[prev in list] [next in list] [prev in thread] [next in thread] 

List:       drbd-user
Subject:    [DRBD-user] Best practices for DRBD in a VMware environment?
From:       Lars Kellogg-Stedman <lars () seas ! harvard ! edu>
Date:       2008-09-22 15:29:17
Message-ID: 146B3E1D-634B-4880-BB2A-A35EAE0A92DD () seas ! harvard ! edu
[Download RAW message or body]

Hello all,

We're using DRBD and Heartbeat to provide filesystem replication and  
failover in a VMware environment.  Heartbeat seems ill-suited to this  
environment, since there aren't many options for alternate  
communication paths (e.g., no serial connectivity), so no dopd.  A  
network interruption invariably ends up with DRBD in a split-brain  
situation, which means that failovers stemming from connectivity loss  
require manual intervention.  From my perspective, there are obvious  
solutions:

- Quorum disks.  This would work great, since we have shared storage,  
but this is tragically not implemented by heartbeat.

- STONITH via VMware.  This might work, but the vmware-vi STONITH  
module distributed with VMware is non-functional.

What are other folks doing?  Automatic split-brain recovery is  
probably our best option in the short term, but is anyone doing  
something more reliable?  What about "cman" and the associated tools  
-- I understand that this has support for quorum disks.  Does this  
offer a similar feature set to Hearbeat?

Thanks for your input,

-- Lars

_______________________________________________
drbd-user mailing list
drbd-user@lists.linbit.com
http://lists.linbit.com/mailman/listinfo/drbd-user
[prev in list] [next in list] [prev in thread] [next in thread]