The one point that has basically been solved is NAT traversal. Thanks to Wire guard, Tailscale and the like. The relevant parts are open source and can be used basically as a library.

First my context: I’m also running multiple Proxmox hosts (personal and professional), and havea paperless-ngx instance (personal/family). I tried Firefly, but the effort required to get it to a point where it would be if use to me was too high, so I dropped it. Haven’t used n8n.

For the setup I’d just use the Proxmox community scripts, if you haven’t heard of them. Makes updates trivial and lowers the bar to just trying something to basically zero.

Paperless-ngx I actually use, cause it means I can find something when i need it. It’s all automatically ocr’d and all you have to do is categorize them. With time, it’ll learn and do this for you. You can (manually) setup your scanner to just directly upload files to the “consume” folder and it just works. PC/server power is near irrelevant, it just means OCR takes slightly longer, otherwise it’s a web server. You can run this just fine on a raspberry pi.

I don’t have any real automation setup, so I can’t really comment on that. My advice is to just install it, see what it does and how it feels. Try to anticipate if and how much automation you need. Many aspects of all this are of the “setup once” variety, where once it’s working, you don’t have to touch it again. Try to gauge if the one time effort is worth it for you, then go from there. As I said, it was fine for paperless for me, but not for Firefly (but I might need to revisit this).

On Linux, running Jellyfin through docker with GPU acceleration works fine, yes. But you need some options/flags to pass access to the GPU to the inside of the container. Guides and/or docker tutorials exist and should contain that, as that’s basically the default setup these days.

As for Bazzite and Docker (I just checked), no it isn’t part of the base image and you can’t easily install it. That’s the downside of an immutable distro. I think podman is available, which is compatible and FOSS, but there may be caveats to using that. There is a bazzite version called bazzite-dx intended for developers, so that one would probably work fine for you out of the box. There shouldn’t be any real downside to using that compared to the mainline image, apart from being slightly larger cause all dev tools are installed, but do check that. My practical experience with Bazzite is limited.

My real recommendation is: just try it. Slap in a small/cheap SSD (~20 bucks) instead of whatever you got in there now, install CachyOS and try it out. Then install Bazzite and try it out. By “Try it out” I do mean setting up a copy of or a test-install of your required services (arr stack, jellyfin, …), to see if everything is as you’d expect. Possibly install more distros to try them out, then make up your mind and actually fully migrate, or if it doesn’t work out go back to your currently installed drive. Installing a linux distro takes like 10 minutes these days, then play around with however long you need. Since you already have it narrowed down to only 2 options anyway, that is most likely the best solution.

There’s a lot of well meaning but not too well informed advice in here. Since one of your goals is gaming, stay away from Mint. It can be made to work (well), but you have to get there. It’s basically the recommendation people gave for decades, but there have been massive improvements through many distros while mint just kinda stood still. There’s still some things they do rather well though.

CachyOS will do what you want it to, and it is what I switched to like 8 months ago. It isn’t maintenance heavy at all if you don’t want it to be. I think I had to intervene once since I started using it, but that intervention was necessary or it wouldn’t have booted after updates. The official updater will tell you when that’s the case, as it lists critical news like that. Otherwise it just works, and it’s pre-configured and optimized for gaming. Under the hood it’s basically Arch, just without the fiddling of getting it to a usable state. Because of that they’re is also an enormous amount of information out there (Arch wiki) on how to do stuff.

Bazzite is a stark contrast in many ways as it’s an immutable distro, but also pre-configured and optimized (maybe not quite as much as CachyOS). It will also do what you want just fine. It is relatively “safe” due to the immutability, and updates are much rarer (and by definition always whole system updates). I don’t know exactly how you’d run your services, but assuming they are dockerized or similar that should be just fine, but please do some searching before if it does contain what you need in the base image (presumably docker and docker compose).

DuckDNS had been unreliable when I used it, but it’s been a while. I swapped over to desec.io but their signups aren’t always open. Can highly recommend them though, and they offer many paths to update the IP, including DynDNS(2) protocol or just ddclient.

