Learning Spark SQL

Using Spark SQL in streaming applications

Streaming applications are getting increasingly complex, because such computations don't run in isolation. They need to interact with batch data, support interactive analysis, support sophisticated machine learning applications, and so on. Typically, such applications store incoming event stream(s) on long-term storage, continuously monitor events, and run machine learning models on the stored data, while simultaneously enabling continuous learning on the incoming stream. They also have the capability to interactively query the stored data while providing exactly-once write guarantees, handling late arriving data, performing aggregations, and so on. These types of applications are a lot more than ...