<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta content="text/html;charset=ISO-8859-1" http-equiv="Content-Type">
</head>
<body bgcolor="#ffffff" text="#000000">
Thanks Perry,<br>
<br>
I would probably go with a separate file server.<br>
I have seen a lot of NAS on the market recently (i.e. Netgear ReadyNAS
NV+ 4Tb X-RAID)<br>
and I am wandering if they might be a good/bad choice (the price is
appealing...).<br>
On the other hand I do not want to see all may cluster idle waiting for
the file server.<br>
The w/r rate transfer for these NAS is ~30-40 MB/sec which probably is
a modest one.<br>
<br>
Do you have any recommendations for the file server? I think I need
lots of RAM and fast <br>
disks. Since SCSI are expensive, are SATA disks OK? Also, can I take
the advantage of multiple <br>
LAN ports on the file server? (I see the Dell PowerEdge SC1435 can
handle 4 NICs for example.)<br>
<br>
Since all my machines are from<br>
DELL maybe I can get a discount from them if I would go buying one file
server.<br>
<br>
<br>
Thanks,<br>
Vlad<br>
<br>
Perry E. Metzger wrote:
<blockquote cite="mid87wslt6lxq.fsf@snark.cb.piermont.com" type="cite">
<pre wrap="">Vlad Manea <a class="moz-txt-link-rfc2396E" href="mailto:vlad@geociencias.unam.mx"><vlad@geociencias.unam.mx></a> writes:
</pre>
<blockquote type="cite">
<pre wrap="">I have a question regarding the head node for a small cluster that I
would like to build:
which are the disadvantages of using the head node as a computation
node? Since I plan to have NFS on it, I guess there will be
problems with comunication which probably will slow down the whole
cluster, but I am not sure since this is my first attempt to built
one...
</pre>
</blockquote>
<pre wrap=""><!---->
As is always the case, "it depends". Your computational load might be
such that doing computation on the head speeds up the overall
computation by adding a needed CPU, or might be such that it slows it
down by robbing the whole cluster of needed file i/o performance. That
depends enormously on the particular task you're doing. The only real
way to know is to test it and see what happens.
In general, if you have enough machines, I don't think there is much
point in doing computation on the head -- it complicates matters and
if you have enough machines you're only adding a tiny percentage of
extra CPU -- but if you have a very small cluster and you need the
extra CPU and your tests say it is fine, then go for it. You'll have
to test to be sure.
Perry
</pre>
</blockquote>
<br>
<br>
<div class="moz-signature">-- <br>
<font family="Arial" color="darkblue" size="4"><b>Dr. Vlad Constantin
Manea</b></font><br>
<font family="Arial" color="darkblue" size="3"><b>Professor of
Geophysics</b></font><br>
<a href="http://www.geociencias.unam.mx/geodinamica">Computational
Geodynamics Lab.</a><br>
<font family="Arial" color="darkblue" size="2">Centro de Geociencias,<br>
Campus UNAM, Juriquilla,<br>
Blvd Juriquilla 3001,<br>
Juriquilla, Querétaro, 76230,<br>
México.<br>
phone: +52 55 5623 4104/ext.133<br>
fax: (55) 5623-4129<br>
<a class="moz-txt-link-freetext" href="http://www.geociencias.unam.mx/geodinamica">http://www.geociencias.unam.mx/geodinamica</a><br>
</font></div>
</body>
</html>