Killing hung/frozen or otherwise unresponsive VM:s

Information found on various places about how to stop a hung/frozen or otherwise unresponsive virtual machine.
Use any SSH-client to connect to the service console of the host running the VM.

Nice way:
List virtual machines and worlds.

vm-support -x


Run support script to shutdown/kill the machine and collect troubleshooting data.

vm-support -X worldid

Note that the parameter (-x -X) is case sensitive.

Other ways:
List virtual machines:

vmware-cmd -l


Try stopping softly

vmware-cmd /vmfs/volumes/datastorename/vmname/vmname.vmx stop trysoft


Try stopping hard

vmware-cmd /vmfs/volumes/datastorename/vmname/vmname.vmx stop hard


Since neither of those ever worked for me in a hung/frozen VM I’ve always had to resort to killing the VM:s process (PID) by hand.
Try:

ps auxfww | grep vmname
kill -9 PID

Where vmname is the name of your VM and PID is the process ID.

If you use the last method you may have to do some manual cleaning. I’ve had to remove .swp file from the VM:folder on a few occasions.

It is also a good thing to do a service mgmt-vmware restart when troubleshooting troublesome virtual machines.

Sysprep on Windows Server 2003 x64

OK,
I recently had to create Vmware ESX template (VI3, Virtual Center) of a Windows Server 2003 x64 machine.
Normally I would have used the ‘clone to template’ and then run a customization on the new machine. This option is not available for x64 OSs so I had to find another way.
I found a thread on VMTN and a fair solution (until customization for x64 is included in VC):

if you run the setupmgr that is included in the deploy.cab (on the x64 windows server cd), it will walk you through the creation of the file.

long story short (your situation may vary):
1. copy everything in that cab to c:\sysprep
2. run setupmgr- select sysprep setup, select your version, fully unattended
4. copy the sysprep.inf file somewhere secure (password is in it) and leave a copy in c:\sysprep (this folder is deleted during the 1st bootup)
5. run sysprep.exe from c:\sysprep – select reseal, leave the defaults
6. once the system shuts down, convert to template

Smart Array 532 with VMware ESX 2.5.3

Problem encountered with a HP Smart Array 532 controller with VMware ESX 2.5.3 when installed in an already configured server.
The controller was found by the service console, but not configured in modules.conf and vmmodules.conf.

* Installed an additional RAID controller.
* LSPCI found the controller, but no disks showed in VM Manager.
* Ran KUDZU and found the controller, but still no luck in VM Manager.
* Added a new CCISS in /etc/modules.conf and in /etc/vmware/vmmodules.conf.
* Rebooted the server and was able to use the added disk.