Hi guys
We've been trying to determine a "silver bullet" for our HA clusters when it comes to high availability. We've encountered several different types of scenarios such as core network switches going down, interfaces that flap etc and whereas the default "Leave Powered On" option was then best suited in this regard. Other instances was brocade failures at the back of chassis'. Whereas some hosts in a cluster retained access to their storage and others did not - Leave powered on in this case did not help because the network was still available. Do you think if we chose "power off", that would have mitigated the event and HA would've kicked in?
In your opinion what is the best HA solution for Active/Active FC SAN storage and assuming that the network would be highly available at all times? Which configuration would be the "silver bullet"?
"leave powered on" on isolation response, with enabling VM monitoring but disabling the poll interval?
your response is much appreciated
thanks guys