How Backup Works

Veeam Backup for AWS does not install agent software inside EC2 instances to retrieve data. To back up EC2 instance data, Veeam Backup for AWS uses capabilities of Amazon EBS snapshots.

During backup, Veeam Backup for AWS uses AWS CLI commands to take snapshots of EBS volumes that are attached to an EC2 instance. Depending on the data protection scenario, Veeam Backup for AWS uses the created snapshots to build a chain of cloud-native snapshots or image-level backups of the EC2 instance.

Creating Cloud-Native Snapshots

Veeam Backup for AWS creates a cloud-native snapshot of an EC2 instance in the following way:

1. Takes snapshots of EBS volumes that are attached to the processed EC2 instance.

EBS snapshots are assigned AWS tags upon creation. Keys and values of the assigned tags contain encrypted metadata. Veeam Backup for AWS treats EBS snapshots with metadata as a cloud-native snapshot of the processed EC2 instance.

2. If the EC2 instance has already been processed by a backup policy, Veeam Backup for AWS checks the number of restore points in the snapshot chain. If the number exceeds the retention limit specified in the backup policy, the earliest restore point is removed from the snapshot chain.

Veeam Backup for AWS does not apply retention for snapshots created manually. You can remove unnecessary snapshots as described in Removing Backups.

Creating Image-Level Backups

Veeam Backup for AWS performs image-level backup of an EC2 instance in the following way:

1. Takes snapshots of EBS volumes that are attached to the processed EC2 instance.

Veeam Backup for AWS uses EBS snapshots as a source for backup. When the backup process is complete, EBS snapshots are removed.

2. Launches a worker instance in an AWS region where the processed EC2 instance resides.

Veeam Backup for AWS launches the worker instance to process EC2 instance data.

3. Creates EBS volumes from temporary EBS snapshots and attaches the created volumes to the worker instance.

The number of created EBS volumes equals the number of volumes attached to the processed EC2 instance.

4. Reads data from EBS volumes on the worker instance, transfers the data to an S3 repository and stores it in the native Veeam format.

During incremental job sessions, Veeam Backup for AWS transfers only those data blocks that have changed since the previous job session. To identify unique data blocks, Veeam Backup for AWS uses backup metadata that is stored in the S3 repository.

5. When the backup process is finished, Veeam Backup for AWS removes temporary EBS snapshots and the worker instance from Amazon EC2.