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-16645 Lustre <p>This was seen multiple time while working on <a href="https://jira.whamcloud.com/browse/LU-16618" title="rmfid miscellaneous fixes" class="issue-link" data-issue-key="LU-16618">LU-16618</a>. Will add more debugging details soon.</p> <p>Given all inputs are correct. lfs rmfid returns unrecogized fid: 1.</p> <div class="preformatted panel" style="border-width: 1px;"><div class="preformattedContent panelContent"> <pre>$ lfs rmfid lustre [0x200000402:0x1:0x0] unrecognized FID: 1</pre> </div></div> <p>Looks like the input itself is already mangled when sent to server. </p> <p>Ouput of strace</p> <div class="preformatted panel" style="border-width: 1px;"><div class="preformattedContent panelContent"> <pre>&lt;snip&gt;   1 execve("/usr/bin/lfs", ["lfs", "rmfid", "lustreT", "1"], 0x7ffe70d14d18 /* 53 vars */) = 0 &lt;snip&gt;</pre> </div></div> <p> </p> <p>client/server version</p> <p> </p> <div class="preformatted panel" style="border-width: 1px;"><div class="preformattedContent panelContent"> <pre>Client: 2.15.54.79 MDS: 2.15.54.79 OSS: 2.15.54.79 </pre> </div></div> <p> </p> <p> </p> <p>Kernel version: 4.18.0-372.9.1.el8.x86_64</p> LU-16645

rmfid returns "unrecognized FID: 1"

Bug Minor Resolved Not a Bug Andreas Dilger Arshad Hussain Fri, 17 Mar 2023 09:19:58 +0000 Tue, 21 Mar 2023 07:24:56 +0000 Tue, 21 Mar 2023 07:24:56 +0000 0 2 <p>Is this because of a file named "1" in the current directory? Using "<tt><span class="error">&#91;0x200000402:0x1:0x0&#93;</span></tt>" around the FID will be resolved by bash equivalent to a regexp of "<tt><span class="error">&#91;0124:x&#93;</span></tt>" so it would match a file named "<tt>1</tt>" (eg. from a badly-directed stdout).</p> <p>Presumably the test would pass if the FID was given without "<tt>[]</tt>" around it. </p> <div class="preformatted panel" style="border-width: 1px;"><div class="preformattedContent panelContent"> <pre>Is this because of a file named "1" in the current directory? Using "[0x200000402:0x1:0x0]" around the FID will be resolved by bash equivalent to a regexp of "[0124:x]" so it would match a file named "1" (eg. from a badly-directed stdout).</pre> </div></div> <p>  Thanks for the inputs. There is no manual input and lustre is mounted fresh every time. (ran as standard ONLY=421aa lustre/tests/sanity.sh). I will check on this further.</p> <p> </p> <div class="preformatted panel" style="border-width: 1px;"><div class="preformattedContent panelContent"> <pre>Presumably the test would pass if the FID was given without "[]" around it. </pre> </div></div> <p>Understood.  Since I am using it within script. Just (lazily) passing the '[ ]' without removing it.</p> <p> </p> <div class="preformatted panel" style="border-width: 1px;"><div class="preformattedContent panelContent"> <pre>Is this because of a file named "1" in the current directory? &lt;snip&gt;</pre> </div></div> <p>Andreas, I now get what you meant. Sorry for not getting it first time.</p> <p>Not an lfs bug. (See Andreas's comments) within ticket.</p> Development Rank 1|i03ghr: Rank (Obsolete) 9223372036854775807 Severity