We're running vCenter 5.5 on a physical server and have most of the default alarms configured. Every few days (there is no set pattern) we get two alarms back to back but there does not appear to be anything wrong. The first is as follows:
[VMware vCenter - Alarm alarm.HostConnectionStateAlarm] alarm.HostConnectionStateAlarm changed status from Green to Red
Previous Status: Green
New Status: Red
([Red state Is equal to notResponding] AND [Red state Not equal to standBy])
Current values for metric/state:
State = Not responding AND State = Unknown
Alarm 'Host connection and power state' on fqhostname changed from Green to Red
The next is:
[VMware vCenter - Alarm alarm.HostConnectivityAlarm] Host fqhostname in TPS is not responding
Stateless event alarm
([Event alarm expression: Cannot connect host - incorrect Ccagent] OR [Event alarm expression: Cannot connect host - network error] OR [Event alarm expression: Cannot connect host - time-out] OR [Event alarm expression: Cannot connect host - time-out] OR [Event alarm expression: Host connection lost])
Host fqhostname in TPS is not responding
Now, I'm sure I can go in and disable these alarms but is that a good thing? I'd like to know if any of my hosts disconnect and are not online, but these "false" warnings are doing little more than desensitizing us to vCenter warnings.
Going into Tasks & Events on the vCenter server reveals that the host shows as disconnected for about 4 seconds and then is back online. which is what I'm assuming triggers the alarms. We do have other backup (I/O) operations running on this vCenter at various times and I guess it's possible that is causing an issue. Maybe. I did not see an option to add duration (> 5 seconds) to the alarms that are being triggered. I should also note that it is not always the same host or hosts that trigger these alarms.
Should I disable these alarms? Or is there a better way?