What should I run and why?

Antaeus@lemmy.world to Selfhosted@lemmy.world – 21 points –

I am in the process of setting up a home server, and I am struggling to decide. I have previously used yunohost but in the meantime, freedom box has matured quite a bit. I have also looked at Tipi.

The use case right now is, running a wireguard server and probably some notes of sorts (to be decided). A web GUI for management and updating would be much desired.

Disclaimer: I don't have too much surplus of energy, due to a hectic life, so I would prefer something easy and without the requirement of docker/kubernettes

I will run on a Gigabyte Brix with:

  • AMD Ryzen 4300U (4 core)
  • 16 maybe 32 GB RAM
  • 512GB SSD

I am open to other suggestions.

P.S. I apologise if this has been debated before, but I have not really found anything.

Thank you in advance

EDIT: I have read your recommendations and arguments, and it is noted, I am watching docker tutorials now :)

37

You are viewing a single comment
View all comments
  • If you don’t have a surplus of time and energy, self hosting is not for you. You’re taking on administration that normal people effectively contract out
  • docker is worth learning and using. It’s one thing to learn, and with that knowledge you can run basically everything. It really does make your life easier

Working with traditional technologies does not take up too much of my time, but all I see in the response, is that what I am planning is wrong. I have taken note of your recommendation for learning Docker. Thank you.

Docker and the docker-compose yaml files. They'll be invaluable. Compose files allows you to create custom networking and run multiple containers.

Super useful and what most people use to run simple docker workloads.

You don't have to understand how to create containers, just understand how they work and the commands to use them effectively.

I was actively avoiding Docker too after I tried (and succeeded) getting Home Assistant running in Docker many years ago.

It seemed like a confusing mess when I did it back then and the resulting Home Assistant container ran like a dream for many years until it didn't and I had no clue how to get it working again.

I ended up just throwing Home Assistant OS on thepi and it was very very simple to set up.

Anyway that was then. This is now.

I bought a mini pc in February and installed Proxmox on it.

Initially I just wanted Home Assistant, Plex and some kind of way of populating Plex with media.

I just ran VMs with bare bones programs installed in Windows. Problem is this took a lot of RAM and was flakey.

Cut to now, where I have a Home Assistant VM, a Linux VM and an OMV VM for my NAS.

The Linux VM has a bunch of Docker containers running that do everything my Windows bare bones VM did, but better.

I can access the containers via Portainer and update them with a button press. I cannot access the VM GUI because I passed through my GPU which knackered the console in Proxmox, and that is absolutely fine, if I need to do anything in the VM I have SSH.

My Linux VM uses less RAM than my Home Assistant VM, which is amazing considering what is running on it.

Docker is where it's at! Takes a little learning but with Portainer installed it's all in one GUI instead of SSH in to create text files and folders.

Yesterday I wanted to give Immich a try. So I found a tutorial on YouTube, went into his notes and found his GitHub and in there, his Docker Compose file.

I LITERALLY JUST COPIED IT AND PASTED IT INTO PORTAINER AND PRESSED GO AND HAD IMMICH RUNNING IN MINUTES.

Now the caveat here is that I've had a few months of playing with Docker now. I've tried to get Immich running a couple times and failed in the past few months. But I watched this guy paste his code in and press go, then start talking about how it works, so I was pretty confident he had taken the time to have a working compose file.

Wall of text to say get acquainted with Portainer and try installing and playing with some stuff. Bear in mind that it probably won't work to start with and don't rely on it until you've proven it out, but tinker with it until it's working. Eventually you'll get a feeling for it and it will become simple to you.

I'm actually diving into Linux and Docker for the first time right now, and the setup you described is exactly what I'm looking to do. But seeing as this is my first time. I was hoping I'd be able to get some confirmation of what my idea is or find faults with my setup.

In terms of hardware I have on-hand, I've got two laptops. One of which is a half-top, meaning it has no screen. The other is my daily driver. Both are running Linux Mint.

The half-top has 1TB of storage and would have Docker installed with Plex/Jellyfin, the *arr suite, and HomeAssistant. On my daily driver, I would be running Portainer or Heimdall to access everything and make changes within the half-top itself. I would also use this to access any of the media I have stored in the half-top.

Does that sound like everything would work?

Yeah I don't see why not. It should be as easy as SSH in to the half top, install Docker and have it run the Portainer client then just bang Portainer on your daily driver and start throwing docker compose files at it.

Have a look at Gluetun for your VPN needs. I've basically got all my Arr in the same stack with Gluetun as the networking for the stack, then have other containers running independently that don't need the VPN, like Adguard and Homarr.

I've got a Gluetun appreciation post up that should get you started with it.