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-5863 Lustre <p>The patch for <a href="https://jira.whamcloud.com/browse/LU-5383" title="New static analysis issues in mount_lustre" class="issue-link" data-issue-key="LU-5383"><del>LU-5383</del></a> fixed a out bound for ldd_svname. It test to see if ldd_svname is at least 8 characters in length but the function in lustre_disk.h server_make_name() sets ldd_svname for the mgs to MGS which is to short. This can prevent the MGS from mounting.</p> LU-5863

Can't mount MGS due to label being less than 8 characters long.

Bug Critical Resolved Fixed James Nunez James A Simmons HB patch Tue, 4 Nov 2014 22:44:45 +0000 Sun, 25 Jan 2015 02:54:00 +0000 Sun, 25 Jan 2015 02:54:00 +0000 Lustre 2.7.0 Lustre 2.7.0 0 8 <p>Patch to fix this is at <a href="http://review.whamcloud.com/#/c/12564" class="external-link" target="_blank" rel="nofollow noopener">http://review.whamcloud.com/#/c/12564</a>. I think it shows up for previously formatted file systems.</p> <p>Thanks for the patch, James.</p> <p>This bug has stopped all testing on the stable filesystem, as the stable filesystem (formatted with older Lustre ) can no longer be mounted. <br/> arg<span class="error">&#91;4&#93;</span> = /dev/mapper/iws11_4<br/> arg<span class="error">&#91;5&#93;</span> = /mnt/mgs<br/> source = /dev/mapper/iws11_4 (/dev/mapper/iws11_4), target = /mnt/mgs<br/> options = rw<br/> checking for existing Lustre data: found<br/> Reading CONFIGS/mountdata<br/> mount.lustre: invalid name 'MGS'</p> <p>I really don't want to nuke the stable side, so a fix would be appreciated</p> <p>And, I can't make a new clean filesystem, either. Hyperion is down<br/> MGSNID="192.168.120.5@o2ib"<br/> MGSDEV="/dev/mapper/iws10_4"<br/> MDSDEV="/dev/mapper/iws10_1"<br/> mkfs.lustre --reformat --mgs --device-size=$((512 * 1048576)) --fsname lustre $MGSDEV<br/> mkfs.lustre --reformat --mdt --index=0 --mkfsoptions='-i 4096' --mgsnid=$MGSNID --device-size=$((2048 * 1048576)) --fsname lustre $MDSDEV</p> <p>mkdir /mnt/mgs<br/> mkdir /mnt/mds</p> <p>mount -t lustre $MGSDEV /mnt/mgs<br/> mount -t lustre $MDSDEV /mnt/mds<br/> ~ <br/> Fails with same error</p> <p>And i have confirmed that we do NOT test internally with a separate MGS - so this was a complete escape.</p> <p>Landed to master (pre-2.7)</p> <p>The patch fixes the core of the problem, but it doesn't add a test. Clearly, we need a conf-sanity test that formats a separate MGS and mounts it, since we can't possibly have had such a test if this bug slipped through.</p> <p>Hi James Simmons,<br/> Are you going to add a regression test case into conf-sanity.sh or would you like me to do this?</p> <p>Could you finish it off. I'm working on a few other tickets.</p> <p>Sure. I'll do.</p> <p>Hi Andreas,</p> <blockquote><p>Clearly, we need a conf-sanity test that formats a separate MGS and mounts it, since we can't possibly have had such a test if this bug slipped through.</p></blockquote> <p>I found that conf-sanity test 21d is a basic test case that starting a separate MGS. However, autotest system always uses the configuration of combined MGT and MDT, and then test 21d has been always skipped. Should we enhance autotest system to add a configuration of separated MGT and MDT, and run a test session with this configuration?<br/> For now and this ticket, I'll just add a new test case that formats $fs2mds_DEV as a separate MGT and mounts it.</p> <p>In case it influences the decision to know what people are doing in the real world, Cray is increasingly using separate MGS and MDS volumes for ease of backup and recovery.<br/> Also for DNE - it's convenient to have MDS volumes all identical.</p> <p>Same here. ORNL has run separate MGS/MDS servers for years. If I remember right running separate MGS/MDS is required for IR to work properly.</p> <p>Jian Yu (jian.yu@intel.com) uploaded a new patch: <a href="http://review.whamcloud.com/13391" class="external-link" target="_blank" rel="nofollow noopener">http://review.whamcloud.com/13391</a><br/> Subject: <a href="https://jira.whamcloud.com/browse/LU-5863" title="Can&#39;t mount MGS due to label being less than 8 characters long." class="issue-link" data-issue-key="LU-5863"><del>LU-5863</del></a> tests: add a separate MGS/MDS test case into conf-sanity<br/> Project: fs/lustre-release<br/> Branch: master<br/> Current Patch Set: 1<br/> Commit: 41231258e22cb35555e17a8b94d80c7bfdc5522d</p> <p>Thank you Patrick and James for the information. The work to enhance autotest system is tracked in TEI-91. For this ticket, I pushed the above patch to add a test case into conf-sanity.sh to verify separate MGS/MDS. The new test case will not be skipped under autotest combined MGS/MDS configuration.</p> <p>Oleg Drokin (oleg.drokin@intel.com) merged in patch <a href="http://review.whamcloud.com/13391/" class="external-link" target="_blank" rel="nofollow noopener">http://review.whamcloud.com/13391/</a><br/> Subject: <a href="https://jira.whamcloud.com/browse/LU-5863" title="Can&#39;t mount MGS due to label being less than 8 characters long." class="issue-link" data-issue-key="LU-5863"><del>LU-5863</del></a> tests: add a separate MGS/MDS test case into conf-sanity<br/> Project: fs/lustre-release<br/> Branch: master<br/> Current Patch Set: <br/> Commit: 589d93f8b01f0a923c0f6ebb64c7ad9935fb05da</p> <p>Landed for 2.7</p> Duplicate LU-5876 Development Rank 1|hzx073: Rank (Obsolete) 16404 Severity