Hello all. I am working on a cluster where we will have two scheduling nodes; a primary and a failover. We want to run MON on the headnode and have it monitory services on the scheduling nodes; easy enough. However, there is a service that the scheduler runs that can only run on one of the nodes; if the service stop I need to try to restart it. If it doesn't restart, I need to try to start the service on the failover node and if it starts I need to force a failover (bring down the eth card). Anyhow, basically I need MON to check for the service on both machines, but if its running on one don't bother checking the other (or at least don't fire off an alert if its not running on the failover, since it shouldn't be).
Does that make any sense? Any ideas? Thanks, Mike
_______________________________________________ mon mailing list [EMAIL PROTECTED] http://linux.kernel.org/mailman/listinfo/mon
