Streaming Databases

Chapter 7. Emergence of Other Hybrid Data Systems

In this chapter, we broaden our focus to include the greater landscape of hybrid systems that have surfaced in response to the growing demands of modern real-time event-driven applications. While these systems are not streaming databases as we defined them in this book, they share qualities and features that bridge between relational, analytical, and streaming workloads. We will explore the motivations behind their development, the innovative techniques they employ, and the specific use cases that make them relevant. More importantly, we will discuss the niches these other hybrid databases cover. This understanding will allow us to uncover the trends that databases are following to provide real-time analytics.

It’s important to acknowledge that a streaming database is also an example of a hybrid system. Hybrid systems take at least two perspectives, and in the streaming database case, the two perspectives are stream processing and the database.

Appreciating the perspectives of hybrid systems will reveal the problems that they try to solve and how. In this book, we define streaming databases from the stream processing perspective as follows: a streaming database is a stream processor that exposes its state store for clients to issue pull queries.

An alternative definition created from the database perspective is as follows: a streaming database is a database that can consume and emit streams as well as execute materialized views ...