I’m trying to use nginx as a reverse proxy in a Kubernetes setup. We are required to verify client certificates against CRL-files, and a sidecar container is responsible for updating the CRL’s. My idea was to use “nginx -s reload” to update the configuration in the nginx-container, but for some reason the container starts a completely new master-process and since the listen-ports aren’t available it shuts down after a few seconds.
It looks like the reload is behaving correctly at first, the logs describes a graceful shutdown and starting new workers and suddenly the container is stopped.
I’ve tried the debugging version with no luck, I can’t find any root cause to the behavior.

I’m using nginx:1.26.0, and I haven’t added any scripts in the docker-entrypoint.d-dir.
We are using a custom uid/gid range, and I´m not able to specify the default 101/101 uid/gid due to security restrictions.