1xfs_quota(8) System Manager's Manual xfs_quota(8)
2
6 xfs_quota - manage use of quota on XFS filesystems
7
9 xfs_quota [ -x ] [ -f ] [ -p prog ] [ -c cmd ] ... [ -d project ] ... [
10 -D projects_file ] [ -P projid_file ] [ path ... ]
11 xfs_quota -V
12
14 xfs_quota is a utility for reporting and editing various aspects of
15 filesystem quota.
16 The options to xfs_quota are:
18 -c cmd xfs_quota commands may be run interactively (the default) or
20 as arguments on the command line. Multiple -c arguments may
21 be given. The commands are run in the sequence given, then
22 the program exits.
23 -p prog Set the program name for prompts and some error messages, the
25 default value is xfs_quota.
26 -x Enable expert mode. All of the administrative commands (see
28 the ADMINISTRATOR COMMANDS section below) which allow modifi‐
29 cations to the quota system are available only in expert
30 mode.
31 -f Enable foreign filesystem mode. A limited number of user and
33 administrative commands are available for use on some foreign
34 (non-XFS) filesystems.
35 -d project
37 Project names or numeric identifiers may be specified with
38 this option, which restricts the output of the individual
39 xfs_quota commands to the set of projects specified. Multiple
40 -d arguments may be given.
41 -D projects_file
43 Specify a file containing the mapping of numeric project
44 identifiers to directory trees. /etc/projects as default, if
45 this option is none.
46 -P projid_file
48 Specify a file containing the mapping of numeric project
49 identifiers to project names. /etc/projid as default, if
50 this option is none.
51 -V Prints the version number and exits.
53 The optional path argument(s) can be used to specify mount points or
55 device files which identify XFS filesystems. The output of the individ‐
56 ual xfs_quota commands will then be restricted to the set of filesys‐
57 tems specified.
58 This manual page is divided into two sections - firstly, information
60 for users of filesystems with quota enabled, and the xfs_quota commands
61 of interest to such users; and then information which is useful only to
62 administrators of XFS filesystems using quota and the quota commands
63 which allow modifications to the quota system.
64 Note that common to almost all of the individual commands described
66 below are the options for specifying which quota types are of interest
67 - user quota (-u), group quota (-g), and/or project quota (-p). Also,
68 several commands provide options to operate on "blocks used" (-b),
69 "inodes used" (-i), and/or "realtime blocks used" (-r).
70 Many commands also have extensive online help. Use the help command for
72 more details on any command.
73
75 In most computing environments, disk space is not infinite. The quota
76 subsystem provides a mechanism to control usage of disk space. Quotas
77 can be set for each individual user on any/all of the local filesys‐
78 tems. The quota subsystem warns users when they exceed their allotted
79 limit, but allows some extra space for current work (hard limit/soft
80 limit). In addition, XFS filesystems with limit enforcement turned off
81 can be used as an effective disk usage accounting system.
82 Users' View of Disk Quotas
84 To most users, disk quotas are either of no concern or a fact of life
85 that cannot be avoided. There are two possible quotas that can be
86 imposed - a limit can be set on the amount of space a user can occupy,
87 and there may be a limit on the number of files (inodes) he can own.
88 The quota command provides information on the quotas that have been set
90 by the system administrators and current usage.
91 There are four numbers for each limit: current usage, soft limit
93 (quota), hard limit, and time limit. The soft limit is the number of
94 1K-blocks (or files) that the user is expected to remain below. The
95 hard limit cannot be exceeded. If a user's usage reaches the hard
96 limit, further requests for space (or attempts to create a file) fail
97 with the "Quota exceeded" (EDQUOT) error.
98 When a user exceeds the soft limit, the timer is enabled. Any time the
100 quota drops below the soft limits, the timer is disabled. If the timer
101 pops, the particular limit that has been exceeded is treated as if the
102 hard limit has been reached, and no more resources are allocated to the
103 user. The only way to reset this condition, short of turning off limit
104 enforcement or increasing the limit, is to reduce usage below quota.
105 Only the superuser (i.e. a sufficiently capable process) can set the
106 time limits and this is done on a per filesystem basis.
107 Surviving When the Quota Limit Is Reached
109 In most cases, the only way for a user to recover from over-quota con‐
110 ditions is to abort whatever activity is in progress on the filesystem
111 that has reached its limit, remove sufficient files to bring the limit
112 back below quota, and retry the failed program.
