A developer exposes a Lambda function through Amazon API Gateway. During testing, API Gateway reports timeouts even though the Lambda function completes within its configured timeout. Which CloudWatch metrics for API Gateway should the developer examine to diagnose the problem? (Choose two.)

When API Gateway reports timeouts despite the Lambda function completing, it indicates a delay occurring either within API Gateway itself or between API Gateway and the Lambda integration. Latency measures the total time between when API Gateway receives a request and when it returns a response to the client. This metric helps identify if the overall request-response cycle is exceeding the client's or API Gateway's timeout. IntegrationLatency specifically measures the time taken for the backend integration (in this case, the Lambda function) to respond to API Gateway. If IntegrationLatency is high, it points to delays in the Lambda invocation or the network path to Lambda, even if the Lambda function's internal execution is fast. CacheHitCount and CacheMissCount are related to API Gateway caching and wouldn't explain a timeout issue unless caching was misconfigured to cause delays, which is less direct than latency metrics. Count simply tracks the number of requests and doesn't provide timing information.