Random Cluster errors

[Oaktech]

I've got a 2 node HA cluster, and I've started getting error 1126 and 1129

 

1126:

Cluster network interface 'SVR-CLS-01 - vEthernet (New Virtual Switch)' for cluster node 'SVR-CLS-01' on network 'Cluster Network 1' is unreachable by at least one other cluster node attached to the network. The failover cluster was not able to determine the location of the failure. Run the Validate a Configuration wizard to check your network configuration. If the condition persists, check for hardware or software errors related to the network adapter. Also check for failures in any other network components to which the node is connected such as hubs, switches, or bridges.

 

1129:

Cluster network 'Cluster Network 1' is partitioned. Some attached failover cluster nodes cannot communicate with each other over the network. The failover cluster was not able to determine the location of the failure. Run the Validate a Configuration wizard to check your network configuration. If the condition persists, check for hardware or software errors related to the network adapter. Also check for failures in any other network components to which the node is connected such as hubs, switches, or bridges.

 

The errors appear at random times, they don't appear to be load related as they sometimes occur in the middle of the night, they only appear 1 or 2 at a time, and only ever for a second or so and then may not occur again for several days. They appear simultaneously, to the nearest millisecond on both hosts. If I run the cluster validation tests for the network it always succeeds. No-one has ever noticed an issue, but I'm not happy that it appears at all.

 

The two servers are connected to the network on 4x1gb static NIC teams to 2x HP 1920-48 switches which are in turn connected together with a 6x1gb LACP group. There are very few errors on the switches and the uptime on the switches is >60 days.

 

Any ideas?

Edited March 15, 2017 by Oaktech