rffmpeg VS outrun

I got rffmpeg for distributed transcoding. Probably I understood Tdarr wrong, Tdarr leaves the files already transcoded or does transcoding on demand just like Rffmpeg? I got a lot of content on 4k that I would prefer to leave it like this for the best clients, and transcode as needed for 1080p computers

I'd love to point out https://github.com/joshuaboniface/rffmpeg

Just having support for a remote transcoder, along the lines of unicorn transcoder or rffmpeg would be a good first step to let us start homebrewing a solution.

In the rare case you need transcoding I would suggest you use a script like rffmpeg to offload transcoding to a more powerful target and/or use my script which creates trascoding workers when need be and deletes them afterwards (could be used with Unmanic as well).

rffmpeg

> Also, remote transcoding of videos seems like a huge win! Makes it so you won't need as beefy as a system to run PeerTube.

You can technically do it if you're willing to hack around. Since it uses the ffmpeg binary you could technically replace it with rffmpeg[1], the setup isn't pretty but it's doable. I haven't done it myself but have seen reddit posts on the jellyfin community that it works pretty well once setup.

[1] https://github.com/joshuaboniface/rffmpeg

If it's a normal Jellyfin install you can just use the rffmpeg python script directly, or you could run docker on Windows or WSL.

Maybe in the future you can use multiple machines to do your transcoding?

rffmpeg is supposed to do that. I haven't used it.

It should hopefully be pretty straightforward. We've got a user who's been working on autoscaling it with Kube and Hesner, so might be worth checking that out (see https://github.com/joshuaboniface/rffmpeg/issues/34 and specifically the last comment; his profile has his repos).

While it's purpose is different it can be used to do distributed compiling, so I'll leave it here.

https://github.com/Overv/outrun

Since I was just going down this rabbit hole recently, I kind of wonder if it's possible to set the filesystem on something more like the BitTorrent protocol so things like the libraries/compilers/headers that are used during compilation dont all need to come from the main pc. It probably wouldn't be useful until you reached a stupid number of computers and you started reaching the limits of the Ethernet wire, but for something stupid that can run on a pi cluster it would be a fun project.

Your mention of compile heavy workloads reminded me of a project called Outrun, it offloads work to another machine. All it seems to require is Python, Fuse3 and ssh.

Awesome! This write up is satisfyingly detailed. Prior work in this space includes Plan9 of course, as well as the python project Outrun, which has it's own RPC-based FUSE FS: https://github.com/Overv/outrun

Other approachs to deployment in particular include the functional package managers Nix and Guix, which can create lightweight application images, and could probably be cobbled together into some sort of remote environment replication even across architectures. As I read on, I thought less about how this compares with Guix in regards to application/environment packaging and more about how these things could be glued together in interesting ways, because I think the intro leads in through slightly off-label examples, if that makes sense. Application packaging isn't what this addresses at the end of the day, but it's no less fascinating for it.