I have split off a clone of a 30TB volume with 27TB used space of which 16TB is a LUN. It is on the backup system which is spinning disks, so not that fast... "volume clone split status" only shows the scanned and updated blocks counting up... the % complete is 0 after 8 hours... is there a way to see how far this process has to yet?
The reason I need to split it off, is that I need it inside another SVM, and you cannot relocate a cloned volume... 🙂
#Clone split progress?
terse forge
terse forge
..update.. we are at 1 pct. after 19 hours... 😦
dense panther
"Included in the output is a
% complete or processed field. This field is based on the count of the inodes split, rather than the count of the blocks split. The value is a whole number and rounded down. If the volume contains a few large files or LUNs, the % field might not be an accurate representation of progress."
next crest
It's quicker to vol move than split for next time fyi
terse forge
yeah.. it's at 2 pct. now.. so I should be able to just create a clone, and then make a vol copy of that volume?
...if it's faster, why do they even do the split the order way? doesn't make sense to me
terse forge
I have now created another clone, and started a "lun move" to another volume created on a SSD aggregate in hope that it will speed up things a but... but it has started up with a process of "Allocation-Map"... and 0% pct 😉
terse forge
...OK FYI, the lun move turns out not to be much faster... We have about 50 spinning disk in the aggregate and it's pegged at 100% util on the disks, so I guess I cannot ask much more from it 😉
next crest
I always do a vol move as it's basically a snapmirror then a cut over. Much quicker than a split. The split is only a good option if you don't have another aggr to vol move to. With a split or vol move your clone is independent of the parent anyway so set and forget. Obvs iops here would be an issue as you got 50 disks. Sata that's 4000 iops you got for that move split plus any other services using it. Do set adv and then statistics aggr show to see how your aggr is doing iops wise. If it's ssd we'll I would just do it and leave it run. Check after a few days.
terse forge
I still need the original volume online (because it's a snapmirror destination), but I guess I could have started a snapmirror... but then I would also mirror the snapshots, of which there are many 😉 So I think the lun move was the best of all evils here... it's at 43 pct. now so with a little luck it will be ready monday 😉