Server Rsync Backup Script

A wrapper around rsync to create backups from remote systems of any set of files on a schedule.

Requirements

Bash >= 5.1
rsync >= 3.2
ssh >= 8.9

Install

Simply clone this repository.

git clone https://github.com/MitMaro/server-rsync-backup

Update

To update, from the project directory, run:

git pull

Usage

Configuration Files

Most of the configuration for the script is stored in a set of configuration files.

All configuration files are contained in a single root configuration directory.

Inside this directory, the primary configuration file, called config contains the following options:

name	default	required	description
target		Yes	The target directory for all backup files.
ident_file		No	The SSH private key to use for SSH connections. Default to no explict key.
verbose	false	No	Set to true to enable verbose logging, false to disable
dry_run	false	No	Do not sync any files. Note some actions are still performed, like creating log directories
relative	true	No	Use the --relative option with rsync, to preserve paths
log_color		No	Set to false to disable logging with colors, true to enable. Defaults to false for file logging, true for stdout
log_to_file	false	No	Log to a file instead of stdout
log_file_root	/var/logs/rsync-backup	No	The directory to store log files
log_file_date_format	+%Y-%m-%d	No	The date format for the log file name. Must work with `date`.

Every directory inside the root config directory contains a backup batch, with a config file and a set of file patterns.

The config file supports the following:

name	default	required	description
id	Name of directory	No	The unique identifier for the backup.
skip	false	No	Ignore this set of configuration files and file patterns.
remote_user	root	No	The remote user for the rsync/ssh connection
remote_host		Yes	The remote host/ip address
ident_file		No	The SSH private key to use for this particlar remote host. Overwrites the script level config

Inside each batch, a directory called files.d contains config files that describe the files to back up. The config file contains:

name	Default	required	description
path		Yes	The remote path to backup
target		No	Target path relative to root target
allow_missing	false	No	Do not error if path is missing on remote
exclude		No	A pattern to pass rsync as an `--exclude`
include		No	A pattern to pass rsync as an `--include`

The include and exclude configurations can be provided multiple times, are are passed to rsync in the order provided in the file.

The root configuration directory can also contain a files.d directory, in which case, these files are backed up for all batches.

Example Directory

backup.d
├── config
└── files.d
│   ├── readme
│   └── ssh-config
├── pihole
│   ├── config
│   └── files.d
│       ├── database
│       └── root
├── uptime
│   ├── config
│   └── files.d
│       └── default
└── vpn
    ├── config
    └── files.d
        ├── configuration
        └── private-keys

Script Arguments

While the configuration files provide the majority of the options for the script, there are a handful of arguments that can be passed directly. These arguments take precedence over their config file counterparts.

server rsync backup

Usage: backup.sh [options] <path-to-config-root>

Options:
  --id <id>         Only sync files for provided id.

  --verbose, -v     Show more verbose output of actions performed.

  --no-color        Disable colored output.

  --dry-run         Run rsync in dry run mode. Providing this options also assumes --verbose.

  --help            Show this usage message and exit.

Use case

This script can be used to automatically back up a list of paths, generally useful in server environments. I use it to back up files from various Proxmox LXCs in my homelab.