Streaming Data Mesh

Chapter 3. Domain Ownership

The first of the four pillars of a streaming data mesh that we will discuss is domain ownership. Defining a data domain can be difficult, especially when data is often shared with multiple domains, or when a domain has a dependency on data from another domain.

However, at its core, data domains are quite simple to define. Domains are the people and systems involved in generating data that is logically grouped. A domain refers to an area of interrelated data pertaining to a common purpose, object, or concept. This data will eventually get shared (and potentially transformed) into published data products. Data domains also have output ports that define how data products will be shared and input ports that define how data is consumed from other domains (see Figure 3-1).

Figure 3-1. A data mesh domain can have many input and output ports

These ports build the interconnections between domains that create the data mesh. The ports represent the production and consumption of data products. In a streaming data mesh, these ports are streaming data products.

The systems and teams internal to the domain do not get exposed to the mesh. They do help support the management and development of the streaming data products. In this chapter we will outline ways to identify domains to ensure that the domain will operate seamlessly in the streaming data mesh.

Identifying ...