<html><head><meta http-equiv="Content-Type" content="text/html charset=windows-1252"></head><body style="word-wrap: break-word; -webkit-nbsp-mode: space; -webkit-line-break: after-white-space;" class="">„schedd_job_info“ does not scale due to its nature (the amount of <div class="">messages per job are depend on the cluster size and for each job</div><div class="">messages are generated). It is also questionable if all scheduler </div><div class="">decisions for each job and resource (queue instances) needs to</div><div class="">be documented temporarily. Hence the recommendation is always</div><div class="">to turn it off (I think we changed the default to that in one of the </div><div class="">last Sun versions). Alternatively you can use "qalter -w p <jobid>“</div><div class="">for figuring out why a job is not scheduled (produces similar messages</div><div class="">but for only one particular job instead).</div><div class=""><br class=""></div><div class="">Daniel<br class=""><div class=""><div class=""><div class="">
<br class=""><div><blockquote type="cite" class=""><div class="">Am 09.02.2015 um 09:43 schrieb Remy Dernat <<a href="mailto:remy.dernat@univ-montp2.fr" class="">remy.dernat@univ-montp2.fr</a>>:</div><br class="Apple-interchange-newline"><div class="">
<meta content="text/html; charset=windows-1252" http-equiv="Content-Type" class="">
<div bgcolor="#FFFFFF" text="#000000" class="">
<br class="">
<div class="moz-cite-prefix">Le 09/02/2015 03:56, Christopher Samuel
a écrit :<br class="">
</div>
<blockquote cite="mid:54D821C8.8050406@unimelb.edu.au" type="cite" class="">
<pre wrap="" class="">On 07/02/15 14:57, Alan Louis Scheinine wrote:
</pre>
<blockquote type="cite" class="">
<pre wrap="" class="">Only problem I've seen is that if a user allocates too much memory,
OOM killer can kill maintenance processes such as a scheduler daemon.
</pre>
</blockquote>
<pre wrap="" class="">This is why we disable overcommit. :-)
</pre>
</blockquote>
<small class="">Hi,<br class="">
<br class="">
I already saw that problem on our master. The scheduler, SGE, runs
out of memory and OOM decided to kill it:</small><br class="">
<br class="">
<meta http-equiv="content-type" content="text/html;
charset=windows-1252" class="">
<span style="color: rgb(74, 85, 113); font-family: Verdana, Geneva,
Arial, Helvetica, sans-serif; font-size: 12px; font-style: normal;
font-variant: normal; font-weight: normal; letter-spacing: normal;
line-height: normal; orphans: auto; text-align: left; text-indent:
0px; text-transform: none; white-space: normal; widows: auto;
word-spacing: 0px; -webkit-text-stroke-width: 0px; display: inline
!important; float: none; background-color: rgb(255, 255, 255);" class="">Dec
1 15:01:07 cluster1 kernel: Out of memory: Kill process 7963
(sge_qmaster) score 948 or sacrifice child<br class="">
<big class=""><br class="">
<font class="">I resolved that issue by disabling
"schedd_job_info" in SGE with "qconf -msconf".<br class="">
<br class="">
However, this setting gives significant informations about our
jobs.<br class="">
<br class="">
How should I adjust OOM killer ? Sould I set </font></big></span><br class="">
<span style="color: rgb(74, 85, 113); font-family: Verdana, Geneva,
Arial, Helvetica, sans-serif; font-size: 12px; font-style: normal;
font-variant: normal; font-weight: normal; letter-spacing: normal;
line-height: normal; orphans: auto; text-align: left; text-indent:
0px; text-transform: none; white-space: normal; widows: auto;
word-spacing: 0px; -webkit-text-stroke-width: 0px; display: inline
!important; float: none; background-color: rgb(255, 255, 255);" class=""><big class=""><font class="">
<meta http-equiv="content-type" content="text/html;
charset=windows-1252" class="">
</font></big></span>
<pre style="margin: 0px 0px 10px; padding: 5px; border: 0px; font-size: 13.6960000991821px; vertical-align: baseline; font-family: Consolas, Menlo, Monaco, 'Lucida Console', 'Liberation Mono', 'DejaVu Sans Mono', 'Bitstream Vera Sans Mono', 'Courier New', monospace, serif; overflow: auto; width: auto; max-height: 600px; word-wrap: normal; font-style: normal; font-variant: normal; font-weight: normal; letter-spacing: normal; line-height: 17.8048000335693px; orphans: auto; text-align: left; text-indent: 0px; text-transform: none; widows: auto; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: rgb(238, 238, 238); background-position: initial initial; background-repeat: initial initial;" class=""><code style="margin: 0px; padding: 0px; border: 0px; font-size: 13.6960000991821px; vertical-align: baseline; font-family: Consolas, Menlo, Monaco, 'Lucida Console', 'Liberation Mono', 'DejaVu Sans Mono', 'Bitstream Vera Sans Mono', 'Courier New', monospace, serif; white-space: inherit; background: rgb(238, 238, 238);" class="">vm.overcomm!
it_memory
= 2<small class="">
</small></code></pre>
<small class="">?<br class="">
<br class="">
Best regards,<br class="">
<br class="">
Rémy<br class="">
</small><br class="">
<pre class="moz-signature" cols="72">--
Rémy Dernat
MBB/ISE-M</pre>
</div>
_______________________________________________<br class="">Beowulf mailing list, <a href="mailto:Beowulf@beowulf.org" class="">Beowulf@beowulf.org</a> sponsored by Penguin Computing<br class="">To change your subscription (digest mode or unsubscribe) visit <a href="http://www.beowulf.org/mailman/listinfo/beowulf" class="">http://www.beowulf.org/mailman/listinfo/beowulf</a><br class=""></div></blockquote></div><br class=""></div></div></div></div></body></html>