So I just installed everything on dietpi yesterday (fresh install), and by everything I mean: Plex, nordvpn, radarr, bazarr, sonarr, transmission and jackett.
Everything was working yesterday, but after reboot the following services are not starting: radarr, bazarr, sonarr.
If I start the service manually:
sudo -u radarr /opt/radarr/Radarr -nobrowser -data=/mnt/dietpi_userdata/radarr
The service starts normally, but using dietpi-services I have the following error:
May 09 10:01:54 DietPi systemd: Started Radarr Daemon (DietPi).
May 09 10:01:54 DietPi systemd: radarr.service: Failed to set up mount namespacing: No such file or directory
May 09 10:01:54 DietPi systemd: radarr.service: Failed at step NAMESPACE spawning /opt/radarr/Radarr: No such file or directory
May 09 10:01:54 DietPi systemd: radarr.service: Main process exited, code=exited, status=226/NAMESPACE
May 09 10:01:54 DietPi systemd: radarr.service: Failed with result 'exit-code'.
(The error is the same on the other 2 services)
I have the dietpi user data moved to an external HDD.
Any help is appreciated.
we had a similar case in the past https://dietpi.com/forum/t/sonarr-lidarr-radarr-not-starting/5197/1
For the user the issue was fixed by reinstall the failed applications. Probably you can create a backup and try it out.
I did that, but again after a restart it stopped working again.
By restart Im using a smart plug which basically cuts the power, I use this to shut down the raspberry pi over night and turn it on in the morning.
I never had the issue before when I was using everything in docker containers.
But then why am I able to start the service manually and not through dietpi-services?
You should definitely switch off the system by running poweroff or shutdown now command and not just toggle power switch as this could cause inconsistencies
Here a solution our end: https://github.com/MichaIng/DietPi/commit/5320442
But in your case it is most likely the log directory, which is not stored to disk as long as you do not do a regular shutdown/restart at least once after Radarr install. So the service would fail anyway, only the error message would have been clearer. So at least when using RAMlog, an ordinary shutdown makes sense to have new or removed files in /var/log synced to disk (not to content but the directories and empty log files to hold metadata) and having it loaded back into RAM disk at boot.
but the error is really misleading in this case, as it looks like /opt/radarr/Radarr was/is missing
Failed at step NAMESPACE spawning /opt/radarr/Radarr: No such file or directory
Yes it is. In this logs it was shown
Probably the newer systemd on Bullseye extended the output.
yes have seen it as well but I was not sure on the meaning.
I also needed to send out the duck to find the hint .