245 is exceeded, the authentication service refuses further
4246 authentication attempts for a specified period of time (the lockout
4247 time). To reenable authentication attempts before the lockout time
4248 expires, an administrator must issue the <emphasis role="bold">kas
4249 unlock</emphasis> command.</para>
4252 <primary>password</primary>
4254 <secondary>checking quality of</secondary>
4258 <primary>kpasswd command</primary>
4262 <primary>commands</primary>
4264 <secondary>kpasswd</secondary>
4268 <primary>kas commands</primary>
4270 <secondary>setpassword</secondary>
4274 <primary>kpwvalid program</primary>
4277 <para>In addition to settings on user's authentication accounts, you
4278 can improve security by automatically checking the quality of new
4279 user passwords. The <emphasis role="bold">kpasswd</emphasis> and
4280 <emphasis role="bold">kas setpassword</emphasis> commands pass the
4281 proposed password to a program or script called <emphasis
4282 role="bold">kpwvalid</emphasis>, if it exists. The <emphasis
4283 role="bold">kpwvalid</emphasis> performs quality checks and returns
4284 a code to indicate whether the password is acceptable. You can
4285 create your own program or modified the sample program included in
4286 the AFS distribution. See the <emphasis
4287 role="bold">kpwvalid</emphasis> reference page in the OpenAFS
4288 Administration Reference.</para>
4290 <para>There are several types of quality checks that can improve
4294 <para>The password is a minimum length</para>
4298 <para>The password is not a word</para>
4302 <para>The password contains both numbers and letters</para>
4308 <sect2 id="HDRWQ70">
4309 <title>Support for Kerberos Authentication</title>
4312 <primary>Kerberos</primary>
4314 <secondary>support for in AFS</secondary>
4318 <primary>commands</primary>
4320 <secondary>klog.krb</secondary>
4324 <primary>commands</primary>
4326 <secondary>pagsh.krb</secondary>
4330 <primary>commands</primary>
4332 <secondary>tokens.krb</secondary>
4336 <primary>klog.krb command</primary>
4340 <primary>pagsh.krb command</primary>
4344 <primary>tokens.krb command</primary>
4347 <para>If your site is using standard Kerberos authentication rather
4348 than the AFS Authentication Server, use the modified versions of the
4349 <emphasis role="bold">klog</emphasis>, <emphasis
4350 role="bold">pagsh</emphasis>, and <emphasis
4351 role="bold">tokens</emphasis> commands that support Kerberos
4352 authentication. The binaries for the modified version of these
4353 commands have the same name as the standard binaries with the
4354 addition of a <emphasis role="bold">.krb</emphasis>
4357 <para>Use either the Kerberos version or the standard command
4358 throughout the cell; do not mix the two versions. AFS Product
4359 Support can provide instructions on installing the Kerberos version
4360 of these four commands. For information on the differences between
4361 the two versions of these commands, see the OpenAFS Administration
4365 <sect1 id="HDRWQ71">
4366 <title>Security and Authorization in AFS</title>
4368 <para>AFS incorporates several features to ensure that only authorized
4369 users gain access to data. This section summarizes the most important
4370 of them and suggests methods for improving security in your
4373 <sect2 id="HDRWQ72">
4374 <title>Some Important Security Features</title>
4377 <primary>security</primary>
4379 <secondary>AFS features</secondary>
4383 <primary>AFS</primary>
4385 <secondary>security features</secondary>
4389 <title>ACLs on Directories</title>
4391 <para>Files in AFS are protected by the access control list (ACL)
4392 associated with their parent directory. The ACL defines which
4393 users or groups can access the data in the directory, and in what
4394 way. See <link linkend="HDRWQ562">Managing Access Control
4395 Lists</link>.</para>
4399 <title>Mutual Authentication Between Client and Server</title>
4401 <para>When an AFS client and server process communicate, each
4402 requires the other to prove its identity during mutual
4403 authentication, which involves the exchange of encrypted
4404 information that only valid parties can decrypt and respond
4405 to. For a detailed description of the mutual authentication
4406 process, see <link linkend="HDRWQ75">A More Detailed Look at
4407 Mutual Authentication</link>.</para>
4410 <para>AFS server processes mutually authenticate both with one
4411 another and with processes that represent human users. After mutual
4412 authentication is complete, the server and client have established
4413 an authenticated connection, across which they can communicate
4414 repeatedly without having to authenticate again until the connection
4415 expires or one of the parties closes it. Authenticated connections
4416 have varying lifetimes.</para>
4419 <title>Tokens</title>
4421 <para>In order to access AFS files, users must prove their
4422 identities to the AFS authentication service by providing the
4423 correct AFS password. If the password is correct, the
4424 authentication service sends the user a token as evidence of
4425 authenticated status. See <link linkend="HDRWQ63">Login and
4426 Authentication in AFS</link>.</para>
4429 <para>Servers assign the user identity <emphasis
4430 role="bold">anonymous</emphasis> to users and processes that do not
4431 have a valid token. The <emphasis role="bold">anonymous</emphasis>
4432 identity has only the access granted to the <emphasis
4433 role="bold">system:anyuser</emphasis> group on ACLs.</para>
4436 <title>Authorization Checking</title>
4438 <para>Mutual authentication establishes that two parties
4439 communicating with one another are actually who they claim to be.
