API call to import (not upload) and existing file on the local filesystem

[AngusThermo-Pyle]

Describe the problem to be solved

I've had troubles uploading large files, large numbers of files, and existing local files. Tweaking ngnix until I get disk speed uploads helps, but then I run into timing issues with large file copies. I have ~18TB of 2-4 hour 2160p content to put online and even if existing upload methods functioned on this content, the time wasted "uploading" content that is already on the local filesystem wastes large amounts of time and resources unnecessarily.

Describe the solution you would like

What would be far faster for sites that have existing content or regularly import content from local sources would be an API call that skips the upload portion of new content creation and starts with a local filesystem path and processes the file in the same way it would after completing an upload. This method could also be extended to the web UI or even a "watch folder" for automated new content uploads.

[Chocobozzz]

Tweaking ngnix until I get disk speed uploads helps, but then I run into timing issues with large file copies.

Can you paste the error you have?

[AngusThermo-Pyle]

The ffmpeg errors are reported separately as a bug in #6757

The feature request is more about bypassing the http upload when a file is already available on the local filesystem in order to be able to batch add large amounts of content without doing a local http upload, which massively increases the time to add large amounts of content.