Probably your HA VM will also be starting from scratch.

Hey mate - so something is going wrong, my cache has filled back up to 95% and I am getting warning notifications - seems to have moved a bunch of my media files over from the array (where they should be) to the cache? The media files are in the data share data/usenet/media/complete/movies etc.

I noticed as I can't install any more apps (or access the apps menu in the GUI) as the cache is all but full. This is pretty much exactly what was happening before the SSD died and why I had moved everything more or less to the array.

New diags attached. Did I just dumb ass set some of these cache-arrays settings wrong on the shares? So far I haven't even configured any of the apps i just installed to the docker.

Edit - mover appears to be running too.

tower-diagnostics-160125.zip

Edited January 16Jan 16 by sihorrocks

10 hours ago, sihorrocks said:

my cache has filled back up to 95% and I am getting warning notifications - seems to have moved a bunch of my media files over from the array (where they should be) to the cache?

Looks like you set things correctly. Not sure what's going on. Maybe something didn't notice the changes you made? Have you rebooted since you removed the Mover Tuning plugin?

See if you can stop mover from the command line with this

mover stop

7 hours ago, trurl said:

Looks like you set things correctly. Not sure what's going on. Maybe something didn't notice the changes you made? Have you rebooted since you removed the Mover Tuning plugin?

See if you can stop mover from the command line with this

mover stop

Worked from console. Mover stopped - great. Can't recall if I rebooted - so tried and it wouldn't actually restart, just seemed to refresh the GUI. Either that or it restarted in one minute. Very strange.

Have performed a full shutdown (took 20 mins to shutdown - guessing because of the full cache?) and restarted the array. New diags attached. Should i retry the mover?

tower-diagnostics-170126.zip

Looks like your appdata and system shares have overflowed to the array. Or maybe whatever caused mover to behave incorrectly with data also made it behave incorrectly with those shares.

Might be better if we cleaned those up after making more room on cache, so for now

Disable Docker and VM Manager

Set appdata, domains, and system shares to

Primary:array; Secondary:none

That will leave only the data share with a setting for Mover action, which is cache->array.

And don't write anything to the data share either.

Then run mover again and see how it goes. If it's still moving to cache instead of from cache, stop it again. If it is working correctly, let it finish. Either way post new diagnostics.

Seeme to be moving files off the cache - you'll see the immediate reduction in size. Seems to be continuously reducing!

20 hours ago, trurl said:

Looks like your appdata and system shares have overflowed to the array. Or maybe whatever caused mover to behave incorrectly with data also made it behave incorrectly with those shares.

Might be better if we cleaned those up after making more room on cache, so for now

Disable Docker and VM Manager

Set appdata, domains, and system shares to

Primary:array; Secondary:none

That will leave only the data share with a setting for Mover action, which is cache->array.

And don't write anything to the data share either.

Then run mover again and see how it goes. If it's still moving to cache instead of from cache, stop it again. If it is working correctly, let it finish. Either way post new diagnostics.

Took all day and night, but mover finished, and all but 3GB moved off of the cache. Back to normal. Really happy.

Next steps? New diags attached.

Diagnostics-180125.zip

Now that you have plenty of room on cache, let's get appdata and system shares moved there where they belong.

Set appdata and system to

Primary:cache; Secondary:array; Mover action:array->cache.

Then run mover, wait for it to complete, post new diagnostics.

Mover settings done

Invoked mover, but didn't really seem to do much, just refreshed the GUI and the mover button wasn't greyed out and no mover is running info, but the appropriate shares seem to be on the cache and navigating into the appdata folders shows the application's data. Cache increased from 3Gb to 8Gb - so looking good - but just seemed to move instantly which was weird.

New diags attached

diagnostics-180126.zip

That looks good. If you want, you can make appdata, domains, system shares exclusive to cache, which will make access to them more efficient.

You would set them to

Primary:cache; Secondary:none

and set Exclusive access:yes

You really don't want these to overflow to the array anyway.

You can enable Docker and VM Manager now and start working with those.

1 hour ago, trurl said:

That looks good. If you want, you can make appdata, domains, system shares exclusive to cache, which will make access to them more efficient.

You would set them to

Primary:cache; Secondary:none

and set Exclusive access:yes

You really don't want these to overflow to the array anyway.

That's all done - thank you very much for your time and patience so far. I'll start putting apps back on - and be in touch when they're all up and running. Probs a couple of days away.

