This topic has been marked solved and closed to new posts due to inactivity. We hope you'll join the conversation by posting to an open topic or starting a new one.
I'm trying to work out why our backup NAS system is reporting such a huge discrepancy between our consumed space and our free space. As per the screenshot below, the consumed space adds up to about 17.6TB. However, we have 21.81TB in total, with only 655.73GB free. We could really use those missing terabytes.
Snapshots are off (Smart -> Never). I've rebooted it (which has been good for clearing up space consumed by aborted/failed backups in the past). I've also defragged and balanced it. I'm yet to do a recent scrub, but the description doesn't make it sound like it will help.
Interestingly, the primary NAS unit has more than twice as much space free, despite backing up via Rsync (with 'delete files' ticked) and storing an extra folder with significant amounts of data. That's still leaving several TB unaccounted for though, and for some reason it doesn't have an available or consumed data column for me to look at, despite both devices running 6.9.1.
I've discontinued the use of these NAS boxes and moved to custom fileservers. We were using about 20.5TB. No doubt recreating the volumes would have helped recover the missing 1-2TB.
How old is the volume? Is it possible that snapshots were turned on at some point? Turning them off doesn't delete existing snapshots; it just stops creating new ones.
Might be worth enabling ssh, and enter btrfs sub list //data - looking for any .snapshot folders in its output.
Ah, that is probably it. They are old volumes and snapshots would have been on at times in the past. I just assumed the GUI would show me if they still exist.
btrfs sub list //data
couldn't find a directory, so I changed it to
btrfs sub list //<nasvolumehere>
and there were 3 .snapshot paths listed. Also one .timemachine if that's relevant.
You should be able to delete them from the web UI (select snapshots from the share list, then select the share. Right-click on one or more snapshots, and you'll see a delete option).
btrfs sub delete <<pathname>> should work too. Just be careful to delete the snapshot, and not the share.
Space isn't freed up instantly, the datablocks are freed in the background.
Hmm, no dice unfortunately. No snapshots were visible in the GUI at any point. I deleted them through SSH, but no space was freed up at all, even after a reboot and balance.
FreeFileSync is reporting 20.7GB of files as I back up to a different file system, so maybe BTRFS is just a bit confused and under reporting space usage in its share breakdown?
Space is freed up in the background, so you might look again (now that the system has been up for a while).
Another approach is that you can make a backup of the shares, and then delete/recreate/restore them one at a time. Or of course start fresh - delete the volume, create a new one, and restore the data.
There's definitely no space freed up, but based on the backup I'm doing to the impending replacement system, even a fresh volume would probably just make it mostly report higher usage on each share.
That said, it will probably be worth trying that route anyway once I'm satisfied I have safe copies of the data. It will take around a week over a gigabit connection though, so this thread will be idle until then.
I've discontinued the use of these NAS boxes and moved to custom fileservers. We were using about 20.5TB. No doubt recreating the volumes would have helped recover the missing 1-2TB.
Space is freed up in the background, so you might look again (now that the system has been up for a while).
Another approach is that you can make a backup of the shares, and then delete/recreate/restore them one at a time. Or of course start fresh - delete the volume, create a new one, and restore the data.
