Complex Event Processing Using Apache Flink

What you'll learn

The Apache Flink library supports complex event processing using the Complex Event Processing or the CEP library which is built on top of the Flink core APIs.

In this course, Complex Event Processing Using Apache Flink, you'll learn about complex event processing involving defining patterns that you want to detect and identify in the input stream, either across the stream as a whole or within a certain time window.

First, you'll be introduced to the Pattern API which allows you to define patterns that you want to detect in input streams. You'll create patterns using singleton patterns that match only one event in the input stream, you'll also use looping patterns created using quantifiers that can match multiple events in the input stream.

Next, you'll see how you can specify different kinds of conditions to accept events into your pattern. Simple conditions only look at the properties of an event to perform matching, iterative conditions can look at the previous matched events and even compute statistics on the input stream to accept matching events. You'll see how more complex conditions can be created using combining operations with logical 'AND' and logical 'OR.' You'll also explore the contiguity constraints that we can define on patterns including strict contiguity, relaxed contiguity, and non-deterministic relaxed contiguity.

Finally, you'll see how you can control how many times a particular event matches a pattern using the after match skip strategy. You'll understand the differences between the no skip, skip to next, skip past last event, skip to first, and skip to last strategies. You'll also see how you can perform pattern matching within a time window, and handle partial timed-out matches.

When you're finished with this course, you'll have the skills and the ability to detect patterns in input streams using the pattern API in the Apache Flink Complex Event Processing library.