So how do I kill a Virtual Machine?

I wrote this post, back in 2009, for ESX 3.5:
http://technoblog.org/2009/01/how-to-kill-a-stuck-or-hung-vm-vmware-esx-35/

Both the “vmware-cmd /.vmx stop hard” method, and the following methods work, sometimes.. :

1. sudo vm-support -x
2. sudo less -S /proc/vmware/vm//cpu/status
3. sudo /usr/lib/vmware/bin/vmkload_app -k 9

But, when both of those fail? Then what?
How to kill a stuck VM when everything else fails?

It’s a bit brutal, but it works, but be warned, you might corrupt your VM.

1. sudo bash (for simplicity)
2. ps -ax | grep 3. Find the first number to the left, in the string with the right VM. This is the PID.
4. kill -9

Brutal? Yes. Efficient? Yes.

If this for some reasons doesnt work (like me accidentaly deleting the .vmx file), you can try the
following:

1. Run the following command (to get a list of running VMs):
sudo vm-support -x

The output will look something like this:

VMware ESX Server Support Script 1.29

Available worlds to debug:

vmid=1126 vm-01
vmid=1151 vm-02
vmid=1272 vm-03
vmid=1291 vm-04
vmid=1150 vm-05
vmid=1420 vm-06
vmid=1433 vm-07

2. Then run the command (remember to replace the number with the “vmid” of your VM):

less -S /proc/vmware/vm/1433/cpu/status

Press the right arrow key.

In the right corner there should be som info about the “group”:
group
vm.1432

3. By running the following command, you can safely kill your VM without risking corrupting it
(remember to replace the number with your “group number”):
sudo /usr/lib/vmware/bin/vmkload_app -k 9 1432

4. If successful, you should see a message like this:
Warning: Jan 06 06:42:49.717: Sending signal ’9′ to world 1432.