Possible misconfiguration in cluster.ascii (ServiceGuard).

[mpua]

Hi gurus ;)

I noticed a weird network configuration in one of our ServiceGuard clusters (11.18) over HP-UX 11.23.

The problem arises from a somewhat odd config in the cluster.ascii file, specifically in the Node Definition part. 

There are three network interfaces defined for both nodes in the cluster, with the nuance that the last "NETWORK INTERFACE" on each has got no "HEARTBEAT_IP" associated.:

NODE_NAME server1
NETWORK_INTERFACE lan4
HEARTBEAT_IP X.X.X.X
NETWORK_INTERFACE lan3
HEARTBEAT_IP X.X.X.X
NETWORK_INTERFACE lan1
# CLUSTER_LOCK_LUN
FIRST_CLUSTER_LOCK_PV /dev/dsk/c4t0d4

This is causing that LAN1 is up on the system with IP address "none". Should I remove the definition of such interface (lan1) from the cluster.ascii file?

lan1      1500 none            none            0       0     0       0     0

Regards.

[Duncan Edmonstone · Accepted solution]

No you probably shouldn't. I expect lan1 is in fact a standby interface for lan3 - i.e. if the lan3 NIC fails, the IP address on it will be moved to lan1 - that's how Serviceguard provides network resilience.

[mpua]

Ok, i was wondering if that was a normal ServiceGuard behaviour or the result of a bad configuration. I'll leave it like that then. Thanks!