I'm having an issue on an NFS volume where it's not reclaiming any space that i have deleted.
The amount of data i deleted was approximately 100GB worth of data.
That 134GB avail has not changed at all.
I deleted all the snapshots and it said there was about 100GB reclaimable.
I have not seen any of that space return to my volume.
Filesystem total used avail capacity Mounted on
/vol/nfs21/ 921GB 786GB 134GB 85% /vol/nfs21/
/vol/nfs21/.snapshot 102GB 0GB 102GB 0% /vol/nfs21/.snapshot
vol options nfs21
nosnap=off, nosnapdir=off, minra=off, no_atime_update=off, nvfail=off,
ignore_inconsistent=off, snapmirrored=off, create_ucode=on,
convert_ucode=off, maxdirsize=28835, schedsnapname=ordinal,
fs_size_fixed=off, compression=off, guarantee=volume, svo_enable=off,
svo_checksum=off, svo_allow_rman=off, svo_reject_errors=off,
no_i2p=off, fractional_reserve=100, extent=off, try_first=volume_grow,
read_realloc=off, snapshot_clone_dependency=off, nbu_archival_snap=off
The snapshot reserve should adjusted, the amount available in the volume won't change until you do.
But if i deleted ~ 100GB data and i also deleted 100GB of reclaimable snapshot space shouldn't some of that have gotten picked back up in the actual volume? that doesn't make sense that none of it gets reclaimed. it's highly unlikely the 100GB of data was using was deduplicated 100%
Deleted files in question - weren't they sparse ("thin") by any chance?
Also, can you check the df output once again to see whether it shows any different figures now?
Not likely they were sparse files since a bunch of it was assets like jpg files html files etc etc. there were also a couple databases which those possibly could've been but i still would've expected like at least 20% return of space.
And no i haven't seen the space change since doing a df again
They could be held in the snapshot space. If you delete the snapshot it takes time to complete the delete, it's considered a background task, check back in a few hours (depending what else is going on).
yeah so i let it sit for the day and it still didn't reclaim anything.
You have snap reserve. So any space in snapshots you deleted is accounted as part of snap reserve, not as part of user visible usable volume space (unless snapshots consumption exceeded snap reserve).
Отправлено с iPhone
03.01.2013, в 23:08, "james tran" <firstname.lastname@example.org<mailto:email@example.com>> написал(а):
Re: Volume space not returning after file deletes
created by james tran<https://communities.netapp.com/people/JAMES.TRAN> in Products & Solutions - View the full discussion<https://communities.netapp.com/message/97605#97605>
yeah so i let it sit for the day and it still didn't reclaim anything.
Reply to this message by replying to this email -or- go to the message on NetApp Community<https://communities.netapp.com/message/97605#97605>
Start a new discussion in Products & Solutions by email<mailto:firstname.lastname@example.org> or at NetApp Community<https://communities.netapp.com/choose-container.jspa?contentType=1&containerType=14&container=2068>
aborzenkov is correct - I understood your OP to say that you deleted about 100G from the filesystem, THEN deleted the snapshots - but if you just deleted 100G in snapshots, the live filesystem space won't reflect any change, and in that case you should have seen the snapshot space go from 100% to 0%.
i am having same issue. pls advise if adjusting the snap reserve would help reclaim the space in the user usage visible volume.
Awaiting your swift response.
Do you still have the df output from before you deleted the files, then before you deleted the snapshots? Do you have OnCommand (DFM) (or other) capacity graphs covering the period in question?
You say it said there was about 100G reclaimable - was that from the snap reclaimable command?
So it should've been approximately 100GB of data I was trying to reclaim on the filesystem by doing an rm –fr on an nfs mount
My assumption is by doing so that shoves it all into the snapshot
So then I deleted the snapshots which said the reclaimable was 80 some odd gigs.
I did notice the snapshot used went from pretty full down to 0 , which I kind of expected but I was also expecting some of the regular volume to get something back. And by something I mean like more than 0
Anyhow I'm pretty much done chasing down space so I'm just going to work around this by shifting some data or reducing my snapshot reserve as mentioned before.
Thanks for all the input anyhow
It's interesting - what you describe should be exactly what happened. It is pretty difficult to deconstruct after the fact though, without space graphs or a df at each point.
Let me see if I have all the facts here:
You have a volume with 921GB of data space and 102GB of snap reserve space. So you have a 1TB volume using 10% snap reserve.
You had 100GB of snapshots.
Those snapshots fit neatly into the 102GB snap reserve.
You deleted said 100GB of snapshots and didn't see the usable data space change.
The thing is, when you made the volume and set aside 10% for snap reserve, it removed that amount from your usable data space and reserved it just for snapshots. Think of it as stowing stuff below the floor tiles in the data center. The room is actually bigger than you see but you raised the floor so you could have a place to keep power cables and such. I can add and remove power cables all day just as long as I don't try to stuff more cables down there than I have crawl space. At that point, the cables would poke up through the floor tiles and start taking up space in the data center, lessening the amount of work space that I had. At that point, I could remove some of the cables and reclaim my old amount of work space in the data center.
Same thing with the snapshots. Snapshots will fill up snap reserved space first. If they run out of snap reserve space, they will start taking up data space and count against your % available. This wasn't the case with you as all the snapshots were still within the snap reserve. If you want to recover some data space, you will have to reduce the snap reserve from 10% to something lower, which will apply the difference to your % available. Keep in mind that you will still be writing snapshots. If you want to lower your snap reserve, you will need to adjust your snapshot retention requirements or you will have snapshots exceeding your snap reserve pretty quickly. Basically, you just lessened the amount of crawl space below the floor tiles in your data center but you still need to leave some space for necessary cables.
I have the same issue
I have the volume of 1 TB with snapshot reserve of 2 %.
My volume was full so I have deleted around 500 GB of data from volume, thin provisioning was not enable, which I have enabled it.
I thought when next time snapshot schedule run it update the volume and I can reclaim space but still the same.
Can someone help and explain me why I cannot see free space available.
To free space you need to delete all snapshots that keep "copy" of deleted file(s). Whether it happens "next time snapshot schedule runs" depends on which snapshots have these files and whether these snapshots are deleted when "snapshot schedule runs".