Opalis 6.3: Checkpoint Cleanup with Opalis and System Center Virtual Machine Manager

So now that you are provisioning VMs and checkpointing them regularly you may find that there is a sudden growth in your storage devices Smile. This is good, and it is important to manage VM sprawl with good processes that identify and remove old VMs. For checkpoints it is useful to have a rolling a checkpoint strategy. So if you have put your checkpointing on a schedule, you can also put your checkpoint cleanup on a schedule. This is a really simple workflow that will keep your checkpoints from growing out of control.

image

Every 7 Days is a Monitor Date/Time activity that runs every 7 days to kick off the workflow.

image

The Calculate 90 days prior activity takes todays date and subtracts 90 days from it. The activity uses the Object End Time of the Every 7 Days activity as its reference point.

image

Now that we have the date, we can use the Get Checkpoint activity to Get Checkpoints older than 90 days by using the returned date from the Calculate 90 days prior.

image

And then the Manage Checkpoint finishes the workflow off and will Remove the Checkpoints that were retrieved by the Get Checkpoints older than 90 days.

image

This is a compact workflow that will keep your storage clean and well maintained.