Engineering Data Mesh in Azure Cloud

5

Security Architecture for Data Meshes

The big promise of data meshes is to decouple producers and consumers, create data products, and allow the producers and consumers to share and consume the output of a product however they want. In the era of centralized data analytics, access control was governed centrally and was more focused on the consumption of the data stored in a central data repository because the serving and consuming layers were one and the same. In the data mesh world, this concept needs to change. Data is owned by the data product team. They are responsible for managing access to it. Additionally, they know which data is sensitive and which is not. While data product owners decide the parameters for access to their data, some ...