Open ESClaus opened 2 years ago
@ESClaus I don't know of an implementation like that but it wouldn't surprise me if someone had already written it. Sorry.
@ESClaus
I think everyone will end up having a different take on this, so your best option is probably to try and come up with a workflow that makes sense to you. Try it, and then adjust it based on the things that actually do or do not work well for you.
In my particular case, I have Plex in a Kubernetes cluster, and the Plex libraries are on an NFS mount. So I have a Kubernetes Job that runs, and I handle the "tracking" with sidecar files.
https://github.com/ttyS0/kubernetes/blob/main/plex/transcoding/transcode.sh
Basically, I put rips directly into the Plex library directories. Then, the transcoding job runs the transcode.sh
script on four NUCs simultaneously. The workflow is such that an .inprogress
file is first made for the source file. The source file is then copied to the NUC, and is checksum checked. Then it's transcoded on the NUC. When that's done, the remote source file is moved into an originals
directory, and the transcoded file is moved into where the remote source file was. Once that's done, the .inprogress
file is replaced with a .transcoded
file. This way, the media that hasn't been transcoded, or isn't in the process of being transcoded, can be known through the lack of a corresponding .inprogress
or .transcoded
sidecar file.
So far, this process has been working well for me, but it's also super specific to the way I have things arranged in my environment.
@ttyS0 Thanks! Your comment is much more helpful than mine. 👍
I was wondering if anyone has implemented a watch folder so that if it sees a file, it will kick off the application and dump it into another folder?
I know there is batch processing and I have implemented something like that in python but before I dive into that for myself I thought I would ask.