This is a demonstration of how to use Sleuth, Zipkin & an ELK stack for tracing requests. You may take a look at the presentation given on Devoxx by bit.ly/sleuth-zipkin-devoxx
To start ELK (Elasticsearch, Logstash, Kibana) & the Zipkin Server, ensure that you have Docker & Docker-compose installed.
In the root you can run ELK & Zipkin with docker-compose up monitoring.
Go to http://localhost:5601 & http://localhost:9411 to check if you can reach respectively Kibana & Zipkin.
The application itself consists of 5 (micro)services. Ensure you have Maven installed on your computer. The Breakfast service, Machine service, Temperature service & Preference service are REST services, while the Energy Service is a SOAP service. To ensure that Sleuth will work with the SOAP client & server, there is a submodule added (sleuthsoapinterceptor).
Go to the sleuthsoapinterceptor directory & perform a mvn clean install.
After this, perform a mvn clean package on the root of the project.
You can start the applications by docker-compose up services. Or use any other way to start the applications you feel comfortable with.
Note: Ensure you have enough RAM assigned to Docker. See also http://elk-docker.readthedocs.io/ why; from which we are using the Dockerfile for our ELK environment.
Go to http://localhost:8080 to go to the frontpage, to play with the toaster, eggs, bacon & coffee.
Go to http://localhost:5601 and go to management. Here you have to create a index pattern; for example logstash-*. After you configured this, you can discover your logs.
Add fields like severity, facility, message to view the logs of the services.
Add X-B3-SpanId, X-B3-TraceId or even X-B3-CONVID to view Sleuth related information.
Go to http://localhost:9411/zipkin/ to view the user interface of Zipkin. Push Find Traces to view the traces. You can click on one trace to view the service calls/spans.