1 <?xml version='1.0' encoding='utf-8'?>
2 <chapter xmlns="http://docbook.org/ns/docbook"
3 xmlns:xl="http://www.w3.org/1999/xlink" version="5.0" xml:lang="en-US"
4 xml:id="configuringquotas">
5 <title xml:id="configuringquotas.title">Configuring and Managing
7 <section xml:id="quota_configuring">
10 <primary>Quotas</primary>
11 <secondary>configuring</secondary>
12 </indexterm>Working with Quotas</title>
13 <para>Quotas allow a system administrator to limit the amount of disk space
14 a user or group can use. Quotas are set by root, and can be specified for
15 individual users and/or groups. Before a file is written to a partition
16 where quotas are set, the quota of the creator's group is checked. If a
17 quota exists, then the file size counts towards the group's quota. If no
18 quota exists, then the owner's user quota is checked before the file is
19 written. Similarly, inode usage for specific functions can be controlled if
20 a user over-uses the allocated space.</para>
21 <para>Lustre quota enforcement differs from standard Linux quota
22 enforcement in several ways:</para>
25 <para>Quotas are administered via the
26 <literal>lfs</literal> and
27 <literal>lctl</literal> commands (post-mount).</para>
30 <para>The quota feature in Lustre software is distributed
31 throughout the system (as the Lustre file system is a distributed file
32 system). Because of this, quota setup and behavior on Lustre is
33 different from local disk quotas in the following ways:</para>
36 <para>No single point of administration: some commands must be
37 executed on the MGS, other commands on the MDSs and OSSs, and still
38 other commands on the client.</para>
41 <para>Granularity: a local quota is typically specified for
42 kilobyte resolution, Lustre uses one megabyte as the smallest quota
46 <para>Accuracy: quota information is distributed throughout
47 the file system and can only be accurately calculated with a completely
48 quite file system.</para>
53 <para>Quotas are allocated and consumed in a quantized fashion.</para>
56 <para>Client does not set the
57 <literal>usrquota</literal> or
58 <literal>grpquota</literal> options to mount. As of Lustre software
59 release 2.4, space accounting is always enabled by default and quota
60 enforcement can be enabled/disabled on a per-file system basis with
61 <literal>lctl conf_param</literal>. It is worth noting that both
62 <literal>lfs quotaon</literal> and
63 <literal>quota_type</literal> are deprecated as of Lustre software
68 <para>Although a quota feature is available in the Lustre software, root
69 quotas are NOT enforced.</para>
71 <literal>lfs setquota -u root</literal> (limits are not enforced)</para>
73 <literal>lfs quota -u root</literal> (usage includes internal Lustre data
74 that is dynamic in size and does not accurately reflect mount point
75 visible block and inode usage).</para>
78 <section xml:id="enabling_disk_quotas">
81 <primary>Quotas</primary>
82 <secondary>enabling disk</secondary>
83 </indexterm>Enabling Disk Quotas</title>
84 <para>The design of quotas on Lustre has management and enforcement
85 separated from resource usage and accounting. Lustre software is
86 responsible for management and enforcement. The back-end file
87 system is responsible for resource usage and accounting. Because of
88 this, it is necessary to begin enabling quotas by enabling quotas on the
89 back-end disk system. Because quota setup is dependent on the Lustre
90 software version in use, you may first need to run
91 <literal>lctl get_param version</literal> to identify
92 <xref linkend="whichversion"/> you are currently using.
95 <title>Enabling Disk Quotas (Lustre Software Prior to Release 2.4)
98 For Lustre software releases older than release 2.4,
99 <literal>lfs quotacheck</literal> must be first run from a client node to
100 create quota files on the Lustre targets (i.e. the MDT and OSTs).
101 <literal>lfs quotacheck</literal> requires the file system to be quiescent
102 (i.e. no modifying operations like write, truncate, create or delete
103 should run concurrently). Failure to follow this caution may result in
104 inaccurate user/group disk usage. Operations that do not change Lustre
105 files (such as read or mount) are okay to run.
