MCS Base Disks Not Getting Deleted On XenServer Storage After Catalog Updates

When updating the Windows 7 VDA images, the SR becomes full as some of the older images never get deleted. The 2008 R2 VDAs are fine. I have put the W7 VDA's on a different Storage Repository, It was on a SR all by itself, so I made a copy of the master image and put it onto the SR with our 2008 R2 VDA's but still ran into the same issue.

Manual removal of disks if LUN gets full. If LUN does not get full, customer needs to double the amount the space of number of targets being created. Since there is a chance of never getting the diff disks updated when system run out of space, it is recommended always to plan accordingly prior to deploy the catalog.??

When LUN runs out of space diff disks are not properly deleted / created --- There is not evidence coalescing fails on hypervisor. On the other hand coalescing seems to work just fine, however, leftovers are left behind.??

Diff Disks are properly deleted / coalesced in XenServer always when the overhead is observed. During the catalog upgrade there is a cleanup process for extra diff disks. Environment may not recover properly when LUN gets full even if system is let to sit for hours. Documentation should be updated to let customer know that overhead is required when upgrading the catalog.
For 27 targets with 40GB system spikes to????2384.9GB and drops down to 1270GB.