When activating a Disaster Recovery Plan (DRP), whether in a test or real-world scenario, one question often arises: why not leave the machines active immediately after they have been restored?
This approach is actually the result of several technical and operational observations. Let’s explore the reasons behind this decision and the benefits it brings.
The reasons for the extinction of restored machines #
Client Side #
- Integrating machines into the domain: Restored machines may encounter difficulties integrating properly into the domain if other critical elements of the infrastructure are not yet ready.
Customer specifics: Some customers have specific constraints regarding the scheduling of their machines’ startup. This may be due to the execution of synchronization scripts between applications or specific activation priority requirements.
Nuabee industrialization #
- Dependency management: The industrialization process restores the client infrastructure based on “dependency management,” which could delay the restart of other machines.
The benefits of this solution #
Benefits for customers #
- In the event of a cyberattack, start servers individually or in batches: During a cyberattack, it is often preferable to restart servers gradually in order to limit the risk of propagation and verify the integrity of each machine before putting it back into service.
- Customer autonomy and choice: By letting customers decide when and in what order to start their machines, customers have greater flexibility and control over their IT environment.
- Optimization of cloud resources: By not leaving all machines on all the time, we avoid excessive consumption of resources and optimize their use according to actual needs.
Internal benefits #
- Better startup and restore management: Since restoration and restarting are important operations, it is essential to control each step of the process to ensure effective recovery.
The improvement implemented #
In light of these findings, the decision was made to systematically shut down restored machines when the DRP is activated (in test or real mode). This approach ensures a smoother transition, optimizes performance, and gives customers greater autonomy over their infrastructure.
In conclusion, shutting down machines after they have been restored is not a constraint, but rather an optimization of operations that ensures a more efficient resumption of activity that is better suited to the needs of businesses. This approach illustrates the importance of careful and controlled management of service continuity in the event of a crisis.