Try booting with the 'Fallback' kernel and look for clues in the boot messages.

Type fallback or fallback-serial at the boot: prompt.

If you are installing to disks that previously had been used in RAID configurations, you may get a kernel panic (hang on boot).

In the case of Intel (integrated) RAID:

1. Configure the SATA in the host BIOS to RAID mode
2. Enter the RAID configuration (Control-I)
3. Delete the RAID volumes
4. If the RAID configuration (metadata) is 'messed up'
   1. Recreate a RAID volume
   2. Reboot and go back into the RAID configuration (Control-I) again
   3. Delete the RAID volume just created
   4. Reboot again
5. Install XenServer normally

You probably won't see the kernel panic (hang on boot) until after installing SP1 (XenServer 6.2).

http://xenserver.org/component/easyblog/entry/xenserver-root-disk-maintenance.html?Itemid=179

Check disk space:

df -h | grep "/$" | head -n 1

Looking for files bigger than a given size (specify path and size):

find / -mount -type f -size +20M -exec ls -lh {} \; | awk '{ print $9 ": " $5 }'

You can pretty much ignore anythin in /var/run/sr-mount/ as it's not on the local filesystem.

Delete rotated log files:

find /var/log/ -type f -name *.gz -exec rm -f {} \;

Delete applied patches:

Works pool-wide if run on the pool master server.

for i in `xe patch-list --minimal|tr "," " "`;do xe patch-pool-clean uuid=$i;done

Delete 'orphaned' patches and backup patches:

Do not delete /var/patch/applied or its contents.

rm /var/patch/*-*-*-*-*

For true emergencies or for patch install problems, you can delete patch backups:

rm -rf /opt/xensource/patch-backup/*-*-*-*-*

On the Pool Masters, we mount an NFS backup share on /mnt/backup.

Unmount the backup share and delete any 'hidden' files:

umount /mnt/backup
rm -rf /mnt/backup/*
mount -a