CM 6.3 Heartbeat Setting

[crossmark76]

Hi Guys,

I have an intermittent problem where my two CM 6.3 VM servers interchange, I know the problem is down to backups/snapshot deletion on the VM platform. I don't really have the capacity to put the CM's on their own datastore so instead if its possible I what to decrease the sensitivity of the heartbeat between CM's as adding an extra heartbeat check or an extra second or two the check would stop these interchanges from happening.

Is this possible? and if so where would I set this?

Many Thanks

 

[kyle555]

Heck no. Stop snapshotting live virtual machines. There's no need.

Think about it. If you restore a snapshot, you're restoring the memory and state of call processing at that time. You might be able to restore a snapshot, but your PBX will puke something awful if you did.

 

[crossmark76]

sorry if I wasn't clear, it is the snapshotting of 'other' servers on the same datastore that is causing the problem, I don't snapshot the CM's themselves.

The snapshotting happens during veeam backups.

 

[dialtonebone]

It sounds like you have installed in an unsupported setup. Like others said, a snapshot for this is useless and I believe the CM 6.x manuals even mentioned it causes stability problems. I don't know of any way to decrease the sensitivity of the duplication link, but it literally wouldn't make sense too. You would have split brain or constant interchanges possibly.
Are these interchanges being causes by IPSI sanity timers? If so that is adjustable, but if it's simply the dup link getting upset you are probably out of luck. Someone else might be able to chime in with a better answer.

Could you carve out some space on the same SAN to make a new datastore that doesn' get snapshotted? Or make exceptions for VMs?