What problem does it solve?
Provides clear guidance and practical code patterns to make network, database, and message operations resilient to transient failures by implementing retries with exponential backoff and jitter while avoiding common pitfalls like retry storms or duplicated side effects.
Core Features & Use Cases
- HTTP resilience: Configure HttpClient pipelines to retry on 5xx and rate-limit responses, respect Retry-After headers, and log retry events for observability.
- Database and queue retries: Use EF Core retry settings or custom resilience pipelines for transient SQL errors and message queue operations with capped delays and jitter.
- Idempotency and safety: Recommend idempotency keys for non-idempotent operations and show when not to retry business errors such as 4xx responses.
Quick Start
Add a retry pipeline with exponential backoff and jitter to your HTTP client configuration and ensure non-idempotent requests include idempotency keys.