Veeam Backup for Google Cloud

When discussing how to back up data for Google Cloud, we’re mostly interested in cloud-native tools that allow us to protect compute engine instances (i.e., where VMs run) with Google-native snapshots.

There are also native services in place that allow you to back up Google Engine instances, and there are many vendors that offer agent-based approaches for Google Cloud backups. However, data protection in the cloud isn't always straightforward, so leveraging limited or legacy backup methods can present challenges.

First, snapshots of Google Compute Engine instances are not real backups and therefore can’t be leveraged for long-term data retention and recovery. Aged snapshots are also expensive to keep in Google Cloud Storage for longer than a few days. Manual scripting for automatic snapshot scheduling for Google Cloud instances can also quickly become very complex, especially in large environments.

Secondly, agent-based approaches for Google Cloud backups aren’t the best options since agents require resource-intensive deployment. They are hard to scale appropriately, and they take additional compute engine space.

Veeam Backup for Google Cloud can automatically create and manage snapshots and backups of compute engine instances (and attached disks), including VM configurations. Backup chains can be created with either image-level backups or snapshots. First, a full snapshot is created and then each subsequent snapshot only contains incremental changes. Veeam does not use agents to back up Google Cloud but instead uses temporary instances, called workers, that are deployed only for the time it takes to generate Veeam backup image data from the snapshot or perform a restore.

Image-level backups can include full and incremental backups of your Google Cloud data.

Snapshots and backups can be tiered and stored in Google Cloud Storage. Please keep in mind that it’s always a good practice to separate your backups from your production environment. Veeam Backup for Google Cloud supports both cross-project and cross-region data tiering, which makes it easy to secure your backups.

In terms of recovery, Veeam Backup for Google Cloud can restore files from both snapshots and backups, which makes file-level recovery fast and on-demand. During recovery, a worker instance will spin up and attach to a point-in-time within the Google Cloud snapshot or backup, providing you with a web console and file system where you can browse and restore any file you want. Recovery within Google Cloud can be performed at the full- and file-level in just seconds.

In addition to ensuring the security and protection of the underlying Google Cloud infrastructure, Google Cloud does provide native methods for performing backups of Google Compute Engine instances. These methods include the functionality to create persistent disk snapshots or machine images. Snapshot chains include full and consequent incremental snapshots. It’s also possible to create a schedule for performing regular snapshots of your Google Cloud data.

Google Cloud-native tools also allow you to create full and incremental image files of a particular VM.

Once snapshots or VM images are created, the (cp) command is used to move these files from production to Google Cloud Storage.

These methods are handy for environments that have very small scale or ones that are designated for testing purposes. However, for bigger production environments where performing against SLAs is critical, it may be necessary to have different tools in place to ensure a robust backup and recovery strategy for Google Cloud. Snapshots and images alone are not an optimal solution when you need longer retention or need to store aged snapshots for more than a few days, since this can get costly quickly. Snapshots are also not automatically isolated from the production environment. This can create security and data loss risk in case of disaster. Lastly, if your Google Cloud environment is set to grow, you will require more profound recovery options that must include file-level recovery and scheduling options that aren’t just for snapshots.