I have a lifetime pass from many years ago when it was cheap. So I’m not in a huge rush to convert and want to do it right. But I am on the path to converting. I decided to make a major change to my home server infrastructure and it’s still in an experimental stage. Moving from a really old standalone computer I’ve used for. HTPC purposes over the years, currently dedicated to Plex combined with a few raspberry pi’s of various generations for the little stuff, and a single, good NUC for my router, to adding two additional NUCs and eventually upgrading the Plex computer with a more modern processor and video card for ML stuff for Immich and a few other systems that I plan to start using. I’m not just moving from Plex, but also a lot of Google and Nest products.
My dilemma has been Docker Swarm vs Kubernetes. I was trying to set up Kubernetes in a way that is easily repeatable and self documenting, but ended up with lots of manual steps required to install things and lots of things that I had to write my own helm charts for as well as the scripts to install and set up Kubernetes itself on each of the servers. Lots of custom stuff. Docker Swarm would be way easier, but the issue is I’m worried about Docker getting so proprietary these days and swarm mode getting so little attention, and Podman quadlets aren’t self balancing across multiple small servers like swarm. So that’s why I haven’t switched to Jellyfin yet.
I wanted the same thing with Kubernetes and ended up using FluxCD. Highly recommend it. It basically syncs a git repo to the cluster, so you just push to github or whatever, and it auto applies the changes you pushed. Also, llm models tend to be good at teaching this topic and even writing yaml files for it, so the initial learning curve was not bad actually.
I have a lifetime pass from many years ago when it was cheap. So I’m not in a huge rush to convert and want to do it right. But I am on the path to converting. I decided to make a major change to my home server infrastructure and it’s still in an experimental stage. Moving from a really old standalone computer I’ve used for. HTPC purposes over the years, currently dedicated to Plex combined with a few raspberry pi’s of various generations for the little stuff, and a single, good NUC for my router, to adding two additional NUCs and eventually upgrading the Plex computer with a more modern processor and video card for ML stuff for Immich and a few other systems that I plan to start using. I’m not just moving from Plex, but also a lot of Google and Nest products.
My dilemma has been Docker Swarm vs Kubernetes. I was trying to set up Kubernetes in a way that is easily repeatable and self documenting, but ended up with lots of manual steps required to install things and lots of things that I had to write my own helm charts for as well as the scripts to install and set up Kubernetes itself on each of the servers. Lots of custom stuff. Docker Swarm would be way easier, but the issue is I’m worried about Docker getting so proprietary these days and swarm mode getting so little attention, and Podman quadlets aren’t self balancing across multiple small servers like swarm. So that’s why I haven’t switched to Jellyfin yet.
I wanted the same thing with Kubernetes and ended up using FluxCD. Highly recommend it. It basically syncs a git repo to the cluster, so you just push to github or whatever, and it auto applies the changes you pushed. Also, llm models tend to be good at teaching this topic and even writing yaml files for it, so the initial learning curve was not bad actually.
Now I’m exploring doing this even better with this template: https://github.com/onedr0p/cluster-template