How to use Bash+rsync to automate your periodical backups on Linux

Linux is all about simplicity. Simple and time-tested tools like iptables, netstat and rsync can be called for help for basic tasks instead of untrusted third-party tools as happens in case of Windows. If you are a desktop user, then taking frequent backups of your data to a pen-drive or external disk is a typical problem to solve.

In this article, I’ll show how I solved this problem using a combination of bash scripting and rsync, the basic tooling available on any linux distro these days, be it Ubuntu, Debian or Fedora.

One of the things you may want to do is determine what folder(s) you want to backup to which device. You may want to copy the source code folder only to your pen drive and your images and documents to only external drive, for instance. Here is where identifying the disk label through bash comes in handy in the backup script:

if [ -d "/media/prahlad/DATA128" ]; then
    device_name="DATA128"
    folder_name="/media/prahlad/DATA128"
elif [ -d "/media/prahlad/extHD" ]; then
    device_name="extHD"
    folder_name="/media/prahlad/extHD"
else
    echo "No Drive Found"
    exit
fi

In the above example, DATA128 is a pen-drive and extHD is an external hard-drive and the above script determines which disk is inserted in the USB drive. You can then use the $folder_name bash variable to dynamically copy to that device instead of hard-coding that path unnecessarily. You can also use $device_name bash variable to include or skip specific folders when running the rsync command:

rsync -va /home/prahlad/source "$folder_name/source"

if [ "$device_name" = "extHD" ]; then
 rsync -va /home/prahlad/Pictures "$folder_name/home/Pictures"
 rsync -va /home/prahlad/Documents "$folder_name/home/Documents"
fi


The rsync command itself is also versatile enough to do a lot of things which are not possible using a simple copy-paste-replace using a file manager. For instance, the “a” or “–archive” option intelligently archives (copies) files while skipping identical ones based on checksum or modification date automatically. Further, the “–delete” option deletes files which are present on the destination backup device, but not on the source device which is typically the case when you want to backup your data. Run “man rsync” to see the full range of options exposed by this wonderful command.

Finally, another advantage of using a script for backup automation is that you can implement custom actions through the script. For example, compressing the mozilla firefox user folder before taking its backup:

echo "backing up firefox..."
tar czf ~/firefox-backup.tar.gz ~/.mozilla/firefox/
rsync -va ~/firefox-backup.tar.gz "$folder_name/home/firefox-backup.tar.gz"

Below is a typical example of how you might implement a script to automate backup of your home folder and set this as a cron job to run say weekly or fortnightly:

#!/bin/sh

if [ -d "/media/prahlad/DATA128" ]; then
 device_name="DATA128"
 folder_name="/media/prahlad/DATA128"
elif [ -d "/media/prahlad/extHD" ]; then
 device_name="extHD"
 folder_name="/media/prahlad/extHD"
else
 echo "No Drive Found"
 exit
fi

echo "Device: $device_name"
echo "Folder: $folder_name"

#start copying 

#home
rsync -va ~/.bashrc "$folder_name/home/" --delete
rsync -va ~/.profile "$folder_name/home/" --delete
rsync -va ~/Downloads/ "$folder_name/home/Downloads/" --delete
rsync -va ~/Documents/ "$folder_name/home/Documents/" --delete
rsync -va ~/Pictures/ "$folder_name/home/Pictures/" --delete
rsync -va ~/.ssh/ "$folder_name/home/.ssh/" --delete
rsync -va ~/.gnupg/ "$folder_name/home/.gnupg/" --delete
rsync -va ~/.thunderbird/ "$folder_name/home/.thunderbird/" --delete
rsync -va ~/.mozilla/ "$folder_name/home/.mozilla/" --delete

echo "Done"

You can customize the above script by adding custom actions or if conditions to skip or include specific folders based on the device label. If you know of any other techniques for backup automation on linux, then please let us know through comments.

2 comments

  1. deviousfusion says:

    You should try rsnapshot. It’s based on rsync and provides incremental backup using hardlinks.

  2. Bill Simons says:

    i’m sure you’re aware, but there are standard packages for this. I like rsnapshot (a more fully featured rsync script), but there’s also duplicity (and duply, its gui front-end) which can even save to cloud-based storage

Leave a Reply

Your email address will not be published. Required fields are marked *