4440 For many functions, AFS server processes also check that the
4441 client whose identity they have verified is also authorized to
4442 make the request. Different requests require different kinds of
4443 privilege. See <link linkend="HDRWQ73">Three Types of
4444 Privilege</link>.</para>
4448 <title>Encrypted Network Communications</title>
4451 <primary>network</primary>
4453 <secondary>encrypted communication in AFS</secondary>
4457 <primary>encrypted network communication</primary>
4461 <primary>security</primary>
4463 <secondary>encrypted network communication</secondary>
4466 <para>The AFS server processes encrypt particularly sensitive
4467 information before sending it back to clients. Even if an
4468 unauthorized party is able to eavesdrop on an authenticated
4469 connection, they cannot decipher encrypted data without the proper
4473 <para>The following AFS commands encrypt data because they involve
4474 server encryption keys and passwords:
4477 <para>The <emphasis role="bold">bos addkey</emphasis> command,
4478 which adds a server encryption key to the <emphasis
4479 role="bold">/usr/afs/etc/KeyFile</emphasis> file</para>
4483 <para>The <emphasis role="bold">bos listkeys</emphasis>
4484 command, which lists the server encryption keys from the
4485 <emphasis role="bold">/usr/afs/etc/KeyFile</emphasis>
4490 <para>The <emphasis role="bold">kpasswd</emphasis> command,
4491 which changes a password in the Authentication Database</para>
4495 <para>Most commands in the <emphasis
4496 role="bold">kas</emphasis> command suite</para>
4501 <para>In addition, the United States edition of the Update Server
4502 encrypts sensitive information (such as the contents of <emphasis
4503 role="bold">KeyFile</emphasis>) when distributing it. Other commands
4504 in the <emphasis role="bold">bos</emphasis> suite and the commands
4505 in the <emphasis role="bold">fs</emphasis>, <emphasis
4506 role="bold">pts</emphasis> and <emphasis role="bold">vos</emphasis>
4507 suites do not encrypt data before transmitting it.</para>
4510 <sect2 id="HDRWQ73">
4511 <title>Three Types of Privilege</title>
4513 <para>AFS uses three separate types of privilege for the reasons
4514 discussed in <link linkend="HDRWQ585">The Reason for Separate
4518 <para>Membership in the <emphasis
4519 role="bold">system:administrators</emphasis> group. Members
4520 are entitled to issue any <emphasis role="bold">pts</emphasis>
4521 command and those <emphasis role="bold">fs</emphasis> commands
4522 that set volume quota. By default, they also implicitly have
4523 the <emphasis role="bold">a</emphasis> (<emphasis
4524 role="bold">administer</emphasis>) and <emphasis
4525 role="bold">l</emphasis> (<emphasis
4526 role="bold">lookup</emphasis>) permissions on every ACL in the
4527 file tree even if the ACL does not include an entry for
4532 <para>The <computeroutput>ADMIN</computeroutput> flag on the
4533 Authentication Database entry. An administrator with this flag
4534 can issue any <emphasis role="bold">kas</emphasis>
4539 <para>Inclusion in the <emphasis
4540 role="bold">/usr/afs/etc/UserList</emphasis> file. An
4541 administrator whose username appears in this file can issue
4542 any <emphasis role="bold">bos</emphasis>, <emphasis
4543 role="bold">vos</emphasis>, or <emphasis
4544 role="bold">backup</emphasis> command (although some <emphasis
4545 role="bold">backup</emphasis> commands require additional
4546 privilege as described in <link linkend="HDRWQ260">Granting
4547 Administrative Privilege to Backup Operators</link>).</para>
4553 <sect2 id="Header_89">
4554 <title>Authorization Checking versus Authentication</title>
4556 <para>AFS distinguishes between authentication and authorization
4557 checking. Authentication refers to the process of proving
4558 identity. Authorization checking refers to the process of verifying
4559 that an authenticated identity is allowed to perform a certain
4562 <para>AFS implements authentication at the level of
4563 connections. Each time two parties establish a new connection, they
4564 mutually authenticate. In general, each issue of an AFS command
4565 establishes a new connection between AFS server process and
4568 <para>AFS implements authorization checking at the level of server
4569 machines. If authorization checking is enabled on a server machine,
4570 then all of the server processes running on it provide services only
4571 to authorized users. If authorization checking is disabled on a
4572 server machine, then all of the server processes perform any action
