Orchestration · Software component

Graceful Degradation Manager

Software componentOrchestrationOrchestration & Toolsarc:GracefulDegradationManager

A control component that selects the highest capability level still supported by healthy dependencies, disabling auxiliary capabilities while preserving core ones and labelling results with degradation status.

Responsibility. Chooses the capability level to serve based on current dependency health.

Also known as: Degradation controller, Capability-level selector, Graceful degradation path, Fallback pattern

Direct neighbourhood (hover for relationship types)

Relationships

is configured by structural

Capability Tier Map Ch2.8

reads dependency

emits telemetry to dynamic

is routed to by dynamic

Tool Error Classifier Ch3.7

is triggered by dynamic

receives data from dynamic

routes to dynamic

sends data to dynamic

Error Presenter Ch2.8

monitors assurance

Reranker abstract Ch6.5
Vector Index Store abstract Ch6.5

Design guidance

MUST be designed upfront by identifying core versus auxiliary capabilities and a fallback implementation for each layer; degradation does not emerge naturally.
MUST return degradation status, reduced confidence and warnings with degraded results, especially in high-stakes domains such as compliance, legal or medical review.
MUST NOT serve empty results, invalidate local state without fallback, or leave inconsistent state during partial failures.
SHOULD let optional tools (enrichment, performance optimization, secondary confirmation) fail with documented reduced capability rather than block the workflow.
MUST NOT degrade essential tools such as payment processing.
SHOULD fall back to cached data when APIs time out and simplify decision logic when complexity exceeds thresholds.
SHOULD acknowledge limitations under error conditions rather than fabricating responses.
SHOULD continue without reranking when the reranker fails and fall back to keyword search when the vector database is unavailable.
SHOULD queue requests with user-visible wait times when LLM rate limits are hit rather than failing outright.
SHOULD substitute cached data or simplified workflows when external dependencies fail.
SHOULD provide graceful degradation and fail-safe defaults as a design-level incident prevention measure.

Quantitative guidance

As stated by the sources; verify before use.

An agent with 3% hallucination under normal tool behaviour may hallucinate 30%+ when tools fail (Ch3.10).

Classification

Patterns: Graceful degradationCore vs. auxiliary capabilitiesLayered capability levels (full/basic/minimal)Explicit degradation signallingBulkhead isolation
Quality attributes: Reliability (ISO/IEC 25010 | NIST AI RMF: valid and reliable)Transparency and accountability (NIST AI RMF: accountable and transparent)Interaction capability (ISO/IEC 25010)
Risks mitigated: Total failure on partial outageOver-reliance on silently degraded resultsInconsistent state during partial failure

Sources

Ch2.8: T. Nguyen, "Error Handling and Resilience," in Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam, 1st ed. 2026, ch. 2.8. ISBN: 9798244538229.
Ch3.7: T. Nguyen, "Tool Usage Auditing," in Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam, 1st ed. 2026, ch. 3.7. ISBN: 9798244538229.
Ch3.10: T. Nguyen, "Efficiency Metrics," in Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam, 1st ed. 2026, ch. 3.10. ISBN: 9798244538229.
Ch4.2: T. Nguyen, "Deployment and Scaling," in Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam, 1st ed. 2026, ch. 4.2. ISBN: 9798244538229.
Ch6.5: T. Nguyen, "Production RAG Systems," in Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam, 1st ed. 2026, ch. 6.5. ISBN: 9798244538229.
Ch6.6: T. Nguyen, "Query Decomposition and Adaptive Retrieval," in Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam, 1st ed. 2026, ch. 6.6. ISBN: 9798244538229.
Ch8.2A: T. Nguyen, "Error Rates and Reliability," in Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam, 1st ed. 2026, ch. 8.2A. ISBN: 9798244538229.
Ref9.08: "Safety Incident Response for AI Systems," unpublished reference note (08-Safety-Incident-Response.md), Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam supplementary materials, 2026. unpublished note