3 =item B<-auditlog> <I<log path>>
5 Turns on audit logging, and sets the path for the audit log. The audit
6 log records information about RPC calls, including the name of the RPC
7 call, the host that submitted the call, the authenticated entity (user)
8 that issued the call, the parameters for the call, and if the call
11 =item B<-audit-interface> (file | sysvmq)
13 Specifies what audit interface to use. The C<file> interface writes audit
14 messages to the file passed to B<-auditlog>. The C<sysvmq> interface
15 writes audit messages to a SYSV message (see L<msgget(2)> and
16 L<msgrcv(2)>). The message queue the C<sysvmq> interface writes to has the
17 key C<ftok(path, 1)>, where C<path> is the path specified in the
22 =item B<-d> <I<debug level>>
24 Sets the detail level for the debugging trace written to the
25 F</usr/afs/logs/FileLog> file. Provide one of the following values, each
26 of which produces an increasingly detailed trace: C<0>, C<1>, C<5>, C<25>,
27 and C<125>. The default value of C<0> produces only a few messages.
29 =item B<-p> <I<number of processes>>
31 Sets the number of threads to run. Provide a positive integer.
32 The File Server creates and uses five threads for special purposes,
33 in addition to the number specified (but if this argument specifies
34 the maximum possible number, the File Server automatically uses five
35 of the threads for its own purposes).
37 The maximum number of threads can differ in each release of OpenAFS.
38 Consult the I<OpenAFS Release Notes> for the current release.
40 =item B<-spare> <I<number of spare blocks>>
42 Specifies the number of additional kilobytes an application can store in a
43 volume after the quota is exceeded. Provide a positive integer; a value of
44 C<0> prevents the volume from ever exceeding its quota. Do not combine
45 this argument with the B<-pctspare> argument.
47 =item B<-pctspare> <I<percentage spare>>
49 Specifies the amount by which the File Server allows a volume to exceed
50 its quota, as a percentage of the quota. Provide an integer between C<0>
51 and C<99>. A value of C<0> prevents the volume from ever exceeding its
52 quota. Do not combine this argument with the B<-spare> argument.
54 =item B<-b> <I<buffers>>
56 Sets the number of directory buffers. Provide a positive integer.
58 =item B<-l> <I<large vnodes>>
60 Sets the number of large vnodes available in memory for caching directory
61 elements. Provide a positive integer.
63 =item B<-s> <I<small nodes>>
65 Sets the number of small vnodes available in memory for caching file
66 elements. Provide a positive integer.
68 =item B<-vc> <I<volume cachesize>>
70 Sets the number of volumes the File Server can cache in memory. Provide a
73 =item B<-w> <I<call back wait interval>>
75 Sets the interval at which the daemon spawned by the File Server performs
76 its maintenance tasks. Do not use this argument; changing the default
77 value can cause unpredictable behavior.
79 =item B<-cb> <I<number of callbacks>>
81 Sets the number of callbacks the File Server can track. Provide a positive
86 Prints the following banner to F</dev/console> about every 10 minutes.
88 File Server is running at I<time>.
92 Prevents the File Server from breaking the callbacks that Cache Managers
93 hold on a volume that the File Server is reattaching after the volume was
94 offline (as a result of the B<vos restore> command, for example). Use of
95 this flag is strongly discouraged.
99 This option slightly changes the error codes reported to clients when an
100 unattached volume is accessed by a client during fileserver startup.
102 Normally, non-DAFS fileservers start accepting requests immediately on startup,
103 but attachment of volumes can take a while. So if a client tries to access a
104 volume that is not attached simply because the fileserver hasn't attached it
105 yet, that client will get an error. With the B<-nobusy> option present, the
106 fileserver will immediately respond with an error code that indicates the
107 server is starting up. However, some older clients (before OpenAFS 1.0) don't
108 understand this error code, and may not function optimally. So the default
109 behavior, without the B<-nobusy> option, is to at first respond with a
110 different error code that is understood by more clients, but is
111 indistinguishable from other scenarios where the volume is busy and not
112 attached for other reasons.
114 There is usually no reason to use this option under normal operation.
116 =item B<-implicit> <I<admin mode bits>>
118 Defines the set of permissions granted by default to the
119 system:administrators group on the ACL of every directory in a volume
120 stored on the file server machine. Provide one or more of the standard
121 permission letters (C<rlidwka>) and auxiliary permission letters
122 (C<ABCDEFGH>), or one of the shorthand notations for groups of permissions
123 (C<all>, C<none>, C<read>, and C<write>). To review the meaning of the
124 permissions, see the B<fs setacl> reference page.
