error-handling-patterns

Exponential backoff retry automatically retries failed API requests with increasing delays between attempts, preventing overwhelming downstream services. Structure your try-catch block to catch failures, increment a retry counter, and calculate delays using powers of 2—for example, wait 1s, then 2s, then 4s—before re-attempting the request.

A circuit breaker is a fault tolerance pattern that stops sending requests to a failing service after a threshold of errors is reached, then periodically tests recovery. Use it to prevent cascading failures when external services become unavailable, allowing them time to recover without your application hammering them with requests.

Promise rejections and async/await errors require structured handling: wrap async functions in try-catch blocks to catch errors thrown by await expressions, and attach .catch() handlers to promises to prevent unhandled rejection crashes. Integrate logging at each catch point to track failure sources.

Yes, error handling patterns are specifically designed for microservices and external API dependencies. Implement retry logic for transient failures, circuit breakers to isolate faults between services, and graceful degradation to serve cached or partial responses when downstream services fail.

Graceful degradation allows applications to continue operating at reduced capacity when components fail—serving cached data or simplified responses. Failing fast stops immediately and surfaces errors to users. Use degradation for non-critical features; use fail-fast for safety-critical operations requiring immediate visibility.

Prevent cascading failures by implementing circuit breakers to isolate faults, bulkheads to limit resource consumption per service, and timeout policies to fail fast rather than hang. Monitor and log failures at integration boundaries so errors don't propagate through the entire system.