106 <literal>lfs quotacheck</literal> performs a scan on all the Lustre
107 targets to calculates the block/inode usage for each user/group. If the
108 Lustre file system has many files,
109 <literal>quotacheck</literal> may take a long time to complete. Several
110 options can be passed to
111 <literal>lfs quotacheck</literal>:</para>
113 # lfs quotacheck -ug /mnt/testfs
118 <literal>u</literal>-- checks the user disk quota information</para>
122 <literal>g</literal>-- checks the group disk quota information</para>
125 <para>By default, quota is turned on after
126 <literal>quotacheck</literal> completes. However, this setting isn't
127 persistent and quota will have to be enabled again (via
128 <literal>lfs quotaon</literal>) if one of the Lustre targets is
130 <literal>lfs quotaoff</literal> is used to turn off quota.</para>
131 <para>To enable quota permanently with a Lustre software release older
132 than release 2.4, the
133 <literal>quota_type</literal> parameter must be used. This requires
135 <literal>mdd.quota_type</literal> and
136 <literal>ost.quota_type</literal>, respectively, on the MDT and OSTs.
137 <literal>quota_type</literal> can be set to the string
138 <literal>u</literal> (user),
139 <literal>g</literal> (group) or
140 <literal>ug</literal> for both users and groups. This parameter can be
142 <literal>mkfs</literal> time (
143 <literal>mkfs.lustre --param mdd.quota_type=ug</literal>) or with
144 <literal>tunefs.lustre</literal>. As an example:</para>
146 tunefs.lustre --param ost.quota_type=ug $ost_dev
149 <literal>mkfs.lustre --param mdd.quota_type=ug</literal> or
150 <literal>tunefs.lustre --param ost.quota_type=ug</literal>, be sure to
151 run the command on all OSTs and the MDT. Otherwise, abnormal results may
155 In Lustre software releases before 2.4, when new OSTs are
156 added to the file system, quotas are not automatically propagated to
157 the new OSTs. As a workaround, clear and then reset quotas for each
158 user or group using the
159 <literal>lfs setquota</literal> command. In the example below, quotas
160 are cleared and reset for user
161 <literal>bob</literal> on file system
162 <literal>testfs</literal>:
164 $ lfs setquota -u bob -b 0 -B 0 -i 0 -I 0 /mnt/testfs
165 $ lfs setquota -u bob -b 307200 -B 309200 -i 10000 -I 11000 /mnt/testfs
169 <section remap="h3" condition="l24">
170 <title>Enabling Disk Quotas (Lustre Software Release 2.4 and
172 <para>Quota setup is orchestrated by the MGS and <emphasis>all setup
173 commands in this section must be run on the MGS</emphasis>. Once setup,
174 verification of the quota state must be performed on the MDT. Although
175 quota enforcement is managed by the Lustre software, each OSD
176 implementation relies on the back-end file system to maintain
177 per-user/group block and inode usage. Hence, differences exist
178 when setting up quotas with ldiskfs or ZFS back-ends:</para>
181 <para>For ldiskfs backends,
182 <literal>mkfs.lustre</literal> now creates empty quota files and
183 enables the QUOTA feature flag in the superblock which turns quota
184 accounting on at mount time automatically. e2fsck was also modified
185 to fix the quota files when the QUOTA feature flag is present.</para>
188 <para>For ZFS backend, accounting ZAPs are created and maintained by
189 the ZFS file system itself. While ZFS tracks per-user and group block
190 usage, it does not handle inode accounting. The ZFS OSD implements
191 its own support for inode tracking. Two options are available:</para>
194 <para>The ZFS OSD can estimate the number of inodes in-use based
195 on the number of blocks used by a given user or group. This mode
196 can be enabled by running the following command on the server
198 <literal>lctl set_param
199 osd-zfs.${FSNAME}-${TARGETNAME}.quota_iused_estimate=1</literal>.</para>
202 <para>Similarly to block accounting, dedicated ZAPs are also
203 created the ZFS OSD to maintain per-user and group inode usage.