113 However, if a user is in the editor and a write fails because of an
114 over quota situation, that is not a suitable course of action. It is
115 most likely that initially attempting to write the file has truncated
116 its previous contents, so if the editor is aborted without correctly
117 writing the file, not only are the recent changes lost, but possibly
118 much, or even all, of the contents that previously existed.
119 There are several possible safe exits for a user caught in this situa‐
120 tion. He can use the editor shell escape command to examine his file
121 space and remove surplus files. Alternatively, using sh(1), he can
122 suspend the editor, remove some files, then resume it. A third possi‐
123 bility is to write the file to some other filesystem (perhaps to a file
124 on /tmp) where the user's quota has not been exceeded. Then after rec‐
125 tifying the quota situation, the file can be moved back to the filesys‐
126 tem it belongs on.
127
129 print Lists all paths with devices/project identifiers. The path list
130 can come from several places - the command line, the mount ta‐
131 ble, and the /etc/projects file.
132 df See the free command.
134 quota [ -g | -p | -u ] [ -bir ] [ -hnNv ] [ -f file ] [ ID | name ] ...
136 Show individual usage and limits, for a single user name or
137 numeric user ID. The -h option reports in a "human-readable"
138 format similar to the df(1) command. The -n option reports the
139 numeric IDs rather than the name. The -N option omits the
140 header. The -v option outputs verbose information. The -f option
141 sends the output to file instead of stdout.
142 free [ -bir ] [ -hN ] [ -f file ]
144 Reports filesystem usage, much like the df(1) utility. It can
145 show usage for blocks, inode, and/or realtime block space, and
146 shows used, free, and total available. If project quota are in
147 use (see the DIRECTORY TREE QUOTA section below), it will also
148 report utilisation for those projects (directory trees). The -h
149 option reports in a "human-readable" format. The -N option omits
150 the header. The -f option outputs the report to file instead of
151 stdout.
152 help [ command ]
154 Online help for all commands, or one specific command.
155 quit Exit xfs_quota.
157 q See the quit command.
159
161 The XFS quota system differs to that of other filesystems in a number
162 of ways. Most importantly, XFS considers quota information as filesys‐
163 tem metadata and uses journaling to provide a higher level guarantee of
164 consistency. As such, it is administered differently, in particular:
165 1. The quotacheck command has no effect on XFS filesystems. The
167 first time quota accounting is turned on (at mount time), XFS
168 does an automatic quotacheck internally; afterwards, the quota
169 system will always be completely consistent until quotas are
170 manually turned off.
171 2. There is no need for quota file(s) in the root of the XFS
173 filesystem.
174 3. XFS distinguishes between quota accounting and limit enforce‐
176 ment. Quota accounting must be turned on at the time of mount‐
177 ing the XFS filesystem. However, it is possible to turn on/off
178 limit enforcement any time quota accounting is turned on. The
179 "quota" option to the mount command turns on both (user) quota
180 accounting and enforcement. The "uqnoenforce" option must be
181 used to turn on user accounting with limit enforcement disabled.
182 4. Turning on quotas on the root filesystem is slightly different
184 from the above. For Linux XFS, the quota mount flags must be
185 passed in with the "rootflags=" boot parameter.
186 5. It is useful to use the state to monitor the XFS quota subsystem
188 at various stages - it can be used to see if quotas are turned
189 on, and also to monitor the space occupied by the quota system
190 itself..
191 6. There is a mechanism built into xfsdump that allows quota limit
193 information to be backed up for later restoration, should the
194 need arise.
195 7. Quota limits cannot be set before turning on quotas on.
197 8. XFS filesystems keep quota accounting on the superuser (user ID
199 zero), and the tool will display the superuser's usage informa‐
200 tion. However, limits are never enforced on the superuser (nor
201 are they enforced for group and project ID zero).
202 9. XFS filesystems perform quota accounting whether the user has
204 quota limits or not.
205 10. XFS supports the notion of project quota, which can be used to
207 implement a form of directory tree quota (i.e. to restrict a
208 directory tree to only being able to use up a component of the
209 filesystems available space; or simply to keep track of the
210 amount of space used, or number of inodes, within the tree).
211
213 path [ N ]
214 Lists all paths with devices/project identifiers or set the cur‐
215 rent path to the Nth list entry (the current path is used by
216 many of the commands described here, it identifies the filesys‐
217 tem toward which a command is directed). The path list can come
218 from several places - the command line, the mount table, and the
219 /etc/projects file.