128 Don't allow writes to this fileserver.
130 =item B<-admin-write>
132 Allows write requests for members of system:administrators on the read-only
133 fileserver in question. The C<-admin-write> option has no effect when the
134 C<-readonly> is not specified.
136 =item B<-hr> <I<number of hours between refreshing the host cps>>
138 Specifies how often the File Server refreshes its knowledge of the
139 machines that belong to protection groups (refreshes the host CPSs for
140 machines). The File Server must update this information to enable users
141 from machines recently added to protection groups to access data for which
142 those machines now have the necessary ACL permissions.
144 =item B<-busyat> <I<< redirect clients when queue > n >>>
146 Defines the number of incoming RPCs that can be waiting for a response
147 from the File Server before the File Server returns the error code
148 C<VBUSY> to the Cache Manager that sent the latest RPC. In response, the
149 Cache Manager retransmits the RPC after a delay. This argument prevents
150 the accumulation of so many waiting RPCs that the File Server can never
151 process them all. Provide a positive integer. The default value is
154 =item B<-rxpck> <I<number of rx extra packets>>
156 Controls the number of Rx packets the File Server uses to store data for
157 incoming RPCs that it is currently handling, that are waiting for a
158 response, and for replies that are not yet complete. Provide a positive
163 Writes a trace of the File Server's operations on Rx packets to the file
164 F</usr/afs/logs/rx_dbg>.
168 Writes a trace of the File Server's operations on Rx events (such as
169 retransmissions) to the file F</usr/afs/logs/rx_dbg>.
171 =item B<-rxmaxmtu> <I<bytes>>
173 Defines the maximum size of an MTU. The value must be between the
174 minimum and maximum packet data sizes for Rx.
178 Allows the server to send and receive jumbograms. A jumbogram is
179 a large-size packet composed of 2 to 4 normal Rx data packets that share
180 the same header. The fileserver does not use jumbograms by default, as some
181 routers are not capable of properly breaking the jumbogram into smaller
182 packets and reassembling them.
186 Deprecated; jumbograms are disabled by default.
190 Force the fileserver to only bind to one IP address.
192 =item B<-allow-dotted-principals>
194 By default, the RXKAD security layer will disallow access by Kerberos
195 principals with a dot in the first component of their name. This is to avoid
196 the confusion where principals user/admin and user.admin are both mapped to the
197 user.admin PTS entry. Sites whose Kerberos realms don't have these collisions
198 between principal names may disable this check by starting the server
203 Sets values for many arguments in a manner suitable for a large file
204 server machine. Combine this flag with any option except the B<-S> flag;
205 omit both flags to set values suitable for a medium-sized file server
210 Sets values for many arguments in a manner suitable for a small file
211 server machine. Combine this flag with any option except the B<-L> flag;
212 omit both flags to set values suitable for a medium-sized file server
215 =item B<-realm> <I<Kerberos realm name>>
217 Defines the Kerberos realm name for the File Server to use. If this
218 argument is not provided, it uses the realm name corresponding to the cell
219 listed in the local F</usr/afs/etc/ThisCell> file.
221 =item B<-udpsize> <I<size of socket buffer in bytes>>
223 Sets the size of the UDP buffer, which is 64 KB by default. Provide a
224 positive integer, preferably larger than the default.
226 =item B<-sendsize> <I<size of send buffer in bytes>>
228 Sets the size of the send buffer, which is 16384 bytes by default.
230 =item B<-abortthreshold> <I<abort threshold>>
232 Sets the abort threshold, which is triggered when an AFS client sends
233 a number of FetchStatus requests in a row and all of them fail due to
234 access control or some other error. When the abort threshold is
235 reached, the file server starts to slow down the responses to the
236 problem client in order to reduce the load on the file server.
238 The throttling behaviour can cause issues especially for some versions
239 of the Windows OpenAFS client. When using Windows Explorer to navigate
240 the AFS directory tree, directories with only "look" access for the
241 current user may load more slowly because of the throttling. This is
242 because the Windows OpenAFS client sends FetchStatus calls one at a
243 time instead of in bulk like the Unix Open AFS client.
245 Setting the threshold to 0 disables the throttling behavior. This
246 option is available in OpenAFS versions 1.4.1 and later.
