Hey there!
Following scenario: We are running three seperate ESX 4.1 servers, all connected to a SAN via fibre channel. A week ago we restarted one ESX server, it has since lost the connection to one of our SAN-LUNs, while the other two worked fine with it. Yesterday we restarted all the ESX servers; now all three have lost their connection to this LUN.
The physical link (hardware) is still there and visible (e.g. via configuration, storage adapters). Rescanning or restarting won't bring the LUN back; manually adding it, the ESX ask to format it. Same problem with our testing servers (ESX 3.5).
Digging into the console, we found out: The LUN is formatted as "Linux", not as "vmfs", as it should be - see the attached screenshot.
Our assumption: While running, the ESX do not monitor the partition tables. When restarting, they reread the partition tables, now with this wrong entry.
We are pretty sure: Our data is still somewhere on that LUN.
What do we do, how do we rescue our data - any ideas? :/
TIA
Steff