[6.12.3] Array stop stuck on "Retry unmounting disk share(s)"

[aurevo]

Hello all,

 

unfortunately I am also one of the people who get the error message "/mnt/cache is busy" when trying to stop the array and shutting down the system.

 

Nothing showed when entering lsof /mnt/cache.

 

umount /dev/loop2 and umount /var/lib/docker not worked.

 

Diagnosis attached.

 

Thanks for you help.

 

tower-diagnostics-20230824-1248.zip

Edited August 24, 2023 by aurevo
missing attachment

[JorgeB]

It appears to be a problem with this container:

 

Aug 24 12:38:30 Tower root: Error response from daemon: cannot stop container: cloudflareddns: tried to kill container, but did not receive an exit event

 

[aurevo]

On 8/24/2023 at 1:30 PM, JorgeB said:

It appears to be a problem with this container:

 

Aug 24 12:38:30 Tower root: Error response from daemon: cannot stop container: cloudflareddns: tried to kill container, but did not receive an exit event

 

 

Curiously, there were actually no problems with the container before. Is it possible to force the termination?

[JorgeB]

Not AFAIK, Unraid is already trying to force it and failing.

[aurevo]

On 8/25/2023 at 4:36 PM, JorgeB said:

Not AFAIK, Unraid is already trying to force it and failing.

 

So I have to kill the container myself before stopping the array or what can I do to shutdown my Tower?

 

The Docker itself was stopped, so don't know how a container can prevent my Tower from shutdown.

 

It is no option to hard shutdown the device each time and do a parity check.

[JorgeB]

6 hours ago, aurevo said:

The Docker itself was stopped

Do you mean that container or the docker service? Docker service is not stopping because of that container:

 

Aug 24 12:38:30 Tower root: stopping dockerd ...
Aug 24 12:38:31 Tower root: waiting for docker to die ...
### [PREVIOUS LINE REPEATED 14 TIMES] ###
Aug 24 12:38:45 Tower root: docker will not die!
Aug 24 12:38:45 Tower emhttpd: shcmd (2833): exit status: 1
Aug 24 12:38:45 Tower emhttpd: shcmd (2834): umount /var/lib/docker
Aug 24 12:38:45 Tower root: umount: /var/lib/docker: target is busy.

 

Suggest posting in that container support thread/discord, assuming one exists, to try and find out why it's not getting killed:

 

 

[aurevo]

On 8/27/2023 at 1:18 AM, JorgeB said:

Do you mean that container or the docker service? Docker service is not stopping because of that container:

 

Aug 24 12:38:30 Tower root: stopping dockerd ...
Aug 24 12:38:31 Tower root: waiting for docker to die ...
### [PREVIOUS LINE REPEATED 14 TIMES] ###
Aug 24 12:38:45 Tower root: docker will not die!
Aug 24 12:38:45 Tower emhttpd: shcmd (2833): exit status: 1
Aug 24 12:38:45 Tower emhttpd: shcmd (2834): umount /var/lib/docker
Aug 24 12:38:45 Tower root: umount: /var/lib/docker: target is busy.

 

Suggest posting in that container support thread/discord, assuming one exists, to try and find out why it's not getting killed:

 

 

 

For now I stopped container by hand and disabled autostart because I don't really need that container.

 

But today I wanted to stop array because of btrfs errors and now I have the same "cache is busy" error messagen.

 

Logs attached, appreciate your help.

tower-diagnostics-20230828-1022.zip

[JorgeB]

It's a different issue now, both pools are busy:

 

Aug 28 10:07:53 Tower emhttpd: shcmd (532): umount /mnt/cache
Aug 28 10:07:53 Tower root: umount: /mnt/cache: target is busy.
Aug 28 10:07:53 Tower emhttpd: shcmd (532): exit status: 32
Aug 28 10:07:53 Tower emhttpd: shcmd (533): umount /mnt/cache_appdata
Aug 28 10:07:53 Tower root: umount: /mnt/cache_appdata: target is busy

 

But cannot see with what, make sure no SSH sessions opened.