VMware VM Occupancy

Sam Kirchoff

VM Occupancy is the actual amount of data the Operating System believes it’s managing in relationship to the actual storage the VM is taking up in the VMware Datastore.

It is related to VM Whitespace and Negative VM Whitespace values.

This concept is specific to VMware due to the availability of VMTools which reports the perspective of the OS in an “Inside-Out” fashion.  

virtualdisk.JPG

Virtual Disk Size:
Is the VM container size provisioned by VMware. In other words, “how thick CAN this VM become.” Unless the VM is thick provisioned to start with, this quantity is not yet using any capacity. It just has to right to take this much capacity as the VM “thickens.”

Virtual Disk Used:
This is the actual capacity VMware thinks this VM is consuming. In other words, “how thick HAS it become.”

Guest VM Disk Capacity:
This is the VMTools/OS perspective of provisioned capacity. This is what the Guest OS running in the VM thinks it has for capacity. This may be lower but is often equal to Virtual Disk Used. When this value increase, the VM becomes “thicker.”

Guest VM Disk Used:
This is the VMTools/OS perspective of the used capacity. This is what the Guest OS running in the VM thinks it has for actual data capacity. This will be the most accurately reported value for used capacity.

Example:

Virtual Disk Capacity

100GBs

Virtual Disk Used

75GBs

Guest VM Disk Capacity

73GBs

Guest VM Disk Used

25GBs

 

In this scenario 25GBs/75GBs = .33 or in other words, the actual data in the VM divided by the actual space the VM is taking from the storage pool.

This shows us that we have a 33% VM Occupancy.

If you are in a license by capacity scenario per VM, this is the value you want to use.

If you are sizing a migration of this VM to a different storage provider, then you must consider its entire size.

You may also consider space optimization maintenance in the form of “re-thinning” the VM for potential space reclamation.

Was this article helpful?
5 out of 5 found this helpful
Article is closed for comments.