I have been torrenting the same primitive way for a long time. Turn on a VPN, open up a browser I only use for this purpose, go to tpb or yts, grab my movie, and shut it all down when the movies over.
I’ve only updated my torrent client in this time, the method has not evolved.
Now im interested in self hosting jellyfin with an arr stack, with the end goal being to share it with friends and family outside my lan. I’d like to use docker containers so its all containerized, and of course keep it safe and secure.
What kind of set up, from hardware to software, would you recommend to get this going? Any guides in particular? I’m especially iffy on allowing remote access for non-tech savvy family (like a roku app), so any tips/guide recs for that would be helpful.
I’ve been searching around some and I’ve found a lot of resources but I’d like to get the opinions of people in here before diving in.
I have some beginner questions, for example: if I have the arr stack running in docker with a vpn, can I browse the internet non-anonymously on that same machine without compromising identifying details, assuming qbittorrent is configured to only move traffic through a VPN? (I’m wondering if I need a dedicated piece of hardware to run everything safely)
Tldr: Suggestions or guides for beginner setting up jellyfin/arr/ remote access for family?
I recently had my first foray into this setup and I have found the arr stack to be all but useless on public trackers. I haven’t setup flaresolvarr so that I can use 1337x and maybe that is why it is so terrible for me but prowlarr couldn’t even find a fast and furious movie that is a popular, well seeded film.
Maybe the addition of 1337 would make it much better but I’m skeptical as to how much difference that one place can make.
I ultimately now just find what I need manually and add the torrent to my server via the web interface then leave it and everything else sorts itself, which I’m fine with doing.
Maybe I’m missing something but I think the arr stack is massively overrated on public trackers. I’m going to look into Usenet next as someone else suggested but it isn’t anything I’ve ever used and I don’t have the money to pay for any right now anyway.
What do you use to clean up the files for jellyfin, or does that happen automatically? I mean things like cover art, consistent title format, etc
I appreciate this take. Maybe I should keep it simple
Don’t get me wrong I don’t regret setting it all up as it was a good, if somewhat steep learning curve for me as someone with no Linux experience beyond installing a couple of distros and using them solely as GUI OS’s. In all honesty though keeping it simple and having full control over the files I’m getting is working absolutely fine for me and I don’t feel the need to make it any more complex with the various arrs, the only exception being I want to look deeper into bazaar as I need subtitles for everything.
Anyway I do absolutely nothing to clean up the files. I copy my magnet links into qbittorrent and set which folder I want it to download to, movies / shows etc. Jellyfin libraries are set to each of these folders by their respective type and then when some new stuff is downloaded I’ll just access my jellyfin dashboard and scan all libraries.
So far with zero file cleaning jellyfin has detected 85% of things correctly regardless, for anything not detected it takes barely any time to go into the identify menu, search for the correct title and choose the relevant correct info.
I’m running proxmox with 3 VMs, one for docker with the arr stack, qbittorrent and gluetun to direct all that traffic via my VPN. One for a Plex server (which hasn’t failed to detect anything so far without file cleaning) and one for the jellyfin server.
I have some beginner questions, for example: if I have the arr stack running in docker with a vpn, can I browse the internet non-anonymously on that same machine without compromising identifying details, assuming qbittorrent is configured to only move traffic through a VPN? (I’m wondering if I need a dedicated piece of hardware to run everything safely)
The answer to this question is you can setup a docker system (or podman) so that all the traffic in that pod (don’t know the docker term for this) will route through the vpn. A good image to accomplish this easily and successfully is
gluetun– and it will only affect the traffic in the containers, not the rest of your computer.Personally, my setup is much more like yours and it works fine for me, except I use a VM. So all the activity gets confined to the VM and that makes a bit idiot-proof. Using automatic management in the torrent client, completed torrents get put in the correct directory. You could combine this with Jellyfin if you desired.
My own problem with Jellyfin is if I ever use it for anything I want direct playback on all relevant devices, because my computer is not good enough for transcoding (and why waste the energy and time on on-demand transcoding, anyway?) so it requires some massaging of the data to get everything right. I only use it infrequently, practically on-demand. I don’t use Jellyfin for myself.
Thanks for taking the time to write up a detailed response I appreciate it. What is your vm workflow/setup?
The VM is Debian Linux with a basic XFCE UI (for a system tray + notification widget) via QEMU/KVM which I run through virt-manager. Most unnecessary packages are removed or not installed in the first place. This is so that I can browse the sites, again, in a fool-proof manner. I share a directory from my host OS to the VM, which mounts it on boot in the fstab. This prevents me from downloading into the guest VM’s disk image and having to keep dealing with that file getting overly big. In the past I’ve done a Samba share but recently I’ve just been using direct shared memory/filesystem and that seems to work OK, too.
As a bonus to this setup, I can use Microsocks in the VM to also proxy a profile in Firefox to get VPN coverage in a specific Firefox profile. I use this when watching on streaming sites instead of trying to watch within the VM, since there is considerable overhead to doing that.
And that’s it, really. My VPN killswitches the VM if it ever experiences a connection interruption. And Qbittorrent is set up to use the VPN interface, as well. I use the aforementioned automatic torrents management feature to sort things when they’re done downloading.
I should state that there are some obvious downsides to this setup. The first is now I have to overcommit disk space and RAM to keep and run a guest VM. You want enough to be able to run updates and the software in the VM without running into a wall. The second is that there does seem to be a CPU penalty when downloading files (maybe it’s because of the way I’m sharing the downloads directory into the VM with virtiofs?)
Does anyone know of a client for AirVPN that supports the split tunnel feature? I’ve been using the Eddie client, but when I enable the VPN, I can’t watch YouTube videos because they require me to sign in, which I’d rather not do. I’m looking for a way to exclude my browser from the VPN.
If AirVPN supports OpenVPN (it probably does), there is a way to have an OpenVPN connection that is not actively used unless you specify it for the app which you want.
For instance with qBittorrent you can select to use the VPN network interface in the settings. This has the nice advantage that it also acts like a kill switch.
It is a little bit finicky to set up for the average user, but if you have some basic tech skills it should be easy with the right guide. If you’re interested I canl try to find it.
