Hello,
I'm seeing a lot of traffic going over my cluster network that may be causing stability issues with my cluster. Communication to nodes would be randomly lost and kicked out of the cluster; Access to fail over cluster manager would stop working on certain nodes due to rpc errors; CSV time out issues.
All member nodes have been fully patched with windows updates and recommended fail over cluster hotfixes.
I noticed on one of my nodes that a large number of network packets were going through a network interface that was designated for cluster heartbeat traffic only. About 7-10k packets/s read and write.
After further investigation and looking at some CSV perfmon counters, I noticed there was a significant amount of redirected WRITE IO occuring even though my CSV volumes were not in redirect mode. Does anybody have any thoughts as to why there is so much redirect IO, and why it's going over my cluster network?
Any help would be appreciated.
Kind Regards,
Norman Phengvath