Policy inheritance and multi-level evaluation
A service is evaluated against all policies that apply to it:
-
Its directly assigned policy (if any).
-
Policies inherited from user journeys it belongs to.
This enables separation of concerns:
-
Central SRE teams set DR policies at the user journey level for governance.
-
Service teams set performance policies at the service level for operational requirements.
-
Both policies apply simultaneously to services in the user journey.
For example:
User Journey: "Path to Trade" Policy: "Trading Platform DR Policy" (Multi-Region, RTO 15 min, RPO 5 min) └── Service: "Order Execution" Policy: "Order Execution Performance Policy" (Availability 99.99%, Error < 0.1%, p99 < 50ms)
The "Order Execution" service is evaluated against both policies. It must meet the DR requirements from the user journey policy and the performance requirements from its service-specific policy.
Conflict resolution
When multiple policies contain conflicting requirements for the same component, Next generation Resilience Hub automatically applies the stricter value. The following example shows how conflicting RTO values are resolved:
| Source | Component | Value |
|---|---|---|
| User journey policy | RTO | 30 minutes |
| Service policy | RTO | 15 minutes |
| Applied value | RTO | 15 minutes (stricter) |