In modern backup environments, the protection ecosystem must ALWAYS be designed assuming that the primary infrastructure will be compromised. This is done so we can guarantee operational continuity and the ability to respond to an isolated problem or an entire disaster in the production environment. With this in mind, I have separated three components that cannot be neglected when designing or reviewing your backup and data protection environment.
1. Immutability
Nowadays, it is no longer enough to guarantee just a single layer of immutable backup. As soon as an attacker obtains administrative credentials for a backup system, the first thing they will try to do is delete local backups or force the expiration of retentions. After that, they will do the exact same thing to the second copies of your data or your cloud backups.
In the past, it was very common to see backup environments with local disk backups and tape backups. In a scenario like that, only the tape backup could be considered air-gapped. This doesn't mean you can't use tape backups within your data protection architecture, but it can no longer be your only point of isolation.
In modern security-focused environments, all backups must be secure and isolated. To achieve this, we must apply immutable backups not only to the first backup layer but also to the second copy of the data, whether that is a physical appliance or a cloud object storage.
2. High-Performance Repositories
One of the key points of a backup policy is the RTO (Recovery Time Objective). The RTO is how the business defines the maximum acceptable downtime for a system or application. However, with all our recent technological advancements, it is increasingly common to see tighter, ever-shorter RTOs.
Because of this, the choice of the repository and how it will be used is a determining factor in meeting the RTO required by the company.
Therefore, it is highly recommended to use low-latency media and modern storage systems. Storage arrays with high-performance LUNs utilizing SSD or NVMe servers are excellent options to meet those applications with the most aggressive RTOs. Servers with RAID 10 arrays or lower can be a solid choice for systems with more lenient RTOs.
Utilizing Instant Recovery at the time of restoration is also a feature that should always be considered. Instant Recovery allows you to initialize virtual machines and databases directly from the backup files. This way, it is possible to make the application available in minutes by mounting it via NFS/iSCSI directly to the hypervisor, all while data migration and backup restoration occur seamlessly in the background. Obviously, for Instant Recovery to perform quickly, the backup repository needs to deliver the highest performance possible.
3. Continuous Restoration Testing
Having a backup job complete with a "Success" status on the dashboard is no guarantee of recovery. The real test is knowing if the operating system will boot, if the database will mount without corruption, and if everything will start up with all application dependencies working properly together. Automating this validation is exactly what transforms an inert data repository into a true business continuity architecture.
Instead of relying on manual, sporadic restore tests that consume a massive amount of your team's time, SureBackup can automate this verification. Through scheduled jobs, Veeam deploys the VMs and mounts the disks directly from the backup files stored in the repository. This entirely takes place within a virtual lab (Virtual Lab), guaranteeing zero impact on your production network.
With all of this in place, you can perform automated checks such as heartbeat, ping, port connectivity, or even run customized scripts that verify and validate the application layer of the server.
Conclusion
Backup is the very last line of defense for any IT infrastructure. However, for it to tank the weight of this responsibility, it must adhere to specific core principles and best practices.
This is exactly why multi-layer backup immutability is implemented to neutralize ransomware attacks. High-performance repositories reduce RTO to mere minutes, guaranteeing the business can continue operating even after a severe problem or disaster. Ultimately, the continuous orchestration of restore tests provides the final guarantee that everything will keep working exactly as intended after a backup recovery.
