Ben Sarmiento

Enhancing Rclone's Bandwidth Efficiency

Added 2024-06-24 21:44:26 +0000 UTC

Hello!

For those utilizing rclone with Zurg, it's important to understand how rclone manages data requests. When Plex accesses a file, rclone requests the file from Zurg via an HTTP range header (learn more about range headers). This header specifies the byte range for the data request, which Zurg forwards to Real-Debrid. Real-Debrid processes this request and logs the data served as traffic used.

By default, rclone requests the first 128MB of data. If you only use 5MB, the remaining 123MB still count against your bandwidth, as Real-Debrid has already prepared this data. This inefficiency increases because rclone doubles the chunk size with each subsequent request—128MB, 256MB, 512MB, and so on—unless you set a limit. Without chunked reading, rclone requests the entire file, consuming vast amounts of bandwidth for large files.

To improve efficiency, consider configuring rclone with the following flags:

--vfs-read-chunk-size 1M: Starts with a 1MB chunk size.
--vfs-read-chunk-size-limit 32M: Caps the chunk size increase at 32MB.

This configuration ensures rclone gradually increases the chunk size but only up to a 32MB limit, significantly reducing wasted bandwidth. Read more about chunked reading here.

Additionally, you can manage bandwidth usage with: