Process Owners

[buckaroogeek]

Not sure if this is an issue or not. This is what I use to start oznu/docker-unms:

docker run -p 80:80 -p 443:443 -e TZ=America/Pacific -e PUID=$uUID -e PGID=$uGID -v $PWD/config:/config:Z --name unms oznu/unms:latest

where uUID=$(id -u) uGID=$(id -g)

when I look in the config directory I see: [bgsmith@pico config]$ ll total 16 drwxr-xr-x. 5 nobody nobody 4096 Mar 23 09:38 logs drwx------. 19 avahi avahi 4096 Mar 23 09:38 postgres drwxr-xr-x. 2 bgsmith bgsmith 4096 Mar 23 09:38 redis drwxr-xr-x. 10 bgsmith bgsmith 4096 Mar 23 09:38 unms

Should logs be set to user nobody and postgres to avahi?

thanks

brad

[oznu]

Hi @buckaroogeek,

This is a bit hard to explain without explaining how linux containers work. What you're seeing is just a side effect of the users inside the container using a UID that happens to be the same as an existing user on the host.

If you run this you will see the correct owners (with bgsmith replaced with abc):

docker exec -it unms ls -la

Or if you run this you'll see all the processes running inside the container and the corresponding users:

docker exec -it unms ps aux

What you're seeing is perfectly normal. Hope this helps.

[buckaroogeek]

Thanks for the response. I had been assuming that setting PGID and PUID would apply to all processes in the container that are writing to the data volume and not just some of them. Perhaps a clarifying statement in the documentation section on User/Group Identifiers would be helpful to others?

Thank you for providing this repackaging of UNMS. I use fedora and this just works like I would expect a docker container to work.

[oznu]

I've actually removed the PGID and PUID options from the next version of the container as they were making this image harder to maintain and required more effort to keep up with the regular updates from Ubiquiti.

https://github.com/oznu/docker-unms/tree/0.12.0-beta.7