user shares turned off and stopping array causes gui crash 6.0 beta10a

[merlyn]

I am not sure if it is just me or anyone with user shares turned off.  I know I cannot stop the array at any time without the gui crashing.

 

it will say unmounting user shares ... retry retry etc ...until it crashes ...

 

tower is still working though since i can ssh in and reboot the server .

 

this has started happening since i went to beta 6 never had a problem before this on 5.

I have one docker running plex with needo as a repository

 

The dirty shutdown that resulted destroyed the plex docker to the point i cannot use it at all anymore ever after reinstall.  see post on that topic top of page 50  .... http://lime-technology.com/forum/index.php?topic=33822.735

 

can anyone with user shares turned off verify if they can stop their array since on my machine it causes gui crash 100 percent of the time.

 

merlyn

[eddiejr]

I am not sure if it is just me or anyone with user shares turned off.  I know I cannot stop the array at any time without the gui crashing. it will say unmounting user shares ... retry retry etc ...until it crashes ... tower is still working though since i can ssh in and reboot the server . this has started happening since i went to beta 6 never had a problem before this on 5. I have one docker running plex with needo as a repository The dirty shutdown that resulted destroyed the plex docker to the point i cannot use it at all anymore ever after reinstall.  see post on that topic top of page 50  .... http://lime-technology.com/forum/index.php?topic=33822.735 can anyone with user shares turned off verify if they can stop their array since on my machine it causes gui crash 100 percent of the time.

 

merlyn

 

I had PLEX working for 2 days.  I went to reboot the server so I hit stop on the array.  The gui stopped responding after that.  whoops ....  gave it a hour and no response so i SSH'd in and rebooted the server. my error should have shutdown the docker apps first before hitting stop. (i guess) Ever since that Dirty shutdown Plex seems to install but i cannot get the UI to come up at all. repository is needo/plex

 

Things i have tried ...

- deleted container ... rebuilt ... UI will not connect.

- deleted image and container rebuilt ... UI will not connect.

- deleted entire disk image ... created new image ... rebuilt plex ... UI will not connect.

 

checking plex logs i see pages and pages of this error

 

WARNING COULDN'T CREATE /config/Library/Application Support, MAKE SURE I HAVE PERMISSON TO DO THAT!

mkdir: cannot create directory '/config/Library': Permission denied

 

can i assume i have some permissions issues?    can anyone help? I am just learning docker so i really don't know what to do next thanks in advance...

 

merlyn

 

I am on 6b12 with users shares off and I had the excellent needo/plex docker app working great without any install errors. The one time I stopped my array for a reboot without stopping the app or docker service, I got an endless loop of ""Retry unmounting user share(s)..." in the unraid gui and could do nothing else from there. Perhaps "Then something is still running that is accessing it", but I can't be sure. I could still PuTTY to unraid and issues a "shutdown -r now" so I wouldn't have to go downstairs and do it. Perhaps I should of checked the syslog before I did that?

 

When the machine came back up it detected an unclean shutdown and a parity check ensued. I was trying to check old system logs but it seems to only have logs from the current booting. I then noticed I could not access the Plex at http://192.168.1.221:32400/web/index.html# page, so I reinstalled the excellent needo plex docker app and got the same errors as you where it could not mkdir /config/Library and /config/Library/Application Support directories. Nothing was working like you and even tried what you tried to recreate the docker image and reinstalling plex without success and with the same errors. My docker image is located at /mnt/disk1/docker02.img but I wonder if there is a better spot for it? I am using the repository at  https://github.com/gfjardim/docker-containers/tree/templates and my /config is the default /mnt/user/appdata/plexmediaserver directory. On top of that, everytime I tried to stop the array, it keep on hanging the unraid gui.

 

I found that if I completely stopped the docker service, I could stop the unraid array and reboot everything just fine. I was still having problems with getting the beloved plex docker app working, but after a few unraid reboots and a bunch of stop/start docker service cycles... something magically got fixed and I was able to once again access the plex gui and things were working fine again. LOL

 

-Eddie

[trurl]

I am on 6b12 with users shares off and I had the excellent needo/plex docker app working great without any install errors. The one time I stopped my array for a reboot without stopping the app or docker service, I got an endless loop of ""Retry unmounting user share(s)..." in the unraid gui and could do nothing else from there. Perhaps "Then something is still running that is accessing it", but I can't be sure. I could still PuTTY to unraid and issues a "shutdown -r now" so I wouldn't have to go downstairs and do it. Perhaps I should of checked the syslog before I did that?

