syslog-ng Operation not permitted

[snqblevj]

Hi, I use the 22.6.0 version of this container on a CentOS 7.9 VM.

I wanted to upgrade my container to 22.7.0 so I deleted my old container(22.6.0), and I did a docker run with the 22.7.0 version.
Unfortunately the new container didn't want to start and it restarted every 30 seconds.
So I looked in docker logs librenms to see if there are any problems and I found this

I tried to delete SYSLOGNG_OPTS="--no-caps" in /etc/default/syslog-ng with docker exec -it librenms bash in a delay of 30 seconds (before the container restart) to resolve the problem but nothing change.
I tried to delete the lign 32 in /etc/my_init.d/10_syslog_ng or to delete the file /etc/my_init.d/10_syslog_ng but nothing change too.

I tried to upgrade to 22.8.0 but it doesn't work.

In the end, I went back to version 22.6.0 and it worked again.

Have you any idea ?

[jarischaefer]

@snqblevj Unfortunately, I have no appropriate test environment at the moment. Perhaps SELinux or some other mechanism is blocking access. You could try running the container with capabilities or privileged if you do not mind the implications.

[snqblevj]

I tried to launch the container with --privileged but nothing change.
I also tried to 'chmod 777 my_init.d' and 'chmod 777 *' inside my_init.d but same issues.
I compared the 22.6.0 version with 22.9.0 --> 22.6.0...22.9.0 to change manually the files '10_syslog-ng' and 'librenms_221_syslog' content but i'm not an expert and it didn't work.
The 22.6.0 version is working perfectly, I'm gonna stay on it for the moment waiting for something.

[neomathews]

Same problem here.
I am trying to upgrade from 21 to 23 but i had to stay on 22.6.0 to avoid the problem.
Ubuntu 20.04.06 and docker-ce 20.10.2
Thanks.

[jarischaefer]

@snqblevj
@neomathews

Please provide the output of docker run --rm jarischaefer/docker-librenms:23.11.0 capsh --print. Make sure you are running the latest docker-ce version (24.0.7): https://docs.docker.com/engine/install/ubuntu/