All in One View

Content from What is Rclone

Last updated on 2026-08-23 | Edit this page

Overview

Questions

  • What is Rclone?
  • Why might you use this application?
  • How might Rclone help you manage your data?

Objectives

  • Basic understanding of Rclone usage
  • Know where to download the software and documentation
  • Know where to get help and examples

Introduction: What is Rclone?

Rclone is a command-line program for managing files on cloud storage. It is primarily used for transferring, syncing, and moving data between local systems and various cloud storage services.

Test Illustration
Test Illustration
Challenge

What can you do with Rclone?

Which statement best describes what Rclone does?

A. Rclone runs continuously in the background, syncing your computer with the cloud in real time, like a Dropbox or OneDrive desktop client.
B. Running rclone copy once is a complete backup strategy: it keeps the destination in sync with every future change to the source.
C. Rclone transfers, syncs, or moves files between local and cloud storage, or between two cloud storage services, each time you run a command.
D. Rclone requires mounting cloud storage as a local drive before it can move any files.

The answer is C.

Rclone is command-driven. It acts when you run rclone copy, sync, or move, not automatically in the background.

  • A is incorrect. Unlike a desktop sync client, rclone does not watch for changes and act on its own. You have to run a command (or schedule one) each time you want it to check for changes.
  • B is incorrect. copy never deletes files at the destination, so the destination can accumulate files that no longer exist at the source. A single copy does not guarantee the two locations match: sync, covered later in this lesson, is designed for that.
  • D is incorrect. Mounting is an optional, more advanced feature. Copy, sync, and move all work directly on cloud storage without mounting anything.
Discussion

How do you think you might use Rclone?

Think of a file-transfer task you currently do by hand: dragging files between a laptop and an external drive, uploading through a browser, or copying the same files to more than one place.

  • Would copy, sync, or move fit that task best, and why?
  • Are you moving files between two machines, two cloud services, or a machine and a cloud service?
  • Is there a point in your current process where you’re not sure whether a transfer actually finished, or whether you’d notice if it didn’t?

Share one example in the chat or with a neighbor.

Rclone command syntax

rclone [command] source:source-folder destination:destination-folder

List of Rclone commands: rclone.org/commands/

Real-World Scenarios for Using Rclone

Understanding how Rclone works can be easier if you relate it to everyday tasks. Here are a few simple examples:

  • Copy:
    Imagine you have a folder of vacation photos on your laptop and want to back them up to Google Drive without removing them from your computer. Using rclone copy will duplicate your pictures to the cloud, leaving the originals intact.

  • Sync:
    Suppose you’re working on a project at home and in the office. You keep a copy of your project folder on your computer and a backup in the cloud. Each time you run rclone sync, it copies any local changes to the cloud folder and removes any cloud files that are no longer present locally, keeping the two folders identical.

  • Move:
    After completing a project, you can free up space on your computer by archiving files to the cloud. The rclone move command transfers your project files to cloud storage and deletes them from your local drive.

These examples show how each rclone command can help you manage your files based on your needs.

Key Points
  • Uses for Rclone
  • Where to find versions for specific operating systems
  • Where to get help

Content from Creating and Configuring Remote Connections

Last updated on 2026-08-23 | Edit this page

Overview

Questions

  • What is an rclone remote, and how is it different from a local path?
  • How do you create, list, and verify a remote connection?

Objectives

  • Understand what a remote is and how it differs from a local connection.
  • Distinguish a local filesystem path from a remote path in the form remote:path.
  • Use rclone config to create a remote connection.
  • Use rclone listremotes and rclone lsd remote: to confirm a remote is configured and reachable.

Creating a remote connection

A remote is a named storage configuration that rclone uses to connect to a storage location, such as a cloud service or another machine. You create a remote once using rclone config, then refer to it by name in commands as remote:path (for example, mybox:documents). This is different from a local filesystem path, such as /Users/name/documents or C:\Users\name\Documents, which refers to storage on your own machine and has no remote: prefix.

