#no snapshots have been created by a volume snapshot policy.

1 messages · Page 1 of 1 (latest)

gray needle
#

We have a 35TB volume and 26TB been used now with no snap reserve configured . The problem is there are no any snapshots ever created since day 1. I created a small size of test volume using the same policy. It worked.

The available aggr size now has 21TB free which is less than 35Tb volume size and also less than 26TB used size. Could it be because the aggr free space is not large enough? Also, considering the aggr has been overcommitted by all other volumes.

Also, what logs I can look into to find details about why the snapshots were not created automatically? I don’t see such information from “event log show”

trail moth
#

what do you mean by "no snapshots are created"? did you check on the CLI/system manager? Or do you just not see the "previous versions" (or ~snapshot dir) in Windows? Did you check that the correct Snapshot Policy is actually attached to the volume? can you post volume show -volume xyz -instance and snapshot policy show -policy xyz -instance for the volume and policy in question?

gray needle
#

There are no any snapshots associated with the volume when I run “snapshot show “ on the volume. I am taking about the cluster not on the clients. A small size of volume with the same snapshot policy works. Can you please let me know your thoughts on my questions above? I am not in front of computer now, will post vol show later

trail moth
#

I would need to see the configuration to analyze. But if there are no snapshots created then either the volume has no policy attached, or the snapshot autodelete settings are totally out of whack and delete the snapshot every time it is created. There is also a per-volume "nosnap" option that could be the culprit but I can't think of anything that would set that (since it's not exposed in clustershell)

gray needle
#

I don't see any possible cases as you mentioned. But, please find the outputs you are interested in.

📎 message.txt
grave leaf
#

Check the "daily" cron schedule.

Also check if you have the snapshot-policy twice, once in the admin SVM and also in your data SVM. I think they can be named the same.

trail moth
#

that's what I thought too but you cannot create a policy with the same name as one already existing in the cluster... I just checked that (on 9.11 but it should be universal)

#

just out of curiosity, try node run -node * -command vol status volume-name and see if it mentions nosnap=on somewhere ...

gray needle
#

Well, as I said , a small volume with the same snapshot policy works fine

#

did your command, no, there is no such option called nosnap=on

#

Is there a log we can check the status on how the automatic snapshots got failed?

trail moth
#

usually this should be in the event log 🤷‍♂️ I have no clue why this doesn't work for you. what ontap version are you running?

gray needle
#

Yeah, already checked event log. 9.11 could be newer but I think it shouldn’t matter

#

The only thing left could be the free aggr space(21TB) maybe too small

grave leaf
#

Doesn't matter, creating a new snapshot doesn't need any newspace.
If you manually create a new snapshot on that volume is there any error message?

gray needle
#

So, I can manually create a snapshot

trail moth
gray needle
#

Correct

#

There is a kb talking about needing to restart something by support engineer. I don’t know if it can be a solution