<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta content="text/html;charset=ISO-8859-1" http-equiv="Content-Type">
<title></title>
</head>
<body bgcolor="#ffffff" text="#000000">
Mark Hahn wrote:
<blockquote
cite="mid:Pine.LNX.4.64.0807311533200.20456@coffee.psychology.mcmaster.ca"
type="cite">
<blockquote type="cite">BTW< where a lot of people are jumping on
the "Get IPMI " bandwagon, I suggest getting PDUs with remote IP
controlled ports is more useful.
<br>
</blockquote>
<br>
the thing I don't like about controlled PDUs is that they're pretty
<br>
harsh - don't you expect a higher failure rate of node PSUs if you go
yanking the power this way?
<br>
</blockquote>
Why?<br>
If nodes shutdown, on commands from the scheduler, that is good.<br>
And, if they do not, how is cutting power by the PDU socket any
different than a power switch on the node?<br>
Obviously we want to avoid "dropping the hammer" on a mounted
filesystem, at least until it has its cache<br>
cleared. That is not hard to accomplish.<br>
<blockquote
cite="mid:Pine.LNX.4.64.0807311533200.20456@coffee.psychology.mcmaster.ca"
type="cite"><br>
I have only seen a handful of different IPMI interfaces, but they all
<br>
were reasonably reliable.
<br>
<br>
</blockquote>
I have used the Supermicro, Tyan, ASUS, and Dell, and they all had some
tendency to choke sometimes.<br>
The thing is, at the nominal cost of $50 to $100 per machine for BMC (
IPMI) cards, one can buy a couple of network controlled PDUs,<br>
with the thermal and humidity sensors.<br>
As you are likely to at least buy "dumb" PDUs, this means the typical
cost per node added by this is usually around<br>
$30 per node, resulting in a tidy savings.<br>
It also means you are "talking" tp only one device pre 10 to 30 nodes,
versus 10 to 30 BMC devices.<br>
<br>
Further, these IPMI cards typically "steal" a GbE port on the nodes.<br>
<br>
<br>
<blockquote
cite="mid:Pine.LNX.4.64.0807311533200.20456@coffee.psychology.mcmaster.ca"
type="cite">
<blockquote type="cite">If you set your machines BIOS to start on
power up, it is trivial to stop and start machines with the PD U power,
and that is definitely reliable.
<br>
</blockquote>
<br>
huh? we're talking about network-attached IPMI, which is fully
independent
<br>
of the controlled motherboard's bios. are you talking about those
hybrid systems where the IPMI controller shares an ethernet port with
the host?
<br>
or IPMI through a kernel driver?
<br>
<br>
</blockquote>
Either.<br>
Most share a port, some have dedicated ports on board.<br>
<br>
<blockquote
cite="mid:Pine.LNX.4.64.0807311533200.20456@coffee.psychology.mcmaster.ca"
type="cite">
<blockquote type="cite">Plus , with a lot of those PDUs you can add
thermal sensors and trigger power off on high temperature conditions.
<br>
</blockquote>
<br>
IPMI normally provides all the motherboard's sensors as well. it seems
like those are far more relevant than the temp of the PDU...
<br>
</blockquote>
I would rather monitor the room temperature at the racks, and shut the
whole works down in a hurry if something is wrong, such as air
conditioning failure.<br>
<br>
<blockquote
cite="mid:Pine.LNX.4.64.0807311533200.20456@coffee.psychology.mcmaster.ca"
type="cite">using lm_sensors is a poor substitute for IPMI.
<br>
</blockquote>
Yes, and no.<br>
For monitoring the temps and fans an such on nodes it is quite
sufficient.<br>
For power control it is useless, of course.<br>
<br>
<br>
<div class="moz-signature">-- <br>
<font color="blue">With our best regards,</font><br>
<br>
<font color="black">
<i><i>Maurice W. Hilarius Telephone: 01-780-456-9771</i><br>
<i>Hard Data Ltd. FAX: 01-780-456-9772</i><br>
<i>11060 - 166 Avenue <a class="moz-txt-link-abbreviated" href="mailto:email:maurice@harddata.com">email:maurice@harddata.com</a></i><br>
<i>Edmonton, AB, Canada <a class="moz-txt-link-freetext" href="http://www.harddata.com/">http://www.harddata.com/</a></i><br>
<i> T5X 1Y3</i><br>
</i></font>
</div>
</body>
</html>