248 =item B<-enable_peer_stats>
250 Activates the collection of Rx statistics and allocates memory for their
251 storage. For each connection with a specific UDP port on another machine,
252 a separate record is kept for each type of RPC (FetchFile, GetStatus, and
253 so on) sent or received. To display or otherwise access the records, use
254 the Rx Monitoring API.
256 =item B<-enable_process_stats>
258 Activates the collection of Rx statistics and allocates memory for their
259 storage. A separate record is kept for each type of RPC (FetchFile,
260 GetStatus, and so on) sent or received, aggregated over all connections to
261 other machines. To display or otherwise access the records, use the Rx
264 =item B<-syslog> [<loglevel>]
266 Use syslog instead of the normal logging location for the fileserver
267 process. If provided, log messages are at <loglevel> instead of the
272 Use MR-AFS (Multi-Resident) style logging. This option is deprecated.
274 =item B<-transarc-logs>
276 Use Transarc style logging features. Rename the existing log file
277 F</usr/afs/logs/FileLog> to F</usr/afs/logs/FileLog.old> when the fileserver is
278 restarted. This option is provided for compatibility with older versions.
282 Offer the SANEACLS capability for the fileserver. This option is
283 currently unimplemented.
287 Prints the online help for this command. All other valid options are
290 =item B<-vhandle-setaside> <I<fds reserved for non-cache io>>
292 Number of file handles set aside for I/O not in the cache. Defaults to 128.
294 =item B<-vhandle-max-cachesize> <I<max open files>>
296 Maximum number of available file handles.
298 =item B<-vhandle-initial-cachesize> <I<initial open file cache>>
300 Number of file handles set aside for I/O in the cache. Defaults to 128.
302 =item B<-vattachpar> <I<number of volume attach threads>>
304 The number of threads assigned to attach and detach volumes. The default
305 is 1. Warning: many of the I/O parallelism features of Demand-Attach
306 Fileserver are turned off when the number of volume attach threads is only
309 This option is only meaningful for a file server built with pthreads
312 =item B<-m> <I<min percentage spare in partition>>
314 Specifies the percentage of each AFS server partition that the AIX version
315 of the File Server creates as a reserve. Specify an integer value between
316 C<0> and C<30>; the default is 8%. A value of C<0> means that the
317 partition can become completely full, which can have serious negative
318 consequences. This option is not supported on platforms other than AIX.
322 Prevents any portion of the fileserver binary from being paged (swapped)
323 out of memory on a file server machine running the IRIX operating system.
324 This option is not supported on platforms other than IRIX.
326 =item B<-offline-timeout> <I<timeout in seconds>>
328 Setting this option to I<N> means that if any clients are reading from a
329 volume when we want to offline that volume (for example, as part of
330 releasing a volume), we will wait I<N> seconds for the clients' request
331 to finish. If the clients' requests have not finished, we will then
332 interrupt the client requests and send an error to those clients,
333 allowing the volume to go offline.
335 If a client is interrupted, from the client's point of view, it will
336 appear as if they had accessed the volume after it had gone offline. For
337 RO volumes, this mean the client should fail-over to other valid RO
338 sites for that volume. This option may speed up volume releases if
339 volumes are being accessed by clients that have slow or unreliable
342 Setting this option to C<0> means to interrupt clients immediately if a
343 volume is waiting to go offline. Setting this option to C<-1> means to
344 wait forever for client requests to finish. The default value is C<-1>.
346 =item B<-offline-shutdown-timeout> <I<timeout in seconds>>
348 This option behaves similarly to B<-offline-timeout> but applies to
349 volumes that are going offline as part of the fileserver shutdown
350 process. If the value specified is I<N>, we will interrupt any clients
351 reading from volumes after I<N> seconds have passed since we first
352 needed to wait for a volume to offline during the shutdown process.
354 Setting this option to C<0> means to interrupt all clients reading from
355 volumes immediately during the shutdown process. Setting this option to
356 C<-1> means to wait forever for client requests to finish during the
359 If B<-offline-timeout> is specified, the default value of
360 B<-offline-shutdown-timeout> is the value specified for
361 B<-offline-timeout>. Otherwise, the default value is C<-1>.
363 =item B<-sync> <always | onclose | none>
365 This option changes how hard the fileserver tries to ensure that data written
366 to volumes actually hits the physical disk.
