The log partition /var/log is filling up quite quickly.
As soon as the partition is full - some functions are no longer working propperly and xenserver responses are quite slow on each operation you do.
If the host is a pool member - a reboot into that pool will fail.
As soon as you have space on that partition - the reboot is sucessful.
Easy to reproduce - just fill up /var/log and try some actions like VM move -> very slow; reboot the pool member -> not able to get back up.