Even though distributed tracing is still a bit of a novelty in the software engineering industry, the open source world is making great strides in making free tracing infrastructure available to anyone, from data gathering via projects like OpenTracing, OpenCensus, and W3C Trace Context, to storing and processing the data via many open source tracing backends like Jaeger, Zipkin, SkyWalking, and Haystack. As the tracing infrastructures become commodities, data mining and data analysis are going to be the areas of the main focus of research and development.
In this chapter, we covered some basic techniques for building data analysis tools on top of the tracing data, including looking at some of the challenges, such as trace completion triggers, which do not yet have perfect solutions.
We ran through an exercise of building a feature extraction framework and a sample span count job that can be used as a foundation for a full-featured platform.
Finally, we reviewed a very promising approach...