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-4397 Lustre <p>A no longer existing OST has been permanently disabled on the MGS using</p> <p>lctl conf_param vegas-OST0059.osc.active=0</p> <p>After this, clients hang on df (after lustre is remounted the next time). strace shows it hanging on a statfs call.</p> <p>Tried with the lustre mount option lazystatfs (undocumented?) which is supposed to work around this issue, but it did not help with lustre 2.4.1, clients still hang on df. This is the standard df, lfs df seems to work OK.</p> RHEL6 2.6.32-358.18.1.el6.x86_64 LU-4397

Permanently disabled OST causes clients to hang on df (statfs)

Bug Major Resolved Fixed WC Triage Wolfgang Baudler Thu, 19 Dec 2013 15:26:17 +0000 Fri, 1 Dec 2017 02:17:37 +0000 Tue, 20 May 2014 13:51:17 +0000 Lustre 2.4.1 Lustre 2.6.0 Lustre 2.5.2 0 7 <p>Seems this may be the same as <a href="https://jira.whamcloud.com/browse/LU-4010" title="Don&#39;t wait for active target if OBD_STATFS_NODELAY flag is set." class="issue-link" data-issue-key="LU-4010"><del>LU-4010</del></a> - patch is <a href="http://review.whamcloud.com/7762" class="external-link" target="_blank" rel="nofollow noopener">http://review.whamcloud.com/7762</a></p> <p>Was this problem resolved with Change, 7762?<br/> Let us know if this ticket can be closed.</p> <p>The change 7762 was landed for 2.5.0, but I found another problem related to this on my test system: <a href="http://review.whamcloud.com/8949" class="external-link" target="_blank" rel="nofollow noopener">http://review.whamcloud.com/8949</a></p> <p>Landed for 2.5.2 and 2.6</p> <p>Hello,</p> <p>We recently upgraded out clients to 2.5.3 and this very issue appears to have manifested itself again?</p> <p>Lustre: setting import RSF1-OST0007_UUID INACTIVE by administrator request<br/> Lustre: Layout lock feature supported.<br/> Lustre: Mounted RSF1-client</p> <p>$ strace -v -f df /RSF1<br/> statfs("/var/lib/nfs/rpc_pipefs", {f_type=0x67596969, f_bsize=4096, f_blocks=0, f_bfree=0, f_bavail=0, f_files=0, f_ffree=0, f_fsid=</p> {0, 0} <p>, f_namelen=255, f_frsize=4096}) = 0<br/> statfs("/RSF1", </p> <p>At which point the df hangs until reboot.</p> <p>Because some of these client re-export Lustre via NFS and Samba the exportfs etc. hand on boot. For now we put a hack in /etc/rc.local as work around</p> <p>/etc/init.d/samba stop<br/> /etc/init.d/nfs stop<br/> mount -t lustre 10.82.0.15@tcp1:/RSF1 /RSF1<br/> /usr/sbin/lctl set_param llite.*.lazystatfs=1<br/> /etc/init.d/nfs start<br/> /etc/init.d/samba start</p> <p>Perhaps we missed something but this seem fit our experience.</p> <p>&#8211; <br/> Eric Kolb<br/> Data Centre Services<br/> University of Victoria<br/> Office: 250-721-7658</p> <p>I can confirm the behaviour described by Eric Kolb above. Problem exists on 2.5.3. I did not get a chance to test it with 2.5.2, so not sure if it was re-introduced or if it was never fixed.</p> <p>Note that it is also possible to set lazystatfs permanently on all clients using:</p> <div class="preformatted panel" style="border-width: 1px;"><div class="preformattedContent panelContent"> <pre>lctl conf_param llite.lazystatfs=1 </pre> </div></div> <p>This seems to work. The syntax is</p> <p>lctl conf_param &lt;fsname&gt;.llite.lazystafs=1</p> <p>So, is this expected behaviour and setting this option is required on any lustre filesystem with permanently deactivated OSTs? Or is it still a bug? The conf_param option does not seem to be documented.</p> Duplicate LU-4010 Related LU-7668 Development Epic client hang Epic/Theme OST client Rank 1|hzwbn3: Rank (Obsolete) 12067 Severity