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-8353 Lustre <p>In DNE Commit-on-Sharing implementation, mdt_reint_unlink() changed code to lookup child before lock parent, as is to find whether current operation is distributed, if so it needs to trigger commit of local changes to eliminate dependencies. But this may cause stale child used in later code. We should always follow lock-before-lookup semantic.</p> LU-8353

mdt unlink should lock parent before lookup child

Bug Critical Resolved Fixed Lai Siyao Lai Siyao dne2 Thu, 30 Jun 2016 02:29:32 +0000 Thu, 28 Jul 2016 17:36:12 +0000 Wed, 13 Jul 2016 18:39:25 +0000 Lustre 2.8.0 Lustre 2.9.0 Fri, 15 Jul 2016 00:00:00 +0000 0 5 <p>Lai Siyao (lai.siyao@intel.com) uploaded a new patch: <a href="http://review.whamcloud.com/21088" class="external-link" target="_blank" rel="nofollow noopener">http://review.whamcloud.com/21088</a><br/> Subject: <a href="https://jira.whamcloud.com/browse/LU-8353" title="mdt unlink should lock parent before lookup child" class="issue-link" data-issue-key="LU-8353"><del>LU-8353</del></a> mdt: mdt unlink should lock before lookup<br/> Project: fs/lustre-release<br/> Branch: master<br/> Current Patch Set: 1<br/> Commit: cd90c77d1216c988126c86ec09de1b9242ab96e7</p> <p>Oleg Drokin (oleg.drokin@intel.com) merged in patch <a href="http://review.whamcloud.com/21088/" class="external-link" target="_blank" rel="nofollow noopener">http://review.whamcloud.com/21088/</a><br/> Subject: <a href="https://jira.whamcloud.com/browse/LU-8353" title="mdt unlink should lock parent before lookup child" class="issue-link" data-issue-key="LU-8353"><del>LU-8353</del></a> mdt: mdt unlink should lock before lookup<br/> Project: fs/lustre-release<br/> Branch: master<br/> Current Patch Set: <br/> Commit: 5e5dabac9eca8bc6ea516fb3ad53b8d74fe80a37</p> <p>Should this ticket be closed? Is it related to <a href="https://jira.whamcloud.com/browse/LU-7117" title="replay-single test_70d: timeout and mkdir/rmdir stopped" class="issue-link" data-issue-key="LU-7117"><del>LU-7117</del></a> and should be linked there, or is this unrelated?</p> <p>I think it's one of the root causes for <a href="https://jira.whamcloud.com/browse/LU-7117" title="replay-single test_70d: timeout and mkdir/rmdir stopped" class="issue-link" data-issue-key="LU-7117"><del>LU-7117</del></a>, but it can pop up w/o 7117, w/o any recovery. basically we were doing lookups w/o LDLM lock and that can result in a corrupted namespace if another "unlink+create" can win race.</p> <p>Patch landed to master for 2.9.0</p> Related LU-7117 LU-6844 Development Epic/Theme dne2 Rank 1|hzyg8v: Rank (Obsolete) 9223372036854775807 Severity