First, let me say that in my dealings with ESX, a snapshot deletion process will rarely timeout if your snapshots are properly managed. Snapshots are meant to be taken before modifying the guest OS, and then removed within 2 or 3 days. Forgotten snapshots can cause a multitude of problems ranging from filling up your datastores to taking hours and hours to commit/delete. In some cases, snapshot removal can cause the hostd process of your ESX host to not respond, which may even disconnect it from virtual center temporarily.
It is also important to keep in mind that commiting multiple snapshots (hitting “delete all”) on multiple VMs will put considerable load on the service console of your host, as well as generate signifigant disk IO.
So, how do you monitor the removal of a snpashot once the task has timed out in virtual center? Easy…
When logged into an ESX server, change to the directory containing your virtual machines vmdk file (and ultimately snapshot delta). Enter the following command:
#watch “ls -Ghtu –full-time *.vmdk”
This will give you a real time view of wether or not the snapshot has been deleted. Once you no longer see the snapshot file, it’s gone!
I would also suggest either using a powershell script (via the vSphere Power CLI) or a tool like RVTools to identify any old snapshots.
