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-5030 Lustre <p>In the upstream Linux kernel there is a requ irement to move files out of /proc/fs/lustre and into /sys/fs/lustre (lnet too). In order to facilitate this move it makes sense to start checking for files in /sys/fs/lustre in addition to the current glob that is used.</p> <p>This will allow moving incrementally (e.g. ET first</p> LU-5030

"lctl {get,set}_param" should also check in /sys/fs/{lnet,lustre}

Bug Critical Resolved Fixed Yang Sheng Andreas Dilger usk Thu, 8 May 2014 09:11:51 +0000 Fri, 31 Mar 2017 11:06:37 +0000 Thu, 4 Feb 2016 13:57:25 +0000 Lustre 2.8.0 0 18 <p>This reminds me that our e2fsprogs uses "/proc/fs/lustre" and has several "/proc/fs/lustre" paths hard coded in the source.</p> <p>Can I ask you a favor. The new guy I have been training at ORNL to be a Lustre developer has been working on this project. He is on vaction but should be back in the next few weeks. Is it okay for him to finish this work.</p> <p>The sysfs port work for LNET has already been done. Please link <a href="https://jira.whamcloud.com/browse/LU-2633" title="Base sysfs support" class="issue-link" data-issue-key="LU-2633"><del>LU-2633</del></a> to this one. </p> <p>James, that would be very much appreciated.</p> <p>I'd ask that the patch to fix "lctl get_param" be done first, so it can go into the released versions earlier, and then the work to actually convert over to sysfs can be done on master and on upstream clients in parallel.</p> <p>This is a patch that I had partially completed in my tree, and is probably a good starting point for this work: <a href="http://review.whamcloud.com/10300" class="external-link" target="_blank" rel="nofollow noopener">http://review.whamcloud.com/10300</a></p> <p>Do you want the work to be based on top of that patch or be added to this patch?</p> <p>This week I looked at the pros and cons of going to sysfs. The one con I see for going to sysfs is we end up with the one page for data issue that we had with the old proc handling. Going to seq_files now allows use to use more than one page for the data buffer.<br/> Now we could work around this by using debugfs. The second con is the very nature of sysfs. Sysfs is designed with a single data item per file. Most of lustre can easily be converted over but we have things like import which would need to be broken up.</p> <p>Chao is back from vacation so I will be meeting with him today to discuss this work. He also has a patch that moves nearly all the test suite from accessing proc directly. The only test we could not port over is sanity 900 due to lctl not being able to handle setting multiple proc entries at the same time. This is the same issue <a href="https://jira.whamcloud.com/browse/LU-3970" title="Add procfs interface for clearing lustre caches in parallel" class="issue-link" data-issue-key="LU-3970"><del>LU-3970</del></a> was created for but in that case it was to handle parallel access to ldlm drop_cache.</p> <p>I think the 10300 patch is a good starting point for changes to the lctl and liblustreapi code to consolidate pathname handling, instead of having it spread throughout the code. Ideally, this would eventually result in usable llapi_get_param() and llapi_set_param() functions which can be used by lctl, lfs, and other applications that hide the details of the interface and the location of the files in /proc or /sys or /debugfs or whatever.