Welcome to the second part of our Snippets series on real-time stream processing using Apache Kafka! To recap, in part one we introduced stream processing and discussed some of the challenges involved, like the stateful nature of aggregations and joins. We’ll discuss the impact of such stateful operations in a while, but first, let’s delve deeper into one key aspect of real-time stream processing: performance.

Apache Kafka is probably primarily known as a messaging middleware with a more flexible structure than queues, but it also empowers teams with its lower entry barrier for real-time data pipelines.