220 report [ -gpu ] [ -bir ] [ -ahntlLNU ] [ -f file ]
222 Report filesystem quota information. This reports all quota
223 usage for a filesystem, for the specified quota type (u/g/p
224 and/or blocks/inodes/realtime). It reports blocks in 1KB units
225 by default. The -h option reports in a "human-readable" format
226 similar to the df(1) command. The -f option outputs the report
227 to file instead of stdout. The -a option reports on all filesys‐
228 tems. By default, outputs the name of the user/group/project. If
229 no name is defined for a given ID, outputs the numeric ID
230 instead. The -n option outputs the numeric ID instead of the
231 name. The -L and -U options specify lower and upper ID bounds to
232 report on. If upper/lower bounds are specified, then by default
233 only the IDs will be displayed in output; with the -l option, a
234 lookup will be performed to translate these IDs to names. The -N
235 option reports information without the header line. The -t
236 option performs a terse report.
237 state [ -gpu ] [ -av ] [ -f file ]
239 Report overall quota state information. This reports on the
240 state of quota accounting, quota enforcement, and the number of
241 extents being used by quota metadata within the filesystem. The
242 -f option outputs state information to file instead of stdout.
243 The -a option reports state on all filesystems and not just the
244 current path.
245 limit [ -g | -p | -u ] bsoft=N | bhard=N | isoft=N | ihard=N | rtb‐
247 soft=N | rtbhard=N -d | id | name
248 Set quota block limits (bhard/bsoft), inode count limits
249 (ihard/isoft) and/or realtime block limits (rtbhard/rtbsoft).
250 The -d option (defaults) can be used to set the default value
251 that will be used, otherwise a specific user/group/project name
252 or numeric identifier must be specified.
253 timer [ -g | -p | -u ] [ -bir ] value [ -d | id | name ]
255 Allows the quota enforcement timeout (i.e. the amount of time
256 allowed to pass before the soft limits are enforced as the hard
257 limits) to be modified. The current timeout setting can be dis‐
258 played using the state command.
259 When setting the default timer via the -d option, or for id 0,
260 or if no argument is given after value the value argument is a
261 number of seconds indicating the relative amount of time after
262 soft limits are exceeded, before hard limits are enforced.
263 When setting any other individual timer by id or name, the value
264 is the number of seconds from now, at which time the hard limits
265 will be enforced. This allows extending the grace time of an
266 individual user who has exceeded soft limits.
267 For value, units of 'minutes', 'hours', 'days', and 'weeks' are
268 also understood (as are their abbreviations 'm', 'h', 'd', and
269 'w').
270 warn [ -g | -p | -u ] [ -bir ] value -d | id | name
272 Allows the quota warnings limit (i.e. the number of times a
273 warning will be send to someone over quota) to be viewed and
274 modified. The -d option (defaults) can be used to set the
275 default time that will be used, otherwise a specific
276 user/group/project name or numeric identifier must be specified.
277 NOTE: this feature is not currently implemented.
278 enable [ -gpu ] [ -v ]
280 Switches on quota enforcement for the filesystem identified by
281 the current path. This requires the filesystem to have been
282 mounted with quota enabled, and for accounting to be currently
283 active. The -v option (verbose) displays the state after the
284 operation has completed.
285 disable [ -gpu ] [ -v ]
287 Disables quota enforcement, while leaving quota accounting
288 active. The -v option (verbose) displays the state after the
289 operation has completed.
290 off [ -gpu ] [ -v ]
292 Permanently switches quota off for the filesystem identified by
293 the current path. Quota can only be switched back on subse‐
294 quently by unmounting and then mounting again.
295 remove [ -gpu ] [ -v ]
297 Remove any space allocated to quota metadata from the filesystem
298 identified by the current path. Quota must not be enabled on
299 the filesystem, else this operation will report an error.
300 dump [ -g | -p | -u ] [ -f file ]
302 Dump out quota limit information for backup utilities, either to
303 standard output (default) or to a file. This is only the lim‐
304 its, not the usage information, of course.
305 restore [ -g | -p | -u ] [ -f file ]
307 Restore quota limits from a backup file. The file must be in
308 the format produced by the dump command.
309 quot [ -g | -p | -u ] [ -bir ] [ -acnv ] [ -f file ]
311 Summarize filesystem ownership, by user, group or project. This
312 command uses a special XFS "bulkstat" interface to quickly scan
313 an entire filesystem and report usage information. This command
314 can be used even when filesystem quota are not enabled, as it is
315 a full-filesystem scan (it may also take a long time...). The -a
316 option displays information on all filesystems. The -c option
317 displays a histogram instead of a report. The -n option displays
318 numeric IDs rather than names. The -v option displays verbose
319 information. The -f option send the output to file instead of
320 stdout.
