Good Patterns¶
The source framework identifies the following examples of architectural patterns that support operational resilience.
Simple identity architecture¶
Reduce unnecessary identity dependencies and ensure critical recovery paths do not depend entirely on the service being recovered.
Few integrations¶
Prefer purposeful integrations over uncontrolled coupling. Each integration should have a clear protection or operational purpose.
Layered recovery¶
Design recovery mechanisms at multiple levels so that failure of one control plane does not eliminate every recovery path.
Offline administration¶
Maintain appropriately controlled recovery or administrative capability that does not depend entirely on normal production services.
Documented dependency mapping¶
Make technical, operational and vendor dependencies visible before they become hidden single points of failure.
Technology consolidation¶
Consolidate where it genuinely reduces duplication and operational burden, while avoiding unacceptable concentration risk.
Graceful degradation¶
Design critical services to continue at an understood reduced capability when a security component is unavailable.