204 This is the default mode which corresponds to
205 <literal>quota_iused_estimate</literal> set to 0.</para>
211 <para>Lustre file systems formatted with a Lustre release prior to 2.4.0
212 can be still safely upgraded to release 2.4.0, but will not have
213 functional space usage report until
214 <literal>tunefs.lustre --quota</literal> is run against all targets. This
215 command sets the QUOTA feature flag in the superblock and runs e2fsck (as
216 a result, the target must be offline) to build the per-UID/GID disk usage
217 database. See <xref linkend="quota_interoperability"/> for further
218 important considerations.</para>
221 <para>Lustre software release 2.4 and later requires a version of
222 e2fsprogs that supports quota (i.e. newer or equal to 1.42.3.wc1) to be
223 installed on the server nodes using ldiskfs backend (e2fsprogs is not
224 needed with ZFS backend). In general, we recommend to use the latest
225 e2fsprogs version available on
226 <link xl:href="http://downloads.hpdd.intel.com/e2fsprogs/">
227 http://downloads.hpdd.intel.com/public/e2fsprogs/</link>.</para>
228 <para>The ldiskfs OSD relies on the standard Linux quota to maintain
229 accounting information on disk. As a consequence, the Linux kernel
230 running on the Lustre servers using ldiskfs backend must have
231 <literal>CONFIG_QUOTA</literal>,
232 <literal>CONFIG_QUOTACTL</literal> and
233 <literal>CONFIG_QFMT_V2</literal> enabled.</para>
235 <para>As of Lustre software release 2.4.0, quota enforcement is thus
236 turned on/off independently of space accounting which is always enabled.
238 <replaceable>on|off</replaceable></literal> as well as the per-target
239 <literal>quota_type</literal> parameter are deprecated in favor of a
240 single per-file system quota parameter controlling inode/block quota
241 enforcement. Like all permanent parameters, this quota parameter can be
243 <literal>lctl conf_param</literal> on the MGS via the following
246 lctl conf_param <replaceable>fsname</replaceable>.quota.<replaceable>ost|mdt</replaceable>=<replaceable>u|g|ug|none</replaceable>
251 <literal>ost</literal> -- to configure block quota managed by
256 <literal>mdt</literal> -- to configure inode quota managed by
261 <literal>u</literal> -- to enable quota enforcement for users
266 <literal>g</literal> -- to enable quota enforcement for groups
271 <literal>ug</literal> -- to enable quota enforcement for both users
276 <literal>none</literal> -- to disable quota enforcement for both users
280 <para>Examples:</para>
281 <para>To turn on user and group quotas for block only on file system
282 <literal>testfs1</literal>, <emphasis>on the MGS</emphasis> run:</para>
283 <screen>$ lctl conf_param testfs1.quota.ost=ug
285 <para>To turn on group quotas for inodes on file system
286 <literal>testfs2</literal>, on the MGS run:</para>
287 <screen>$ lctl conf_param testfs2.quota.mdt=g
289 <para>To turn off user and group quotas for both inode and block on file
291 <literal>testfs3</literal>, on the MGS run:</para>
292 <screen>$ lctl conf_param testfs3.quota.ost=none
294 <screen>$ lctl conf_param testfs3.quota.mdt=none
299 <primary>Quotas</primary>
300 <secondary>verifying</secondary>
301 </indexterm>Quota Verification</title>
302 <para>Once the quota parameters have been configured, all targets
303 which are part of the file system will be automatically notified of the
304 new quota settings and enable/disable quota enforcement as needed. The
305 per-target enforcement status can still be verified by running the
306 following <emphasis>command on the MDS(s)</emphasis>:</para>
308 $ lctl get_param osd-*.*.quota_slave.info
309 osd-zfs.testfs-MDT0000.quota_slave.info=
310 target name: testfs-MDT0000
314 conn to master: setup
315 user uptodate: glb[1],slv[1],reint[0]
316 group uptodate: glb[1],slv[1],reint[0]
321 <section xml:id="quota_administration">
324 <primary>Quotas</primary>
325 <secondary>creating</secondary>
326 </indexterm>Quota Administration</title>
327 <para>Once the file system is up and running, quota limits on blocks
328 and inodes can be set for both user and group. This is <emphasis>
329 controlled entirely from a client</emphasis> via three quota
332 <emphasis role="bold">Grace period</emphasis>-- The period of time (in
