DX NetOps

I'm looking for information about the causes of the VMWARE VCENTER SERVER PROXY LOST alarm, beyond what is included in the alarm information.  We're getting these alarms, and the teams in charge of the VMware environment say there is nothing wrong.  It seems like the alarms are coming in some time after the VM has been moved to a new ESX host.  The time between vmotion and the alarm though can be an hour or so.  The alarms also usually clear within 5 minutes, but sometimes take 10 minutes or so.  We don't have any access to the VMware environment, so I'd like to find out what might be going on behind the scenes to get this event/alarm.

Thanks

Steve Moore

You usually see these when contact is lost with the VC AIM that’s communicating with your vCenter server.  Make sure you VC AIM can still talk to vCenter and that Spectrum can talk to the server that’s running your VC AIM.

As a side note, I exclude that alert from all my SANM filters EXCEPT the one that notifies the Spectrum admin.  These tend to show up for VMs that are being monitored so they trigger the alerts for those VMs when this is not a symptom of anything that’s wrong with the VMs.  It’s a problem with communication along the information path from Spectrum to VC AIM to vCenter.

Thanks, your information is helpful, and your reason for excluding it from SANM notifications is something we're trying to deal with.

I somewhat confused though.  Our VNM is running on a standalone server, and communicating to vCenter.  We've got over 1000 VMs, but we only get this alarm on a few at a time.  Yesterday, 4/10/16, we didn't get any, and today, we got 17.  I've excluded the alarm from some SANM filters, but not all.  If we were having a communication issue between vCenter and the VNM that is running the VC AIM, I should expect to get this alarm for a lot more devices shouldn't I?

Again, thanks for your information, it is helpful.

Steve Moore

I have a similar scenario, with one VCAIM managing 4 vCenter servers, and very often (2 a day at least) we get these alarms for the whole VHM tree, and there's no real communication issue (as far as we could see). 

Is there any useful log to analyze?