Linux heartbeat and long-haul networks

[Richard Pieri]

Does anyone have experience with Linux heartbeat over long-haul networks?  I'm looking at setting up a failover between two sites.  Specifically:

I have two Xen dom0 servers currently in the same rack, using DRBD-backed logical volumes for domU images.  They currently use a manual switchover procedure.  What I want to do is move one of the two dom0 servers to an off-site data center and set up an automatic failover for my domUs.  The latter should be easy with Pacemaker but my concern is heartbeat latency causing spurious failover.

--Rich P.