333 seconds) within which users are allowed to exceed their soft limit. There
334 are four types of grace periods:</para>
337 <para>user block soft limit</para>
340 <para>user inode soft limit</para>
343 <para>group block soft limit</para>
346 <para>group inode soft limit</para>
349 <para>The grace period applies to all users. The user block soft limit is
350 for all users who are using a blocks quota.</para>
352 <emphasis role="bold">Soft limit</emphasis> -- The grace timer is started
353 once the soft limit is exceeded. At this point, the user/group can still
354 allocate block/inode. When the grace time expires and if the user is still
355 above the soft limit, the soft limit becomes a hard limit and the
356 user/group can't allocate any new block/inode any more. The user/group
357 should then delete files to be under the soft limit. The soft limit MUST be
358 smaller than the hard limit. If the soft limit is not needed, it should be
359 set to zero (0).</para>
361 <emphasis role="bold">Hard limit</emphasis> -- Block or inode allocation
363 <literal>EDQUOT</literal>(i.e. quota exceeded) when the hard limit is
364 reached. The hard limit is the absolute limit. When a grace period is set,
365 one can exceed the soft limit within the grace period if under the hard
367 <para>Due to the distributed nature of a Lustre file system and the need to
368 maintain performance under load, those quota parameters may not be 100%
369 accurate. The quota settings can be manipulated via the
370 <literal>lfs</literal> command, executed on a client, and includes several
371 options to work with quotas:</para>
375 <varname>quota</varname> -- displays general quota information (disk
376 usage and limits)</para>
380 <varname>setquota</varname> -- specifies quota limits and tunes the
381 grace period. By default, the grace period is one week.</para>
386 lfs quota [-q] [-v] [-h] [-o obd_uuid] [-u|-g <replaceable>uname|uid|gname|gid</replaceable>] <replaceable>/mount_point</replaceable>
387 lfs quota -t <replaceable>-u|-g</replaceable> <replaceable>/mount_point</replaceable>
388 lfs setquota <replaceable>-u|--user|-g|--group</replaceable> <replaceable>username|groupname</replaceable> [-b <replaceable>block-softlimit</replaceable>] \
389 [-B <replaceable>block_hardlimit</replaceable>] [-i <replaceable>inode_softlimit</replaceable>] \
390 [-I <replaceable>inode_hardlimit</replaceable>] <replaceable>/mount_point</replaceable>
392 <para>To display general quota information (disk usage and limits) for the
393 user running the command and his primary group, run:</para>
395 $ lfs quota /mnt/testfs
397 <para>To display general quota information for a specific user ("
398 <literal>bob</literal>" in this example), run:</para>
400 $ lfs quota -u bob /mnt/testfs
402 <para>To display general quota information for a specific user ("
403 <literal>bob</literal>" in this example) and detailed quota statistics for
404 each MDT and OST, run:</para>
406 $ lfs quota -u bob -v /mnt/testfs
408 <para>To display general quota information for a specific group ("
409 <literal>eng</literal>" in this example), run:</para>
411 $ lfs quota -g eng /mnt/testfs
413 <para>To display block and inode grace times for user quotas, run:</para>
415 $ lfs quota -t -u /mnt/testfs
417 <para>To set user or group quotas for a specific ID ("bob" in this
418 example), run:</para>
420 $ lfs setquota -u bob -b 307200 -B 309200 -i 10000 -I 11000 /mnt/testfs
422 <para>In this example, the quota for user "bob" is set to 300 MB
423 (309200*1024) and the hard limit is 11,000 files. Therefore, the inode hard
424 limit should be 11000.</para>
425 <para>The quota command displays the quota allocated and consumed by each
426 Lustre target. Using the previous
427 <literal>setquota</literal> example, running this
428 <literal>lfs</literal> quota command:</para>
430 $ lfs quota -u bob -v /mnt/testfs
432 <para>displays this command output:</para>
434 Disk quotas for user bob (uid 6000):
435 Filesystem kbytes quota limit grace files quota limit grace
436 /mnt/testfs 0 30720 30920 - 0 10000 11000 -
437 testfs-MDT0000_UUID 0 - 8192 - 0 - 2560 -
438 testfs-OST0000_UUID 0 - 8192 - 0 - 0 -
439 testfs-OST0001_UUID 0 - 8192 - 0 - 0 -
440 Total allocated inode limit: 2560, total allocated block limit: 24576
