Whamcloud Community JIRA https://jira.whamcloud.com This file is an XML representation of an issue en-us 9.4.14 940014 05-12-2023 [LU-107] Lustre init scripts with heartbeat v1 integration https://jira.whamcloud.com/browse/LU-107 Lustre <p>This issue is to request that Lustre initialization scripts developed at LLNL be reviewed for inclusion in Lustre. A Gerrit submission for review is on its way.</p> LU-107 Lustre init scripts with heartbeat v1 integration Improvement Minor Resolved Fixed Oleg Drokin Ned Bass Wed, 2 Mar 2011 15:55:35 +0000 Thu, 25 Oct 2012 02:01:22 +0000 Thu, 27 Sep 2012 16:31:33 +0000 Lustre 2.1.0 Lustre 2.3.0 0 10 <p>See <a href="http://review.whamcloud.com/290" class="external-link" target="_blank" rel="nofollow noopener">http://review.whamcloud.com/290</a></p> <p>Updated gerrit with a couple of bug fixes in ldev script.</p> <ul class="alternate" type="square"> <li>missing implementation of -r flag to query raidtab field</li> <li>incorrectly parsed block device names containing :</li> </ul> <p>Integrated in <span class="image-wrap" style=""><img src="http://build.whamcloud.com/images/16x16/blue.png" style="border: 0px solid black" /></span> <a href="http://build.whamcloud.com/job/reviews-centos5/385/" class="external-link" target="_blank" rel="nofollow noopener">reviews-centos5 #385</a><br/> <a href="https://jira.whamcloud.com/browse/LU-107" title="Lustre init scripts with heartbeat v1 integration" class="issue-link" data-issue-key="LU-107"><del>LU-107</del></a> Add scripts for implementing heartbeat v1 failover</p> <p>Ned Bass : <a href="http://git.whamcloud.com/gitweb/?p=fs/lustre-release.git&amp;a=commit&amp;h=33e3a53d63cfaa85da01c0c8b1032704f5c745d9" class="external-link" target="_blank" rel="nofollow noopener">33e3a53d63cfaa85da01c0c8b1032704f5c745d9</a><br/> Files : </p> <ul> <li>lustre/scripts/haconfig</li> <li>lustre/scripts/lhbadm</li> <li>lustre/scripts/lustre.in</li> <li>lustre/autoconf/lustre-core.m4</li> <li>lustre/doc/lhbadm.8</li> <li>lustre/conf/lustre</li> <li>lustre/doc/nids.5</li> <li>lustre/scripts/Makefile.am</li> <li>build/autoconf/lustre-build.m4</li> <li>lustre/doc/ldev.8</li> <li>lustre.spec.in</li> <li>lustre/scripts/lustre</li> <li>lustre/conf/ldev.conf</li> <li>lustre/scripts/Lustre</li> <li>lustre/doc/ldev.conf.5</li> <li>lustre/scripts/ldev</li> <li>lustre/conf/Makefile.am</li> <li>lustre/scripts/lnet</li> <li>lustre/doc/Makefile.am</li> </ul> <p>Integrated in <span class="image-wrap" style=""><img src="http://build.whamcloud.com/images/16x16/blue.png" style="border: 0px solid black" /></span> <a href="http://build.whamcloud.com/job/reviews-centos5/389/" class="external-link" target="_blank" rel="nofollow noopener">reviews-centos5 #389</a><br/> <a href="https://jira.whamcloud.com/browse/LU-107" title="Lustre init scripts with heartbeat v1 integration" class="issue-link" data-issue-key="LU-107"><del>LU-107</del></a> Add scripts for implementing heartbeat v1 failover</p> <p>Ned Bass : <a href="http://git.whamcloud.com/gitweb/?p=fs/lustre-release.git&amp;a=commit&amp;h=c0c1013e14d5bf99df874cd52fa747962b4441d3" class="external-link" target="_blank" rel="nofollow noopener">c0c1013e14d5bf99df874cd52fa747962b4441d3</a><br/> Files : </p> <ul> <li>lustre/scripts/lnet</li> <li>lustre/scripts/lustre.in</li> <li>build/autoconf/lustre-build.m4</li> <li>lustre/doc/ldev.8</li> <li>lustre/doc/nids.5</li> <li>lustre/conf/ldev.conf</li> <li>lustre/conf/lustre</li> <li>lustre/doc/lhbadm.8</li> <li>lustre/scripts/haconfig</li> <li>lustre.spec.in</li> <li>lustre/scripts/Lustre</li> <li>lustre/conf/Makefile.am</li> <li>lustre/scripts/lustre</li> <li>lustre/doc/ldev.conf.5</li> <li>lustre/scripts/ldev</li> <li>lustre/autoconf/lustre-core.m4</li> <li>lustre/scripts/Makefile.am</li> <li>lustre/doc/Makefile.am</li> <li>lustre/scripts/lhbadm</li> </ul> <p>Oleg</p> <p>Could you please assess whether this is safe to include in 2.1</p> <p>Thanks</p> <p>Peter</p> <p>This is currently breaking the build so it needs some updating. I've also asked Brian to inspect this.</p> <p>Hi Robert,</p> <p>Thanks for your comments. I replaced lustre with lustre.in in EXTRA_DIST in lustre/scripts/Makefile.am and got good results on my end (i.e. 