Latest 1.14 docker image fails to start (1.14.5+1)

[tobias-b-gmx]

I've seen that you updated the docker images and since in the last 1.14.5 release I manually had to mount the 99-custom.ini file to the proper php8 conf.d directory to make it work and as you switched the packages to php81 in 1.14.5+1, I thought I can just pull the latest image and it will work fine without my work-around, however, with the latest image, humhub doesn't want to start at all.

Probably related to the new Alpine version?

humhub | /docker-entrypoint.sh: Configuration complete; ready for start up
humhub | /docker-entrypoint.sh: Entrypoint finished! Launching ...
humhub | tail: can't open '/var/www/localhost/htdocs/protected/runtime/logs/app.log': No such file or directory
humhub | tail: no files
humhub | 2024-01-07 15:29:38,195 WARN exited: humhub-log (exit status 1; not expected)
humhub | 2024-01-07 15:29:38,195 WARN exited: humhub-log (exit status 1; not expected)
humhub | Traceback (most recent call last):
humhub | File "/usr/bin/supervisord", line 33, in
humhub | sys.exit(load_entry_point('supervisor==4.2.5', 'console_scripts', 'supervisord')())
humhub | ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
humhub | File "/usr/lib/python3.11/site-packages/supervisor/supervisord.py", line 359, in main
humhub | go(options)
humhub | File "/usr/lib/python3.11/site-packages/supervisor/supervisord.py", line 369, in go
humhub | d.main()
humhub | File "/usr/lib/python3.11/site-packages/supervisor/supervisord.py", line 78, in main
humhub | self.run()
humhub | File "/usr/lib/python3.11/site-packages/supervisor/supervisord.py", line 94, in run
humhub | self.runforever()
humhub | File "/usr/lib/python3.11/site-packages/supervisor/supervisord.py", line 244, in runforever
humhub | self.reap()
humhub | File "/usr/lib/python3.11/site-packages/supervisor/supervisord.py", line 281, in reap
humhub | process.finish(pid, sts)
humhub | File "/usr/lib/python3.11/site-packages/supervisor/process.py", line 585, in finish
humhub | self.config.options.logger.warn(msg)
humhub | File "/usr/lib/python3.11/site-packages/supervisor/loggers.py", line 331, in warn
humhub | self.log(LevelsByName.WARN, msg, **kw)
humhub | File "/usr/lib/python3.11/site-packages/supervisor/loggers.py", line 345, in log
humhub | handler.emit(record)
humhub | File "/usr/lib/python3.11/site-packages/supervisor/loggers.py", line 227, in emit
humhub | self.doRollover()
humhub | File "/usr/lib/python3.11/site-packages/supervisor/loggers.py", line 264, in doRollover
humhub | if not (self.stream.tell() >= self.maxBytes):
humhub | ^^^^^^^^^^^^^^^^^^
humhub | OSError: [Errno 29] Invalid seek
humhub exited with code 1

[tobias-b-gmx]

Note: same error happens with 1.15.2 ...

[infiniteAppsUG]

I had the same issue, and I dont think its related to the HumHub Version itself. IMPORTANT: I made a clean install to a new database, so there was no need to migrate. So please be careful if you want to upgrade from a previous install, because those steps probably affect your current deployment! However, what worked for me to run HumHub with 1.15.2 was:

docker-compose.yml:
[...]
humhub:latest
[...]

List all containers:
docker ps -a

Stop and remove the humhub related containers (for me it was sufficient to stop only humhub/mriedmann since I work with an external DB)
docker stop container_name
docker rm container_name

Show docker images and remove the humhub related images
docker images
docker rmi image_name

show and remove the mounted volumes
I think this is the most important part, which leads to the error because after unmounting, updating and remounting everything works as expected. Though the previous steps are necessary to unmount successfully

docker volume ls
docker volume rm volume_name

EDIT: Removing the volumes is sufficient, so stopping the container and removing them. However HumHub installation will start from the beginning. So probably one needs to backup the three volumes first and re-integrate them afterwards again and include the database in the compose-file so it will be skipped on initial start.

However the problem occurs with every restart of the container.

[surface0]

I added logfile_maxbytes=0 to supervisord.conf and it fixed it.

[supervisord]
user=root
nodaemon=true
loglevel=warn
pidfile=/var/run/supervisord.pid
logfile=/proc/self/fd/2
logfile_maxbytes=0
stdout_logfile=/proc/self/fd/2
stdout_logfile_maxbytes=0

[infiniteAppsUG]

hey @surface0 : thanks for the hint. I created a local supervisord.conf file and a Dockerfile to make it persistent :

COPY /etc/supervisord.conf /etc/supervisord.conf

• docker compose down
• docker compose build
• docker compose up -d

-> works! thank you!

[tobias-b-gmx]

I added logfile_maxbytes=0 to supervisord.conf and it fixed it.

Thanks a lot, that helped!

[github-actions]

This issue is stale because it has been open for 30 days with no activity.

[github-actions]

This issue is stale because it has been open for 30 days with no activity.

[github-actions]

This issue was closed because it has been inactive for 14 days since being marked as stale.