How to Restore Your Instance Data from a Backup using Snapshots on AWS EC2/EBS

You are writing all these backups as individual EC2 snapshots and/or full AMIs on AWS EC2 but how can you actually use them, should disaster befall your data?

Here is how to do that on Linux.

Step 1. Create a volume from a snapshot

In the AWS Management Console, find the snapshot that holds you backup.

AWS EC2: create an EBS volume from a snapshot

Next, create a volume from it and attach the volume to your running instance.

AWS EC2: attaching an EBS volume to an instance

Add a name label to your volume to help avoid a mishap.

Step 2. Identify your backup volume in Linux

Now is the time to have a look at the currently attached file systems:

fdisk -l

Compare this with the contents of fstab:

cat /etc/fstab

Make a note of the address of the volume.

Step 3. Replace the UUID and adjust the label

Read out the UUID of the supplemental drive:

xfs_admin -u /dev/xvdg1

Replace the UUID with a newly generated one:

xfs_admin -U generate /dev/xvdg1

Change the label:

xfs_admin -L "Backup" /dev/xvdg1

You should see output resembling this one:

writing all SBs

new label = "Backup"

Step 4. Mount the file system

Before you can access the volume, you need to mount the file system:

mount /dev/xvdg1 /var/www/backup

Should you encounter problems, see:

dmesg | tail

for leads on trooubleshooting.

Step 5. Do your thing with the backup

Next, switch to the source directory containing your backup files:

cd /path/to/your/backup

Then copy or move the files.

Step 6. Always clean up after yourself

Exit the directory and unmount the volume:

cd /

umount /dev/xvdg1

Now feel free to detach the volume and destroy it (you still get to keep the original snapshot!).