Ocfs2 users - Mar 2011 - OCFS2 shared volume getting slow when you add more nodes

Hello,
I have a cluster with two nodes, with SLES10 as base system. First I powered
on one node, and the system is working just fine. Then, when a second node
was added, the performance came down pretty bad.

Any hints or ideas about this behaviour?

TIA,
M

-- 
Saludos,

Mauro Parra-Miranda
Consultor Senior Novell - mparra at novell.com
openSUSE Developer - mauro at openSUSE.org
BB PIN - 22600AE9
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
http://oss.oracle.com/pipermail/ocfs2-users/attachments/20110301/9493e2e5/attachment.html

Two things to check would be the latency in the interconnect
and the io thruput of the shared device.

Use ping -I to determine the latency of the interconnect. A gige
in our lab shows approx 150 microsecs.

IO thruput is trickier. One needs to hammer the disk concurrently
from both the nodes. See if the thruput falls as compared to just
one node.

On 02/28/2011 11:03 PM, Mauro Parra wrote:
> Hello,
> I have a cluster with two nodes, with SLES10 as base system. First I 
> powered on one node, and the system is working just fine. Then, when a 
> second node was added, the performance came down pretty bad.
>
> Any hints or ideas about this behaviour?
>
> TIA,
> M
>
> -- 
> Saludos,
>
> Mauro Parra-Miranda
> Consultor Senior Novell - mparra at novell.com <mailto:mparra at
novell.com>
> openSUSE Developer - mauro at openSUSE.org
> BB PIN - 22600AE9
>
>
> _______________________________________________
> Ocfs2-users mailing list
> Ocfs2-users at oss.oracle.com
> http://oss.oracle.com/mailman/listinfo/ocfs2-users
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
http://oss.oracle.com/pipermail/ocfs2-users/attachments/20110301/6aca1fec/attachment.html