'make rpms' still works). Removing Lustre from EXTRA_DIST breaks 'make rpms' so I left it in.</p> <p>However, I don't think this is what is breaking the ubuntu build. To confirm, I submitted an unmodified master to Hudson and it fails in the same way:</p> <p><a href="http://build.whamcloud.com/job/reviews-ubuntu/229" class="external-link" target="_blank" rel="nofollow noopener">http://build.whamcloud.com/job/reviews-ubuntu/229</a></p> <p>reverting patch 0032-<a href="https://jira.whamcloud.com/browse/LU-111" title="mount_lustre::parse_options() does not pass force* options down" class="issue-link" data-issue-key="LU-111"><del>LU-111</del></a>-Fix-force-options-parsing from ./ ... failed.<br/> make<span class="error">&#91;1&#93;</span>: *** <span class="error">&#91;clean&#93;</span> Error 1</p> <p>Thanks,<br/> Ned</p> <p>Net Bass said ...</p> <blockquote> <p>However, I don't think this is what is breaking the ubuntu build. To confirm, I submitted an unmodified master to Hudson and it fails in the same way:</p> <p><a href="http://build.whamcloud.com/job/reviews-ubuntu/229" class="external-link" target="_blank" rel="nofollow noopener">http://build.whamcloud.com/job/reviews-ubuntu/229</a></p></blockquote> <p>I've had a look at this build attempt. It's based on quite an old revision of master from back at the end of December 2010. I have landed at least one fix to the debian (and therefore ubuntu) build code since then.</p> <p>I'd be willing to bet that if you rebase your changes to the most recent master, this issue will go away.</p> <p>Ned, your master branch is very old. It looks like you are still based on the Oracle tree, and the ubuntu build is broken in that version.</p> <p><a href="http://git.whamcloud.com/?p=fs/lustre-release.git;a=log;h=f537233800d39a456d318815578aaafecc974fde" class="external-link" target="_blank" rel="nofollow noopener">http://git.whamcloud.com/?p=fs/lustre-release.git;a=log;h=f537233800d39a456d318815578aaafecc974fde</a></p> <p>Please rebase your branch with the current master in fs/lustre-repository and push your request again. </p> <p>Integrated in <span class="image-wrap" style=""><img src="http://build.whamcloud.com/images/16x16/blue.png" style="border: 0px solid black" /></span> <a href="http://build.whamcloud.com/job/reviews-centos5/405/" class="external-link" target="_blank" rel="nofollow noopener">reviews-centos5 #405</a><br/> <a href="https://jira.whamcloud.com/browse/LU-107" title="Lustre init scripts with heartbeat v1 integration" class="issue-link" data-issue-key="LU-107"><del>LU-107</del></a> Add scripts for implementing heartbeat v1 failover</p> <p>Ned Bass : <a href="http://git.whamcloud.com/gitweb/?p=fs/lustre-release.git&amp;a=commit&amp;h=cdd6bbe6152647db5bb8d388313ec03f35fcb080" class="external-link" target="_blank" rel="nofollow noopener">cdd6bbe6152647db5bb8d388313ec03f35fcb080</a><br/> Files : </p> <ul> <li>lustre/autoconf/lustre-core.m4</li> <li>lustre/conf/lustre</li> <li>lustre/doc/ldev.8</li> <li>lustre/scripts/lhbadm</li> <li>lustre/scripts/Lustre</li> <li>lustre/conf/ldev.conf</li> <li>lustre/scripts/lnet</li> <li>lustre/scripts/ldev</li> <li>lustre/scripts/lustre.in</li> <li>lustre.spec.in</li> <li>lustre/scripts/haconfig</li> <li>lustre/conf/Makefile.am</li> <li>build/autoconf/lustre-build.m4</li> <li>lustre/doc/nids.5</li> <li>lustre/scripts/Makefile.am</li> <li>lustre/scripts/lustre</li> <li>lustre/doc/Makefile.am</li> <li>lustre/doc/ldev.conf.5</li> <li>lustre/doc/lhbadm.8</li> </ul> <p>My apologies--I was accidentally using review/lustre instead of review/fs/lustre-release.</p> <p>I rebased and resubmitted and the ubuntu build still fails, but I think I understand why now. My patch removes lustre/scripts/lustre and replaces it with lustre/scripts/lustre.in (due to the name of the tune2fs executable being determined at configure time). So lustre/scripts/lustre gets auto-generated when configure is run. But following configure the build runs</p> <p>fakeroot debian/rules clean</p> <p>which reverts all the patches. Reverting my patch tries to recreate lustre/scripts/lustre, but this fails because it already exists (it was created by configure). I suppose one way to fix this is to separate out the removal of lustre/scripts/lustre as a separate patch. Thoughts?