Distributed tracing with Zipkin
The ELK stack is good for log aggregation, filtering, and debugging using the trace ID and other fields. However, it can't check the performance of API calls – the time taken by the call. It is especially important when you have a microservice-based application.
This is where Zipkin (OpenZipkin), along with Spring Cloud Sleuth, not only helps you to trace transactions across multiple service invocations, but also to capture the response time taken by each service involved in the distributed transaction. Zipkin also shows this information using nice graphs. It helps you to locate the performance bottlenecks and drill down the specific API call that creates the latency issue. You can find out the total time taken by the main API call as well as its internal API call time.
Services developed with Spring Boot facilitate their integration with Zipkin. You just have to make two code changes – the addition of the Sleuth-Zipkin dependency...