VMware VVol snapshots on Nimble

VMware VVols on Nimble are great, but there are some stuff to be aware of to keep the array neat and tidy.
Unmanaged snapshots, these are basically all the snapshot not created by the current protection policy for the VM. If you change the protection policy all the existing snapshots will become unmanaged and have to be cleaned up manually. Which is good, wouldn’t like them to be automatically deleted…

I find CLI the easiest to manage this, so open a putty session to the array. Run the command: snap –list –unmanaged –all | awk ‘$4 == “Yes”‘
This lists all managed snapshots which are online, hopefully this is none.
To be safe you can swap it and list only the ones which are offline with no instead:
snap –list –unmanaged –all | awk ‘$4 == “No”‘
Copy the output to a txt-file and import it to Excel, then in a new sheet use the concatenate function; =CONCATENATE(“snap –delete “;snaps!B1;” –vol “;snaps!A1;)
Do any last minute removal of snapshots in the list which you don’t want to delete, and paste the lines back into CLI.
Avoid pasting more than about 50 lines at a time, at least my putty session has some hiccups then.

Some other umanaged snapshots you will probably encounter are on vvol configs:
after-create – these are automatically created by Nimble to aid in the recovery of a deleted VM. If you delete a VM using regular vCenter delete function it uses this snapshot to recover the vvol config.
before-delete – these are snapshots Nimble create when you delete VMs using their vCenter plugin. This config snapshot is done at the time of deletion
The after-create snapshot might be missing some changes you did to the VM config, so make it a habit to delete VMs using the plugin if you think you ever need to recover deleted VM before it is purged from the array.

Snapshots taken in vCenter will be named vvol-snapcoll.X on the array. So if you see any of these check first if the VM has an active snapshot in vCenter.
Be aware that if you do a snapshot in vCenter and have replication configured on the array the snapshot might be replicated to the downstream array. When the snapshot is removed in vCenter it does not remove it from the downstream array.
I had a backup solution which did a VM snapshot in vCenter and removed it afterwards, but the Nimble array managed to replicate it first. So I had lots of snapshots on the downstream array which I didn’t understand what triggered at first.
When deleting these snapshots you will need to add –force to the command since something external created those snapshots.

When hot adding memory of expanding a disk you will also get an unmanaged snapshot. I talked with support and there is a feature request to have a expiration timer on these.

Finally, if you feel courageous, this command deletes all unmanaged snapshots which are offline. Including the ones needed to undelete a VM.
snap –list –unmanaged –all | awk ‘{if (NR>4 && $4 == “No”) system(“snap –delete “$2” –vol “$1″; sleep 0.1”)}’

NimOS used here was 5.0.6.

There is also a thread on the community forum about this, https://community.hpe.com/t5/Array-Performance-and-Data/Delete-all-orphaned-snapshots-at-once/td-p/6983368

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s