Common remote types include cloud storage services such as Box, Google Drive, and Amazon S3, as well as other servers accessed over protocols such as SFTP.

flowchart LR
    accTitle: Rclone connects a local machine to one or more named remotes
    accDescr {
        A local machine connects through rclone to three example remotes, named mybox, mygdrive, and mys3, each labeled with the storage provider it points to. A dashed line shows that rclone can also transfer files directly between two remotes without passing through the local machine.
    }
    L["Local machine<br>/Users/you/documents"]
    B["mybox:<br>(Box)"]
    G["mygdrive:<br>(Google Drive)"]
    S["mys3:<br>(Amazon S3)"]

    L -- "copy / sync / move" --> B
    L -- "copy / sync / move" --> G
    L -- "copy / sync / move" --> S
    B <-. "copy (remote-to-remote)" .-> G
Callout

Custom client IDs: Google Drive vs. Box

By default, rclone uses a shared client_id for both Google Drive and Box, so most learners do not need to set up their own credentials to complete this lesson.

For Google Drive, rclone’s shared client ID is being retired and will stop working during 2026. Create and configure your own client_id and client_secret to avoid an interruption. See Making your own client ID in the rclone documentation for details.

For Box, the Box documentation notes that custom client credentials can normally be left blank. Most users do not need to set up their own client_id for Box; only do so if you hit rate limits or your organization requires it.

Configuring a remote with rclone config

The rclone config command starts an interactive session where you create, edit, and manage remotes.

BASH 

rclone config

The first time you run it, with no remotes configured yet, you’ll see:

OUTPUT 

No remotes found, make a new one?
n) New remote
s) Set configuration password
q) Quit config
n/s/q>

Choose n for a new remote. Rclone asks for a name, then the storage type from a long list of supported providers:

OUTPUT 

Enter name for new remote.

Option Storage.
Type of storage to configure.
Choose a number from below, or type in your own value.
 1 / 1Fichier
   \ (fichier)
 2 / Akamai NetStorage
   \ (netstorage)
...
 7 / Box
   \ (box)
...
36 / Local Disk
   \ (local)
...

Type mybox for the name and box for the type. Accept the defaults for client_id and client_secret unless the callout above tells you otherwise. For a remote that needs authorization, like Box or Google Drive, rclone will next either open a browser window for you to sign in, or, on a machine without a browser available, print a URL and code to enter on another device — see Configuring using rclone authorize in the rclone documentation for that case.

Once you’ve stepped through the provider-specific prompts, rclone shows a summary and asks you to confirm:

OUTPUT 

Configuration complete.
Options:
- type: box
Keep this "mybox" remote?
y) Yes this is OK (default)
e) Edit this remote
d) Delete this remote

Answer y, then q to quit the config session. Your new remote is saved.

Checking your configuration

Once a remote exists, use these commands to confirm it’s there and working. None of them transfer or change any files, so they’re safe to run at any time.

List every remote rclone knows about:

BASH 

rclone listremotes

OUTPUT 

mybox:

Find out where the configuration file itself lives — worth knowing, since it’s the file that holds your credentials (see the callout above):

BASH 

rclone config file

OUTPUT 

Configuration file is stored at:
/Users/you/.config/rclone/rclone.conf

List the contents of the remote itself:

BASH 

rclone lsd mybox:

OUTPUT 

          64 2026-08-23 13:27:28        -1 rclone-workshop

lsd lists directories only and never transfers a file, which makes it a safe way to confirm a remote is reachable before you run a real transfer command.

Rclone command flags

There are numerous command flags but these three are especially worth remembering:

  • -n, –dry-run Do a trial run with no permanent changes
  • -i, –interactive Enable interactive mode
  • -v, –verbose count Print lots more stuff (repeat for more) - useful when debugging
Challenge

Configure and verify your own remote

Using rclone config, create a remote connected to your own Box account (or whichever storage your instructor has set up for this workshop). Then:

  1. Confirm it appears in rclone listremotes.
  2. Run rclone lsd remote: to see its contents.

If rclone lsd returns an error, run rclone listremotes first — a typo in the remote name is the most common cause.

BASH 

rclone config

Follow the prompts: n for a new remote, a name such as mybox, box for the type, defaults for the client ID and secret, and complete the authorization step. Confirm with y, then q to quit.

BASH 

rclone listremotes

OUTPUT 

mybox:

BASH 

rclone lsd mybox:

If your Box account has no folders yet, this returns nothing — that’s expected, and still confirms the remote is reachable.

Documentation specific to each remote

Rclone includes extensive documentation that is specific to particular remotes.

Examples:

See the Learner Reference page for a full command and documentation reference.

Key Points
  • A remote is a named storage configuration; a local path has no remote: prefix.
  • Use rclone config to create, edit, and manage remotes.
  • Use rclone listremotes, rclone config file, and rclone lsd remote: to confirm a remote is set up and reachable, without transferring anything.
  • Google Drive’s shared client ID is being retired in 2026 and needs your own credentials; Box’s client ID can normally stay blank.

