Selfhosted

For me it's been about slowly learning, building, and re-building what I have, as I have gotten more knowledgable.

I started with a Windows laptop running 24/7. Learned about Docker Desktop and ran Jellyfin and the *arr stack on it.

I had Linux running on my PC, so as I got more comfortable with CLI, I decided to migrate the server and everything on it to Ubuntu Server. Learned to navigate on SSH better, learned more about networking and expanded the server with more services like Nginx Revers Proxy Manager and exposing services online.

Then I upgraded my hardware and installed Proxmox on the new machine, to better manage that and the old laptop. Slowly getting better at backing up at different levels, heightening security and so on.

Guides are good, but I find experience and steady knowledge building is a better way to grow and not get overwhelmed or burned out.

Try this

https://tinyurl.com/FUTOselfhost

For an absolute beginner I would recommend Linux tutorials first so you can navigate the Linux cli.
After that watch some docker tutorials as a significant chunk of self hosting is done with docker or a kube variant but you won't need to lean about kube yet.

LearnLinuxTV on YouTube has some great playlists for that.

The true plug-and-play user friendly way is to buy a NAS device and install all apps there.

i think this space will actually get bigger because of everything going on. these nas devices will ve the way to go for your average user.

------- what i did for ciriosity ------

i have everything running on proxmox. mostly as lxcs. there are proxmox helper scripts so it installs nice. i use proxmox for a lot of services so ut all wotks. also, for hardware rendering , intel cpus work better because of intel qsv.

i bought a ugreen nas device. bought an nvme. installed truenas on the nvme, because i like it and its open source. tge ugreen device has a built in boot drive with ugreen os which you can not download so i did not want to overwrite it.

now i have 27 years professionally as a software dev, so i had a foundation to learn this stuff. i want to give a heads up if you use an llm for learning. always ask it to explain and question things. look up any new terms. because of my experience, i understood most of the answers it gave and questioned when i didnt. i remember one time it said to do one thing (i dont remember) and i asked for more info. when i understood i knew it was missing something. i said "but i have x, y, and z" which in alm fairness i never mentioned and it did change the answer. my long-windwd point? dont run anything, unless you fully understand why. this advice goes for anyone not just using ai. people will be wrong too. know what you are doing

I've always recommended the Linux Upskill Challenge. It takes you step by step from the cli to scripting, incrementally. It's also a good bookmark to reference in the future.

There isn't one. Build incrementally, and adapt and modify as it grows. Starting out, you have no idea where you will eventually end up. It's a journey and a learning experience bespoke to you. Start with a basic dedicated device as a server that's cheap. Add a service and get it safe and secure. Once you are happy with it and it is reliable and solid, add another service... Rinse and repeat.

That was a lot of engagement to wake up to, I appreciate you all! Thank you very much!

For anyone in the same position as me that may find this in the future, clearly the first step is to get acquainted with Linux and learn Linux commands. I feel I may have more to learn than I initially thought, but hey, that's what free time is for right?

An ultimate guide could be a few tomes long not because you need to know that much, but because there are so many options and solutions. I'd recommend learning basics: core utils (i.e. command line tools), docker, network security basics (vpn, firewall).

Then as you start adding more services, you can identify gaps in your knowledge and close them (e.g. what is a database, how to manage it, how to back it up and then restore if it fails).

If you manage to get Docker Compose installed on that Mac (I don’t know Dockers limitations with Macs) create a docker-compose.yaml file with the contents of;

services:

  jellyfin:

    image: jellyfin/jellyfin

    container_name: jellyfin

    # Optional - specify the uid and gid you would like Jellyfin to use instead of root

    user: uid:gid

    ports:

      - 8096:8096/tcp

      - 7359:7359/udp

    volumes:

      - /path/to/config:/config

      - /path/to/cache:/cache

      - type: bind

        source: /path/to/media

        target: /media

      - type: bind

        source: /path/to/media2

        target: /media2

        read_only: true

      # Optional - extra fonts to be used during transcoding with subtitle burn-in

      - type: bind

        source: /path/to/fonts

        target: /usr/local/share/fonts/custom

        read_only: true

    restart: 'unless-stopped'

    # Optional - alternative address used for autodiscovery

    environment:

      - JELLYFIN_PublishedServerUrl=http://example.com/

    # Optional - may be necessary for docker healthcheck to pass if running in host network mode

    extra_hosts:

      - 'host.docker.internal:host-gateway'

Edit the volumes so that the drive you want to use is exposed to the container and then run;

sudo docker compose up -d

In a shell while in the folder with the docker-compose.yaml file you just created. If everything works then the server should be running on port :8086 open your browser and make an http:// request to the ip of the server followed by the port.

Example; http://127.0.0.1:8096/

Documentation.

Edit: removed unnecessary \

Edit 2: I really should have read the post more thoroughly, while my example above works fine for a Jellyfin server (which you apparently already run) it also can be adapted for other services. Read and interpret the documentation as best as you can.

Mr Frankenstein’s blog helped me out early on. I also found a write up that dovetailed into it using dockerproxy & swag instead of traefik or caddy. Unfortunately, I don’t have those links handy. Should be able to search for the first one pretty easily, though.

I'd start with learning some networking basics.
Subnets, default gateways, DNS, routes, VLANs.

Then get another computer to install proxmox on.
Learn VMs, SSH, firewalls, basically Linux. Probably VPNs as well.

Then set up a VM and play with docker & docker-compose.

And that's most of the tooling to be able to self host anything. A VM with docker, set up the docker compose of what you want, boom: done.
Some projects are more of a "full of" kinda thing (like home assistant), which would be its own VM

Self-hosting is a very individual journey - everyone wants different things and finds they own way to meet their own requirements. So there isn't really one guide that covers everything.

Anyway, as a general road map:

1. Create a Virtual Machine on your PC. Install Linux inside the VM.

2. Play around in the VM to learn Linux basics. When you break the OS you can just wipe the VM and reinstall.

3. In the VM, try some docker containers until you're comfortable-ish with docker.

4. Maybe try Yunohost in the VM. You might find Yunohost saves you a lot of time and hassle.

5. Get hardware suitable for your goals.

6. Install Linux, configure networking and docker containers on real hardware.

You'll need networking basics for anything you do (there are lots of courses on youtube).

After that, you can just follow wherever you curiosity carries you :)

Acronyms, initialisms, abbreviations, contractions, and other phrases which expand to something larger, that I've seen in this thread:

[Thread #49 for this comm, first seen 9th Jul 2026, 21:50] [FAQ] [Full list] [Contact] [Source code]

well what do you want to do?

i found that using alpine linux made things a lot clearer. i'm not a newbie but i gut ovurwhelmed by comylicated setups, and alpine is as simple as can be. just scripts that start and stop services, and configuration in /etc/<service-name>. where it started getting complex was with exposing services to the internet.

I started by running Jellyfin on my PC with so-so success. Next move was to get a raspberry pi to screw around with. That went kind of poorly (hosted a website that was quickly ransomwared, lessons learned). Ditched the pi and got some used PC parts on FBM to spin up an Unraid server. I've been using that for years now, with some upgrades here and there to hardware and storage. There are a couple YT channels with spectacular Unraid content for self hosting. Trash guides for the arr stack. Years later and I now handle multiple servers using either Unraid or Portainer.

As others have said, it's a very personal journey. Best of luck, happy to try and answer questions you may have.