[NTLUG:Discuss] High Availability Storage Solution

[George Lass]

----- Original Message ----

From: Wayne Walker <wwalker at solid-constructs.com>
To: NTLUG Discussion List <discuss at ntlug.org>
Sent: Wed, March 2, 2011 7:38:24 PM
Subject: Re: [NTLUG:Discuss] High Availability Storage Solution

On Wed, Mar 02, 2011 at 10:30:12AM -0800, George Lass wrote:
> Greetings,
> 
> I am trying to figure out an inexpensive storage solution to support the needs
> of 2 different high availability clusters.  One of the HA clusters will be 
> running
> Suse Enterprise Linux 11, and the other CentOS 5.x. I have been reading up
> on CentOS clustering, and GFS and GNBD with directly connected storage:
> 
>http://www.centos.org/docs/5/html/5.2/Cluster_Suite_Overview/s2-ov-economy-CSO.html#fig-gfs-gnbd-storage-CSO
>O

You could also look into DRBD, but it sounds like GFS and GNBD may be
better for your solution.

-- 

Wayne,

DRBD might just be what I need for my CentOS HA cluster.  It will be running a 
couple of
different web servers as well as 2 custom real time java serverlet apps with 
some mysql
access requirements.  From what I can tell, I can run DRBD on the same 2 
physical
machines that will make up that cluster, saving me the expense of a separate set 
of
machines for the storage cluster.

The Suse machines will be running Oracle, and we may just resort to Oracle's 
native
redundancy methodology.

I'm still trying to figure out if GFS and GNBD using directly connected storage 
does mirroring
across nodes, so if any one knows anything about that, please let me know.

Thanks for the tip,

George