How HACMP Handles Network Failures on the Local Node

A local network failure occurs when all interfaces of a specific cluster network on a node fail. For example, if you have nodes A and B, and networks net1 and net2, and all interfaces of network net1 on node A fail, then a network_down event runs for net1 with node A as the event node. You can see this in the /tmp/hacmp.out file. This is also called a local network failure.
In this case, the Cluster Manager takes selective recovery actions for resource groups containing a service IP label connected to that network. The Cluster Manager attempts to recover only the resource groups affected by the local network failure event.