[OpenAFS-devel] fileserver parameters

Sven Oehme oehmes@de.ibm.com
Mon, 20 Jun 2005 12:46:14 +0200


This is a multipart message in MIME format.
--=_alternative 003B26D2C1257026_=
Content-Type: text/plain; charset="US-ASCII"

Hi,

we run XFS and we use Kernel 2.6 , SMP and we have AFS Performance >40 
MB/sec over Gigabit Ethernet.

Sven




Niklas.Edmundsson@hpc2n.umu.se 
Sent by: openafs-devel-admin@openafs.org
20/06/2005 10:19

To
Roland Kuhn <rkuhn@e18.physik.tu-muenchen.de>
cc
openafs-devel@openafs.org
Subject
Re: [OpenAFS-devel] fileserver parameters






On Sat, 18 Jun 2005, Tom Keiser wrote:

>> So, can anybody comment on these numbers? Those are dual Opteron
>> boxes with enough RAM, so please make some suggestions what options I
>> should try to get more like the real performance of a fileserver...

What kernel? What underlying filesystem?

The performance figures you mentioned when talking directly to the 
raid, are they when doing IO on a filesystem that are similar to the 
AFS IO pattern (ie accessing a large number of small files)? Streaming 
IO figures are more or less uninteresting when doing AFS in our 
experience.

If using ext*, go for ext2 and boot the machine with SMP disabled and 
se what happens. This improved the situation somewhat for us, but this 
was on a 2.4 kernel tho (albeit 2.6 is said to still have issues wrt 
ext3 and SMP).

Filesystem-wise, xfs seems to be the best performing (and most robust) 
filesystem on Linux nowadays, but I don't know whether it's supported 
to have your /vice* on Linux-XFS...

/Nikke
-- 
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
  Niklas Edmundsson, Admin @ {acc,hpc2n}.umu.se     | nikke@hpc2n.umu.se
---------------------------------------------------------------------------
  All hope abandon, ye who enter messages here.
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
_______________________________________________
OpenAFS-devel mailing list
OpenAFS-devel@openafs.org
https://lists.openafs.org/mailman/listinfo/openafs-devel


--=_alternative 003B26D2C1257026_=
Content-Type: text/html; charset="US-ASCII"


<br><font size=2 face="sans-serif">Hi,</font>
<br>
<br><font size=2 face="sans-serif">we run XFS and we use Kernel 2.6 , SMP
and we have AFS Performance &gt;40 MB/sec over Gigabit Ethernet.</font>
<br>
<br><font size=2 face="sans-serif">Sven<br>
</font>
<br>
<br>
<br>
<table width=100%>
<tr valign=top>
<td width=40%><font size=1 face="sans-serif"><b>Niklas.Edmundsson@hpc2n.umu.se</b>
</font>
<br><font size=1 face="sans-serif">Sent by: openafs-devel-admin@openafs.org</font>
<p><font size=1 face="sans-serif">20/06/2005 10:19</font>
<td width=59%>
<table width=100%>
<tr>
<td>
<div align=right><font size=1 face="sans-serif">To</font></div>
<td valign=top><font size=1 face="sans-serif">Roland Kuhn &lt;rkuhn@e18.physik.tu-muenchen.de&gt;</font>
<tr>
<td>
<div align=right><font size=1 face="sans-serif">cc</font></div>
<td valign=top><font size=1 face="sans-serif">openafs-devel@openafs.org</font>
<tr>
<td>
<div align=right><font size=1 face="sans-serif">Subject</font></div>
<td valign=top><font size=1 face="sans-serif">Re: [OpenAFS-devel] fileserver
parameters</font></table>
<br>
<table>
<tr valign=top>
<td>
<td></table>
<br></table>
<br>
<br>
<br><font size=2><tt>On Sat, 18 Jun 2005, Tom Keiser wrote:<br>
<br>
&gt;&gt; So, can anybody comment on these numbers? Those are dual Opteron<br>
&gt;&gt; boxes with enough RAM, so please make some suggestions what options
I<br>
&gt;&gt; should try to get more like the real performance of a fileserver...<br>
<br>
What kernel? What underlying filesystem?<br>
<br>
The performance figures you mentioned when talking directly to the <br>
raid, are they when doing IO on a filesystem that are similar to the <br>
AFS IO pattern (ie accessing a large number of small files)? Streaming
<br>
IO figures are more or less uninteresting when doing AFS in our <br>
experience.<br>
<br>
If using ext*, go for ext2 and boot the machine with SMP disabled and <br>
se what happens. This improved the situation somewhat for us, but this
<br>
was on a 2.4 kernel tho (albeit 2.6 is said to still have issues wrt <br>
ext3 and SMP).<br>
<br>
Filesystem-wise, xfs seems to be the best performing (and most robust)
<br>
filesystem on Linux nowadays, but I don't know whether it's supported <br>
to have your /vice* on Linux-XFS...<br>
<br>
/Nikke<br>
-- <br>
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-<br>
 &nbsp;Niklas Edmundsson, Admin @ {acc,hpc2n}.umu.se &nbsp; &nbsp; | &nbsp;
&nbsp;nikke@hpc2n.umu.se<br>
---------------------------------------------------------------------------<br>
 &nbsp;All hope abandon, ye who enter messages here.<br>
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=<br>
_______________________________________________<br>
OpenAFS-devel mailing list<br>
OpenAFS-devel@openafs.org<br>
https://lists.openafs.org/mailman/listinfo/openafs-devel<br>
</tt></font>
<br>
--=_alternative 003B26D2C1257026_=--