<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8">
</head>
<body bgcolor="#FFFFFF" text="#000000">
<p>You might try using Partition QoS's, those can do a bunch of neat
features.</p>
<p>-Paul Edmon-<br>
</p>
<br>
<div class="moz-cite-prefix">On 05/04/2018 09:59 AM, Liam Forbes
wrote:<br>
</div>
<blockquote type="cite"
cite="mid:A488CD4C-FDEC-4D6D-86BD-1485067D9EAF@alaska.edu">
<meta http-equiv="content-type" content="text/html; charset=utf-8">
<div><span style="background-color: rgba(255, 255, 255, 0);">We
have three "big memory" nodes. We'd like to limit the number
of jobs that run per node in two partitions that share these
nodes. Jobs in these two partitions are limited to a single
node max. We'd only like 8 or fewer jobs from either partition
to run per node. So at most only 16 jobs should be allowed to
share a given node. <br>
<br>
Currently, we have <br clear="all">
</span>
<div><span style="background-color: rgba(255, 255, 255, 0);">
SelectType=select/cons_res<br>
SelectTypeParameters=CR_CPU<br>
</span></div>
<span style="background-color: rgba(255, 255, 255, 0);">in our
slurm.conf<br>
<br>
The nodes are defined as:<br>
NodeName=n[144-146] NodeAddr=10.50.50.[144-146] CPUs=56
Sockets=2 CoresPerSocket=14 ThreadsPerCore=2 RealMemory=<a
href="tel:1500000" dir="ltr" x-apple-data-detectors="true"
x-apple-data-detectors-type="telephone"
x-apple-data-detectors-result="0" moz-do-not-send="true">1500000</a> State=UNKNOWN<br>
<br>
</span></div>
<span style="background-color: rgba(255, 255, 255, 0);">The two
partitions are defined as:<br>
</span>
<div>
<div><span style="background-color: rgba(255, 255, 255, 0);">PartitionName=analysis
Nodes=n[144-146] MaxTime=4-0:0 MaxNodes=1 State=UP
AllowGroups=all Priority=100 OverSubscribe=FORCE:4 Hidden=NO
Default=NO<br>
PartitionName=bio Nodes=n[144-146] MaxTime=14-0:0 MaxNodes=1
State=UP AllowGroups=all Priority=100 OverSubscribe=FORCE:4
Hidden=NO Default=NO<br>
<br>
</span></div>
<div><span style="background-color: rgba(255, 255, 255, 0);">We
discovered the hard way this means users can run 4 jobs per
each of the 56 CPUs/threads on each node. Oops! Not what we
intended.<br>
<br>
</span></div>
<div><span style="background-color: rgba(255, 255, 255, 0);">All
our other compute nodes are defined as exclusive, and we
don't allow multiple jobs to run on them.<br>
</span></div>
<div><span style="background-color: rgba(255, 255, 255, 0);"><br>
</span></div>
<div><span style="background-color: rgba(255, 255, 255, 0);">Any
recommendations how to implement the 8 jobs per partition
per node limit we'd like? Should we switch our
SelectTypeParameters to CR_Socket or CR_Socket_Memory, for
example?<br>
</span></div>
<div><span style="background-color: rgba(255, 255, 255, 0);"><br>
</span></div>
<div><span style="background-color: rgba(255, 255, 255, 0);">-- <br>
</span>
<div class="gmail_signature">
<div dir="ltr"><span style="background-color: rgba(255, 255,
255, 0);">Regards,<br>
-liam<br>
<br>
-There are uncountably more irrational fears than
rational ones. -P. Dolan<br>
Liam Forbes <a href="mailto:loforbes@alaska.edu"
target="_blank" moz-do-not-send="true">loforbes@alaska.edu</a> ph: <a
href="tel:907-450-8618" dir="ltr"
x-apple-data-detectors="true"
x-apple-data-detectors-type="telephone"
x-apple-data-detectors-result="2"
moz-do-not-send="true">907-450-8618</a> fax: <a
href="tel:907-450-8601" dir="ltr"
x-apple-data-detectors="true"
x-apple-data-detectors-type="telephone"
x-apple-data-detectors-result="3"
moz-do-not-send="true">907-450-8601</a><br>
UAF Research Computing Systems Senior HPC Engineer
CISSP</span></div>
</div>
</div>
</div>
</blockquote>
<br>
</body>
</html>