We have been having a particular problem with one of our hyper-v hosts, the production LAN goes into a suspect state and takes the server offline. This server is a member of a two node cluster and should failover all of the guests to the second node during this event. We have seen an issue where not all of the guests are failed over and they continue to run on the host that has no network comms (they are up and running but solarwinds marks them as down as they can't be 'pinged').
I have been asked to alert our on call response team if this event happens but not fire an alert if the host fails and all of the VMs move properly as this would be handled during normal business hours. I would like to configure my alert with something along these lines:
Host system status is equal to down AND (any guests on the host) status is equal to down then fire an alert.
I don't want to add all of the guest names to the alert as this adds a management overhead.
Any advice or pointers would be appreciated.