Content from File Transfers, Listing, and Verification

Last updated on 2026-08-23 | Edit this page

Overview

Questions

  • How are files moved or copied?
  • What does the rclone sync do?
  • How do you see what is already in the destination?

Objectives

  • Understand the difference between copy and sync
  • Be able to list what is already in the destination
  • Compare rclone command syntax between Linux/macOS and Windows (including WSL2)

Moving files around

Rclone is most frequently used to move files, individually or as a group from one place to another.

Specifying source and destination paths

BASH 

source:folder  destination:folder

Moving, syncing and knowing what is already there

Lists contents of a remote:

BASH 

rclone ls remote:path

Copy local files to remote:

BASH 

rclone copy /local/path remote:path # copies /local/path to the remote

Sync local files to remote:

BASH 

rclone sync --interactive /local/path remote:path # syncs /local/path to the remote

Filtering

Many rclone commands resemble their Unix counterparts, such as ls. Rclone selects which files a command applies to using flags such as --include and --exclude, rather than shell wildcard expansion, so quote the pattern (for example, --include "*.txt") to keep your shell from expanding it first.

Some Examples:

Copying local files to an external drive (Windows using the Linux subsystem)

BASH 

rclone copy /mnt/d/work-related /mnt/f/work-related-backup

Copying files from Google Drive and filtering for *.txt

BASH 

rclone copy rclone-intro-google:rclone-intro-google rclone-intro-box:rclone-intro --include "*.txt"

Checking a result before running by using the -n flag

BASH 

rclone sync rclone-intro-google:rclone-intro-google rclone-intro-box:rclone-intro -n

OUTPUT 

2025/02/09 14:52:11 NOTICE: Beans, Snap and Italian – Pieces, Green and Wax - National Center for Home Food Preservation.pdf: Skipped copy as --dry-run is set (size 91.581Ki)
2025/02/09 14:52:11 NOTICE: For Safety's Sake - National Center for Home Food Preservation.pdf: Skipped copy as --dry-run is set (size 69.982Ki)
2025/02/09 14:52:11 NOTICE: Preserving_Food__Using_Pressure_Canners.pdf: Skipped copy as --dry-run is set (size 3.839Mi)
2025/02/09 14:52:11 NOTICE: Selecting the Correct Processing Time - National Center for Home Food Preservation.pdf: Skipped copy as --dry-run is set (size 123.316Ki)
2025/02/09 14:52:11 NOTICE: test01.txt.docx: Skipped copy as --dry-run is set
2025/02/09 14:52:11 NOTICE: USDA-Complete-Guide-to-Home-Canning-2015-revision.pdf: Skipped copy as --dry-run is set (size 16.518Mi)
2025/02/09 14:52:11 NOTICE: Potatoes, Sweet – Pieces or Whole - National Center for Home Food Preservation.pdf: Skipped copy as --dry-run is set (size 87.602Ki)
2025/02/09 14:52:11 NOTICE:
Transferred:       21.079 MiB / 21.079 MiB, 100%, 0 B/s, ETA -
Checks:                 3 / 3, 100%
Transferred:            8 / 8, 100%
Elapsed time:         2.0s
Challenge

Predict, then verify

You have a disposable rclone-workshop folder with these files:

OUTPUT 

source/notes.txt
source/data.csv
source/photo.jpg
dest/old-report.pdf

dest/old-report.pdf does not exist in source.

Before running anything, predict the answers:

  1. Will rclone copy source dest --include "*.txt" --dry-run touch old-report.pdf?
  2. Will rclone sync source dest --dry-run touch old-report.pdf? If so, how?

Then run both commands against your own disposable folder and check your predictions against the real output.

--dry-run never changes anything on disk — it only reports what would happen, which makes it safe to run as many times as you like while you’re predicting.

BASH 

rclone copy source dest --include "*.txt" --dry-run -v

OUTPUT 

2026/08/23 13:37:13 NOTICE: notes.txt: Skipped copy as --dry-run is set (size 14)
2026/08/23 13:37:13 NOTICE:
Transferred:   	         14 B / 14 B, 100%, 0 B/s, ETA -
Checks:                 0 / 0, -, Listed 1
Transferred:            1 / 1, 100%
Elapsed time:         0.0s

copy never mentions old-report.pdf — it only ever adds or updates files, so anything already at the destination but missing from the source is simply left alone.

