Check for sufficient free space in the datastore prior to taking a vm snapshot

[wilrodriguez]

If you're in the situation where you can't run with Storage DRS turned on, you're very likely to find that there's not sufficient space on your datastore to keep a snapshot around during an activity as potentially likely to make a snapshot grow real big such as patching. So, it's probably a good idea to add some simple checks to the snapshot task to make sure that there's enough space for the snapshot to be created, and also that there's sufficient space to grow a bit during the course of patching.

[nmaludy]

Great idea!

I'm thinking of having a task that finds all of the VMs then performs the check you described.

Worst case scenario, the VM could grow to 2x the size of its original disks. This might not be ideal though because then all of your datastores will need to be, worst case, 50% consumed.

I'm thinking a more generalized approach could be used where we allow the user to specify an estimation factor the be used when calculating the required space. This would be a number >= 1.0.

For example, if i wanted to account for 2x the space i would use a factor of: 2.0

If i wanted to use something a little less i might specify a factor of: 1.5 or maybe 1.25

This would expect the VM to grow, but not double in size.

The other caveat here is that VMs and disks can potentially live on different datastores but the snapshot is always created on the datastore where the VMX file lives, need to take that into account.