</p> <p>Thanks,<br/> Ned</p> <blockquote> <p>My apologies--I was accidentally using review/lustre instead of review/fs/lustre-release.</p></blockquote> <p>No worries. Glad you figured out what it was.</p> <blockquote> <p>I rebased and resubmitted and the ubuntu build still fails, but I think I understand why now. My patch removes lustre/scripts/lustre and replaces it with lustre/scripts/lustre.in (due to the name of the tune2fs executable being determined at configure time). So lustre/scripts/lustre gets auto-generated when configure is run. But following configure the build runs</p> <p>fakeroot debian/rules clean</p> <p>which reverts all the patches. Reverting my patch tries to recreate lustre/scripts/lustre, but this fails because it already exists (it was created by configure).</p></blockquote> <p>Nice catch. I also discovered the same yesterday since I was looking at why this build was failing also. It is a "perfect storm" of conditions that causes this. <img class="emoticon" src="https://jira.whamcloud.com/images/icons/emoticons/sad.png" height="16" width="16" align="absmiddle" alt="" border="0"/></p> <blockquote> <p> I suppose one way to fix this is to separate out the removal of lustre/scripts/lustre as a separate patch. Thoughts?</p></blockquote> <p>I don't think that will fix it, but my changeset for <a href="https://jira.whamcloud.com/browse/LU-120" title="bug in creating debian patches" class="issue-link" data-issue-key="LU-120"><del>LU-120</del></a> does. Perhaps you can cherry-pick that change and put it in front of yours and see if it fixes it. It does for me, locally.</p> <p>That worked! Thanks Brian.</p> <p><a href="http://build.whamcloud.com/job/reviews-ubuntu/256/" class="external-link" target="_blank" rel="nofollow noopener">http://build.whamcloud.com/job/reviews-ubuntu/256/</a></p> <p>Integrated in <span class="image-wrap" style=""><img src="http://build.whamcloud.com/images/16x16/blue.png" style="border: 0px solid black" /></span> <a href="http://build.whamcloud.com/job/reviews-centos5/416/" class="external-link" target="_blank" rel="nofollow noopener">reviews-centos5 #416</a><br/> <a href="https://jira.whamcloud.com/browse/LU-107" title="Lustre init scripts with heartbeat v1 integration" class="issue-link" data-issue-key="LU-107"><del>LU-107</del></a> Add scripts for implementing heartbeat v1 failover</p> <p>Ned Bass : <a href="http://git.whamcloud.com/gitweb/?p=fs/lustre-release.git&amp;a=commit&amp;h=a7de9d4b241454fa54e8e4638594240fec3bc82d" class="external-link" target="_blank" rel="nofollow noopener">a7de9d4b241454fa54e8e4638594240fec3bc82d</a><br/> Files : </p> <ul> <li>lustre/scripts/lustre.in</li> <li>lustre/conf/ldev.conf</li> <li>lustre/scripts/lnet</li> <li>lustre/doc/ldev.conf.5</li> <li>lustre/autoconf/lustre-core.m4</li> <li>lustre/doc/nids.5</li> <li>lustre/scripts/lhbadm</li> <li>lustre/scripts/haconfig</li> <li>lustre/doc/lhbadm.8</li> <li>lustre/conf/Makefile.am</li> <li>lustre/conf/lustre</li> <li>lustre.spec.in</li> <li>build/autoconf/lustre-build.m4</li> <li>lustre/scripts/lustre</li> <li>lustre/doc/Makefile.am</li> <li>lustre/doc/ldev.8</li> <li>lustre/scripts/Lustre</li> <li>lustre/scripts/ldev</li> <li>lustre/scripts/Makefile.am</li> </ul> <p>Integrated in <span class="image-wrap" style=""><img src="http://build.whamcloud.com/images/16x16/blue.png" style="border: 0px solid black" /></span> <a href="http://build.whamcloud.com/job/reviews-centos5/418/" class="external-link" target="_blank" rel="nofollow noopener">reviews-centos5 #418</a><br/> <a href="https://jira.whamcloud.com/browse/LU-107" title="Lustre init scripts with heartbeat v1 integration" class="issue-link" data-issue-key="LU-107"><del>LU-107</del></a> Add scripts for implementing heartbeat v1 failover</p> <p>Ned Bass : <a href="http://git.whamcloud.com/gitweb/?p=fs/lustre-release.git&amp;a=commit&amp;h=ffdb6f1830fead020e79aab7c32b4633e6bdb179" class="external-link" target="_blank" rel="nofollow noopener">ffdb6f1830fead020e79aab7c32b4633e6bdb179</a><br/> Files : </p> <ul> <li>lustre/scripts/haconfig</li> <li>lustre/autoconf/lustre-core.m4</li> <li>lustre/doc/ldev.conf.5</li> <li>build/autoconf/lustre-build.m4</li> <li>lustre/doc/lhbadm.8</li> <li>lustre/scripts/Lustre</li> <li>lustre/doc/Makefile.am</li> <li>lustre/scripts/lustre.in</li> <li>lustre/doc/ldev.8</li> <li>lustre/conf/ldev.conf</li> <li>lustre/scripts/lhbadm</li> <li>lustre/doc/nids.5</li> <li>lustre/scripts/Makefile.am</li> <li>lustre/conf/lustre</li> <li>lustre/conf/Makefile.am</li> <li>lustre/scripts/ldev</li> <li>lustre/scripts/lustre</li> <li>lustre/scripts/lnet</li> <li>lustre.spec.in</li> </ul> <p>Ned, Brian, are there any implications to an existing system if /etc/init.d/lustre and /etc/init.d/lnet are suddenly added to an existing system?</p> <p>What we don't want is that someone upgrades to Lustre 2.4 from 2.1 and suddenly their system is unusable until they generate an /etc/ldev.conf or something. My assumption is that nobody ever reads the manual or release notes when upgrading, so if it doesn't work correctly "out of the box" then something was done incorrectly by the code(r).</p> <p>Hi Andreas,</p> <p>If a site hasn't configured /etc/ldev.conf then adding the /etc/init.d/lustre script should have no effect. That is, it won't start any services or interfere with whatever method was used to start lustre before updating. Also it not run by the init system by default, but rather it is intended to by started by a HA/failover mechanism such as heartbeat.</p> <p>/etc/init.d/lnet could start lnet sooner than was the case before updating, but I wouldn't expect that to break anything.</p> <p>Doug,<br/> could you please have a look at how the patch in <a href="http://review.whamcloud.com/290" class="external-link" target="_blank" rel="nofollow noopener">http://review.whamcloud.com/290</a> complements or conflicts with your proposed changes to LNET configuration. I believe you were already planning to start with an /etc/init.d/lnet startup file. Hopefully by landing this now, it will give you a starting point for your configuration changes, and users will already be aware of this script, so your changes will be transparent to them.</p> <blockquote><p>If a site hasn't configured /etc/ldev.conf then adding the /etc/init.d/lustre script should have no effect.</p></blockquote> <p>That is unless there was already a site-specific /etc/init.d/lustre script in place and we overwrite it. The safest thing to do may be to mark it %config(noreplace) in the spec file.</p> <p>When we do our packaging on SLES11 SP1/2, we run into the following problems:</p> <p>1. no LSB header information:</p> <p>E: File `lnet' without LSB header found in /var/tmp/cray-lustre-cray_gem_c-2.3_3.0.34_0.7.9_1.0000.6718.11.1-root/etc/init.d/<br/> E: File `lustre' without LSB header found in /var/tmp/cray-lustre-cray_gem_c-2.3_3.0.34_0.7.9_1.0000.6718.11.1-root/etc/init.d/</p> <p>2. need sysconfig.lustre in /var/adm/fillup-templates:</p> <p>cray-lustre-cray_gem_c: "/etc/sysconfig/lustre" is not allowed anymore in SuSE Linux.</p> <p>3. if failover is only for servers, it should probably be excluded from client build</p> <p>Hi Wally,</p> <p>Thanks for reporting these problems. We knew the init scripts would probably need work to properly support non-redhat distros. I don't currently have a SLES system to test on, but when I get a chance I'll try to bring up a VM to look into this.</p> <p>Are you just using 'make rpm'? </p> <p>We have our own make/spec to build for our environment but I think you should run into the same problem using 'make rpm' in SLES11. </p> <p>Please reopen this ticket if there is outstanding work to do.</p> <p>Ned Bass was right, people who have a preexisting /etc/init.d/lustre will be in trouble.</p> <blockquote> <p>That is unless there was already a site-specific /etc/init.d/lustre script in place and we overwrite it. The safest thing to do may be to mark it %config(noreplace) in the spec file.</p></blockquote> <p>That wasn't done when this landed for 2.3, so it has broken Cray's environment. (besides the other issues that Wally raised)<br/> We can probably live with it, but maybe the Ops Manual should be updated to warn the installer.</p> <p>Cory and/or Ned, could you please submit a patch to resolve this issue.</p> Bugzilla ID 20165.0 Development Rank 1|hzv6av: Rank (Obsolete) 4555