rsync for Mere Mortals

rsync is the standard tool for making one directory tree match another — on the same disk, or across a network. It is not a plain copier. A copier sends everything, every time; rsync works out what actually changed and sends only that. A second sync of a large tree usually moves almost nothing.

That one idea — synchronise, don't copy — explains both its speed and its dangers. A tool that makes trees match will happily delete files to make them match, if you ask it to. This guide teaches the habits that keep that power pointed the right way: dry runs before deletions, itemised output you can actually read, and snapshot backups that cost almost no disk.

Every command in this guide was run for real (rsync 3.2.7) against throwaway trees in a lab — including the mistakes: the trailing-slash slip, a deletion caught by a dry run, a transfer interrupted mid-file and resumed, and a daemon that refused to serve until its permissions made sense. The few things that genuinely could not be run there are marked (documented) wherever they appear.

It assumes you use a terminal and know what files, permissions, and a shell prompt are. Nothing here needs more than one ordinary machine; the network chapters run over the loopback address, which works on any machine, connected or not.

What this guide covers

• What rsync is — a synchroniser, not a copier; the delta idea; the three ways it moves files
• Your first sync — the command's anatomy, and the trailing slash that decides where everything lands
• Archive mode and friends — what -a really switches on, and what it deliberately leaves out
• Dry runs and reading changes — --dry-run as a way of life, and decoding --itemize-changes
• Deleting, carefully — --delete and exactly when it bites, plus the safety nets
• Including and excluding — patterns, anchoring, and list files
• How rsync decides what to send — the quick check, its blind spot, --checksum, and resuming an interrupted transfer
• rsync over the network — ssh transport, the daemon mode, compression and bandwidth limits, all measured
• A worked example — one tree through its whole life
• rsync as a backup engine — hardlink snapshots with --link-dest, and what rsync is not
• Troubleshooting — symptom → fix, with the real error messages and exit codes
• Glossary — the terms, collected in one place

What it doesn't cover

The rsync daemon as a public service (authentication setups, secrets file management beyond the basics), the wire protocol itself, and GUI front ends. For backup strategy — what to keep, retention, testing restores — see the companion guide, Backups for Mere Mortals; this guide covers the rsync mechanics that such a strategy can be built on.

How to use this guide

Read the first two chapters, run the worked example, and you have the everyday tool. The middle chapters are reference for the day a sync does something you didn't expect — which, with rsync, is a day that comes for everyone.

---

Part of the series

This guide is one of the For Mere Mortals guides: plain-language guides to the tools programmers and makers use.

Related guides:

• Backups for Mere Mortals
• rclone for Mere Mortals
• SSH for Mere Mortals

License

The writing in this guide is licensed under CC BY 4.0. Code examples and lab scripts are licensed under the MIT License.