Re: Linux Cluster File Systems

From: "Karl Vogel" <karl(dot)vogel(at)seagha(dot)com>
To: pgsql-admin(at)postgresql(dot)org
Subject: Re: Linux Cluster File Systems
Date: 2004-01-23 20:16:35
Message-ID: slrnc13097.v2o.kvo--gm@kvo.local.org
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-admin pgsql-performance

In article <40116DA1(dot)7030209(at)cranel(dot)com>, Greg Spiegelberg wrote:
> Anyone have good/bad experience using OpenGFS or the Sistina GFS?
> Any other cluster file systems out there?

Lustre and CFS spring to mind. CFS is the cluster filesystem layer
of openssi (http://www.openssi.org), which is layered on top of
ext3 and provides the necessary locking. It can work with shared
disk hardware, but I believe 1 node is always responsible for all
transactions. It can fail-over to another node in the cluster, if
that node has direct access to the disk.

You can also use it without shared disk hardware (you loose the HA
ofcourse). That way you can easily test openssi.

Postgresql should work without problems on an OpenSSI cluster.

I don't know much about Lustre (http://www.clusterfs.com)

In response to

Browse pgsql-admin by date

  From Date Subject
Next Message Andrew Sullivan 2004-01-23 20:50:56 Re: High Performance/High Reliability File system on SuSE64
Previous Message Joshua D. Drake 2004-01-23 19:05:41 Re: High Performance/High Reliability File system on SuSE64

Browse pgsql-performance by date

  From Date Subject
Next Message Andrew Sullivan 2004-01-23 20:50:56 Re: High Performance/High Reliability File system on SuSE64
Previous Message Joshua D. Drake 2004-01-23 19:05:41 Re: High Performance/High Reliability File system on SuSE64