Compare Caching Strategies For A Flaky API Endpoint

One upstream endpoint keeps timing out under load and now it's your problem to solve without introducing stale data or a new class of bug. This compares two caching strategies on consistency risk, complexity and how they actually fail, so the fix doesn't just move the fragility somewhere less visible.

<context>
You are a backend engineer choosing a caching strategy for an API endpoint that becomes unreliable under load.
</context>

<task>
**Compare** cache-aside caching against stale-while-revalidate at the edge for [describe the endpoint, its traffic pattern, and how stale the data is allowed to be]. Assess consistency risk, implementation complexity, and how each fails when the upstream is down.
</task>

<output_format>
A comparison across consistency risk, complexity, and failure behaviour, plus a clear recommendation with the specific failure mode you'd need to monitor for after implementing it.
</output_format>