as part of trying to improve the collection of logs during an ost failure, we would like to collect server side information (from CI node) on job's failures.
These are the common things that I think we need and would help in debugging:
1. libvirt log
2. qemu vms log
3. running processes