VMWare ESX and a VM stuck in an inconsistent state
Last week I managed to put a VM into a state where it thought it was on and yet the host thought it was off, presumably by asking for it to be reset at an inconvenient time.I've managed to kill it today thanks to an article in the VMWare communities forums. The key thing I gleaned from this post was the existence of the following command to get the VMID:
Before running that however I ran vm-support -h and one of the options is:
So having run vm-support -x to get the VMID I ran the vm-support -X command. It asks a couple of questions:
The key thing here for me was question 3, because after about 5 minutes I got a notification from VirtualCentre that the VM had been switched off.YMMV.
vm-support -x
Before running that however I ran vm-support -h and one of the options is:
-X <wid> grab debug info for a hung VM
So having run vm-support -x to get the VMID I ran the vm-support -X command. It asks a couple of questions:
Can I include a screenshot of the VM xxxx? [y/n]:
Can I send an NMI (non-maskable interrupt) to the VM xxxx? This might crash the VM, but could aid in debugging [y/n]:
Can I send an ABORT to the VM xxx? This will crash the VM, but could aid in debugging [y/n]:
The key thing here for me was question 3, because after about 5 minutes I got a notification from VirtualCentre that the VM had been switched off.YMMV.
0 comments
Leave a comment...
To leave a comment on this posterous, please login by clicking one of the following.
