I am currently configuring dependencies between Nagios services. Setting up one service depending on another is straight forward, but is it possible to have one service failing when all master services are failing?
My example as hand is a load balanser for failover. The load balancer will work correctly as long as any of its backends are functioning correctly.
Read the documentation on Monitoring server and host clusters, which describes I think exactly addresses your question. Also see the man page for the check_cluster plugin. The
check_cluster
plugin can be configured to raise an alert based on the status of groups of other hosts or services.