• 0 Posts
  • 22 Comments
Joined 1 year ago
cake
Cake day: June 19th, 2023

help-circle












  • Don’t worry, most of this is about learning :) You could also put Docker on the Linux, install Jellyfin as a Docker container, and only fore it up when you are using Jellyfin. It’s a little less convenient, than having it available all the time, but storing all media in H264 AC3 (most common video and audio codecs) costs a lot of storage (H265 and AV1 are far more efficient). Another pro of Docker is, when you decide to move your server, you just shut down the container, copy it’s data and start the container on the new device




  • I’d go for the consumer case with as much drive space as possible, but in the end it’s up to your taste. With rackmount full professional server stuff, you have to keep in mind, it’s not only quite large in terms of space, but also noisy. This is something for the basement or something. Another option would be a mini pc like a NUC combined with a NAS - here is the limit the upgradability of the hardware. Like you can usually upgrade the RAM a little and the system drive, but that’s it. Especially for media stuff (Plex, Emby, Jellyfin) it can be interesting to have a dedicated GPU, like an Intel card, to have more transcoding capabilities. I’m currently running most things from a SFF PC, but also have a Pi and an old laptop in use for various tasks (main reason is lazyness to move the services to the SFF lol)




  • I started with a Pi 4 8GB. It was not able to keep my RAID 1 (constantly loosing one of the two drives, degraded when added back). Transcoding of movies, also Jellyfin, just not possible. So I bit the bullet and took some old hardware (i7-4790K, 1070, 16GB RAM, PSU, CPU-Cooler) and bought some more used (SFF case, mini-itx board, Bluray-Drive). The case wasn’t the cheapest (56€), because I wanted one with enough room for 5.25" drive and 4 3.5" HDDs. Finding a mini itx board for such an old platform was hard and cost 120€ (I didn’t sell the ATX board I originally used with the CPU, yet). But now it’s a great system, running 19 containers atm In other words, don’t underestimate the cost of old hardware. You may find a cheaper, used prebuilt somewhere