</p> <p>As for sanity.sh test_900, I'm not clear on why <tt>cancel_lru_locks()</tt> is not usable for the MGC. Ah, it seems that the filename is <tt>/proc/fs/lustre/ldlm/namespaces/MGC192.168.20.1@tcp/lru_size</tt> which has '.' in the filename, which clashes with <tt>lctl get_param</tt> and <tt>lctl set_param</tt> using '.' as a separator. One solution would be to have "lctl get_param" print these parameter names with the '.' escaped like '<br class="atl-forced-newline" />.' but that also can have problems if it is being evaluated multiple times (e.g. via remote shell). Alternately, the MGC proc name could be printed without '.' in the filename, though that wouldn't be consistent with other /proc files that print a NID.</p> <p>Perhaps I'm missing something, why does cancel_lru_locks in test-framework.sh do this:</p> <div class="code panel" style="border-width: 1px;"><div class="codeContent panelContent"> <pre class="code-java"> <span class="code-keyword">for</span> d in `lctl get_param -N ldlm.namespaces.*.lru_size | egrep -i $1`; <span class="code-keyword">do</span> $LCTL set_param -n $d=clear done </pre> </div></div> <p>instead of just doing an "lctl set_param" with a wildcard? Like this:</p> <div class="code panel" style="border-width: 1px;"><div class="codeContent panelContent"> <pre class="code-java">$LCTL set_param ldlm.namespaces.*$1*.lru_size </pre> </div></div> <p>Sure, cancel_lru_locks would no longer be case insensitive, but is that really important? If it is important, the input to cancel_lru_locks() could be coerced into the right letter case beforehand.</p> <p>I update the base patch but there is much more work to do. Currently I'm thinking of using llapi_target_iterate as the base for llapi_<span class="error">&#91;g|s&#93;</span>set_param. I'm also looking at using it for get_param_obdvar.</p> <p>Hi, James, Could you please give a status update for this ticket? TIA.</p> <p>In the latest version of patch 10300 the function llapi_get_param was defined. This function returns a 2D array of data the user can the parse for what he/she needs. This is a good step forward and can be used to define the needed functions for listing parameters and setting parameters. A few bugs remain in the patch which I have handed off to the other Lustre developer at ORNL, Chao. He will be carrying the rest of the work forward for me. Chao can you give a update were you are at?</p> <p>There are some bugs with the latest patch (patch set 7) of patch 10300 that I'm working on right now. After these bugs are fixed, would work on the functions of llapi_set_param and llapi_list_param. After then, need to fix the hard-coded /proc/fs/lustre in a few .c files by replacing them with the calls to these new functions.</p> <p>James Simmons (uja.ornl@yahoo.com) uploaded a new patch: <a href="http://review.whamcloud.com/14675" class="external-link" target="_blank" rel="nofollow noopener">http://review.whamcloud.com/14675</a><br/> Subject: <a href="https://jira.whamcloud.com/browse/LU-5030" title="&quot;lctl {get,set}_param&quot; should also check in /sys/fs/{lnet,lustre}" class="issue-link" data-issue-key="LU-5030"><del>LU-5030</del></a> utils: create a libcfs function to handle tunable paths<br/> Project: fs/lustre-release<br/> Branch: master<br/> Current Patch Set: 1<br/> Commit: cd1cf0b72ddcdaaf32ef7687cd0061b3b860a33c</p> <p>I revised <a href="http://review.whamcloud.com/14675/" class="external-link" target="_blank" rel="nofollow noopener">http://review.whamcloud.com/14675/</a> patch. Now it should work with /proc, /sys and /sys/kernel/debug file systems. I think it's an intermediate step forward. This patch aggregates all information about proc tree in one function, but we also have other parameters that can not be reached for now. For example, "/sys/module/libcfs/parameters/*". I'd like to use them also but we have explicit name "lnet/" or "lustre/" in pattern for all requests to cfs_get_procpath(). While we elaborating 10300 patch could we take this into account and remove those names? Could we append those name always in llapi_get_param() or we should have a separate API for lustre and lnet parameters? Does it make sense to have a separate API for them at all or all parameters should be common? What do you think about this?