Large snapshot copies filling up snapshot reserve

[reckonerx]

I have two NetApp AFF A400 clusters running version 9.11.1P9.

I have volumes on them being used for vSphere NFSv3 datastores which are in datastore clusters. For those volumes I have snapshot copies enabled with the default 5% snapshot reserve. I also have 'Automatically delete older snapshot copes' enabled.

I am constantly seeing alarms for 'Volume Snapshot Reserve Days Until Full' and 'Volume Snapshot Reserve Space Full.' Right now I have three volumes using 168%, 204%, and 237% of the snapshot reserve.

-

What could be causing these snapshots to be so huge? Just VMs with a lot of deltas? How can I investigate?

Why aren't snapshots being deleted even though I have 'Automatically delete older snapshot copes' enabled?

[Ontapforrum]

volume snapshot autodelete feature has few parameters which governs it. What is "Trigger" set to ?

Could you check yours and share it with us:
::> volume snapshot autodelete show -volume <volume_name>

[reckonerx]

Hello. The output is below. None of the snapshots are locked for SnapMirror or clones.

Enabled true
Commitment try
Trigger snap_reserve
Target Free Space 20%
Delete Order oldest_first
Defer Delete user_created
Defer Delete Prefix (not specified)
Destroy List none

[Ontapforrum]

You might be hitting a following BUG.

Workaround: Check and delete, if possible, the oldest and largest snapshots manually until the snapshot reserve is below 100%.

https://kb.netapp.com/?title=onprem/ontap/os/snap_reserve_snapshots_not_autodeleted_when_over_100%2525

Please review the above kb.

[reckonerx]

That seems to describe it, even though it says it's fixed in my version.

I'm considering adding some space to the volumes and then either...

Increasing the snapshot reserve and then changing the autodelete trigger to volume, or...

Setting the snapshot reserved to 0% and then changing the autodelete trigger to volume. I remember seeing that 0% isn't recommended for NFS vols though?

[Ontapforrum]

I think the recommended action would be to simply set the Trigger: volume

When the "volume" option is selected, Snapshot Autodelete will begin deleting snapshots when the volume reaches the threshold capacity AND the volume space reserved for snapshots is exceeded. If the space reserved for snapshots is zero, then exceeding the reserved snapshot space is always true and only the volume capacity threshold is used.

It's up to you if you want to keep snap-reserve or not, either case it is fine. Infact just stay with your existing figures, just change the Trigger to Volume.

Hence, setting the trigger to volume will do the job and will make sure volume does not run out of space.

https://kb.netapp.com/onprem/ontap/os/How_to_use_Snapshot_Autodelete