How it works

How it works

Every episode (a film counts as one) goes through the same stages, and the web page shows the current one as a chip:

Shown as	What happens
Creating links	Each hoster link goes to Debrid-Link, which answers with a direct download link.
Downloading	The archive parts are fetched into archives/. An interrupted part continues where it stopped.
Downloaded	Only when unpacking was switched off for the download: the parts are on disk, Unpack goes on.
Unpacking	7-Zip looks at the archive, finds the password and unpacks it into output/.
Password needed	None of the known passwords opens the archive. The episode waits until you enter one.
Sorting files	The video, its .nfo and an info.txt are moved out of output/ and renamed.
Done	The files are in place.
Failed	Something went wrong, and the episode shows what. Retry starts it again.
Canceled	The episode was canceled. Retry starts it again.
Paused	The download was paused. Resume carries on.

A few rules follow from that:

• Links are created one after the other, with the pause Debrid_Request_Delay between two requests. An episode starts downloading as soon as all of its links exist, while the other episodes keep getting theirs.
• At most Concurrent_Downloads archive parts are downloaded at the same time, counted over all downloads together.
• As many archives are unpacked at the same time as the server has CPU threads, because 7-Zip unpacks one archive on one thread. An episode that finds every place taken waits for one. This is not a setting, and a Docker CPU limit lowers it.
• After sorting, the files of no use are deleted: .md5, .download, .url and .uri files, and an image that is named like the video (see What ends up on disk).
• When every episode is done, archives/ and output/ are removed (see What ends up on disk).

The settings are described on the Configuration page.