Also works with certbot for Let’s encrypt certificates using dns challenge.

Never run something like Vaultwarden with unencrypted traffic. Throwing in a self signed cert is basically free insurance. You never know when even in your “trusted network” something starts listening in. Just why risk it?

Don’t buy anything by crucial, as you’ll have trouble getting a 2nd stick in the future. They are shutting down their end user business.

Ssh over Internet is fine as long as it’s properly setup (no password auth, root not allowed, etc.). Obviously a VPN is even better.

Or if you have separated your devices into subnets/VLANs. Which becomes more important as your get more hardware that you don’t really trust.

Multiple times a day and many times a day isn’t necessarily the same thing. Also just having a 1-2 hours long timeout might still be a viable option preventing repeated spin ups, but still allowing spin down during longer unused periods.

While probably not worth it for your particular case, it might will be for others reading this. Ideally, one could observe the access patterns for a while a find a suitable timeout setting.

While that is still true, unless you spin them up many times a day, it’s a non-issue. Set your timeout to at least 30 minutes. Most jellyfin servers are not gonna be used fir many hours at night, for example. Or when everyone is at work/school.

These days, power is expensive for have people. If I keep my drives spinning 24/7, it’ll cost me around 150€ per year. If they spin when needed, it’ll cost me whatever percentage they are spinning, so in practice they are on for like 1-3 hours a day. So let’s say 20€ per year.

Tailscale is WireGuard under the hood, if you didn’t know. It’s an overlay network that uses WireGuard to make the actual connections, and has some very clever “stuff” to get the clients actually to connect, even if behind firewalls without needing port forwarding.

Using WireGuard directly basically just changes the app you use, which may or may not help with your issues. But the connecting technology is the exact same.

I assume you mean unsolicited phone calls with this? I haven’t gotten any of those in about a decade, if not more. And those were isolated cases as well. We have laws against that sort of thing. It’s not been a problem for a very very long time (early 2000s or so).

I’d suggest looking into TeamSpeak, like others have mentioned. Trivial to self host, too.

Edit: to be clear, this would cover the voice call aspect of discord, not the chat channels and other community tools. While it’s can do text chat, it’s more of a side feature rather than core. I didn’t think it does images or video, but it’s been a hot minute.

but you can do everything without it.

yes but why would you? There’s a reason we use GUIs, especially when new to a field (like virtualization).

Didn’t notice anything like this, not even slightly.

Mine are of course also on a VLan but with no Internet access unless they need it for everyday operation (like a radio, or the amplifier that can play Spotify).

We don’t use the manufacturer apps at all. Everything is integrated into (fully local) home assistant. No need to open a specific app to operate a switch, or a light. Everything in one place. Trivial and incredibly clear. Things that can be are of course automated.

Just because it’s a “smart” service doesn’t mean it has to connect to the Internet or a server or the manufacturer. If it does neither, it can’t be turned off by them.

All my devices run local-only protocols. Nothing leaves my house. The devices that would be proprietary were reflashed to tasmota (fully open source, local only). Others are either Zigbee or Shelly. While Shelly has a cloud connection, it’s fully optional and disabled by default (including automatic updates). The hardware is also supported by tasmota, and reflashing is always just 5 minutes of effort away.

There is absolutely nothing that any manufacturer has to do to keep my stuff working. I have to do a little something (keep my tiny server on, basically). But more importantly there is nothing any manufacturer can do to stop my stuff from working.

While it’s fantastic software, it’s probably a relative cannon to shoot at his problem. Maybe there’s a way around this, but I’ve found the necessary management, curation and bookkeeping that was necessary for it too be useful to be just way too much to be worth it. I mean it’s fun for some, including me to a degree, but not too this extent.

UnRaid doesn’t provide anything I am interested in, at all. Currently running TrueNAS for main storage and proxmox for virtualization, both ZFS based. If TrueNAS ever enshittifies, I’d run some bare metal Linux with ZFS. My workstations also run ZFS as the file system, making backups trivial. VM snapshots and backups of any system are trivial and take seconds (including network transfers).

I never understood why I’d even consider UnRaid for anything.