Hello!
We have a 7-host cluster which has been working fine, but for the third time in 2 months a random host has gone "not responding" and the other hosts have HA Agent errors.
The HA Agent errors are resolved easily by "reconfigure HA.." option, but, the "not responding" host doesnt come back even after restarting the mgmt-vmware service, killing the relevant the processes or restarting the service console network service. A restart of the host itself is the only fix. The guests are fine but because they are disconnected we cant vmotion but we can RDP and shut down cleanly. (not ideal in healthcare environment).
A couple of questions; the log doesnt show anything obvious, but some pointers of what to look for would help.
i havent found any common factors to the 3 instances; one at night, one at weekend, one during the working week. no work being done to any host at the time. The host are in no way over-worked or at maximum resource.
alternatively has anyone seen similar symptoms and resolved or found the issue? is the HA Agent drop out coincidence or relevant?
My current thinking revolves around the network for the vmotion/service console/heart beat, but I am not sure if I am on the right track or how to test this.
The ESX Hosts are at 3.5 Update 4.
Many Thanks in advance.