368 Normally, when the fileserver writes to disk, the underlying filesystem or
369 Operating System may delay writes from actually going to disk, and reorder
370 which writes hit the disk first. So, during an unclean shutdown of the machine
371 (if the power goes out, or the machine crashes, etc), file data may become lost
372 that the server previously told clients was already successfully written.
374 To try to mitigate this, the fileserver will try to "sync" file data to the
375 physical disk at numerous points during various I/O. However, this can result
376 in significantly reduced performance. Depending on the usage patterns, this may
377 or may not be acceptable. This option dictates specifically what the fileserver
378 does when it wants to perform a "sync".
380 There are several options; pass one of these as the argument to -sync. The
381 default is C<onclose>.
387 This causes a sync operation to always sync immediately and synchronously.
388 This is the slowest option that provides the greatest protection against data
389 loss in the event of a crash.
391 Note that this is still not a 100% guarantee that data will not be lost or
392 corrupted during a crash. The underlying filesystem itself may cause data to
393 be lost or corrupt in such a situation. And OpenAFS itself does not (yet) even
394 guarantee that all data is consistent at any point in time; so even if the
395 filesystem and OS do not buffer or reorder any writes, you are not guaranteed
396 that all data will be okay after a crash.
398 This was the only behavior allowed in OpenAFS releases prior to 1.4.5.
402 This causes a sync to do nothing immediately, but causes the relevant file to
403 be flagged as potentially needing a sync. When a volume is detached, volume
404 metadata files flaged for synced are synced, as well as data files that have
405 been accessed recently. Events that cause a volume to detach include:
406 performing volume operations (dump, restore, clone, etc), a clean shutdown
407 of the fileserver, or during DAFS "soft detachment".
409 Effectively this option is the same as C<never> while a volume is attached and
410 actively being used, but if a volume is detached, there is an additional
411 guarantee for the data's consistency.
413 After the removal of the C<delayed> option after the OpenAFS 1.6 series, this
414 option became the default.
418 This causes all syncs to never do anything. This is the fastest option, with
419 the weakest guarantees for data consistency.
421 Depending on the underlying filesystem and Operating System, there may be
422 guarantees that any data written to disk will hit the physical media after a
423 certain amount of time. For example, Linux's pdflush process usually makes this
424 guarantee, and ext3 can make certain various consistency guarantees according
425 to the options given. ZFS on Solaris can also provide similar guarantees, as
426 can various other platforms and filesystems. Consult the documentation for
427 your platform if you are unsure.
431 This option used to exist in OpenAFS 1.6, but was later removed due to issues
432 encountered with data corruption during normal operation. Outside of the
433 OpenAFS 1.6 series, it is not a valid option, and the fileserver will fail to
434 start if you specify this (or any other unknown option). It caused syncs to
435 occur in a background thread, executing every 10 seconds.
437 This was the only behavior allowed in OpenAFS releases starting from 1.4.5 up
438 to and including 1.6.2. It was also the default for the 1.6 series starting in
443 Which option you choose is not an easy decision to make. Various developers
444 and experts sometimes disagree on which option is the most reasonable, and it
445 may depend on the specific scenario and workload involved. Some argue that
446 the C<always> option does not provide significantly greater guarantees over
447 any other option, whereas others argue that choosing anything besides the
448 C<always> option allows for an unacceptable risk of data loss. This may
449 depend on your usage patterns, your platform and filesystem, and who you talk
452 =item B<-logfile> <I<log file>>
454 Sets the file to use for server logging. If logfile is not specified and
455 no other logging options are supplied, this will be F</usr/afs/logs/FileLog>.
456 Note that this option is intended for debugging and testing purposes.
457 Changing the location of the log file from the command line may result
458 in undesirable interactions with tools such as B<bos>.
460 =item B<-vhashsize <I<size>>
462 The log(2) of the number of volume hash buckets. Default is 8 (i.e., by
463 default, there are 2^8 = 256 volume hash buckets). The minimum that can
464 be specified is 6 (64 hash buckets). In OpenAFS 1.5.77 and earlier, the
465 maximum that can be specified is 14 (16384 buckets). After 1.5.77, the
466 maximum that can be specified is 28 (268435456 buckets).
468 =item B<-config> <I<configuration directory>>
470 Set the location of the configuration directory used to configure this
471 service. In a typical configuration this will be F</usr/afs/etc> - this
472 option allows the use of alternative configuration locations for testing