442 <para>Global quota limits are stored in dedicated index files (there is one
443 such index per quota type) on the quota master target (aka QMT). The QMT
444 runs on MDT0000 and exports the global indexes via /proc. The global
445 indexes can thus be dumped via the following command:
447 # lctl get_param qmt.testfs-QMT0000.*.glb-*
448 </screen>The format of global indexes depends on the OSD type. The ldiskfs OSD
449 uses an IAM files while the ZFS OSD creates dedicated ZAPs.</para>
450 <para>Each slave also stores a copy of this global index locally. When the
451 global index is modified on the master, a glimpse callback is issued on the
452 global quota lock to notify all slaves that the global index has been
453 modified. This glimpse callback includes information about the identifier
454 subject to the change. If the global index on the QMT is modified while a
455 slave is disconnected, the index version is used to determine whether the
456 slave copy of the global index isn't up to date any more. If so, the slave
457 fetches the whole index again and updates the local copy. The slave copy of
458 the global index is also exported via /proc and can be accessed via the
461 lctl get_param osd-*.*.quota_slave.limit*
464 <para>Prior to 2.4, global quota limits used to be stored in
465 administrative quota files using the on-disk format of the linux quota
466 file. When upgrading MDT0000 to 2.4, those administrative quota files are
467 converted into IAM indexes automatically, conserving existing quota
468 limits previously set by the administrator.</para>
471 <section xml:id="quota_allocation">
474 <primary>Quotas</primary>
475 <secondary>allocating</secondary>
476 </indexterm>Quota Allocation</title>
477 <para>In a Lustre file system, quota must be properly allocated or users
478 may experience unnecessary failures. The file system block quota is divided
479 up among the OSTs within the file system. Each OST requests an allocation
480 which is increased up to the quota limit. The quota allocation is then
481 <emphasis role="italic">quantized</emphasis> to reduce the number of
482 quota-related request traffic.</para>
483 <para>The Lustre quota system distributes quotas from the Quota Master
484 Target (aka QMT). Only one QMT instance is supported for now and only runs
485 on the same node as MDT0000. All OSTs and MDTs set up a Quota Slave Device
486 (aka QSD) which connects to the QMT to allocate/release quota space. The
487 QSD is setup directly from the OSD layer.</para>
488 <para>To reduce quota requests, quota space is initially allocated to QSDs
489 in very large chunks. How much unused quota space can be hold by a target
490 is controlled by the qunit size. When quota space for a given ID is close
491 to exhaustion on the QMT, the qunit size is reduced and QSDs are notified
492 of the new qunit size value via a glimpse callback. Slaves are then
493 responsible for releasing quota space above the new qunit value. The qunit
494 size isn't shrunk indefinitely and there is a minimal value of 1MB for
495 blocks and 1,024 for inodes. This means that the quota space rebalancing
496 process will stop when this minimum value is reached. As a result, quota
497 exceeded can be returned while many slaves still have 1MB or 1,024 inodes
498 of spare quota space.</para>
499 <para>If we look at the
500 <literal>setquota</literal> example again, running this
501 <literal>lfs quota</literal> command:</para>
503 # lfs quota -u bob -v /mnt/testfs
505 <para>displays this command output:</para>
507 Disk quotas for user bob (uid 500):
508 Filesystem kbytes quota limit grace files quota limit grace
509 /mnt/testfs 30720* 30720 30920 6d23h56m44s 10101* 10000 11000
511 testfs-MDT0000_UUID 0 - 0 - 10101 - 10240
512 testfs-OST0000_UUID 0 - 1024 - - - -
513 testfs-OST0001_UUID 30720* - 29896 - - - -
514 Total allocated inode limit: 10240, total allocated block limit: 30920
516 <para>The total quota limit of 30,920 is allocated to user bob, which is
517 further distributed to two OSTs.</para>
518 <para>Values appended with '
519 <literal>*</literal>' show that the quota limit has been exceeded, causing
520 the following error when trying to write or create a file:</para>
523 $ cp: writing `/mnt/testfs/foo`: Disk quota exceeded.
