I am running esxi 4.0.0, so apparently, I can't install plug ins to use the "maps" feature and see if anything is still pointing to a stuck datastore that won't delete.
The raid drives for the datastore, internal to the host, first lost one drive, then another, so the raid drive is degraded and the data has been lost. I am not trying to recover that data. All VM's have been recreated on an external NAS drive, using iscsi LUN, and aside from occasionally having that datastore disappear and then reappear after a reboot all is going well.
I have edited the settings of each VM I have, to make sure that none of the CD drives are pointing to an ISO on the stuck datastore. They have all been moved to ISO's of the iscsi datastore.
When I try to browse this datastore, there is one folder, but then vsphere disconnects as it tries to browse to the files. I suspect because the raid group is degraded or dead.
When I try to delete this datastore, I get the "resource vim.datastore is in use".
Since I am on 4.0.0 and the free version, i don't seem to have the CLI on the vsphere client.
I am not familiar with the CLI commands for esxi and only use the GUI, so need some help here.
Welcome to the Community,
basically you can access the CLI on the host even with the free edition, it's only the API which is limited. Anyway, you may run RVTools (http://www.robware.net/) to get an overview. Does any of the VM's have active snapshots which may have been created while there was a connection to the datastore in question?
André
There have been no snapshots. None of the VM's from that datastore even exist any more. They were all lost when the raid drive was lost. So there are no VM's in vsphere to check for ghost files belonging to that datastore. The only thing remaining in vsphere related to those old VM's, is the datastore, which is stuck. I am not trying to recover anything from this datastore. There are no VM's still alive from that datastore. I just want to kill that datastore.
Maybe there's still a VM registered somehow. You may check this by running the command vim-cmd vmsvc/getallvms from the ESXi host's Tech Support Mode console (see http://kb.vmware.com/kb/1003677 for how to login).
André
OK, that was fun. I'm using the CLI for the first time. All VM's point to my iscsi datastore, so there are no ghost vm's still pointing to the old datastore on the raid controller. is there a way to look at, browse, and/or kill that datastore from the cli?
Yes, you can view/browse the datastores in /vmfs/volumes/.... However, deleting the datastore from the command line may not work if the GUI doesn't allow it either.
André
i see pretty much the same as the vsphere gui. i have a datastore called datastore2, which was created on the raid drive that is internal to the host and that failed. it has one folder visible on it, called "xyz" for our purposes. on the gui, when i try to browse that folder, i get kicked out of vsphere altogether. can we try to remove that folder from datastore2 and then try to remove the datastore2 altogether?
Yon can certainly try to delete any files or folders on that datastore from the command line to see whether this helps.
André
Andre, I suspect that the underlying issue is that any attempts to read or write to the datastore, then are attempts to connect to a dead raid array, so it fails. can you tell me how to remove a folder xyz in datastore 2 from the command line? its worth a try.
