Like (Elasticsearch, Fluentd, and Kibana (EFK)), Jaeger is a complex system that collects and processes quite a bit of information. It is shown here:
This is the architecture of how Jaeger worked at Uber as of 2017. The configuration uses the side-car pattern we mentioned earlier, with each container running a nearby container that collects the spans from the instrumentation using UDP, and then forwards those spans into a storage system based on Cassandra. Setting up a Cassandra cluster, as well as the individual collectors and the query engine, is far more than can be easily created in a local development ...