<html>
<head>
<meta content="text/html; charset=utf-8" http-equiv="Content-Type">
</head>
<body text="#000000" bgcolor="#FFFFFF">
<div class="moz-cite-prefix">It makes perfect sense. My BG/P nodes
don't swap and only have 512 MB/core. But before I do it myself, I
want to see if anyone else did this with negative effects. <br>
<br>
Truly diskless nodes would have to be swapless, but a lot of
people use 'diskless' to mean no OS installed locally, but there's
still a disk in the node for local scratch or swap. <br>
<br>
I did some googling searching this afternoon, and I saw a lot of
people recommending it for HPC nodes, but it sounded like they
were talking based on theory rather than actual practice which is
why I'm looking for actual practitioners. <br>
<br>
One person already replied off list and says he does this without
any issues. <br>
<pre class="moz-signature" cols="72">Prentice</pre>
On 02/06/2015 05:38 PM, Novosielski, Ryan wrote:<br>
</div>
<blockquote
cite="mid:27177836-0668-4CAA-9FFD-0016C8CC1605@ca.rutgers.edu"
type="cite">
<meta http-equiv="content-type" content="text/html; charset=utf-8">
<div>It does seem like a good idea on the face of it. I would have
to check if we do that, but I know that the jobs we tend to run
don't happen to come close to using all of the physical memory
on the machines anyway, so there ought not really be an impact,
I'd think. <br>
<br>
<span style="background-color: rgba(255, 255, 255, 0);">____
*Note: UMDNJ is now Rutgers-Biomedical and Health Sciences*<br>
|| \\UTGERS
|---------------------*O*---------------------<br>
||_// Biomedical | Ryan Novosielski - Senior Technologist<br>
|| \\ and Health | <a moz-do-not-send="true"
href="mailto:novosirj@rutgers.edu"
x-apple-data-detectors="true"
x-apple-data-detectors-type="link"
x-apple-data-detectors-result="3">novosirj@rutgers.edu</a>-
973/972.0922 (2x0922)<br>
|| \\ Sciences | OIRT/High Perf & Res Comp - MSB C630,
Newark<br>
`'</span></div>
<div><br>
On Feb 6, 2015, at 17:34, Prentice Bisbal <<a
moz-do-not-send="true"
href="mailto:prentice.bisbal@rutgers.edu">prentice.bisbal@rutgers.edu</a>>
wrote:<br>
<br>
</div>
<blockquote type="cite">
<div><span>Do any of you disable swap on your compute nodes?</span><br>
<span></span><br>
<span>I brought this up in a presentation I gave last night on
HPC system </span><br>
<span>administration, but realized I never actually did this,
or know of </span><br>
<span>anyone who has. I would tweak the vm.overcommit_memory
setting, but </span><br>
<span>that's not the same as disabling swap altogether. I'd
like to try doing </span><br>
<span>this in the future, but I prefer to learn from someone
else's mistakes </span><br>
<span>first.</span><br>
<span></span><br>
<span></span><br>
<span>-- </span><br>
<span>Prentice</span><br>
<span>_______________________________________________</span><br>
<span>Beowulf mailing list, <a moz-do-not-send="true"
href="mailto:Beowulf@beowulf.org">Beowulf@beowulf.org</a>
sponsored by Penguin Computing</span><br>
<span>To change your subscription (digest mode or unsubscribe)
visit <a moz-do-not-send="true"
href="http://www.beowulf.org/mailman/listinfo/beowulf">http://www.beowulf.org/mailman/listinfo/beowulf</a></span><br>
</div>
</blockquote>
</blockquote>
<br>
</body>
</html>