[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

RE: [Linux-cluster] GFS slower than NFS ???

I would stay away from blanket terms like GFS being slower than NFS or
vice versa. Undoubtedly, stat calls "should" perform better on NFS with
attribute caching enabled, but I would first go down the route out
insure that the iSCSI is pumping out as much io as it should before
getting too carried away on blaming GFS. 

Actually, more info on the setup would make it easier for us to make an
educated guess on what could be causing the problem, for instance, what
do you mean by "StorageWorks via 2GB Fiber - iSCSI" is that iscsi over a
fibre Ethernet device? That alone would be an odd setup.

- raf

-----Original Message-----
From: linux-cluster-bounces redhat com
[mailto:linux-cluster-bounces redhat com] On Behalf Of Robinson Maureira
Sent: Tuesday, July 10, 2007 8:23 AM
To: linux clustering
Subject: Re: [Linux-cluster] GFS slower than NFS ???

Hash: SHA1

Rainer Duffner wrote:
> GFS is supposed to have a smaller overhead, compared to NFS.
> However, I'm not sure this pays out in case a maildir-mailstorage is
> clustered.

In my personal experience using GFS on RHEL4 vs NFS, is that GFS
outperforms NFS on a mail system, both using maildir and mbox style

The email software we're using is CommuniGate Pro, which doesn't do any
locking at filesystem level.

Under heavy use, in a 5TB (split on 5 mountpoints) filesystem, we
experienced a drop on WIO from ~90% to 60% using maildir, and then to
~45% when we switched to mbox.

One important tip with GFS is to disable quota (noquota mount flag) if
you don't need it, it saves a good amount of resources.

Best regards,
Version: GnuPG v1.4.7 (GNU/Linux)
Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org


Linux-cluster mailing list
Linux-cluster redhat com

[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]