527 <para>It is very important to note that the block quota is consumed per
528 OST and the inode quota per MDS. Therefore, when the quota is consumed on
529 one OST (resp. MDT), the client may not be able to create files
530 regardless of the quota available on other OSTs (resp. MDTs).</para>
531 <para>Setting the quota limit below the minimal qunit size may prevent
532 the user/group from all file creation. It is thus recommended to use
533 soft/hard limits which are a multiple of the number of OSTs * the minimal
536 <para>To determine the total number of inodes, use
537 <literal>lfs df -i</literal>(and also
538 <literal>lctl get_param *.*.filestotal</literal>). For more information on
540 <literal>lfs df -i</literal> command and the command output, see
541 <xref linkend="dbdoclet.50438209_35838" />.</para>
542 <para>Unfortunately, the
543 <literal>statfs</literal> interface does not report the free inode count
544 directly, but instead reports the total inode and used inode counts. The
545 free inode count is calculated for
546 <literal>df</literal> from (total inodes - used inodes). It is not critical
547 to know the total inode count for a file system. Instead, you should know
548 (accurately), the free inode count and the used inode count for a file
549 system. The Lustre software manipulates the total inode count in order to
550 accurately report the other two values.</para>
552 <section xml:id="quota_interoperability">
555 <primary>Quotas</primary>
556 <secondary>Interoperability</secondary>
557 </indexterm>Quotas and Version Interoperability</title>
558 <para>The new quota protocol introduced in Lustre software release 2.4.0
559 <emphasis role="bold">is not compatible</emphasis> with previous
560 versions. As a consequence,
561 <emphasis role="bold">all Lustre servers must be upgraded to release 2.4.0
562 for quota to be functional</emphasis>. Quota limits set on the Lustre file
563 system prior to the upgrade will be automatically migrated to the new quota
564 index format. As for accounting information with ldiskfs backend, they will
565 be regenerated by running
566 <literal>tunefs.lustre --quota</literal> against all targets. It is worth
568 <literal>tunefs.lustre --quota</literal> is
569 <emphasis role="bold">mandatory</emphasis> for all targets formatted with a
570 Lustre software release older than release 2.4.0, otherwise quota
571 enforcement as well as accounting won't be functional.</para>
572 <para>Besides, the quota protocol in release 2.4 takes for granted that the
573 Lustre client supports the
574 <literal>OBD_CONNECT_EINPROGRESS</literal> connect flag. Clients supporting
575 this flag will retry indefinitely when the server returns
576 <literal>EINPROGRESS</literal> in a reply. Here is the list of Lustre client
577 version which are compatible with release 2.4:</para>
580 <para>Release 2.3-based clients and later</para>
583 <para>Release 1.8 clients newer or equal to release 1.8.9-wc1</para>
586 <para>Release 2.1 clients newer or equal to release 2.1.4</para>
590 <section xml:id="granted_cache_and_quota_limits">
593 <primary>Quotas</primary>
594 <secondary>known issues</secondary>
595 </indexterm>Granted Cache and Quota Limits</title>
596 <para>In a Lustre file system, granted cache does not respect quota limits.
597 In this situation, OSTs grant cache to a Lustre client to accelerate I/O.
598 Granting cache causes writes to be successful in OSTs, even if they exceed
599 the quota limits, and will overwrite them.</para>
600 <para>The sequence is:</para>
603 <para>A user writes files to the Lustre file system.</para>
606 <para>If the Lustre client has enough granted cache, then it returns
607 'success' to users and arranges the writes to the OSTs.</para>
610 <para>Because Lustre clients have delivered success to users, the OSTs
611 cannot fail these writes.</para>
614 <para>Because of granted cache, writes always overwrite quota limitations.
615 For example, if you set a 400 GB quota on user A and use IOR to write for
616 user A from a bundle of clients, you will write much more data than 400 GB,
617 and cause an out-of-quota error (
618 <literal>EDQUOT</literal>).</para>
620 <para>The effect of granted cache on quota limits can be mitigated, but
621 not eradicated. Reduce the maximum amount of dirty data on the clients
622 (minimal value is 1MB):</para>
626 <literal>lctl set_param osc.*.max_dirty_mb=8</literal>
632 <section xml:id="lustre_quota_statistics">
635 <primary>Quotas</primary>
636 <secondary>statistics</secondary>
637 </indexterm>Lustre Quota Statistics</title>
638 <para>The Lustre software includes statistics that monitor quota activity,
639 such as the kinds of quota RPCs sent during a specific period, the average
640 time to complete the RPCs, etc. These statistics are useful to measure
641 performance of a Lustre file system.</para>
642 <para>Each quota statistic consists of a quota event and
643 <literal>min_time</literal>,
644 <literal>max_time</literal> and