321 project [ -cCs [ -d depth ] [ -p path ] id | name ]
323 The -c, -C, and -s options allow the directory tree quota mecha‐
324 nism to be maintained. -d allows one to limit recursion level
325 when processing project directories and -p allows one to specify
326 project paths at command line ( instead of /etc/projects ). All
327 options are discussed in detail below.
328
330 The project quota mechanism in XFS can be used to implement a form of
331 directory tree quota, where a specified directory and all of the files
332 and subdirectories below it (i.e. a tree) can be restricted to using a
333 subset of the available space in the filesystem.
334 A managed tree must be setup initially using the -s option to the
336 project command. The specified project name or identifier is matched to
337 one or more trees defined in /etc/projects, and these trees are then
338 recursively descended to mark the affected inodes as being part of that
339 tree. This process sets an inode flag and the project identifier on
340 every file in the affected tree. Once this has been done, new files
341 created in the tree will automatically be accounted to the tree based
342 on their project identifier. An attempt to create a hard link to a
343 file in the tree will only succeed if the project identifier matches
344 the project identifier for the tree. The xfs_io utility can be used to
345 set the project ID for an arbitrary file, but this can only be done by
346 a privileged user.
347 A previously setup tree can be cleared from project quota control
349 through use of the project -C option, which will recursively descend
350 the tree, clearing the affected inodes from project quota control.
351 Finally, the project -c option can be used to check whether a tree is
353 setup, it reports nothing if the tree is correct, otherwise it reports
354 the paths of inodes which do not have the project ID of the rest of the
355 tree, or if the inode flag is not set.
356 Option -d can be used to limit recursion level (-1 is infinite, 0 is
358 top level only, 1 is first level ... ). Option -p adds possibility to
359 specify project paths in command line without a need for /etc/projects
360 to exist. Note that if projects file exists then it is also used.
361
364 Enabling quota enforcement on an XFS filesystem (restrict a user to a
365 set amount of space).
366 # mount -o uquota /dev/xvm/home /home
368 # xfs_quota -x -c 'limit bsoft=500m bhard=550m tanya' /home
369 # xfs_quota -x -c report /home
370 Enabling project quota on an XFS filesystem (restrict files in log file
372 directories to only using 1 gigabyte of space).
373 # mount -o prjquota /dev/xvm/var /var
375 # echo 42:/var/log >> /etc/projects
376 # echo logfiles:42 >> /etc/projid
377 # xfs_quota -x -c 'project -s logfiles' /var
378 # xfs_quota -x -c 'limit -p bhard=1g logfiles' /var
379 Same as above without a need for configuration files.
381 # rm -f /etc/projects /etc/projid
383 # mount -o prjquota /dev/xvm/var /var
384 # xfs_quota -x -c 'project -s -p /var/log 42' /var
385 # xfs_quota -x -c 'limit -p bhard=1g 42' /var
386
388 XFS implements delayed allocation (aka. allocate-on-flush) and this has
389 implications for the quota subsystem. Since quota accounting can only
390 be done when blocks are actually allocated, it is possible to issue
391 (buffered) writes into a file and not see the usage immediately
392 updated. Only when the data is actually written out, either via one of
393 the kernels flushing mechanisms, or via a manual sync(2), will the
394 usage reported reflect what has actually been written.
395 In addition, the XFS allocation mechanism will always reserve the maxi‐
397 mum amount of space required before proceeding with an allocation. If
398 insufficient space for this reservation is available, due to the block
399 quota limit being reached for example, this may result in the alloca‐
400 tion failing even though there is sufficient space. Quota enforcement
401 can thus sometimes happen in situations where the user is under quota
402 and the end result of some operation would still have left the user
403 under quota had the operation been allowed to run its course. This
404 additional overhead is typically in the range of tens of blocks.
405 Both of these properties are unavoidable side effects of the way XFS
407 operates, so should be kept in mind when assigning block limits.
408
410 Quota support for filesystems with realtime subvolumes is not yet
411 implemented, nor is the quota warning mechanism (the Linux warnquota(8)
412 tool can be used to provide similar functionality on that platform).
413
415 /etc/projects Mapping of numeric project identifiers to directo‐
416 ries trees.
417 /etc/projid Mapping of numeric project identifiers to project
418 names.
419
422 df(1), mount(1), sync(2), projid(5), projects(5). xfs(5). warn‐
423 quota(8),
424 xfs_quota(8)