This is a nasty, nasty restriction on 9.11.1P6. I have a volume with a large footprint that's not tiering fast enough on a slow filer. One workaround that I've used in the past is to bounce the volume to the other head with -tiering-policy all to force the data to tier, and then reset the tiering policy after the move is done.
I tried a vol move today and got the error that the destination aggregate was too full with over 50TB free. vol move appears to not be checking that I've requested -tiering-policy all which would allow the volume to easily fit on the destination aggregate. Sigh...
#vol move bug (unnecessary restriction)
ocean jungle
vast cradle
Hmm. Intresting I'll have to check into this later.
vast cradle
@ocean jungle do you have the exact error so it is easier to find information on?
ocean jungle
@vast cradle
`[5/3/2023 10:24:27] cluster::*> vol move start -vserver svm -volume foo_dest -destination-aggregate aggr1 -tiering-policy all
Error: command failed: There is 56.03TB of available space on the aggregate aggr1 which is not enough to accommodate a volume.`
vast cradle
It's a hard limit where the aggregate needs 10% of space, but it should work.
What is the space guarantee of the volume?
ocean jungle
space guarantee is none. The destination 114TB aggregate is 50% full. I am guessing that it's looking at the footprint in the performance tier to make the decision.
` Feature Used Used%
-------------------------------- ---------- -----
Volume Data Footprint 69.89TB 61%
Footprint in Performance Tier 65.57TB 93%
Footprint in xxxx-fpool1 4.61TB 7%
Volume Guarantee 0B 0%
Flexible Volume Metadata 500.3GB 0%
Delayed Frees 288.7GB 0%
Total Footprint 70.66TB 62%`
The kb refers a move from a non-FP aggregate to an FP aggregate. In this case, both source and destination are FP. By the math, I should need 10% of about 70TB or 7TB. I have 56TB free so the command should have worked.