Docker compose fails after reboot
Part of this problem is not Tipi. I am running this on VmWare and when I shut the Ubuntu server down I think the system does not allow enough time for docker to shut down leading to file errors. I fixed the file errors with fsck but after reboot the server starts, tipi starts, other tipi apps start but plex does not start with this error
cat error.log
worker: 2024-05-22T00:21:56.969Z - error > An error occurred: Command failed: docker-compose --env-file /app-data/plex/app.env --project-name plex -f /data/apps/plex/docker-compose.yml -f /data/repos/29ca930bfdaffa1dfabf5726336380ede7066bc53297e3c0c868b27c97282903/apps/docker-compose.common.yml up --detach --force-recreate --remove-orphans --pull always
time="2024-05-22T00:21:56Z" level=warning msg="/data/apps/plex/docker-compose.yml:
version is obsolete"
time="2024-05-22T00:21:56Z" level=warning msg="/data/repos/29ca930bfdaffa1dfabf5726336380ede7066bc53297e3c0c868b27c97282903/apps/docker-compose.common.yml: version is obsolete"
unable to get image 'lscr.io/linuxserver/plex:1.40.2': Error response from daemon: layer does not exist
How do I fix the docker compose and how do I allow more time for the system to shut down docker?10 Replies
Restarted docker and same result
Pulled latest plex version with sudo docker pull lscr.io/linuxserver/plex:1.40.2
and same result: plex does not start
Try docker image rm
Remove all traces of plex on your system
docker image remove indicated plex not found
docker prune -a got rid of a lot of images but after stopping tipi and updating to newest version still could not start plex
uninstalled all apps from tipi and rebooted
tried to install plex and got this from the error.log
then I removed all plex files from app-data and more and tried to reinstall plex and got same error
This answer on Stack Overflow indicates deleting the docker folder should do the job but still....
Stack Overflow
Docker run results in error: layer does not exist
I'm trying to setup redmine (with postgres) on my raspberry pi 3 using docker-compose. It already worked once, but then I tried to install plugins and somehow managed to bork my system.
Now it won'...
Ok, so this was caused by file corruption of the docker plex image. No amount of prune or rm -f /var/lib/docker worked. I had to use the handy initramfs that comes up after reboot and run it several times to fix the file errors then delete the corrupted layer
How can I increase the amount of time that the system allocates to tipi to shut down the docker images? (That was what caused the problem)
You could run a script before reboot to stop runtipi using the cli
Yes, that would work. Also, delightful surprise that deleting the plex application does not remove the user data. That was not apparent to me from the manual.
@Nicolas After prolonged testing I now use the runtipi CLI to shut down Tipi first before powering off a system. By doing this there are no file errors.
Glad to hear it! This is the way to go