Gluster users - Nov 2011 - small cluster question

Hi Folks,

I'm in the process of expanding a 2-node, high-availability cluster to 4 
nodes.  Of course that means that my current approach to mirroring data 
(DRBD) breaks, and I need to use either a SAN or a cluster file system - 
and GlusterFS sure looks like it might fit the bill.

Two question:

1. I'm running a hypervisor (Xen) on each node, and my goal is to 
support VM migration for both load leveling and failover.  As I 
understand it, earlier versions of Gluster weren't particularly friendly 
to VMs - with self-healing hanging client access.  Am I correct in 
understanding that newer releases don't have this problem?

2. It looks like the standard Gluster configuration separates storage 
bricks from client (compute) nodes.  Is it feasible to run virtual 
machines on the same servers that are hosting storage?  (I'm working 
with 4 multi-core servers, each with 4 large drives attached - I'm not 
really in a position to split things up.)

Any comments, advice, suggestions are most welcome.

Thanks very much,

Miles Fidelman

-- 
In theory, there is no difference between theory and practice.
In<fnord>  practice, there is.   .... Yogi Berra

Hi,

I can answer point 1.  GlusterFS 3.3 (still in beta), does finer locking during
self-heal, which is what the VM images like.

Gerald

----- Original Message -----
> From: "Miles Fidelman" <mfidelman at meetinghouse.net>
> To: gluster-users at gluster.org
> Sent: Tuesday, November 1, 2011 6:31:35 PM
> Subject: [Gluster-users] small cluster question
> 
> Hi Folks,
> 
> I'm in the process of expanding a 2-node, high-availability cluster
> to 4
> nodes.  Of course that means that my current approach to mirroring
> data
> (DRBD) breaks, and I need to use either a SAN or a cluster file
> system -
> and GlusterFS sure looks like it might fit the bill.
> 
> Two question:
> 
> 1. I'm running a hypervisor (Xen) on each node, and my goal is to
> support VM migration for both load leveling and failover.  As I
> understand it, earlier versions of Gluster weren't particularly
> friendly
> to VMs - with self-healing hanging client access.  Am I correct in
> understanding that newer releases don't have this problem?
> 
> 2. It looks like the standard Gluster configuration separates storage
> bricks from client (compute) nodes.  Is it feasible to run virtual
> machines on the same servers that are hosting storage?  (I'm working
> with 4 multi-core servers, each with 4 large drives attached - I'm
> not
> really in a position to split things up.)
> 
> Any comments, advice, suggestions are most welcome.
> 
> Thanks very much,
> 
> Miles Fidelman
> 
> --
> In theory, there is no difference between theory and practice.
> In<fnord>  practice, there is.   .... Yogi Berra
> 
> 
> _______________________________________________
> Gluster-users mailing list
> Gluster-users at gluster.org
> http://gluster.org/cgi-bin/mailman/listinfo/gluster-users
>