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-14642 Lustre <p>There are cases that layout version has not been transferred to OST object after mirror extend/split/resync which makes following sync hang.</p> <p>OFD will compare the layout version from client with on-disk object's in ofd_verify_layout_version(), as the client's version increased with mirror extend/split/resync, the sync IO will keep loop with EINPROGRESS reply from the OFD.</p> <p> </p> LU-14642

transfer layout version to OST objects in layout change

Bug Critical Resolved Fixed Zhenyu Xu Zhenyu Xu Mon, 26 Apr 2021 12:44:27 +0000 Fri, 13 Jan 2023 19:27:39 +0000 Sat, 17 Sep 2022 12:33:32 +0000 Lustre 2.16.0 0 9 <p>Add a FLR mode in fsx.c to test various IO upon mirrored file. </p> <p>Bobi Jam (bobijam@hotmail.com) uploaded a new patch: <a href="https://review.whamcloud.com/43472" class="external-link" target="_blank" rel="nofollow noopener">https://review.whamcloud.com/43472</a><br/> Subject: <a href="https://jira.whamcloud.com/browse/LU-14642" title="transfer layout version to OST objects in layout change" class="issue-link" data-issue-key="LU-14642"><del>LU-14642</del></a> flr: transfer layout version on layout change<br/> Project: fs/lustre-release<br/> Branch: master<br/> Current Patch Set: 1<br/> Commit: 1cc3b747b73cb5818746f676a9defd49748597b6</p> <p>Bobi Jam (bobijam@hotmail.com) uploaded a new patch: <a href="https://review.whamcloud.com/43473" class="external-link" target="_blank" rel="nofollow noopener">https://review.whamcloud.com/43473</a><br/> Subject: <a href="https://jira.whamcloud.com/browse/LU-14642" title="transfer layout version to OST objects in layout change" class="issue-link" data-issue-key="LU-14642"><del>LU-14642</del></a> test: add flrmode in fsx<br/> Project: fs/lustre-release<br/> Branch: master<br/> Current Patch Set: 1<br/> Commit: d12684cce04a008ba3a6779744178bf1bbc01d53</p> <p>Oleg Drokin (green@whamcloud.com) merged in patch <a href="https://review.whamcloud.com/43472/" class="external-link" target="_blank" rel="nofollow noopener">https://review.whamcloud.com/43472/</a><br/> Subject: <a href="https://jira.whamcloud.com/browse/LU-14642" title="transfer layout version to OST objects in layout change" class="issue-link" data-issue-key="LU-14642"><del>LU-14642</del></a> flr: transfer layout version on layout change<br/> Project: fs/lustre-release<br/> Branch: master<br/> Current Patch Set: <br/> Commit: c0f8f6b5001ec1a58493cde790da50dd2876030c</p> <p>Bobijam, is there a reason why the MDS <b>must</b> update the layout version on the OST object before the client RPC can complete? If the client has a <b>newer</b> layout version than the OST, it would make sense that the client knows the new layout still includes the OST object, and the OST could just update the layout version directly. </p> <p>This is very different from the case where the client sends an RPC with <b>old</b> layout version and may need to update the layout in order to generate the correct RPC. That may mean the client is flushing old data without the layout lock being revoked, and it needs to contact the MDS to get the new file layout.</p> <p>Not only would this avoid the requirement for the MDS to send an urgent RPC to the OSS to avoid the client IO being blocked, but it seems like the MDS may not need to send any RPC at all, or at least not a distributed transaction RPC, if the client is still actively writing to the object. That would also avoid the race condition between the file layout being changed on the MDT and it being updated on all of the OST objects (which may be up to 2000 objects for a very wide striped layout). If a client sends an RPC with a new layout version to the OST first, then it can immediately update the object version and continue, unlike the current case of the client being blocked and waiting for a long time for the OST object version to be updated.