4573 for anyone. Obviously, disabling authorization checking is an
4574 extreme security exposure. For more information, see <link
4575 linkend="HDRWQ123">Managing Authentication and Authorization
4576 Requirements</link>.</para>
4579 <sect2 id="HDRWQ74">
4580 <title>Improving Security in Your Cell</title>
4583 <primary>security</primary>
4585 <secondary>suggestions for improving</secondary>
4588 <para>You can improve the level of security in your cell by
4589 configuring user accounts, server machines, and system administrator
4590 accounts in the indicated way.</para>
4593 <title>User Accounts</title>
4598 <para>Use an AFS-modified login utility, or include the
4599 <emphasis role="bold">-setpag</emphasis> flag to the
4600 <emphasis role="bold">klog</emphasis> command, to associate
4601 the credential structure that houses tokens with a PAG
4602 rather than a UNIX UID. This prevents users from inheriting
4603 someone else's tokens by assuming their UNIX identity. For
4604 further discussion, see <link linkend="HDRWQ64">Identifying
4605 AFS Tokens by PAG</link>.</para>
4609 <para>Encourage users to issue the <emphasis
4610 role="bold">unlog</emphasis> command to destroy their tokens
4611 before logging out. This forestalls attempts to access
4612 tokens left behind kernel memory. Consider including the
4613 <emphasis role="bold">unlog</emphasis> command in every
4614 user's <emphasis role="bold">.logout</emphasis> file or
4622 <title>Server Machines</title>
4627 <para>Disable authorization checking only in emergencies or
4628 for very brief periods of time. It is best to work at the
4629 console of the affected machine during this time, to prevent
4630 anyone else from accessing the machine through the
4635 <para>Change the AFS server encryption key on a frequent and
4636 regular schedule. Make it difficult to guess (a long string
4637 including nonalphabetic characters, for instance). Unlike
4638 user passwords, the password from which the AFS key is
4639 derived can be longer than eight characters, because it is
4640 never used during login. The <emphasis role="bold">kas
4641 setpassword</emphasis> command accepts a password hundreds
4642 of characters long. For instructions, see <link
4643 linkend="HDRWQ355">Managing Server Encryption
4648 <para>As much as possible, limit the number of people who
4649 can login at a server machine's console or remotely.
4650 Imposing this limit is an extra security precaution rather
4651 than a necessity. The machine cannot serve as an AFS client
4652 in this case.</para>
4656 <para>Particularly limit access to the local superuser
4657 <emphasis role="bold">root</emphasis> account on a server
4658 machine. The local superuser <emphasis
4659 role="bold">root</emphasis> has free access to important
4660 administrative subdirectories of the <emphasis
4661 role="bold">/usr/afs</emphasis> directory, as described in
4662 <link linkend="HDRWQ53">AFS Files on the Local
4666 <primary>root superuser</primary>
4668 <secondary>limiting logins</secondary>
4673 <para>As in any computing environment, server machines must
4674 be located in a secured area. Any other security measures
4675 are effectively worthless if unauthorized people can access
4676 the computer hardware.</para>
4683 <title>System Administrators</title>
4688 <para>Limit the number of system administrators in your
4689 cell. Limit the use of system administrator accounts on
4690 publicly accessible workstations. Such machines are not
4691 secure, so unscrupulous users can install programs that try
4692 to steal tokens or passwords. If administrators must use
4693 publicly accessible workstations at times, they must issue
4694 the <emphasis role="bold">unlog</emphasis> command before
4695 leaving the machine.</para>
4699 <para>Create an administrative account for each
4700 administrator separate from the personal account, and assign
4701 AFS privileges only to the administrative account. The
4702 administrators must authenticate to the administrative
4703 accounts to perform duties that require privilege, which
4704 provides a useful audit trail as well.</para>
4708 <para>Administrators must not leave a machine unattended
4709 while they have valid tokens. Issue the <emphasis
4710 role="bold">unlog</emphasis> command before leaving.</para>
4714 <para>Use the <emphasis role="bold">-lifetime</emphasis>
4715 argument to the <emphasis role="bold">kas
4716 setfields</emphasis> command to set the token lifetime for
4717 administrative accounts to a fairly short amount of time.
