----- "Norman Gaywood" ngaywood@une.edu.au wrote:

...

If your guest's state is corrupted, you can't rely on its behavior.

For that reason, you should set up a watchdog and only rely on the panic behavior as a preliminary measure.

I figured that I had setup a watchdog with the hung_task_panic=1 guest parameter. Is there another way of setting up a watchdog in this case?

Internal watchdogs are seldom as effective as you need them to be. I meant an external one. There was talk in the past about a Xen watchdog driver, but I don't know where that ended up. (I use existing code that works for me, so I haven't kept up.)

There are various ways to set one up. You can do it on your own with a couple of scripts and xenstore, if you're so inclined. If not, there are a lot of existing examples and projects that will handle it for you. This one that uses QEMU and unfortunately only KVM is decent and should give you enough ideas on how to implement one: http://rwmj.wordpress.com/2010/03/03/what-is-a-watchdog/