Dear all,
I would like to submit to your attention a warning that I can't fix in our two-node Windows Server 2012 Failover cluster :
We have 6 "NICs" per node:
- 2 NICs for iSCSI (PT0/1; 10.0.0.x and 10.0.2.x; intracluster communication not allowed)
- 2 teamed NICs and its relative vSwitch for intradomain communication (CIGS; internal allowed, clients can connect through this network);
- 1 teamed NIC (for possible further deployment) connected to the other node with a crossover cable for intracluster communication (Cluster; 192.168.1.x, IPv6 enabled; internal allowed, clientscannot connect through this network);
- 1 teamed NIC and its relative vSwitch for heartbeat and Hyper-V replica traffic towards a 3rd hypervisor which is not clustered (HB; 192.168.2.x, IPv6 enabled; internal allowed, clients can connect through this network).
Name State Role AutoMetric Metric
---- ----- ---- ---------- ------
CIGS Up 3 False 21000
Cluster Up 1 False 1000
HB Up 3 False 1100
iSCSI_0 Up 0 False 1200
iSCSI_2 Up 0 False 1300
According to cluster validation, the output states that, in current configuration, we have a single point of failure. Actually, AFAICT, we granted communication to the cluster to all the networks but the iSCSI one, although the validation report states that Cluster and Heartbeat networks are disable... all pings requests ended up properly.
Thanks for your kind attention, please find the validation report enclosed:
Validate Network Communication
includes network interface HYPERV1.xxx - vEthernet (vSwitch_HB). This network interface may be a single point of
failure for communication within the cluster. Please verify that this network interface is highly available or consider adding additional networks or network interfaces to the cluster.