Building an Event-Driven Data Mesh

Chapter 11. Bringing It All Together

Data mesh is founded on four principles, based on hard-learned lessons of dealing with difficult-to-use data, often provided as no more than a form of exhaust emitted by operational applications. Mechanisms for obtaining and accessing data outside of its original source are typically added as an afterthought, bolted onto the source system in an ad hoc manner.

Data mesh promotes data to a product with the same rigor, ownership, and feature management of any other product in your business. The free-for-all, “figure it out yourself” data access is replaced with purpose-built, maintained, and supported modes. It is as much a social shift as it is a technological shift and requires both top-down and bottom-up buy-in. We reevaluate how we own, create, discover, and access data, and then build out the processes, governance, and technology required to make it work.

Event streams provide the number one option for making data mesh a reality. They provide a singular mechanism for both historical and real-time data communications, forming the data mesh’s basic building blocks for consumers to use, mix, and match to their needs.

Data mesh is based on four main principles: domain ownership, data as a product, federated governance, and self-service platform. Together, these principles help us focus on communicating important business data across the entire organization safely, effectively, and at-scale:

Domain ownership

Those who know the data best are ...