Tried looking around for more info and I might just be uninformed but atm got 2 nvmes sitting in cache, as btrfs, with data system and metadata converted to raid1 (done years ago)

 

I cleaned out my cache and noticed 250 gigs being in use despite /mnt/cache being empty.

 

looked around found 256gig partition (data) being mostly used up, this doesn't seem right? is there a way to reclaim this space? is it misconfigured somehow? and what will happen once it fills up with whatever data is in there?

 

Edited March 25, 20251 yr by Mizerka
cant word good

30 minutes ago, Mizerka said:

found 256gig partition (data) being mostly used up

Not sure what you mean by this, do you mean you've found what is using that space? Also, please post the diagnostics.

Ill post diags in a minute but yeah empty mnt cache, but 250giv are used of 1tb dual cache. Leaving 750 free. Only thing that lines up is this data used space found inside cache disk view. I'm not sure what it's doing

I should do more reading on btrfs tbh, cli output says its got the data across both disks but im none the wiser what its doing and why its almost full. cache seems to work fine otherwise btw;

 

Yeah I still dont get it haha, but it was a nice read. with the state it's in im considering just dropping cache and recreating it as zfs/xfs volume instead, seems like most people are running xfs. if I understand it, unraid takes no advantage from snapshots and its resiliency just relies on copy on write functions, probably why balancing is required sometimes. would be nice to get 2nd opinion though.

Edited March 25, 20251 yr by Mizerka

Something is using data space, if you don't know what it is, you can use the mover to move the share(s) to the array, for example, then move back one by one and check the used space after.

thanks, to clarify I ran mover, saw 300gigs used, it had some open files in the /mnt/cache, shut offending containers and moved the data out until /mnt/cache was empty, no files or folders, 250 gigs remained "used". that's how I got here, as I can see, this data RAID1 volume created by btrfs is using all this data but I cant figure out what its doing, best guess its some subvolume for btrfs functions but its odd its full and so big. subsequent mover just quit instantly.

 

Thoughts on removing cache pool and recreating in xfs? that seems general consensus from what I've read online.

Edited March 25, 20251 yr by Mizerka

19 minutes ago, Mizerka said:

Thoughts on removing cache pool and recreating in xfs?

Xfs only supports single drive pools.   Multi-drive pools have to be btrfs or ZFS.

yeah you're right, okay well what could go wrong, I'll get cache emptied again and get fs changed and see how it goes. And I'll just assume that brfs eating 25% of volume is black magic and forget about it. Cheers both.