645 <literal>sum_time</literal> values for the event.</para>
646 <informaltable frame="all">
648 <colspec colname="c1" colwidth="50*" />
649 <colspec colname="c2" colwidth="50*" />
654 <emphasis role="bold">Quota Event</emphasis>
659 <emphasis role="bold">Description</emphasis>
668 <emphasis role="bold">sync_acq_req</emphasis>
672 <para>Quota slaves send a acquiring_quota request and wait for
679 <emphasis role="bold">sync_rel_req</emphasis>
683 <para>Quota slaves send a releasing_quota request and wait for
690 <emphasis role="bold">async_acq_req</emphasis>
694 <para>Quota slaves send an acquiring_quota request and do not
695 wait for its return.</para>
701 <emphasis role="bold">async_rel_req</emphasis>
705 <para>Quota slaves send a releasing_quota request and do not wait
706 for its return.</para>
712 <emphasis role="bold">wait_for_blk_quota
713 (lquota_chkquota)</emphasis>
717 <para>Before data is written to OSTs, the OSTs check if the
718 remaining block quota is sufficient. This is done in the
719 lquota_chkquota function.</para>
725 <emphasis role="bold">wait_for_ino_quota
726 (lquota_chkquota)</emphasis>
730 <para>Before files are created on the MDS, the MDS checks if the
731 remaining inode quota is sufficient. This is done in the
732 lquota_chkquota function.</para>
738 <emphasis role="bold">wait_for_blk_quota
739 (lquota_pending_commit)</emphasis>
743 <para>After blocks are written to OSTs, relative quota
744 information is updated. This is done in the lquota_pending_commit
751 <emphasis role="bold">wait_for_ino_quota
752 (lquota_pending_commit)</emphasis>
756 <para>After files are created, relative quota information is
757 updated. This is done in the lquota_pending_commit
764 <emphasis role="bold">wait_for_pending_blk_quota_req
765 (qctxt_wait_pending_dqacq)</emphasis>
769 <para>On the MDS or OSTs, there is one thread sending a quota
770 request for a specific UID/GID for block quota at any time. At
771 that time, if other threads need to do this too, they should
772 wait. This is done in the qctxt_wait_pending_dqacq
779 <emphasis role="bold">wait_for_pending_ino_quota_req
780 (qctxt_wait_pending_dqacq)</emphasis>
784 <para>On the MDS, there is one thread sending a quota request for
785 a specific UID/GID for inode quota at any time. If other threads
786 need to do this too, they should wait. This is done in the
787 qctxt_wait_pending_dqacq function.</para>
793 <emphasis role="bold">nowait_for_pending_blk_quota_req
794 (qctxt_wait_pending_dqacq)</emphasis>
798 <para>On the MDS or OSTs, there is one thread sending a quota
799 request for a specific UID/GID for block quota at any time. When
800 threads enter qctxt_wait_pending_dqacq, they do not need to wait.
801 This is done in the qctxt_wait_pending_dqacq function.</para>
807 <emphasis role="bold">nowait_for_pending_ino_quota_req
808 (qctxt_wait_pending_dqacq)</emphasis>
812 <para>On the MDS, there is one thread sending a quota request for
813 a specific UID/GID for inode quota at any time. When threads
814 enter qctxt_wait_pending_dqacq, they do not need to wait. This is
815 done in the qctxt_wait_pending_dqacq function.</para>
821 <emphasis role="bold">quota_ctl</emphasis>
825 <para>The quota_ctl statistic is generated when lfs
826 <literal>setquota</literal>,
827 <literal>lfs quota</literal> and so on, are issued.</para>
833 <emphasis role="bold">adjust_qunit</emphasis>
837 <para>Each time qunit is adjusted, it is counted.</para>
844 <title>Interpreting Quota Statistics</title>
845 <para>Quota statistics are an important measure of the performance of a
846 Lustre file system. Interpreting these statistics correctly can help you
847 diagnose problems with quotas, and may indicate adjustments to improve
848 system performance.</para>
849 <para>For example, if you run this command on the OSTs:</para>
851 lctl get_param lquota.testfs-OST0000.stats
853 <para>You will get a result similar to this:</para>
855 snapshot_time 1219908615.506895 secs.usecs
856 async_acq_req 1 samples [us] 32 32 32
857 async_rel_req 1 samples [us] 5 5 5
858 nowait_for_pending_blk_quota_req(qctxt_wait_pending_dqacq) 1 samples [us] 2\
860 quota_ctl 4 samples [us] 80 3470 4293
861 adjust_qunit 1 samples [us] 70 70 70
864 <para>In the first line,
865 <literal>snapshot_time</literal> indicates when the statistics were taken.
866 The remaining lines list the quota events and their associated
868 <para>In the second line, the
869 <literal>async_acq_req</literal> event occurs one time. The
870 <literal>min_time</literal>,
871 <literal>max_time</literal> and
872 <literal>sum_time</literal> statistics for this event are 32, 32 and 32,
873 respectively. The unit is microseconds (μs).</para>
874 <para>In the fifth line, the quota_ctl event occurs four times. The
875 <literal>min_time</literal>,
876 <literal>max_time</literal> and
877 <literal>sum_time</literal> statistics for this event are 80, 3470 and
878 4293, respectively. The unit is microseconds (μs).</para>