</p> <p>Dmitry Eremin (dmitry.eremin@intel.com) uploaded a new patch: <a href="http://review.whamcloud.com/14887" class="external-link" target="_blank" rel="nofollow noopener">http://review.whamcloud.com/14887</a><br/> Subject: <a href="https://jira.whamcloud.com/browse/LU-5030" title="&quot;lctl {get,set}_param&quot; should also check in /sys/fs/{lnet,lustre}" class="issue-link" data-issue-key="LU-5030"><del>LU-5030</del></a> utils: create llapi_get_param function<br/> Project: fs/lustre-release<br/> Branch: master<br/> Current Patch Set: 1<br/> Commit: f8d7ae7a6e58a1fa870c320af53284287558bef9</p> <p>I started to work on this again. Will have a update in a few days.</p> <p>I tried the current patch <a href="http://review.whamcloud.com/#/c/14675/" class="external-link" target="_blank" rel="nofollow noopener">http://review.whamcloud.com/#/c/14675/</a> and it works with current staging tree.</p> <p>Due to the way llite/lov was reimplemented (it used to be llite/lov/common_name that contained lov name which is way too convluluted, so tha was replaced with llite/lov being a symlink to the proper lov dir) we either need to adopt this same change in our tree and update the test scripts or just update the test scripts with something like (this is what I am running now):</p> <div class="preformatted panel" style="border-width: 1px;"><div class="preformattedContent panelContent"> <pre>diff --git a/lustre/tests/sanity.sh b/lustre/tests/sanity.sh index 0a9e496..b25dd83 100644 --- a/lustre/tests/sanity.sh +++ b/lustre/tests/sanity.sh @@ -95,6 +95,10 @@ assert_DIR MDT0=$($LCTL get_param -n mdc.*.mds_server_uuid | awk '{ gsub(/_UUID/,""); print $1 }' | head -n1) LOVNAME=$($LCTL get_param -n llite.*.lov.common_name | tail -n 1) +if [ -z "$LOVNAME" ] ; then +LOVNAME=$(readlink /sys/fs/lustre/llite/*/lov) +LOVNAME=$(basename $LOVNAME) +fi OSTCOUNT=$($LCTL get_param -n lov.$LOVNAME.numobd) STRIPECOUNT=$($LCTL get_param -n lov.$LOVNAME.stripecount) STRIPESIZE=$($LCTL get_param -n lov.$LOVNAME.stripesize) </pre> </div></div> <p>(similar patch is also needed for sanity-quota.sh)</p> <p>That is still a complicated way to do things. Plus if you have more than one file system LOVNAME might fetch the wrong one since only the first answer is respected. Why not set LOVNAME=lov.$FSNAME-clilov-* or just plain LOVNAME="lov-<b><del>clilov</del></b>" when you don't care about which file system. Looking at the proc/sys tree IMO I don't see a need for lov or lmv in the llite directory tree. The lov/lmv subdirectory in llite and llite's uuid all match. All we have it common_name which is unique but we can use wildcards to the lctl command to do the same thing. I can update my patch with this change.</p> <p>I updated the patch to remove the last bits of glob being used directly. This was especially true for lustre_cfg.c which defines the lctl <span class="error">&#91;s|g&#93;</span>et_params functionality. Only place left to move away from glob is the perl script utilities. Also more clean ups can be done for the liblustreapi internal parameter helper functions. Also for some reason /sys/module/&#42;/parameters/&#42; still don't work. I'm debugging to figure out why. Feedback welcomed.</p> <p>"llite.*.lov.common_name" would potentially match wrong fs too, I think, so it's ok to either way, no?