Seedbox to Google Drive with rclone

The most common seedbox problem is not speed, it is space. You fill the disk, then spend evenings deciding what to delete. Syncing to cloud storage solves that: the box becomes fast working space, and the cloud becomes the archive.

The reason to do it from the box rather than from your computer is bandwidth. Uploading 500 GB to Google Drive over a home connection takes days and saturates your line. From a seedbox on a 10 Gbps port it is a background task you never notice.

What rclone is

rclone is a command-line tool that moves files between local storage and roughly seventy cloud providers, using one consistent set of commands. Think of it as rsync that speaks Google Drive, Dropbox, OneDrive, Backblaze B2, S3 and most things like them.

It handles resuming, parallel transfers, checksums and bandwidth limits — the things that make an unattended overnight sync survive a dropped connection.

Setting up a remote

rclone calls each configured destination a remote. Create one interactively:

rclone config

It walks you through picking a provider, naming the remote and authorising it. On a headless box the OAuth step needs one extra move: rclone prints a command to run on your own machine, you authorise in a browser there, then paste the token back.

Once done, check it works:

rclone lsd gdrive:

If that lists your folders, you are connected.

The three commands worth knowing

Command	What it does
`rclone copy`	Copies new and changed files. Never deletes anything. The safe default.
`rclone sync`	Makes the destination match the source exactly — including deleting files not present at the source.
`rclone move`	Copies, then removes the local original. Useful for freeing disk after archiving.

Use copy until you are certain. sync will happily delete a terabyte of cloud archive if you point it at an empty folder — always dry-run first with --dry-run.

A typical archive run:

rclone copy /data/media gdrive:seedbox-archive \
  --transfers 8 \
  --bwlimit 200M \
  --log-file /var/log/rclone.log \
  --log-level INFO

--transfers sets how many files move at once (8 is a sensible start; too many and some providers rate-limit you). --bwlimit caps throughput so the sync does not compete with your seeding for the port.

Running it automatically

Once a manual run works, schedule it. A cron entry that archives every night at 4am:

0 4 * * * /usr/bin/rclone copy /data/media gdrive:seedbox-archive \
  --transfers 8 --bwlimit 200M --log-file /var/log/rclone.log

Two habits that save trouble later: always write a log file so you can see what happened, and use flock so a long run is never started twice on top of itself.

0 4 * * * /usr/bin/flock -n /tmp/rclone.lock /usr/bin/rclone copy ...

Encrypting what you store

Cloud providers can see what you upload, and several scan for copyrighted material. rclone's crypt remote wraps another remote and encrypts both file contents and filenames before anything leaves the box.

rclone config          # create a new remote, type: crypt
                       # point it at gdrive:seedbox-archive
rclone copy /data/media gcrypt:

The trade-off is real: keep the password safe, because losing it means losing the archive. Store it somewhere outside the box.

When Syncthing is the better tool

rclone is for pushing to cloud storage. If what you actually want is a folder on your laptop that mirrors a folder on the box, Syncthing is the better fit — continuous two-way sync between your own devices, no cloud provider in the middle, and nothing to schedule.

A common setup uses both: Syncthing for the handful of things you want on your laptop now, rclone for the bulk archive you want off the box entirely. Both deploy in one click from the Cresdock panel.

Practical notes

Watch provider limits. Google Drive caps uploads at roughly 750 GB per day per account. Exceed it and transfers fail until the window resets — --bwlimit 8.5M keeps you comfortably under it on a continuous sync.
Do not sync your download folder. Archive the library, not in-progress torrents. Partial files waste transfers and confuse the destination.
Keep seeding. If you move files off the box you stop seeding them, which will hurt your ratio on private trackers. Use copy for anything still seeding.

Common questions

Does syncing to cloud use my home bandwidth?

No. The transfer happens entirely between the seedbox and the cloud provider, both in datacentres. Your connection is not involved.

Can I stream directly from the cloud?

Yes — rclone can mount a remote as a local folder, and media servers can read from it. Playback depends on the provider's speed and rate limits, so it is less reliable than streaming from local disk.

Will I stop seeding if I archive?

Only if you use move, which deletes the local copy. Use copy to keep seeding while a second copy lives in the cloud.

Sync your seedbox to cloud storage with rclone