Network Interface Failure

The HACMP software handles failures of network interfaces on which a service IP label is configured. Types of such failures are:
•Out of two network interfaces configured on the same HACMP node and network, the network interface with a service IP label fails, but an additional “backup” network interface card remains available. In this case, the Cluster Manager removes the service IP label from the failed network interface, and recovers it, via IP aliasing, on the “backup” network interface. Such a network interface failure is transparent to you except for a small delay while the system reconfigures the network interface on the node.
•Out of two network interfaces configured on a node, an additional or a “backup” network interface fails, but the network interface with a service IP label configured on it remains available. In this case, the Cluster Manager detects a (backup) network interface failure, logs the event, and sends a message to the system console. The application continues to be highly available. If you want additional processing, you can customize the processing for this event.
•If the service IP label that is part of a resource group cannot be recovered on a local node, HACMP moves the resource group with the associated IP label to another node, using IP aliasing as the mechanism to recover the associated service IP label.