we have a serious problem after Update our cluster nodes to sp5.
After a few days or up to twice a day on cluster node stopps communicating,
the failover does not work correctly, because the IP address is still in use.

a manual failover a few minutes later works.

It seems that the cluster node is loosing it's primary ipaddress, but the secondary are still alive for a few moments.

So we downgraded the nodes back to Sp 4a and the server is now running for a week.

We are using an intel server board with dual NIC Intel Pro 1000 Server and iSCSI connections to NW 65 sp4 Storage Fileservers
with mirrored sbd and nss partitions.

Has anyone similary experiences ?

thanks for any advices.