BASH 

rclone sync source dest --dry-run -v

OUTPUT 

2026/08/23 13:37:14 NOTICE: notes.txt: Skipped copy as --dry-run is set (size 14)
2026/08/23 13:37:14 NOTICE: photo.jpg: Skipped copy as --dry-run is set (size 6)
2026/08/23 13:37:14 NOTICE: data.csv: Skipped copy as --dry-run is set (size 6)
2026/08/23 13:37:14 NOTICE: old-report.pdf: Skipped delete as --dry-run is set (size 34)
2026/08/23 13:37:14 NOTICE:
Transferred:   	         26 B / 26 B, 100%, 0 B/s, ETA -
Checks:                 1 / 1, 100%, Listed 4
Deleted:                1 (files), 0 (dirs), 34 B (freed)
Transferred:            3 / 3, 100%
Elapsed time:         0.0s

sync reports old-report.pdf as a file it would delete, because sync makes the destination an exact mirror of the source — this is exactly why sync exercises always need --dry-run or --interactive first.

Different operating systems have slightly different syntax

Windows syntax:

BASH 

rclone ls C:\Users\jjamison\rclone
rclone ls "C:\Users\jjamison\My Documents\rclone-workshop"
  • Linux and macOS:

BASH 

rclone copy rclone-intro-box:rclone-intro rclone-intro-google:rclone-intro-google    
rclone copy rclone-intro-box:rclone-intro rclone-intro-google:rclone-intro-google -n

Windows Subsystem for Linux (WSL2) syntax:

BASH 

rclone ls /mnt/c/Users/jjamison/rclone

See the Learner Reference page for valid remote names, subcommand syntax, and platform-specific notes.

Key Points
  • Difference between Copy and Sync
  • See what is already in the destination
  • --dry-run shows what a command would do, including deletions, without changing anything on disk

Content from Choosing the Right Rclone Command: Copy, Sync, or Move

Last updated on 2026-08-23 | Edit this page

Overview

Questions

  • What is the difference between copy, sync and move?
  • When should you choose one command over the others?

Objectives

  • Be able to pick the most appropriate command to move file(s)
  • Determine which command is most appropriate for different file management scenarios.

Copy, Sync, or Move

When managing your files with rclone, you have three primary commands: copy, sync, and move. Each command handles your data differently, so it’s important to know their unique behaviors to choose the right one for your task.

Command Changes source Adds/updates at destination Deletes at destination Safest for…
copy No Yes No Adding or updating files without any risk of deleting something
sync No Yes Yes — removes anything not in the source Making the destination an exact, up-to-date mirror of the source
move Removes files after a successful transfer Yes No Relocating files to free up space at the source

Copy

The copy command copies files from the source to the destination. It compares files by size, modification time, or checksum, and transfers only those that are new or have changed. Existing destination files that aren’t present at the source are left in place — copy does not delete them. Because of this, copy does not guarantee the destination matches the source, so running it once is not, by itself, a complete backup strategy.

Syntax:

BASH 

rclone copy source:sourcepath dest:destpath

Notes:

  • Only the contents of a directory are copied — not the directory itself.
  • Use the copyto command for copying single files.
  • If the destination path does not exist, it will be created.

Sync

The sync command makes the destination an exact mirror of the source. It copies new or updated files and deletes files in the destination that are not present in the source. Use sync when you need both locations to be identical, but be cautious as it can remove files from the destination.

Syntax:

BASH 

rclone sync source:path dest:path

Move

The move command transfers files from the source to the destination and then deletes them from the source after a successful transfer. This is useful when you want to relocate files rather than keep copies in both places.

Important Note: Since this can cause data loss, test first with the –dry-run or the –interactive/-i flag.

Syntax:

BASH 

rclone move source:path dest:path [flags]
Challenge

Choose the right command

For each scenario, decide whether copy, sync, or move is the safest choice, and explain why.

  1. You want your Box photos folder to always exactly match a folder on your laptop, including removing anything you’ve deleted locally.
  2. You’ve finished a project and want to free up space on your laptop by relocating the files to the cloud.
  3. You want to hand a colleague a copy of your files without touching anything already in their folder, or removing anything from yours.

Ask two questions for each scenario: should anything be deleted anywhere, and should the files still exist in both places afterward?

  1. sync — an exact mirror, including deletions, is exactly what sync does. Run it with --dry-run or --interactive first, since it can delete files.
  2. move — moves the files and removes them from your laptop only after the transfer to the cloud succeeds.
  3. copy — adds and updates files at the destination without deleting anything at either end.

