https://jira.whamcloud.com This file is an XML representation of an issue en-us 9.4.14 940014 05-12-2023 https://jira.whamcloud.com/browse/LU-12353 Lustre <p>We can slightly improve quota usage accounting times by avoiding unnecessary ondisk updates.</p> <p>A patch and explanation will be added.</p> LU-12353

optimizations for ldiskfs quota updates

Improvement Minor Resolved Fixed Andrew Perepechko Andrew Perepechko patch performance Wed, 29 May 2019 17:07:04 +0000 Wed, 31 May 2023 20:19:28 +0000 Thu, 3 Dec 2020 14:48:42 +0000 Lustre 2.14.0 Lustre 2.16.0 Lustre 2.15.3 0 5 <p>Andrew Perepechko (c17827@cray.com) uploaded a new patch: <a href="https://review.whamcloud.com/34992" class="external-link" target="_blank" rel="nofollow noopener">https://review.whamcloud.com/34992</a><br/> Subject: <a href="https://jira.whamcloud.com/browse/LU-12353" title="optimizations for ldiskfs quota updates" class="issue-link" data-issue-key="LU-12353"><del>LU-12353</del></a> ldiskfs: speedup quota journalling<br/> Project: fs/lustre-release<br/> Branch: master<br/> Current Patch Set: 1<br/> Commit: 44215f0c452da18577475ca4c8a50e147e9e6a2b</p> <p>Why this patch improves performance:</p> <p>Multiple threads will not block on dqio_mutex in dquot_commit().</p> <p>If thread A is writing dquot and another thread B is waiting for dqio_mutex, other threads can just skip dquot update. Thread B is guaranteed to write the latest data. Once it starts writing, it will clear DQ_MOD_B and new threads entering ldiskfs_mark_dquot_dirty() will have to queue at least one dquot update.</p> <p>So we are moving from the following worst case (threads ordered by the time they called ldiskfs_mark_dquot_dirty()):<br/> thread 1) executing dquot_commit()<br/> thread 2) waiting in dquot_commit()<br/> ...<br/> thread N) waiting in dquot_commit()</p> <p>to the following case:</p> <p>thread 1) executing dquot_commit()<br/> thread 2) waiting in dquot_commit()<br/> thread 3) can exit ldiskfs_mark_dquot_dirty() immediately since thread 2 is guaranteed to write the latest data to disk<br/> ...<br/> thread N) can exit ldiskfs_mark_dquot_dirty() immediately since thread 2 is guaranteed to write the latest data to disk</p> <p>Although, in the first case, a subset of thread 2, thread 3, ... thead N may find that DQ_MOD_B is already cleared so they can exit dquot_commit() without updating the disk buffer, each of them will still have to wait for dqio_mutex to do that.</p> <p>Atomicity concerns:</p> <p>The updates are guaranteed to be atomic since we always call ldiskfs_mark_dquot_dirty() with a transaction handle. And if multiple threads enter this code at the same time, they share handles to the same transaction by jbd2 design.</p> <p>Jan Kara mentioned in LKML that this is not the case for dqctl calls which can call ext4_mark_dquot_dirty() without a jbd2 handle, but this case is not relevant for Lustre.</p> <p>Significant dqio_mutex redesign was planned, so this patch was rejected in LKML, but we can at least slightly improve performance without any kernel fs/quota modification for the older kernels.</p> <p>Shuichi, would you be able to test out the referenced patch to see if it improves performance for us? This is <b>not</b> urgent, since I don't think e.g. we have quota enabled for IO500, but if we always have quota enabled on customer systems then it would be nice to add a 10-15% improvement (as mentioned in the comments in Gerrit on this patch) "for free" if this 2-line patch helps.</p> <p>Acutally, this could help IO500 since we have quota accounting by default.