Which fault tolerance pattern protects against an HTTP API that frequently returns 502 status codes?

• A. Use a simple retry policy
• B. Use the First Successful router configured with a fallback API route that can be used if the primary API fails
• C. Switch to a circuit breaker with a single path
• D. Use a rate limiter

Answer: (B)

When an HTTP API frequently returns 502 errors, you want a fault-tolerant path that keeps the user experience smooth by switching to a working backend. The First Successful router with a fallback API route fits this need brilliantly. It tries the primary API first, and if that path doesn’t produce a successful response, it automatically routes the request to a known-good fallback API. This provides a transparent failover so callers get a valid response instead of getting stuck on a persistent 502.

This approach is better than simply retrying the same failing API. Retrying repeats the same problem and can waste resources and increase latency, especially when the upstream gateway is flaky. A circuit breaker with a single path wouldn’t have a viable alternate to route to, so once failures rise, traffic could be blocked for longer without offering a functioning option. A rate limiter helps control load but doesn’t address the root issue of a failing downstream service or provide a backup backend to fulfill requests.

So the fallback-first strategy enables continuous availability by automatically leveraging an alternate API route when the primary API is unreliable, which is precisely what’s needed when 502s are common.