CONCEPT Cited by 1 source

Dependency discovery

Definition

Dependency discovery is the automated process of identifying all components an application relies on — including hidden, undocumented, and transitive dependencies — by programmatically querying cloud APIs, analyzing infrastructure-as-code templates, and scanning source code for connection strings, timeout configurations, and retry logic.

Why it matters

Manual dependency mapping is impractical in continuously-deployed distributed systems: documentation lags behind infrastructure changes, and "quick fixes" leave single points of failure undocumented. Automated discovery addresses this by completing in hours what previously took weeks of cross-team cataloging (Source: sources/2026-06-22-aws-architecting-ai-powered-resilience-framework-on-aws).

Two levels of discovery

1. Infrastructure-level — queries cloud service APIs (EC2, RDS, Lambda, S3, DynamoDB, ELB) for service topology, endpoints, and Multi-AZ configuration.
2. Code-level — scans repositories for hard-coded endpoints, connection strings, timeout values, circuit breaker configurations, and retry logic that infrastructure-level discovery alone cannot detect.

Operational numbers

• Initial mapping: 2–4 hours for single-account environments with thousands of resources.
• Subsequent runs: incremental, processing only changes tracked by configuration management (e.g., AWS Config).

Seen in

• sources/2026-06-22-aws-architecting-ai-powered-resilience-framework-on-aws — foundation layer of the five-layer resilience framework