Physical machine reboots without any errors in system logs


Not sure if this the correct forum to ask this question, since the issue is weird -

I have Ubuntu 9.04 running on 20 physical machines, each with kvm_intel module enabled. Virtualization of instances is managed through Eucalyptus cloud controller. Eucalyptus is configured to KVM hypervisor. Ubuntu 9.04 runs with KVM version 0.84

dpkg --list | grep kvm

ii kvm 1:84+dfsg-0ubuntu12.3 Full virtualization on i386 and amd64 hardwa

Since the deployment of the cluster, few of the physical machines get rebooted every few days without any errors in system logs or any kernel panic. There is no trace available to identify the cause of the issue.

Is there any debugging in KVM which can be enabled to get more details to identify whether its a problem with KVM ?