← Cloud projects

Cloud · Architecture Concept

RPO, RTO & BCDR Design

Translate business impact into measurable recovery objectives, then choose availability, backup, replication, failover, and testing patterns that can meet them.

PlatformCloud
DomainResilience
LevelFoundation
Last reviewed2026-09-19
Use whenDefine measurable recovery objectives before selecting backup, replication, or secondary-region technology.
Key decisionStart with business impact
Key conceptsBusiness impact · RPO · RTO · Recovery tiers
Cost focusSecondary capacity
On this page

Building blocks

Business impactRPORTORecovery tiersBackupReplicationFailover testing

Design goal

Define recovery outcomes before selecting technology so data-loss tolerance and recovery-time tolerance drive the design instead of being inferred afterward.

Architecture

Business services are classified by impact and assigned recovery objectives. Architecture choices are then mapped to those objectives and validated through recovery testing.

Architecture decisions

Decision

Start with business impact

Why

Technology should follow required recovery outcomes, not the other way around.

Trade-off

Outcome-first recovery prevents overengineering, but requires business owners to quantify tolerable downtime and data loss before technology is selected.

Decision

Treat RPO and RTO separately

Why

Data-loss tolerance and recovery-time tolerance drive different architecture decisions.

Trade-off

Separating data-loss and recovery-time objectives produces better designs, but may require different technologies and costs to meet each target.

Security

Protect the control and data paths deliberately. Recovery access, privileged accounts, encryption keys, and emergency procedures must remain available during a primary-environment outage.

Availability

Design for the failure domain that must be survived. High availability reduces some local failures but does not replace backup or disaster recovery.

Disaster recovery

Treat regional recovery as a separate operating state. Define failover authority, recovery order, dependencies, communication, and failback as part of the design.

Cost drivers

  • Secondary capacity
  • replication
  • backup storage/retention
  • cross-region transfer
  • recovery testing

Design assumptions

  • Business owners can classify workload criticality

Implementation plan

  1. Run a business-impact assessment to identify the services, data, and business processes whose loss creates material impact.
  2. Define recovery tiers with explicit RPO and RTO targets, and document who is authorized to accept exceptions or additional cost.
  3. Map upstream and downstream dependencies so recovery order reflects the complete service rather than individual components.
  4. Select backup, replication, standby, redeployment, and traffic-recovery patterns that meet each tier without overengineering lower-criticality workloads.
  5. Document ownership and runbooks, then run tabletop and technical recovery exercises that measure actual RPO/RTO.

Validate the design

  • Run a tabletop exercise and confirm decision owners, dependencies, and escalation paths are understood.
  • Perform a technical recovery test for each recovery tier rather than validating documentation only.
  • Measure actual recovery point and recovery time against the approved RPO/RTO targets.
  • Track every gap found during testing to an owner and closure date, then retest material fixes.

Architecture basis

Continue a learning path