</p> <p>Also don't bother with module parameters just yet - I am installing debugfs symlinks there so we should be fine (assuming this is accepted, see the patch series that just went to the lists).</p> <p>the symlinks pointing to module parameters did get accepted</p> <p>So we can take them for granted assuming debugfs is enabled, and if it's not do we care enough about those configurations?</p> <p>I needed to update one of my MythTV clients to run the vanilla 4.2.3 kernel to support a new motherboard chipset/CPU. The good news is that the vanilla 4.2.3 Lustre client mounts my 2.5.3 server w/o problem, the bad news is that this patch isn't landed yet so get/set the tunable parameters work, not even conf_param settings from the MGS config log, which is a bit strange (I didn't think <tt>lctl conf_param</tt> tunables depended on "lctl" working, only <tt>lctl set_param -P</tt> tunables?). I'll be able to test this patch when I have some spare time.</p> <p>Sorry all our test resources have been devoted to DNE2 and 2.8 release testing. I'm hoping a new test system will be brought up very soon so this work can be completed.</p> <p>Andreas Dilger (andreas.dilger@intel.com) uploaded a new patch: <a href="http://review.whamcloud.com/17081" class="external-link" target="_blank" rel="nofollow noopener">http://review.whamcloud.com/17081</a><br/> Subject: <a href="https://jira.whamcloud.com/browse/LU-5030" title="&quot;lctl {get,set}_param&quot; should also check in /sys/fs/{lnet,lustre}" class="issue-link" data-issue-key="LU-5030"><del>LU-5030</del></a> utils: add -R parameter to lctl get_param<br/> Project: fs/lustre-release<br/> Branch: master<br/> Current Patch Set: 1<br/> Commit: a86b456de9fd9232c34155a266447e88de937001</p> <p>Andreas Dilger (andreas.dilger@intel.com) uploaded a new patch: <a href="http://review.whamcloud.com/17082" class="external-link" target="_blank" rel="nofollow noopener">http://review.whamcloud.com/17082</a><br/> Subject: <a href="https://jira.whamcloud.com/browse/LU-5030" title="&quot;lctl {get,set}_param&quot; should also check in /sys/fs/{lnet,lustre}" class="issue-link" data-issue-key="LU-5030"><del>LU-5030</del></a> tests: remove /proc paths from tests<br/> Project: fs/lustre-release<br/> Branch: master<br/> Current Patch Set: 1<br/> Commit: 6fcb5eeaef446ce2fc03ccc12256382038a66770</p> <p>James Simmons (uja.ornl@yahoo.com) uploaded a new patch: <a href="http://review.whamcloud.com/17459" class="external-link" target="_blank" rel="nofollow noopener">http://review.whamcloud.com/17459</a><br/> Subject: <a href="https://jira.whamcloud.com/browse/LU-5030" title="&quot;lctl {get,set}_param&quot; should also check in /sys/fs/{lnet,lustre}" class="issue-link" data-issue-key="LU-5030"><del>LU-5030</del></a> libcfs: create cfs_get_paths() function<br/> Project: fs/lustre-release<br/> Branch: master<br/> Current Patch Set: 1<br/> Commit: 352e0c0d6e7ed879c3ae2310fe9aa8b3a096fc3e</p> <p>James Simmons (uja.ornl@yahoo.com) uploaded a new patch: <a href="http://review.whamcloud.com/17462" class="external-link" target="_blank" rel="nofollow noopener">http://review.whamcloud.com/17462</a><br/> Subject: <a href="https://jira.whamcloud.com/browse/LU-5030" title="&quot;lctl {get,set}_param&quot; should also check in /sys/fs/{lnet,lustre}" class="issue-link" data-issue-key="LU-5030"><del>LU-5030</del></a> utils: migrate lustre utils to us cfs_get_paths()<br/> Project: fs/lustre-release<br/> Branch: master<br/> Current Patch Set: 1<br/> Commit: 26f0ebd953be3d1c3b1dffa81bd83568c7c98ba7</p> <p>James Simmons (uja.ornl@yahoo.com) uploaded a new patch: <a href="http://review.whamcloud.com/17465" class="external-link" target="_blank" rel="nofollow noopener">http://review.whamcloud.com/17465</a><br/> Subject: <a href="https://jira.whamcloud.com/browse/LU-5030" title="&quot;lctl {get,set}_param&quot; should also check in /sys/fs/{lnet,lustre}" class="issue-link" data-issue-key="LU-5030"><del>LU-5030</del></a> snmp: migrate lustre SNMP utilites to use cfs_get_paths()<br/> Project: fs/lustre-release<br/> Branch: master<br/> Current Patch Set: 1<br/> Commit: fadfc7e2159137933bd79b1bdd404af155f8bd78</p> <p>James Simmons (uja.ornl@yahoo.com) uploaded a new patch: <a href="http://review.whamcloud.com/17466" class="external-link" target="_blank" rel="nofollow noopener">http://review.whamcloud.com/17466</a><br/> Subject: <a href="https://jira.whamcloud.com/browse/LU-5030" title="&quot;lctl {get,set}_param&quot; should also check in /sys/fs/{lnet,lustre}" class="issue-link" data-issue-key="LU-5030"><del>LU-5030</del></a> snmp: migrate lctl params functions to use cfs_get_paths()<br/> Project: fs/lustre-release<br/> Branch: master<br/> Current Patch Set: 1<br/> Commit: 81c06b9fbf4c40971d7ace55764de77694a63547</p> <p>James Simmons (uja.ornl@yahoo.com) uploaded a new patch: <a href="http://review.whamcloud.com/17468" class="external-link" target="_blank" rel="nofollow noopener">http://review.whamcloud.com/17468</a><br/> Subject: <a href="https://jira.whamcloud.com/browse/LU-5030" title="&quot;lctl {get,set}_param&quot; should also check in /sys/fs/{lnet,lustre}" class="issue-link" data-issue-key="LU-5030"><del>LU-5030</del></a> util: migrate liblusterapi to use cfs_get_paths()<br/> Project: fs/lustre-release<br/> Branch: master<br/> Current Patch Set: 1<br/> Commit: 5d58eaf7da2f0aa7e46c2ef28ff927e4f778764b</p> <p>Oleg Drokin (oleg.drokin@intel.com) merged in patch <a href="http://review.whamcloud.com/17081/" class="external-link" target="_blank" rel="nofollow noopener">http://review.whamcloud.com/17081/</a><br/> Subject: <a href="https://jira.whamcloud.com/browse/LU-5030" title="&quot;lctl {get,set}_param&quot; should also check in /sys/fs/{lnet,lustre}" class="issue-link" data-issue-key="LU-5030"><del>LU-5030</del></a> utils: add -R parameter to lctl get_param<br/> Project: fs/lustre-release<br/> Branch: master<br/> Current Patch Set: <br/> Commit: 8c63d9141c0556bf9c1bfe245a8ad68c1c1f4980</p> <p>Andreas Dilger (andreas.dilger@intel.com) uploaded a new patch: <a href="http://review.whamcloud.com/17700" class="external-link" target="_blank" rel="nofollow noopener">http://review.whamcloud.com/17700</a><br/> Subject: <a href="https://jira.whamcloud.com/browse/LU-5030" title="&quot;lctl {get,set}_param&quot; should also check in /sys/fs/{lnet,lustre}" class="issue-link" data-issue-key="LU-5030"><del>LU-5030</del></a> tests: delete old quota test script<br/> Project: fs/lustre-release<br/> Branch: master<br/> Current Patch Set: 1<br/> Commit: 9acb182b4437511aea7358bdb0ffc54680e2abc0</p> <p>Oleg Drokin (oleg.drokin@intel.com) merged in patch <a href="http://review.whamcloud.com/17700/" class="external-link" target="_blank" rel="nofollow noopener">http://review.whamcloud.com/17700/</a><br/> Subject: <a href="https://jira.whamcloud.com/browse/LU-5030" title="&quot;lctl {get,set}_param&quot; should also check in /sys/fs/{lnet,lustre}" class="issue-link" data-issue-key="LU-5030"><del>LU-5030</del></a> tests: delete old quota test script<br/> Project: fs/lustre-release<br/> Branch: master<br/> Current Patch Set: <br/> Commit: 36f730111b64cde8ccbcb0a528f17cee1c1194be</p> <p>Oleg Drokin (oleg.drokin@intel.com) merged in patch <a href="http://review.whamcloud.com/17459/" class="external-link" target="_blank" rel="nofollow noopener">http://review.whamcloud.com/17459/</a><br/> Subject: <a href="https://jira.whamcloud.com/browse/LU-5030" title="&quot;lctl {get,set}_param&quot; should also check in /sys/fs/{lnet,lustre}" class="issue-link" data-issue-key="LU-5030"><del>LU-5030</del></a> libcfs: create cfs_get_paths() function<br/> Project: fs/lustre-release<br/> Branch: master<br/> Current Patch Set: <br/> Commit: 4aed5234f1123efc06c5c7e702085a461a8aae12</p> <p>James Simmons (uja.ornl@yahoo.com) uploaded a new patch: <a href="http://review.whamcloud.com/17900" class="external-link" target="_blank" rel="nofollow noopener">http://review.whamcloud.com/17900</a><br/> Subject: <a href="https://jira.whamcloud.com/browse/LU-5030" title="&quot;lctl {get,set}_param&quot; should also check in /sys/fs/{lnet,lustre}" class="issue-link" data-issue-key="LU-5030"><del>LU-5030</del></a> utils: fix lnet/utils/debug.c compile issue<br/> Project: fs/lustre-release<br/> Branch: master<br/> Current Patch Set: 1<br/> Commit: 667afb125441e7b1f2f87fb74ee958f2b21131a0</p> <p>James Simmons (uja.ornl@yahoo.com) uploaded a new patch: <a href="http://review.whamcloud.com/17904" class="external-link" target="_blank" rel="nofollow noopener">http://review.whamcloud.com/17904</a><br/> Subject: <a href="https://jira.whamcloud.com/browse/LU-5030" title="&quot;lctl {get,set}_param&quot; should also check in /sys/fs/{lnet,lustre}" class="issue-link" data-issue-key="LU-5030"><del>LU-5030</del></a> util: delete no longer functional lltrack_stats application<br/> Project: fs/lustre-release<br/> Branch: master<br/> Current Patch Set: 1<br/> Commit: 71c0199406aade8f0e6e8645262e276cab889f24</p> <p>Oleg Drokin (oleg.drokin@intel.com) merged in patch <a href="http://review.whamcloud.com/14675/" class="external-link" target="_blank" rel="nofollow noopener">http://review.whamcloud.com/14675/</a><br/> Subject: <a href="https://jira.whamcloud.com/browse/LU-5030" title="&quot;lctl {get,set}_param&quot; should also check in /sys/fs/{lnet,lustre}" class="issue-link" data-issue-key="LU-5030"><del>LU-5030</del></a> tests: handle missing common_name for upstream client<br/> Project: fs/lustre-release<br/> Branch: master<br/> Current Patch Set: <br/> Commit: 9b09e38b5008af30c86854e77930fc85323079c5</p> <p>Oleg Drokin (oleg.drokin@intel.com) merged in patch <a href="http://review.whamcloud.com/17900/" class="external-link" target="_blank" rel="nofollow noopener">http://review.whamcloud.com/17900/</a><br/> Subject: <a href="https://jira.whamcloud.com/browse/LU-5030" title="&quot;lctl {get,set}_param&quot; should also check in /sys/fs/{lnet,lustre}" class="issue-link" data-issue-key="LU-5030"><del>LU-5030</del></a> utils: fix lnet/utils/debug.c compile issue<br/> Project: fs/lustre-release<br/> Branch: master<br/> Current Patch Set: <br/> Commit: 6b06230e87aa939b33e70001e216028680c992a4</p> <p>Oleg Drokin (oleg.drokin@intel.com) merged in patch <a href="http://review.whamcloud.com/17904/" class="external-link" target="_blank" rel="nofollow noopener">http://review.whamcloud.com/17904/</a><br/> Subject: <a href="https://jira.whamcloud.com/browse/LU-5030" title="&quot;lctl {get,set}_param&quot; should also check in /sys/fs/{lnet,lustre}" class="issue-link" data-issue-key="LU-5030"><del>LU-5030</del></a> util: delete no longer functional lltrack_stats application<br/> Project: fs/lustre-release<br/> Branch: master<br/> Current Patch Set: <br/> Commit: 832772079aa9bc16ae37524e74be04ef230ec1f0</p> <p>Oleg Drokin (oleg.drokin@intel.com) merged in patch <a href="http://review.whamcloud.com/17468/" class="external-link" target="_blank" rel="nofollow noopener">http://review.whamcloud.com/17468/</a><br/> Subject: <a href="https://jira.whamcloud.com/browse/LU-5030" title="&quot;lctl {get,set}_param&quot; should also check in /sys/fs/{lnet,lustre}" class="issue-link" data-issue-key="LU-5030"><del>LU-5030</del></a> util: migrate liblustreapi to use cfs_get_paths()<br/> Project: fs/lustre-release<br/> Branch: master<br/> Current Patch Set: <br/> Commit: 8813fdf2a4f2055e4867df653644f12ac4c78b15</p> <p>Oleg Drokin (oleg.drokin@intel.com) merged in patch <a href="http://review.whamcloud.com/17082/" class="external-link" target="_blank" rel="nofollow noopener">http://review.whamcloud.com/17082/</a><br/> Subject: <a href="https://jira.whamcloud.com/browse/LU-5030" title="&quot;lctl {get,set}_param&quot; should also check in /sys/fs/{lnet,lustre}" class="issue-link" data-issue-key="LU-5030"><del>LU-5030</del></a> tests: remove /proc paths from tests<br/> Project: fs/lustre-release<br/> Branch: master<br/> Current Patch Set: <br/> Commit: 072d6b3156bf3b5d1738b43aadcba5c378c59ee9</p> <p>Oleg Drokin (oleg.drokin@intel.com) merged in patch <a href="http://review.whamcloud.com/17462/" class="external-link" target="_blank" rel="nofollow noopener">http://review.whamcloud.com/17462/</a><br/> Subject: <a href="https://jira.whamcloud.com/browse/LU-5030" title="&quot;lctl {get,set}_param&quot; should also check in /sys/fs/{lnet,lustre}" class="issue-link" data-issue-key="LU-5030"><del>LU-5030</del></a> utils: migrate most lustre utils to use cfs_get_paths()<br/> Project: fs/lustre-release<br/> Branch: master<br/> Current Patch Set: <br/> Commit: 81aa70a14c7dcadd593ed987d3f92bd05d120a0f</p> <p>Oleg Drokin (oleg.drokin@intel.com) merged in patch <a href="http://review.whamcloud.com/17465/" class="external-link" target="_blank" rel="nofollow noopener">http://review.whamcloud.com/17465/</a><br/> Subject: <a href="https://jira.whamcloud.com/browse/LU-5030" title="&quot;lctl {get,set}_param&quot; should also check in /sys/fs/{lnet,lustre}" class="issue-link" data-issue-key="LU-5030"><del>LU-5030</del></a> snmp: migrate lustre SNMP utilites to use cfs_get_paths<br/> Project: fs/lustre-release<br/> Branch: master<br/> Current Patch Set: <br/> Commit: ddb28333258d23bd40b21278901841499dd4f198</p> <p>Oleg Drokin (oleg.drokin@intel.com) merged in patch <a href="http://review.whamcloud.com/17466/" class="external-link" target="_blank" rel="nofollow noopener">http://review.whamcloud.com/17466/</a><br/> Subject: <a href="https://jira.whamcloud.com/browse/LU-5030" title="&quot;lctl {get,set}_param&quot; should also check in /sys/fs/{lnet,lustre}" class="issue-link" data-issue-key="LU-5030"><del>LU-5030</del></a> util: migrate lctl params functions to use cfs_get_paths()<br/> Project: fs/lustre-release<br/> Branch: master<br/> Current Patch Set: <br/> Commit: 85cbe1a3ee6940f7468884bca43cd736a5365694</p> <p>Landed all patches for 2.8.</p> <p>Kernel side /sys work will be done under <a href="https://jira.whamcloud.com/browse/LU-4423" title="Tracking of patches from upstream kernel to Lustre client" class="issue-link" data-issue-key="LU-4423"><del>LU-4423</del></a></p> <p>Andreas Dilger (andreas.dilger@intel.com) uploaded a new patch: <a href="http://review.whamcloud.com/18338" class="external-link" target="_blank" rel="nofollow noopener">http://review.whamcloud.com/18338</a><br/> Subject: <a href="https://jira.whamcloud.com/browse/LU-5030" title="&quot;lctl {get,set}_param&quot; should also check in /sys/fs/{lnet,lustre}" class="issue-link" data-issue-key="LU-5030"><del>LU-5030</del></a> utils: continue on errors in lctl </p> {get,set} <p>_param<br/> Project: fs/lustre-release<br/> Branch: master<br/> Current Patch Set: 1<br/> Commit: 38b6939a82949bc351db1098f4a0a26c3196a432</p> Related LU-2633 LU-6215 LU-7437 LU-4768 LU-7757 LU-5330 LU-5330 Development Rank 1|hzwm3r: Rank (Obsolete) 13911 Severity