4718 The default lifetime for AFS tokens is 25 hours, but 30 or
4719 60 minutes is possibly a more reasonable lifetime for
4720 administrative tokens. The tokens for administrators who
4721 initiate AFS Backup System operations must last somewhat
4722 longer, because it can take several hours to complete some
4723 dump operations, depending on the speed of the tape device
4724 and the network connecting it to the file server machines
4725 that house the volumes is it accessing.</para>
4729 <para>Limit administrators' use of the <emphasis
4730 role="bold">telnet</emphasis> program. It sends unencrypted
4731 passwords across the network. Similarly, limit use of other
4732 remote programs such as <emphasis role="bold">rsh</emphasis>
4733 and <emphasis role="bold">rcp</emphasis>, which send
4734 unencrypted tokens across the network.</para>
4741 <sect2 id="HDRWQ75">
4742 <title>A More Detailed Look at Mutual Authentication</title>
4745 <primary>mutual authentication</primary>
4749 <primary>distributed file system</primary>
4751 <secondary>security issues</secondary>
4755 <primary>shared secret</primary>
4759 <primary>server encryption key</primary>
4761 <secondary>defined</secondary>
4764 <para>As in any file system, security is a prime concern in AFS. A
4765 file system that makes file sharing easy is not useful if it makes
4766 file sharing mandatory, so AFS incorporates several features that
4767 prevent unauthorized users from accessing data. Security in a
4768 networked environment is difficult because almost all procedures
4769 require transmission of information across wires that almost anyone
4770 can tap into. Also, many machines on networks are powerful enough
4771 that unscrupulous users can monitor transactions or even intercept
4772 transmissions and fake the identity of one of the
4773 participants.</para>
4775 <para>The most effective precaution against eavesdropping and
4776 information theft or fakery is for servers and clients to accept the
4777 claimed identity of the other party only with sufficient proof. In
4778 other words, the nature of the network forces all parties on the
4779 network to assume that the other party in a transaction is not
4780 genuine until proven so. Mutual authentication is the means through
4781 which parties prove their genuineness.</para>
4783 <para>Because the measures needed to prevent fakery must be quite
4784 sophisticated, the implementation of mutual authentication
4785 procedures is complex. The underlying concept is simple, however:
4786 parties prove their identities by demonstrating knowledge of a
4787 shared secret. A shared secret is a piece of information known only
4788 to the parties who are mutually authenticating (they can sometimes
4789 learn it in the first place from a trusted third party or some other
4790 source). The party who originates the transaction presents the
4791 shared secret and refuses to accept the other party as valid until
4792 it shows that it knows the secret too.</para>
4794 <para>The most common form of shared secret in AFS transactions is
4795 the encryption key, also referred to simply as a key. The two
4796 parties use their shared key to encrypt the packets of information
4797 they send and to decrypt the ones they receive. Encryption using
4798 keys actually serves two related purposes. First, it protects
4799 messages as they cross the network, preventing anyone who does not
4800 know the key from eavesdropping. Second, ability to encrypt and
4801 decrypt messages successfully indicates that the parties are using
4802 the key (it is their shared secret). If they are using different
4803 keys, messages remain scrambled and unintelligible after
4806 <para>The following sections describe AFS's mutual authentication
4807 procedures in more detail. Feel free to skip these sections if you
4808 are not interested in the mutual authentication process.</para>
4810 <sect3 id="Header_92">
4811 <title>Simple Mutual Authentication</title>
4813 <para>Simple mutual authentication involves only one encryption
4814 key and two parties, generally a client and server. The client
4815 contacts the server by sending a challenge message encrypted with
4816 a key known only to the two of them. The server decrypts the
4817 message using its key, which is the same as the client's if they
4818 really do share the same secret. The server responds to the
4819 challenge and uses its key to encrypt its response. The client
4820 uses its key to decrypt the server's response, and if it is
4821 correct, then the client can be sure that the server is genuine:
4822 only someone who knows the same key as the client can decrypt the
4823 challenge and answer it correctly. On its side, the server
