A cloud architect is designing a disaster recovery (DR) strategy for a critical application that requires an Recovery Time Objective (RTO) of less than 15 minutes and a Recovery Point Objective (RPO) of less than 5 minutes. The application runs on virtual machines, utilizes a relational database, and stores user-uploaded content in object storage. Which DR strategy would best meet these stringent requirements while minimizing data loss and downtime?
- AWarm Standby deployment to a secondary region.
- BMulti-Region Active/Active deployment.
- CBackup and Restore to a secondary region.
- DPilot Light deployment to a secondary region.
Show answer & explanationAnswer & explanation
Correct answer: B. Multi-Region Active/Active deployment.
An RTO of less than 15 minutes and RPO of less than 5 minutes are extremely stringent. Multi-Region Active/Active deployment involves running the application concurrently in multiple regions, ensuring that if one region fails, traffic can be immediately routed to the other active region with virtually no downtime and minimal data loss, thus meeting both RTO and RPO requirements. This is the only option that can realistically achieve such low RTO/RPO.
Why the other options are wrong
- A. Warm Standby has a fully functional but scaled-down environment, resulting in RTOs of minutes to hours, and RPOs of minutes to tens of minutes, which might be too long for '<5 minutes'.
- C. Backup and Restore typically has RTOs of hours or days and RPOs of minutes to hours, far exceeding the requirements.
- D. Pilot Light involves a minimal set of resources in the DR region, requiring significant manual or automated effort to scale up, leading to RTOs of tens of minutes to hours.
Multi-Region Active/Active DR
A disaster recovery strategy where an application runs simultaneously in multiple geographically separate regions, actively serving traffic, providing the lowest RTO and RPO.
- Application is fully operational in all regions.
- Traffic is distributed across regions.
- Near-zero RTO and RPO.
- Highest cost due to duplicate resources.
Memory trick: Active-Active for Zero Downtime, Zero Data Lost, Zero Excuses.