</p> <p>Andreas, Sure. I will add my Todo list and test this patch too. Just an confirmation. Do we expect performance improvement even no quota enforcement, but just quota accounting on quota slave? quota accounting is always enabled by default, right?</p> <p><a href="https://jira.whamcloud.com/secure/ViewProfile.jspa?name=sihara" class="user-hover" rel="sihara">sihara</a>,</p> <blockquote><p>Do we expect performance improvement even no quota enforcement, but just quota accounting on quota slave?</p></blockquote> <p>Yes.</p> <p>Oleg Drokin (green@whamcloud.com) merged in patch <a href="https://review.whamcloud.com/34992/" class="external-link" target="_blank" rel="nofollow noopener">https://review.whamcloud.com/34992/</a><br/> Subject: <a href="https://jira.whamcloud.com/browse/LU-12353" title="optimizations for ldiskfs quota updates" class="issue-link" data-issue-key="LU-12353"><del>LU-12353</del></a> ldiskfs: speedup quota journalling<br/> Project: fs/lustre-release<br/> Branch: master<br/> Current Patch Set: <br/> Commit: dad25f258e50895b4bd5fce30765599a7a490aa0</p> <p>Landed for 2.14</p> <p>"Jian Yu &lt;yujian@whamcloud.com&gt;" uploaded a new patch: <a href="https://review.whamcloud.com/c/fs/lustre-release/+/50853" class="external-link" target="_blank" rel="nofollow noopener">https://review.whamcloud.com/c/fs/lustre-release/+/50853</a><br/> Subject: <a href="https://jira.whamcloud.com/browse/LU-12353" title="optimizations for ldiskfs quota updates" class="issue-link" data-issue-key="LU-12353"><del>LU-12353</del></a> ldiskfs: add ext4-dquot-commit-speedup patch to more series<br/> Project: fs/lustre-release<br/> Branch: master<br/> Current Patch Set: 1<br/> Commit: 8843159caf504c8fc22402ad5b06594da378bae7</p> <p>"Jian Yu &lt;yujian@whamcloud.com&gt;" uploaded a new patch: <a href="https://review.whamcloud.com/c/fs/lustre-release/+/50983" class="external-link" target="_blank" rel="nofollow noopener">https://review.whamcloud.com/c/fs/lustre-release/+/50983</a><br/> Subject: <a href="https://jira.whamcloud.com/browse/LU-12353" title="optimizations for ldiskfs quota updates" class="issue-link" data-issue-key="LU-12353"><del>LU-12353</del></a> ldiskfs: add ext4-dquot-commit-speedup patch to more series<br/> Project: fs/lustre-release<br/> Branch: b2_15<br/> Current Patch Set: 1<br/> Commit: 2a33165775f7fefa9682432869e920ead225352c</p> <p>"Oleg Drokin &lt;green@whamcloud.com&gt;" merged in patch <a href="https://review.whamcloud.com/c/fs/lustre-release/+/50983/" class="external-link" target="_blank" rel="nofollow noopener">https://review.whamcloud.com/c/fs/lustre-release/+/50983/</a><br/> Subject: <a href="https://jira.whamcloud.com/browse/LU-12353" title="optimizations for ldiskfs quota updates" class="issue-link" data-issue-key="LU-12353"><del>LU-12353</del></a> ldiskfs: add ext4-dquot-commit-speedup patch to more series<br/> Project: fs/lustre-release<br/> Branch: b2_15<br/> Current Patch Set: <br/> Commit: 5305f6efdd3aae1b1034eb6a9d880f6db8559dbf</p> <p>"Oleg Drokin &lt;green@whamcloud.com&gt;" merged in patch <a href="https://review.whamcloud.com/c/fs/lustre-release/+/50853/" class="external-link" target="_blank" rel="nofollow noopener">https://review.whamcloud.com/c/fs/lustre-release/+/50853/</a><br/> Subject: <a href="https://jira.whamcloud.com/browse/LU-12353" title="optimizations for ldiskfs quota updates" class="issue-link" data-issue-key="LU-12353"><del>LU-12353</del></a> ldiskfs: add ext4-dquot-commit-speedup patch to more series<br/> Project: fs/lustre-release<br/> Branch: master<br/> Current Patch Set: <br/> Commit: ec8f93afa0559ef6bfcdf701f4c1a50207901ef2</p> Related LU-10034 Development Rank 1|i00h5j: Rank (Obsolete) 9223372036854775807