Don't forget about Previous Apps on the Apps page. It looks like you have the templates for your dockers so shouldn't take long at all to reinstall those. Without their appdata they would be new installs though.

On 1/19/2026 at 2:56 AM, trurl said:

Don't forget about Previous Apps on the Apps page. It looks like you have the templates for your dockers so shouldn't take long at all to reinstall those. Without their appdata they would be new installs though.

Do you know your way around the arr suite and sabnzbd - I am getting an error (which I have received before) across the arr suite related to folder / share access

but last time i was able to fix it by correctly setting up my categories in sabnzbd. This fix is not working this time.

So I have my arr suite share access set up like this, where 1 is my library in /usenet/media/movies etc and 2 is where sabnzbd downloads files to usenet/complete/movies etc using the relevant categories

The descriptions are now making me second guess - is /data for my library and /media for newly downloaded files to be renamed and moved to my library, or is it the other way round?

Sabnzbd is setup like this

This is manifesting by the arr suite being unable to find the downloaded media - they can access my library just fine, but downloaded media folders look empty to the arr suite - so they cannot import and rename new media. Sab downloads okay, but arr only sees an empty folder in usenet/complete/movies - if i browse to the folder the media file is there.

I've tried every variation to the /media Path I can think of, going higher or lower in the path, but nothing will make it work. Prior to the rebuild I had this error, but setting up my categories in sab seemed to fix it. Any ideas of what I should change on these two paths? Spent a couple of hours pulling my hair out here. The Arr documentation is just confusing me further.

On 1/19/2026 at 2:56 AM, trurl said:

Don't forget about Previous Apps on the Apps page. It looks like you have the templates for your dockers so shouldn't take long at all to reinstall those. Without their appdata they would be new installs though.

7 hours ago, sihorrocks said:

've tried every variation to the /media Path I can think of, going higher or lower in the path, but nothing will make it work

Post docker run for SAB and for some 'arr that is giving you trouble.

15 hours ago, trurl said:

Post docker run for SAB and for some 'arr that is giving you trouble.

Here's sab -
docker run
  -d
  --name='sabnzbd'
  --net='bridge'
  --pids-limit 2048
  -e TZ="Pacific/Auckland"
  -e HOST_OS="Unraid"
  -e HOST_HOSTNAME="Tower"
  -e HOST_CONTAINERNAME="sabnzbd"
  -e 'PRIVOXY_ENABLED'='false'
  -e 'UNBOUND_ENABLED'='false'
  -e 'VPN_ENABLED'='false'
  -e 'VPN_CONF'='wg0'
  -e 'VPN_PROVIDER'='generic'
  -e 'VPN_LAN_NETWORK'=''
  -e 'VPN_EXPOSE_PORTS_ON_LAN'=''
  -e 'VPN_AUTO_PORT_FORWARD'='true'
  -e 'VPN_AUTO_PORT_FORWARD_TO_PORTS'=''
  -e 'VPN_KEEP_LOCAL_DNS'='false'
  -e 'VPN_FIREWALL_TYPE'='auto'
  -e 'VPN_PIA_USER'=''
  -e 'VPN_PIA_PASS'=''
  -e 'VPN_PIA_PREFERRED_REGION'=''
  -e 'VPN_PIA_DIP_TOKEN'='no'
  -e 'VPN_PIA_PORT_FORWARD_PERSIST'='false'
  -e 'PUID'='99'
  -e 'PGID'='100'
  -e 'UMASK'='002'
  -l net.unraid.docker.managed=dockerman
  -l net.unraid.docker.webui='http://[IP]:[PORT:8080]'
  -l net.unraid.docker.icon=''
  -p '8081:8080/tcp'
  -v '/data/usenet':'/config':'rw'
  -v '/mnt/user/data/':'/data':'rw'
  --hostname=sabnzbd.internal
  --cap-add=NET_ADMIN
  --sysctl="net.ipv4.conf.all.src_valid_mark=1"
  --sysctl="net.ipv6.conf.all.disable_ipv6=1" 'ghcr.io/hotio/sabnzbd:latest'

8fc8a932682b5e7e57e250997bb53b74ed4896d5b7637a300c2c35e158fdad48

The command finished successfully!

