I am working on writing up some quality help/instructions for Lemmy instance admins. This is targeted toward those that are newer to Lemmy and Docker, but even those that know a thing or two might learn something from me. I have been a sysadmin for over 20 years, so I know some things.
I thought I’d share the first complete page I have written. I’d love some feedback if you have any to share.
Be seeing you.
P.S. I frequent the Lemmy Admin matrix chat and I moderate the Install Support channel. Stop by if you have questions! https://matrix.to/#/#lemmy-support-new-instance:discuss.online
Thanks for your comment, but I don’t see much value in pulling a new copy of the docker-compose.yml from the Lemmy GitHub. The only things I would be updating when Lemmy updates is the tag/version. If they added new environment variables some time in the future I could certainly take a look at their updated compose file to see the changes but I wouldn’t want to pull it down and replace my custom compose.
I specifically don’t care for their (Lemmy devs) choices for logging, docker networking, and the built in nginx, so removing and simplifying all that was my main goal. Everyone has their own way of doing things, and this is mine.
I will probably take a look at your Traefik configs and add them as a separate document for those that don’t want to use NPM. My goal is to add a subsection for most of the current revproxy choices.
Yeah. I found the official compose… let’s say leaves a lot to desire… so we had slightly different approaches to similar problems. I don’t want their built in nginx so I override it with a simple alpine that quits. But I also chose to use override for exactly the reason you mentioned (in event if they add new config stuff into compose).
Glad to see I’m not the only one finding there are rooms for improvement. Thanks for sharing your thoughts with the community!