Hey,
I know it sucks to rely on cloud services but it is what it is. I use Apple iCloud, Bitwarden and GitHub.
Technically, I could self-host all three but I want my backups not at my place or at least have them in both places cloud/at home.
I do have two spare Raspberry Pi Zero 2W and one small computer with an old i5 / 8 GB RAM.
What do you personally self-host?
You found a solution and you are looking for a problem.
This is not the way it goes. You identify a problem, and if the solution involves self-hosting then you solve it with self-hosting.
This.
I have three servers running these days.
One is a NAS that hosts the .arr suite and my torrent client. This is just to keep the media management in one place.
A N100 NUC that runs a lot of stuff in Proxmox, like Jellyfin, Heimdal, HomeAssistant, PiHole, Tailscale. I hope to add Caddy to this in the future, but I’ve never played with a reverse proxy before so I’m a tiny bit scared hehe.
Lastly is a inudstrial PC I got from work that hosts game servers. Right now it’s down as we haven’t had time to game, but usually it’s either a Minecraft server or Valheim.For backup I have one copy on the NAS and I upload the most critical data to a cloud service I trust and pay for. This is now Proton.
My dream is finding a tech friend with his/her own NAS so we can set up a encrypted partition on each others NASes for automatic backup. I give you 1 TB, you give me 1 TB, life’s good!I self host jellyfin, free anime for me and my closest friends
Just get a vps it still counts as self hosting if you run all the software yourself.
I have an expensive VPS at AWS which I mainly use to host my projects :D
Get a cheap one on lowendspirit.com instead.
Try netcup instead.
I have a couple of friends with nextcloud, and I have nextcloud too. Low tech ish? But we just host our files on Nextcloud and then copy backups to the other machines every now and then.
My NC uses about 6gb of RAM, and it is really badly optimized, since it’s been running forever and isn’t a container, or even a server deploy. (It’s a snap running in desktop Ubuntu since 2016.)
Anyone could do better, I just can’t be bothered.
My buddy has his running on 1.5GB of RAM in a container.
I also host a bunch of other stuff. Navidrome and freshrss get the most use, other than Nextcloud. Immich, searx-ng, jellyfin, guacamole.
I never became friends with Nextcloud for some reason :/ Long story short: I have lots of files but not large ones. For nextcloud there’s no deep system integration like Apple iCloud has (basically just enter your credentials and everything else is handled and it’s fast + end-to-end encrypted). I’d set up an instance if I had frens who would use it as well
deep system integration
You mean remote-mounting a filesystem? That’s possible; you don’t need NC for it.
deleted by creator
I used proxmox and its made backing things up pretty simple. Once a week I have backups copied up to filen for remote storage.
I host: Heimdall - easy homepage for my wife and I to access services. Gitea - git stuff. Photoprism - running into a number of bugs and considering switching to immich. HomeBox - asset tracking with associated documentation NGINX reverse proxy - what it says on the tin BookStack - simple book style wiki Portainer - docker container management Paperless-ngx - Helps organize documents. Jellyfin - media server
I’d love to have my own movie collection with Jellyfin. How’s your experience so far? Do you use torrents and how do you protect your IP?
As with most things: it depends…
If you’re in a country where ISP’s freely give out user info, I’d say you should have a VPN.
If you’re on a private tracker, you might not need it, but you never know if the people hunting pirates managed to get in there too.
I don’t use one as our ISP’s mostly throw those letters in the trash and I’m in a private tracker, but your mileage may vary.To get started, you only need a server (like Jellyfin or Plex) and a torrent client. Then you can automate it with the .arr stack, such as Radarr and Sonarr, race others with autodl-issri/Autobrr, share your media with friends and family with open ports (not recommended) or Tailscale/Netbird…
It gets as advanced as you yourself want it to be.Feel free to ask if you have any questions, not just about piracy but how to set things up in general.
Good luck, and remember to have fun while doing it. If you don’t, you won’t bother keeping it updated and working in the future./c/piracy would probably be a better resource for you. The few comments from that side I’ve seen mostly indicate torrenting through VPN and using an arr stack.
I mostly go garage sale hunting combined with refusing to buy anything with DRM. My library probably is not as extensive as what you would have if you are a movie or TV aficionado. I use it for movies, tv shows, and music. It’s been great so far. Paired with Finamp as a player on my phone and Supersonic on the desktop I can get around the otherwise somewhat awkward music library management. I love the show and movie default interface.
Note, I believe Jellyfin is also starting to do books. I have not tried it in that application yet.
I don’t do a whole lot right now, partially because I’m just not sure what I want to run or how much risk I want to take on in terms of security (for exposed services) and data recovery. I started with pihole to get better ad blocking on my tv, and recently started hosting a dinner vtt server for my d&d group. Some time in the future, I might spin up a nextcloud instance to replace my dropbox or immich for google photos
lmao, people out here on VPSs acting like they went from renting to owning, when really they just rented the motherboard, the case, the fans 🤣🤣
OP get an end-to-end encrypted remote backup app.
I own my data. I own my installation. That’s what I care about.
Why would I want to own the hardware, when it’s in an inaccessible building far away.
I guess for me it kind of depends on your definition of “self host” as 90% of what I host is a hetzner server running out of finland. because well that’s off site backups lol.
my setup is.
Local: Frigate (CCTV manager), Homeassistant (home automation), Matrix (chat).
Remote: Mealie (recipe collection), Vaultwarden (works with bitwarden clients), Nextcloud (files and documents), Freshrss, gitea (github alternative)
Now in terms of wanting an offsite backup, you are probably right, assuming you don’t have something offsite that you can syncronize with, and assuming you don’t have any major privacy fears of what is hosted, those things are probably best to use cloud for, assuming you are more worried of losing everything in a house fire, than you are of say the stuff being spied on by a 3rd party or caught by hackers.
So yeah I’d say, personally in things I like to have self hosted… on site, probably I’d say a local messanger is good if you’d like a reasonably private communication for friends/family etc… Niche things like RSS readers, or recipe books, really anything strange niche you can probably search for some program to self host it.
I think the only thing that makes me worry a little bit is that Microsoft has my source code. It’s not “bad” but sometimes I think about it. So I probably need a gitea instance then.
Edit: it’s interesting that you mentioned private messenger. I just built one (using Bluetooth Mesh routing with a basic encryption for 1-on-1 chats) and am waiting for App Store approval. :)
Forgejo is what you should use nowadays: https://lwn.net/Articles/963095/
Microsoft has my source code
FWIW, other public git instances exist. As in alternatives to github. No need to selfhost.
You can host git anywhere you have ssh or https. I don’t know if you can push over https though.
I know that. My point is: you don’t have to.
Encrypt before backing it up remotely.
Not sure that really works for git though… at least with regards to it’s primary usage.
git isn’t just a backup… it’s about version control.
IE the point is if you know what you are doing, you realize this function isn’t working in this edge case, you can search through and find out, when did this part of this file change… and what was it before, and it will basically find exactly that.
If you encrypted it so that git couldn’t actually read the contents, then you basically reduced a crazy powerful tool, into a glorified dropbox. (IE yeah you could revert back to previous versions… but you’d basically be counting on your memory for what you changed when, if the git server can’t read the files).
Good point and teaches me to be too quick to respond. Cheers!
