Hi everyone,

I’m currently experiencing an issue with my Unraid setup (version 6.12.10). My 1TB cache drive is suddenly full, even though I only have Plex installed as a Docker container, which is using around 60GB of space. The following shares are set to use the cache:

appdata

domains

isos

system

Here's what I've tried so far:

Running Mover manually – but it didn’t free up any space.

Checked Plex cache and transcode directories – no abnormal sizes.

Reviewed Docker image size (20GB) – nothing unusual here.

VM Manager is disabled (no VMs).

I also used the du -h command on the /mnt/cache directory but couldn’t find any large files. I’ve checked Docker container sizes, and they seem normal (binhex-plexpass is around 2GB, others are minimal). I also ran balance and scrub on the cache drive with no success in recovering space.

Additionally, during file size calculations, my CPU and RAM were fully utilized, causing the server to become unresponsive until a hard reset. I've installed a 1TB cache drive to replace my original 250GB cache, but it is still filling up quickly, even though I’m not uploading any new files.

Does anyone have suggestions for what I can check next or how to free up space? Could there be hidden files or something else I’m missing?

Any help would be appreciated!

You are likely to get more informed feedback if you attach your system's diagnostics zip file to your next post in this thread.  
It is always a good idea to do this to allow us to see the current state of your system and so we can see logs.

Sure, thanks for the advice.plexmediaserver-diagnostics-20240927-1059.zip

Delete the trial.key from the flash drive, since it's spamming the log, /config folder, then go to Shares and click "Compute All", post a screenshot of the results.

8 hours ago, snowboardjoe said:

I'm experiencing similar issues

Since this thread is not resolved, please start your own, or it can get confusing.

Thanks, I have deleted the trial.key. Unfortunately, I can't compute the appdata folder size. I tried multiple times for two hours.

 

I checked every single folder within the appdata directory. There are four folders that can't be calculated. Three of them are located in user/appdata/binhex-plexpass/transcode, but when I open them, there's nothing inside. The fourth folder is called "8" and is located in user/appdata/binhex-plexpass/Plex Media Server/Media/localhost.

 

Unfortunately, there are 480 folders inside, and I don't want to check them all manually.

 

In user/appdata/binhex-plexpass/Plex Media Server/Media/localhost, there are folders named 0-9 and a-f, for a total of 16 folders, which coincides with the number of disks I have installed. Could there be any connection?

The usage the GUI is showing is correct, there's a lot of metadata, suggesting there are a lot of small files, if you cannot find which folder(s) are using the space, you can use the mover to move some shares to the array to find out what is using it, also just moving them to the array and back, may save you some space, since btrfs usage can grow with defragmentation.

Thanks for the help. I tried moving both the appdata and system folders to the array, but even after stopping all apps and running the mover, nothing was moved. I also attempted to use the mv and rsync commands in the terminal, but unfortunately, they didn't work either.

 

At this point, the only option I see is to back up my Plex media folder, remove and reinstall the cache drive, and then monitor whether the issue persists when the system starts writing to the cache again.

 

Any advice or suggestions on this would be greatly appreciated.

57 minutes ago, Lauren112 said:

but even after stopping all apps and running the mover, nothing was moved.

Enable the mover logging, run the mover, post the diagnostics.

Thanks, I have enabled mover logging and ran the mover again. I've attached the diagnostics ZIP file.

plexmediaserver-diagnostics-20240930-1230.zip

Sep 29 12:39:35 plexmediaserver kernel: SQUASHFS error: Failed to read block 0xb15bd4: -5
Sep 29 12:39:35 plexmediaserver kernel: SQUASHFS error: xz decompression failed, data probably corrupt
Sep 29 12:39:35 plexmediaserver kernel: SQUASHFS error: Failed to read block 0xb15bd4: -5
Sep 29 12:39:35 plexmediaserver kernel: SQUASHFS error: xz decompression failed, data probably corrupt

 

These error usually mean a flash drive problem, recommend replacing it, also make sure to delete the trial.key from there, it's corrupt and spamming the log.

I've installed a new USB drive, invoked the mover, and downloaded the diagnostics. Unfortunately, the mover still didn't move any files.

However, I did notice in the Main tab that it's reading at around 30MB/s.

plexmediaserver-diagnostics-20241001-1230.zip

What share are you trying to move? They need to be configured correctly, appdata for example isn't.

None of the shares appear to be configured correctly to tell mover to take action.

Yes, I'm trying to move appdata. The primary storage is currently set to the array, and I have not configured any secondary storage. The mover is set to "Not used" and is greyed out, so it can't be triggered.

 

Should I configure the cache as the primary storage and the array as secondary storage, and then invoke the mover? Or is there another configuration that would allow me to move appdata from the cache to the array?

 

Thanks for your help!

22 minutes ago, Lauren112 said:

Should I configure the cache as the primary storage and the array as secondary storage, and then invoke the mover?

Yes, check that mover action is from pool to array.

 

You also need to disable the docker service to move appdata.

It is worth pointing out that one typically wants the ‘system’ and ‘appdata’ shares to be completely on a pool to maximise performance and to allow array drives to be spun down.

Thanks, it's working now, and it managed to move about 40GB. However, for the last hour, it's only been reading/writing at around 500KB/s. Is that normal behavior, or could there be an issue slowing it down?

Appdata, and any other small files, will take a long time to move.

Since the mover still didn't move anything, I ended up erasing and reinstalling my cache drive. Now I'm encountering an issue while trying to delete Plex. The problem is that it seems impossible to open or delete a specific folder within the Plex directory under the media folder (/mnt/user0/appdata/binhex-plexpass/Plex Media Server/Media/localhost/8/3f9ce283d14fbf587252437b080cf6356bf99f3.bundle/Contents/Indexes/tmp).

 

I can't delete it via terminal or even uninstall Plex—this leads to a server error. When I browse to the folder via the Main tab, it just keeps loading indefinitely or won't show any files inside. Opening the folder through Windows Explorer takes several minutes, and when I finally see the contents, they are just .jpg files. However, attempting to delete even one picture results in extreme slowness, and eventually, an error stating that the file no longer exists.

Do you have any ideas on how I can delete this folder and its files?

plexmediaserver-diagnostics-20241002-1457.zip

Edited October 2, 20241 yr by Lauren112

Try rebooting, according to the diags, appdata share does not exist.

The folder is not located on the cache, but on Disk 14 in the array. I want to delete it so I can uninstall Plex and reinstall it on the cache. However, I can't delete the Plex folder at the moment. The only option I see is to invoke the mover, but I fear it might cause an infinite loop like it did yesterday.

I've already tried rebooting, but that didn't help. Does anyone have any other suggestions?

I finally managed to delete the files. Currently, the appdata, domains, isos, and system shares are located on the array. The primary storage is set to the cache, and no secondary storage is configured. How should I configure the shares to move them to the cache? The mover isn't moving them to the cache as expected.

https://docs.unraid.net/unraid-os/release-notes/6.12.0/#share-storage-conceptual-change

33 minutes ago, Lauren112 said:

I finally managed to delete the files. Currently, the appdata, domains, isos, and system shares are located on the array. The primary storage is set to the cache, and no secondary storage is configured. How should I configure the shares to move them to the cache? The mover isn't moving them to the cache as expected.

You need both primary and secondary storage configured for a share before mover will take any action.