Reactor monitoring demo

Preparation

You will need Java 8+, Docker and Docker Compose installed.

Quick start

1. Start Prometheus & Grafana with Docker Compose:

   $ docker-compose up -d

   It will start Grafana in port 3000 and Prometheus on port 9090.

2. Open Grafana and explore the dashboards:

   • "Schedulers" - dashboard for various metrics reported by Reactor's schedulers
   • "Business" - dashboard with a few metrics related to the business logic based on Reactor's metrics mechanism

3. Start com.example.demo.DemoApplication

   The application is a simple Spring Boot application that listens to Wikipedia's change stream and exposes the latest change via HTTP endpoint GET /latestChange.

   It enables Reactor schedulers' metrics by calling Schedulers.enableMetrics().

   The processing pipeline converts the received change into a human readable form. Note that it does not know how to process the updates from bots and reports an error.

   There are two "reactive" metrics in the pipeline:

   1. recentchange for every event sent by the Wikipedia stream
   2. processing for every result of the processing

4. Once the application is started, you should start seeing the metrics in Grafana.

   Try hitting various endpoints of the app to observe the changes on the "Schedulers" dashboard. For instance, you can run some load on GET /latestChange or GET /actuator/health with tools like WRK:

   $ wrk -t12 -c1000 -d30s http://localhost:8080/actuator/health
   Running 30s test @ http://localhost:8080/actuator/health
     12 threads and 1000 connections
     Thread Stats   Avg      Stdev     Max   +/- Stdev
       Latency    24.40ms   51.39ms   2.00s    99.10%
       Req/Sec     3.34k   651.18     5.31k    70.00%
     1197831 requests in 30.07s, 130.23MB read
     Socket errors: connect 0, read 1081, write 0, timeout 14
   Requests/sec:  39834.50
   Transfer/sec:      4.33MB

Where to find the dashboards' definitions

If you want to import them into your running Grafana, see dashboards/ folder for the Grafana dashboard definitions in JSON format.

One can also export the definition from the UI by clicking on "Share dashboard -> Export -> View JSON".