Setting up alert for losing NFS datastore connection?

[ace008]

Hi,

How do I setup an alert so that if a NFS mount point/datastore goes down, I would receive an email notification?  I have setup the "Cannot connect to storage alarm with send an email notification.  We brought down the datastores, but did not get an alert or email.  In tasks and events, I do see the event:

Lost connection to server (IP) mount
point (path) mounted as
2e0636b3-34724cfe-0000-000000000000
(datastore)
error

I would like to create an alert for this specific event.

Thanks,

Steven

[DSTAVERT]

You should be able to create a custom alarm. http://communities.vmware.com/docs/DOC-12145

[ace008]

How does one create a custom alarm?  I couldn't find the event trigger for the event based on the spreadsheet.

[DSTAVERT]

The PDF file is a good source of information. There is, if I remember, a bit of information about a storage failure and why it doesn't generate the expected response. You might consider loss of network connectivity to the NFS device or ??

[DSTAVERT]

You can use an event to trigger an alarm so if you still have the failure look into that. Make sure that the alarms are defined and the response is set at the appropriate level.

[ace008]

Okay, so I did create a custom alert and did find the corresponding Event ID in the excel spreadsheet.  I set the conditions of the trigger to event ID is equal to 420 for when the datastore goes down and gave it a status of warning.  I also set another one for when I get an event ID 421 when the connection restores and set that status to normal. Now my question is how does it know when the status is green/normal. I would imagine that normal means the datastore is connected/working, but how does it monitor that it is working? It only sends an email when going from green to yellow and yellow to green.

[DSTAVERT]

I don't know what mechanism is used. If you look through the logs you may be able to determine something. Is the alert doing what you need or is something still missing? I use separate monitoring tools on a physical machine to verify connectivity to hosts and NFS datastores and other services. In some cases it is just a simple ping but others a protocol specific scan.