Cache filling up all of a sudden, and unable to run mover

[faulty_lamp]

In the past few days I noticed that my cache has began filling up all of a sudden.  I haven't made any changes in the past few days that I would think would coincide with this.  It seems like a few containers are taking up quite a bit of space, mainly unifi-controller.

 

 

cd /mnt/cache/appdata

du -sh * | sort -rn

Using the above commands I can see that my appdata folder has grown to 322G.  Below is a snip of the folders in appdata.  Unifi-controller is the most concerning with 280G.  However a few others, sabnzbd - 22G, binhex-krusader - 8.4G, PlexMediaServer - 7.1G (may be due to metadata and thumbnails), are also fairly large.

 

 

I have attempted to run mover to try to free some initial space, but it seems to start and then stop immediately.  When running mover this is all that I see in the logs

emhttpd: shcmd (97): /usr/local/sbin/mover &> /dev/null &

 

I also noticed a few other errors related to space in the logs:
 

Nov  2 17:44:27 Tower kernel: blk_update_request: critical space allocation error, dev loop2, sector 39389248 op 0x1:(WRITE) flags 0x1800 phys_seg 105 prio class 0
Nov  2 17:44:27 Tower kernel: blk_update_request: critical space allocation error, dev loop2, sector 39913536 op 0x1:(WRITE) flags 0x1800 phys_seg 105 prio class 0
Nov  2 17:44:27 Tower kernel: BTRFS: error (device loop2) in btrfs_commit_transaction:2438: errno=-5 IO failure (Error while writing out transaction)
Nov  2 17:44:27 Tower kernel: BTRFS info (device loop2): forced readonly
Nov  2 17:44:27 Tower kernel: BTRFS warning (device loop2): Skipping commit of aborted transaction.
Nov  2 17:44:27 Tower kernel: BTRFS: error (device loop2) in cleanup_transaction:2011: errno=-5 IO failure
Nov  2 17:44:28 Tower kernel: docker0: port 7(veth89f500d) entered disabled state

 

At this point I'm unable to start any containers, or run mover.  I'm at a loss as to how to clear any space and unsure of what actually caused the cache to begin filling up in the first place.

 

Any thoughts or suggestions?

 

tower-diagnostics-20221102-1731.zip

[jmztaylor]

Plex usage seems about normal. 

 

As for sabnzbd, are you doing anything with the completed downloads?  If you are then sab folder should not be that large.  You can check completed and incomplete folders for sab and delete what you don't need.

 

Pihole seems quite large.  Might be log files.  Can safely delete those if not needed.

 

Unifi might be log files again.  I would poke around that appdata folder and see what is making it so large.

[trurl]

Plex will transcode to appdata by default so you have to specify a mapping for that. SAB must be downloading to appdata also so make sure you have that setup correctly.

 

You should stop anything writing to cache until you get this resolved or you may get corruption.

[faulty_lamp]

Here is how I have sabnzbd configured, does anything look incorrect?

 

 

[trurl]

2 minutes ago, faulty_lamp said:

sabnzbd configured

Doesn't look like you need or are using Host Path 3, shouldn't cause any problem though