I have a cluster of esx servers in a cluster managed by a virtual center server. The servers share two datastore hosted on two LUNS on a Clariion CX-300. One of the datastore became unavailable due to disk problems on the LUN. Needless to say, the vms on that datastore crashed because their virtual disks became unavailable.
Neither ESX nor virtual center sent any notification about loosing access to the store. How do u monitor you environment to proactively spot this as they happen and not after the fact
ESX does not have storage monitoring feature (in version 3.x).
I suggest to use EMC Navisphere and set the mail alarm there (is the last tab in the web interface).
Or enable SNMP on both SP and use ad external tool.
if these hosts are in a cluster, you need to verify that they have HA enabled - If you they did, the VMs would have been started on the other host.
Also, you can set some alerts up for your environmnt in vCentre - have a look at the 'alarms' tab. You may need to set some definitions. In addition, vCentre 4 has far more granular monitoring. The monitors in vCentre will allow you to sent emails ro SNMP traps on any issues.
Also, you can monitopr your ESX hosts using any other 3rd party topol For example, if you have HP servers and montior them with HPSIM, you can get HPSIM to monitor your ESX hosts - down to storage, using the CIM providers that come with the HP customised ESX install