See the Learner Reference page for a full command and documentation reference.

Key Points
  • Different ways to move or copy files: copy, sync, move
  • Understand the difference between copy and sync
  • sync is the only one of the three that can delete files at the destination

Content from Optional: Configuring rclone on a Headless Machine

Last updated on 2026-08-23 | Edit this page

Overview

Questions

  • What is a headless machine, and why does it need special configuration?
  • How do you authorize a remote when no browser is available?

Objectives

  • Identify when a machine is headless and needs a browser-free authorization flow.
  • Use rclone authorize to complete configuration on a headless machine.
Callout

This episode is optional

Skip this episode if you’re setting up rclone on your own laptop or desktop — you have a browser, and the standard rclone config flow from Creating and Configuring Remote Connections is all you need. Come back to this episode only if you need to configure rclone on a machine with no browser available, such as an HPC cluster, a cloud VM, or the Windows Subsystem for Linux (WSL2).

What is a headless machine?

A headless machine is one with no browser available to complete authorization in. For example, you might be using the Windows Subsystem for Linux (WSL2), or you’ve logged into a remote machine (an HPC cluster, a cloud VM, etc.) over SSH.

Some rclone configurations — those involving OAuth2, such as Box and Google Drive — normally open a browser to complete authorization. On a headless machine, there’s no browser to open, so rclone offers an alternative: authorize from a different machine that does have a browser.

Authorizing without a browser

When rclone config reaches the authorization step on a headless machine, it prints a command to run instead of opening a browser, using rclone authorize:

BASH 

rclone authorize box

Run that command on a different machine — one that does have a browser, such as your laptop — and it walks you through the same sign-in flow, then prints a token to paste back into the headless machine’s rclone config session.

Rclone’s documentation includes a full walkthrough of this exact process: Configuring using rclone authorize.

Challenge

Headless or not?

For each situation, decide whether you’d need rclone authorize, or whether the standard rclone config flow is enough.

  1. Configuring rclone on your own laptop, which has a web browser installed.
  2. Configuring rclone after logging into a university HPC cluster over SSH.
  3. Configuring rclone inside the Windows Subsystem for Linux (WSL2) on your own PC.
  1. Standard rclone config — your laptop has a browser, so authorization happens automatically.
  2. rclone authorize — an HPC cluster accessed over SSH has no browser of its own.
  3. rclone authorize — WSL2 is a separate Linux environment with no browser, even though it’s running on a PC that has one.

See the Learner Reference page for a full command and documentation reference.

Key Points
  • A headless machine has no browser, so OAuth-based remotes (Box, Google Drive) need a separate authorization step.
  • Run rclone authorize <backend> on a machine with a browser, then paste the resulting token back into the headless machine’s config session.
  • Most learners working on their own laptop can skip this episode entirely.

Content from Getting Help

Last updated on 2026-08-23 | Edit this page

Overview

Questions

  • Where do I get help using rclone?
  • What are some key resources for troubleshooting and learning advanced features?

Objectives

  • Know where to find comprehensive documentation and support for rclone.
  • Identify community resources and forums that can assist with troubleshooting.

Getting Help with Rclone

Rclone offers extensive documentation and community support to help you solve issues and learn advanced features. Here are some key resources:

Official Documentation

Visit the rclone documentation website for detailed guides on installation, configuration, and usage of various commands. Use the search feature to look up specific topics such as configuration commands or remote storage names.

Community Forums and Support

  • Check out the rclone GitHub repository for bug reports, feature requests, and discussions.
  • Explore community forums and mailing lists where other users share tips and solutions.
  • Use platforms like StackOverflow to ask questions and find answers from experienced users.

Command Help

For a list of every available command, run: bash rclone help or get details on a particular command, including its flags, by adding --help after the command name: bash rclone copy --help

These resources are invaluable whether you’re troubleshooting an issue or looking to explore advanced functionality.

Challenge

Find a flag

Run rclone sync --help and find the flag that lets you preview what a sync would do without actually deleting or transferring anything.

You’ve already used this flag several times earlier in the lesson.

-n, --dry-run — the same flag introduced in Creating and Configuring Remote Connections and used throughout the transfer episodes. rclone <command> --help is the fastest way to rediscover a flag you’ve forgotten, without leaving the terminal.

Key Points
  • Rclone’s official documentation is available at rclone.org.
  • Community support is accessible through GitHub, forums, and StackOverflow.
  • Use rclone help for quick command-line assistance.