 

When the machine came back up it detected an unclean shutdown and a parity check ensued. I was trying to check old system logs but it seems to only have logs from the current booting. I then noticed I could not access the Plex at http://192.168.1.221:32400/web/index.html# page, so I reinstalled the excellent needo plex docker app and got the same errors as you where it could not mkdir /config/Library and /config/Library/Application Support directories. Nothing was working like you and even tried what you tried to recreate the docker image and reinstalling plex without success and with the same errors. My docker image is located at /mnt/disk1/docker02.img but I wonder if there is a better spot for it? I am using the repository at  https://github.com/gfjardim/docker-containers/tree/templates and my /config is the default /mnt/user/appdata/plexmediaserver directory. On top of that, everytime I tried to stop the array, it keep on hanging the unraid gui.

 

I found that if I completely stopped the docker service, I could stop the unraid array and reboot everything just fine. I was still having problems with getting the beloved plex docker app working, but after a few unraid reboots and a bunch of stop/start docker service cycles... something magically got fixed and I was able to once again access the plex gui and things were working fine again. LOL

 

-Eddie

I have never run without user shares so I don't know the answer to this question: Is there a /mnt/user/appdata folder if you don't enable user shares?

 

Maybe it would get created, I don't know, but if it did get created, where would it be?

 

I would think you would have to use something like /mnt/disk1/appdata.

 

[eddiejr]

thanks for the reply...

 

My docker image was located /disk1/dockerXX.img but I have since moved it to /cache/dockerXX.img    ...I was using the 10gb default size, but now have upped it to 20gb.

 

I used the default: host path /mnt/user/appdata/plexmediaserver is mapped to container volume /config in docker ...i figure appdata physically resides inside my dockerXX.img?

 

other than that, i don't know where it physically creates appdata as I can not see it anywhere residing in cache, disk1, or disk2; however if I PuTTY in I can see it in /mnt/user where under /mnt i can see cache, disk1, disk2, & user.

 

Perhaps that is the problem? Maybe appdata needs to be explicitly on a data or cache drive?

 

-Eddie

 

 

 

 

 

[eddiejr]

 

Update: Still have a problem with "Stop AVAHI...Stop SMB...Spinning up all drives...Sync filesystems...Retry unmounting user share(s)...Retry unmounting user share(s)...Retry unmounting user share(s)...Retry unmounting user share(s)...Retry unmounting user share(s)...Retry unmounting user share(s)..." ad infinitum even after completely stopping needo/plex app (my only docker app) and docker service when trying to stop array.

 

I'm on 6.0-beta12 and am not using user shares. Any thoughts?

[JonathanM]

Any thoughts?

After you start the shutdown process, try pulling the network cable out of the server. It sounds like something on your network is still accessing the server.

[eddiejr]

Any thoughts?

After you start the shutdown process, try pulling the network cable out of the server. It sounds like something on your network is still accessing the server.

 

I understand now, thanks I will try it next time. 

[jonp]

If you are logged in via ssh under any folder under /mnt, you will get these errors.

[jonp]

thanks for the reply...

 

My docker image was located /disk1/dockerXX.img but I have since moved it to /cache/dockerXX.img    ...I was using the 10gb default size, but now have upped it to 20gb.

 

I used the default: host path /mnt/user/appdata/plexmediaserver is mapped to container volume /config in docker ...i figure appdata physically resides inside my dockerXX.img?

 

other than that, i don't know where it physically creates appdata as I can not see it anywhere residing in cache, disk1, or disk2; however if I PuTTY in I can see it in /mnt/user where under /mnt i can see cache, disk1, disk2, & user.

 

Perhaps that is the problem? Maybe appdata needs to be explicitly on a data or cache drive?

 

-Eddie

 

The large bolded part is your problem.  You are using Docker containers that are, by default, configured to store data under /mnt/user/appdata.  /mnt/user is the user shares.  If you have user shares disabled, but start a container that's appdata points to /mnt/user, you could be causing a problem.

 

If you're not using user shares, you need to change any reference to "/mnt/user" to "/mnt/disk#" where # is the disk number you want to store it on.

[drewdog]

@jonp  Thanks for the tip on changing /mnt/user to /mnt/disk#.  I plex webui is working again.  i don't have user shared enabled.