4824 concludes that the client is genuine because the challenge message
4825 made sense when the server decrypted it.</para>
4827 <para>AFS uses simple mutual authentication to verify user
4828 identities during the first part of the login procedure. In that
4829 case, the key is based on the user's password.</para>
4832 <sect3 id="HDRWQ76">
4833 <title>Complex Mutual Authentication</title>
4835 <para>Complex mutual authentication involves three encryption keys
4836 and three parties. All secure AFS transactions (except the first
4837 part of the login process) employ complex mutual
4838 authentication.</para>
4841 <primary>ticket-granter</primary>
4845 <primary>server encryption key</primary>
4849 <primary>tokens</primary>
4851 <secondary>data in</secondary>
4854 <para>When a client wishes to communicate with a server, it first
4855 contacts a third party called a ticket-granter. The ticket-granter
4856 and the client mutually authenticate using the simple
4857 procedure. When they finish, the ticket-granter gives the client a
4858 server ticket (or simply ticket) as proof that it (the
4859 ticket-granter) has preverified the identity of the client. The
4860 ticket-granter encrypts the ticket with the first of the three
4861 keys, called the server encryption key because it is known only to
4862 the ticket-granter and the server the client wants to contact. The
4863 client does not know this key.</para>
4865 <para>The ticket-granter sends several other pieces of information
4866 along with the ticket. They enable the client to use the ticket
4867 effectively despite being unable to decrypt the ticket
4868 itself. Along with the ticket, the items constitute a token:
4871 <para>A session key, which is the second encryption key
4872 involved in mutual authentication. The ticket-granter
4873 invents the session key at random as the shared secret
4874 between client and server. For reasons explained further
4875 below, the ticket-granter also puts a copy of the session
4876 key inside the ticket. The client and server use the session
4877 key to encrypt messages they send to one another during
4878 their transactions. The ticket-granter invents a different
4879 session key for each connection between a client and a
4880 server (there can be several transactions during a single
4884 <primary>session key</primary>
4889 <para>The name of the server for which the ticket is valid
4890 (and so which server encryption key encrypts the ticket
4895 <para>A ticket lifetime indicator. The default lifetime of
4896 AFS server tickets is 100 hours. If the client wants to
4897 contact the server again after the ticket expires, it must
4898 contact the ticket-granter to get a new ticket.</para>
4903 <para>The ticket-granter seals the entire token with the third key
4904 involved in complex mutual authentication--the key known only to
4905 it (the ticket-granter) and the client. In some cases, this third
4906 key is derived from the password of the human user whom the client
4909 <para>Now that the client has a valid server ticket, it is ready
4910 to contact the server. It sends the server two things:
4913 <para>The server ticket. This is encrypted with the server
4914 encryption key.</para>
4918 <para>Its request message, encrypted with the session
4919 key. Encrypting the message protects it as it crosses the
4920 network, since only the server/client pair for whom the
4921 ticket-granter invented the session key know it.</para>
4926 <para>At this point, the server does not know the session key,
4927 because the ticket-granter just created it. However, the
4928 ticket-granter put a copy of the session key inside the
4929 ticket. The server uses the server encryption key to decrypts the
4930 ticket and learns the session key. It then uses the session key to
4931 decrypt the client's request message. It generates a response and
4932 sends it to the client. It encrypts the response with the session
4933 key to protect it as it crosses the network.</para>
4935 <para>This step is the heart of mutual authentication between
4936 client and server, because it proves to both parties that they
4937 know the same secret:
4940 <para>The server concludes that the client is authorized to
4941 make a request because the request message makes sense when
4942 the server decrypts it using the session key. If the client
4943 uses a different session key than the one the server finds
4944 inside the ticket, then the request message remains
4945 unintelligible even after decryption. The two copies of the
4946 session key (the one inside the ticket and the one the
4947 client used) can only be the same if they both came from the
4948 ticket-granter. The client cannot fake knowledge of the
