RPO
Definition
Recovery Point Objective - maximum acceptable amount of data loss measured in time, essential for data protection and disaster recovery strategies.
Use Cases
- Netflix: Reduce data loss risk for globally distributed, high-availability streaming platform services. — Uses a multi-region architecture and data replication patterns (commonly discussed in Netflix engineering) to limit potential data loss during regional failures, selecting replication and consistency approaches per datastore based on acceptable RPO. (Improved resilience during infrastructure failures and reduced potential data loss by aligning replication strategies to service-specific recovery objectives.)
- GitHub: Protect source code and operational data against outages while maintaining service continuity. — Publicly shared incident write-ups describe the use of database replication and backups; RPO targets are met by ensuring recent data is replicated and recoverable if a primary system fails. (Faster recovery with reduced data loss exposure compared to relying on periodic backups alone, improving reliability for developers.)
- Capital One: Maintain recoverability for regulated financial workloads in the cloud with strict data-loss limits. — Uses cloud-native backup and replication capabilities and designs systems with defined recovery objectives; RPO is achieved through frequent backups and/or near-real-time replication depending on the system’s criticality. (Better alignment with regulatory expectations and business continuity goals by explicitly engineering and validating recovery objectives.)
Frequently Asked Questions
- What’s the difference between RPO and RTO?
- RPO is how much data you can afford to lose, measured in time (e.g., “up to 15 minutes of data loss”). RTO (Recovery Time Objective) is how long you can afford the system to be down (e.g., “service must be back within 1 hour”). You can have a low RPO (little data loss) but a higher RTO (takes longer to restore), or vice versa, depending on your design.
- When should I define an RPO?
- Define an RPO for any system where data matters—especially customer transactions, orders, payments, healthcare records, or operational logs needed for audits. Start by asking: if we lose the last X minutes/hours of updates, what is the business impact (revenue loss, compliance risk, customer trust)? Use that to set an RPO per application, not one blanket number for everything.
- How much does a lower RPO cost?
- Lower RPO usually costs more because it requires more frequent backups, continuous backup, or near-real-time replication. Cost drivers include: extra storage for backups and replicas, cross-region data transfer/egress charges, write amplification and performance overhead on databases, additional standby infrastructure, and operational complexity (testing, monitoring, runbooks). The cheapest approach is typically periodic backups (higher RPO), while the most expensive is synchronous or near-synchronous replication across zones/regions (very low RPO).
Category: software
Difficulty: intermediate
Related Terms
See Also