Thorough alarms

[khughes]

Last week our director wanted to enable the alarms configuration (which should've been done a while ago anyways). From our origional discussion, he just wanted to know when the host or a vm goes up or down so I configured the host/vm state alarms. Coming in this morning apparently he told my boss that over the weekend he decided he wanted very detailed and thorough alarms. In his mind he wants to see an alarm if a VM is restarted, if a VM is hung etc. etc..

I'm in the process of doing testing with Nagios for some extra monitoring but my origional thought is VMware doesn't offer that detailed of alarms. Is there a way to do this or extra software that might be able to do this? Seems a bit overkill to me but it's kind of you give a mouse a cookie and everyone knows the rest of the story.

[kjb007 · Accepted solution]

Your intial thoughts are pretty much right on. VMware does offer some monitoring capability, as well as alarms, but they are very basic. Also, in my experience, the alarms are not all that reliable. In the absence of any other monitoring tool, they do an ok job. But if you read in the forums, the alarms are very finicky. They'll get set off, and depending on how quickly the alarm condition has recovered, the alarm does not always clear. I've had to go back, disable, and then re-enable the alarm, to clear the condition, even though the vm state is not in alarm status.

As I said, you can do basic checks, memory/cpu usage, etc. But, if you have another product that does only monitoring and alerting, then use it to perform those tasks, and use the VC to do what it does best, which is manage the environment.

Hope that helps.

-KjB