4949 session key because it cannot look inside the ticket, sealed
4950 as it is with the server encryption key known only to the
4951 server and the ticket-granter. The server trusts the
4952 ticket-granter to give tokens only to clients with whom it
4953 (the ticket-granter) has authenticated, so the server
4954 decides the client is legitimate.</para>
4956 <para>(Note that there is no direct communication between
4957 the ticket-granter and the server, even though their
4958 relationship is central to ticket-based mutual
4959 authentication. They interact only indirectly, via the
4960 client's possession of a ticket sealed with their shared
4965 <para>The client concludes that the server is genuine and
4966 trusts the response it gets back from the server, because
4967 the response makes sense after the client decrypts it using
4968 the session key. This indicates that the server encrypted
4969 the response with the same session key as the client
4970 knows. The only way for the server to learn that matching
4971 session key is to decrypt the ticket first. The server can
4972 only decrypt the ticket because it shares the secret of the
4973 server encryption key with the ticket-granter. The client
4974 trusts the ticket-granter to give out tickets only for
4975 legitimate servers, so the client accepts a server that can
4976 decrypt the ticket as genuine, and accepts its
4985 <sect1 id="HDRWQ77">
4986 <title>Backing Up AFS Data</title>
4988 <para>AFS provides two related facilities that help the administrator
4989 back up AFS data: backup volumes and the AFS Backup System.</para>
4991 <sect2 id="Header_95">
4992 <title>Backup Volumes</title>
4994 <para>The first facility is the backup volume, which you create by
4995 cloning a read/write volume. The backup volume is read-only and so
4996 preserves the state of the read/write volume at the time the clone
4999 <para>Backup volumes can ease administration if you mount them in
5000 the file system and make their contents available to users. For
5001 example, it often makes sense to mount the backup version of each
5002 user volume as a subdirectory of the user's home directory. A
5003 conventional name for this mount point is <emphasis
5004 role="bold">OldFiles</emphasis>. Create a new version of the backup
5005 volume (that is, reclone the read/write) once a day to capture any
5006 changes that were made since the previous backup. If a user
5007 accidentally removes or changes data, the user can restore it from
5008 the backup volume, rather than having to ask you to restore
5011 <para>The OpenAFS User Guide does not mention backup volumes, so
5012 regular users do not know about them if you decide not to use
5013 them. This implies that if you <emphasis role="bold">do</emphasis>
5014 make backup versions of user volumes, you need to tell your users
5015 about how the backup works and where you have mounted it.</para>
5017 <para>Users are often concerned that the data in a backup volume
5018 counts against their volume quota and some of them even want to
5019 remove the <emphasis role="bold">OldFiles</emphasis> mount point. It
5020 does not, because the backup volume is a separate volume. The only
5021 amount of space it uses in the user's volume is the amount needed
5022 for the mount point, which is about the same as the amount needed
5023 for a standard directory element.</para>
5025 <para>Backup volumes are discussed in detail in <link
5026 linkend="HDRWQ201">Creating Backup Volumes</link>.</para>
5029 <sect2 id="Header_96">
5030 <title>The AFS Backup System</title>
5032 <para>Backup volumes can reduce restoration requests, but they
5033 reside on disk and so do not protect data from loss due to hardware
5034 failure. Like any file system, AFS is vulnerable to this sort of
5037 <para>To protect your cell's users from permanent loss of data, you
5038 are strongly urged to back up your file system to tape on a regular
5039 and frequent schedule. The AFS Backup System is available to ease
5040 the administration and performance of backups. For detailed
5041 information about the AFS Backup System, see <link
5042 linkend="HDRWQ248">Configuring the AFS Backup System</link> and
5043 <link linkend="HDRWQ283">Backing Up and Restoring AFS
5048 <sect1 id="HDRWQ79">
5049 <title>Accessing AFS through NFS</title>
5051 <para>Users of NFS client machines can access the AFS filespace by
5052 mounting the <emphasis role="bold">/afs</emphasis> directory of an AFS
5053 client machine that is running the NFS/AFS Translator. This is a
5054 particular advantage in cells already running NFS who want to access
5055 AFS using client machines for which AFS is not available. See <link
5056 linkend="HDRWQ595">Appendix A, Managing the NFS/AFS
5057 Translator</link>.</para>