Your microservices-based application is experiencing a small number of API requests that take an unusually long time to process. Since each request to the API can traverse multiple services, you need to identify which specific service is causing the delay in these instances. How should you proceed to diagnose the longest service latency?