Here's lidarr -
docker run
  -d
  --name='binhex-lidarr'
  --net='bridge'
  --pids-limit 2048
  -e TZ="Pacific/Auckland"
  -e HOST_OS="Unraid"
  -e HOST_HOSTNAME="Tower"
  -e HOST_CONTAINERNAME="binhex-lidarr"
  -e 'PUID'='99'
  -e 'PGID'='100'
  -e 'UMASK'='000'
  -l net.unraid.docker.managed=dockerman
  -l net.unraid.docker.webui='http://[IP]:[PORT:8686]'
  -l net.unraid.docker.icon=''
  -p '8687:8686/tcp'
  -v '/mnt/user/appdata/binhex-lidarr':'/config':'rw'
  -v '/mnt/user/data/usenet/':'/data':'rw'
  -v '/mnt/user/data/usenet/complete/':'/media':'rw'
  --restart=always 'ghcr.io/binhex/arch-lidarr'

7554ae58efcff9ef8d8422de7b0c8576045148dacc99e48a51c9e22215adccc1

The command finished successfully!

Looks like you have the wrong /config mapping for SAB. That would result in your SAB appdata being stored at /data/usenet, which is not actual mounted storage on the host, so would be in rootfs (RAM) and so your SAB would be starting like a new install each time you rebooted.

Which application specifically is giving you this error?

On 1/20/2026 at 2:50 AM, sihorrocks said:

I am getting an error (which I have received before) across the arr suite related to folder / share access

Classic 'a little bit of knowledge is a dangerous thing' error there. All the arrs give the exact same error.

Well that would make sense! Changed the /config path to appdata

docker run
  -d
  --name='sabnzbd'
  --net='bridge'
  --pids-limit 2048
  -e TZ="Pacific/Auckland"
  -e HOST_OS="Unraid"
  -e HOST_HOSTNAME="Tower"
  -e HOST_CONTAINERNAME="sabnzbd"
  -e 'PRIVOXY_ENABLED'='false'
  -e 'UNBOUND_ENABLED'='false'
  -e 'VPN_ENABLED'='false'
  -e 'VPN_CONF'='wg0'
  -e 'VPN_PROVIDER'='generic'
  -e 'VPN_LAN_NETWORK'=''
  -e 'VPN_EXPOSE_PORTS_ON_LAN'=''
  -e 'VPN_AUTO_PORT_FORWARD'='true'
  -e 'VPN_AUTO_PORT_FORWARD_TO_PORTS'=''
  -e 'VPN_KEEP_LOCAL_DNS'='false'
  -e 'VPN_FIREWALL_TYPE'='auto'
  -e 'VPN_PIA_USER'=''
  -e 'VPN_PIA_PASS'=''
  -e 'VPN_PIA_PREFERRED_REGION'=''
  -e 'VPN_PIA_DIP_TOKEN'='no'
  -e 'VPN_PIA_PORT_FORWARD_PERSIST'='false'
  -e 'PUID'='99'
  -e 'PGID'='100'
  -e 'UMASK'='002'
  -l net.unraid.docker.managed=dockerman
  -l net.unraid.docker.webui='http://[IP]:[PORT:8080]'
  -l net.unraid.docker.icon=''
  -p '8080:8080/tcp'
  -v '/mnt/cache/appdata/sabnzbd/':'/config':'rw'
  -v '/mnt/user/data/':'/data':'rw'
  --hostname=sabnzbd.internal
  --cap-add=NET_ADMIN
  --sysctl="net.ipv4.conf.all.src_valid_mark=1"
  --sysctl="net.ipv6.conf.all.disable_ipv6=1" 'ghcr.io/hotio/sabnzbd:latest'

316ebf06c2cd887bdf0fb2f1e90049af4745d4eb341d1b0427d836f0344d7dac

The command finished successfully!

This is several years old but still relevant:

I created a wholly new share /downloads and set Sabnzbd to download to that new share. Seems to have fixed it, maybe because both downloads and the media library were on the same data share. Didn't have that problem in the past - but it seems to be working now, so very good - thanks.

Just got to check its all working, then restart the VM for HA.

I think the mover is working correctly too, files that were downloaded to the cache seem to have been moved over to the array overnight.

@trurl have you had any luck with Immich? I can't for the life of me get it running - tried every quick fix I can find on reddit etc - webUI will not run.

12 minutes ago, sihorrocks said:

Immich

Not using that.

You can go directly to the correct support thread for that docker by clicking its icon in your Unraid webUI and selecting Support from the popup.

Thanks. It actually links to the github to raise an issue. But I'll see if i can find the correct support thread in here.