</p> <p>sound reasonable, OST only allows newer version write from the client and that does not need MDS's notification to OST about the layout version change.</p> <p>Thanks for the update bobijam</p> <p>"Bobi Jam &lt;bobijam@hotmail.com&gt;" uploaded a new patch: <a href="https://review.whamcloud.com/45443" class="external-link" target="_blank" rel="nofollow noopener">https://review.whamcloud.com/45443</a><br/> Subject: <a href="https://jira.whamcloud.com/browse/LU-14642" title="transfer layout version to OST objects in layout change" class="issue-link" data-issue-key="LU-14642"><del>LU-14642</del></a> flr: abolish MDS transfer layout version to OST<br/> Project: fs/lustre-release<br/> Branch: master<br/> Current Patch Set: 1<br/> Commit: c9b87ca7ecb87ebfa22d5d3af736b2a42d603680</p> <p>"Oleg Drokin &lt;green@whamcloud.com&gt;" merged in patch <a href="https://review.whamcloud.com/43473/" class="external-link" target="_blank" rel="nofollow noopener">https://review.whamcloud.com/43473/</a><br/> Subject: <a href="https://jira.whamcloud.com/browse/LU-14642" title="transfer layout version to OST objects in layout change" class="issue-link" data-issue-key="LU-14642"><del>LU-14642</del></a> test: add fsx mirror file test mode<br/> Project: fs/lustre-release<br/> Branch: master<br/> Current Patch Set: <br/> Commit: 90ba8b4ac360b1987178445bd2ccd64f7958d912</p> <p>"Andreas Dilger &lt;adilger@whamcloud.com&gt;" uploaded a new patch: <a href="https://review.whamcloud.com/47567" class="external-link" target="_blank" rel="nofollow noopener">https://review.whamcloud.com/47567</a><br/> Subject: <a href="https://jira.whamcloud.com/browse/LU-14642" title="transfer layout version to OST objects in layout change" class="issue-link" data-issue-key="LU-14642"><del>LU-14642</del></a> tests: skip sanity-flr/100 for old servers<br/> Project: fs/lustre-release<br/> Branch: master<br/> Current Patch Set: 1<br/> Commit: 8c8cea505e046853dcebb6d414f54ce260fb9a88</p> <p>"Oleg Drokin &lt;green@whamcloud.com&gt;" merged in patch <a href="https://review.whamcloud.com/47567/" class="external-link" target="_blank" rel="nofollow noopener">https://review.whamcloud.com/47567/</a><br/> Subject: <a href="https://jira.whamcloud.com/browse/LU-14642" title="transfer layout version to OST objects in layout change" class="issue-link" data-issue-key="LU-14642"><del>LU-14642</del></a> tests: skip sanity-flr/100 for old servers<br/> Project: fs/lustre-release<br/> Branch: master<br/> Current Patch Set: <br/> Commit: 9e25e70d78f3f5fdb7489f4f9841b0931927f10c</p> <p><a href="https://jira.whamcloud.com/secure/ViewProfile.jspa?name=bobijam" class="user-hover" rel="bobijam">bobijam</a> </p> <p>what's going on with this fix?</p> <p>"Oleg Drokin &lt;green@whamcloud.com&gt;" merged in patch <a href="https://review.whamcloud.com/45443/" class="external-link" target="_blank" rel="nofollow noopener">https://review.whamcloud.com/45443/</a><br/> Subject: <a href="https://jira.whamcloud.com/browse/LU-14642" title="transfer layout version to OST objects in layout change" class="issue-link" data-issue-key="LU-14642"><del>LU-14642</del></a> flr: allow layout version update from client/MDS<br/> Project: fs/lustre-release<br/> Branch: master<br/> Current Patch Set: <br/> Commit: fa6574150b6f745a668fe69b2d6d970068a4cff1</p> <p>Landed for 2.16</p> Related LU-14647 LU-14646 Development Epic Link EX-4394 Rank 1|i01t33: Rank (Obsolete) 9223372036854775807 Severity