I have a two node cluster running SQL server, and am trying to set up CA's Unicenter monitoring product to monitor for a failover. I need to know what error in what log truly signifies that a node has failed, and that I need to check to see that it failed over. I can scrape any event log for any eventID, but I can't figure out which eventID really and truly means a node failure occured. The goal would be to get as few false alarms as possible. Any help would be appreciated.