We had to shut our 2 ESX servers down last night to move them. We got the recommend Starwind shutdown procedure from tech support. We performed that as given. An hour later when we powered the servers on according to the recommended procedures, which is to bring the Priority: First server online first, bu only 2 of 3 of our vmware datastores came back online. We waited for everything to sync successfully in case that would fix the issue, but it still didn't bring the 3rd one back online even though it said synchronized about an hour later. Multi-pathing shows both paths in ESX. I can see the Starwind device that has the data for this 3rd target under devices in storage adapters but it doesn't show a VMFS partition. It's like it's a brand new drive to ESX.
Looking more into the Starwind logs and interface I notice something very disturbing. On Starwind server #1 in Statistics it says User Data Size 780.05GB and on Starwind server #2 it says User Data Size 444.83GB. It also says Synchronization Status as Synchronized. How can they be synchronized if almost half of the data is missing? The other 2 have identical values.
How can such a controlled and guided reboot process go so horribly wrong? I have over 1TB free on both servers so I didn't run out of space. The directory with the LSFS files is also different sizes.
I was up until 2:30AM on Friday trying to fix this. It's Saturday and I opened a Starwind case and nobody is around. I'm now trying to fix this on my own and I fear there is nothing I can do to get my data back.
The Latest Gartner® Magic Quadrant™Hyperconverged Infrastructure Software