From 4a51de501153840fc937d83286bfbc7b7ef4b58e Mon Sep 17 00:00:00 2001 From: Morris Jette <jette@schedmd.com> Date: Wed, 29 Oct 2014 10:17:14 -0700 Subject: [PATCH] Describe how to clear Slurm/ALPS node state info --- doc/html/cray_alps.shtml | 14 +++++++++++++- 1 file changed, 13 insertions(+), 1 deletion(-) diff --git a/doc/html/cray_alps.shtml b/doc/html/cray_alps.shtml index 8eb5065a586..cc31a6c8dab 100644 --- a/doc/html/cray_alps.shtml +++ b/doc/html/cray_alps.shtml @@ -684,8 +684,20 @@ Cluster Compatibility Mode.</p> <p>By default the launch plugin on a Cray is on set to <i>launch/aprun</i>. Nothing extra is needed to enable it.</p> + +<h3>Node State</h3> + +<p>Slurm gets node state information from ALPS. +Use the Cray <i>xtprocadmin</i> command to set node state <i>up</i> or <i>down</i>. +If a node state is <i>down</i> in Slurm and setting it back up using the +<i>xtprocadmin</i> command fails, it may be necessary to destroy the +<i>node_state*</i> files in your <i>StateSaveLocation</i> directory (as +configured in Slurm), which will remove all of Slurm's node state information +and force Slurm to rely completely upon ALPS for all node state information. +Stop the <i>slurmctld</i> daemon, delete the files, and restart the daemon.</p> + <p class="footer"><a href="#top">top</a></p> -<p style="text-align:center;">Last modified 3 October 2013</p> +<p style="text-align:center;">Last modified 29 October 2014</p> <!--#include virtual="footer.txt"--> -- GitLab