Data Governance in a Data Mesh or Data Fabric Architecture
Reading Time: 2 minutes

Data mesh is a modern, distributed data architecture in which different domain based data products are owned by different groups within an organization. And data fabric is a self-service data layer that is supported in an orchestrated fashion to serve an entire organization. Both architectures can work together, in tandem. However, may people wonder, what is the best way to establish effective data governance across either or both of these architectures?

The Power of a Logical Approach to Data Management

The most sensible approach is to implement a logical data management layer that performs not only data management, but also data integration, data delivery, and data governance. Such a layer can be established by the Denodo Platform, with its data virtualization capabilities.

Let’s take a look at how this works.

People sometimes engage with data from a metadata perspective, and sometimes they begin with a query. The Denodo Platform starts by managing the metadata in a data mesh or data fabric architecture, by normalizing the views of disparate data sources into a logical or semantic layer. In that layer, stakeholders can build multiple different views that lead up to a more canonicalized view.

In a data mesh architecture, different groups are responsible for defining different business entities, such as products or supplies, as well as departmental entities such as HR or finance. With the Denodo Platform, they can build these definitions into the logical models, even though the relevant data sources are disparate, and then expose and share the canonical definitions. Similarly, in a data fabric, multiple data services can be exposed in this fashion.

The Denodo Platform can access business metadata from other third-party data governance tools, such as those from IBM and Informatica, and expose the combined metadata to a data catalog, so anyone with the right credentials can easily explore the metadata. But this is not just a static data catalog; it enables stakeholders to both query the data and monitor access. In this way, the catalog becomes a gateway to an organization’s data, one that can be easily and flexibly secured and governed. 

The Denodo Platform connects the metadata layer to execution, because it can leverage tags and classifications of the data either generated from the technical metadata or the business metadata to enforce, for example, policy based security and access controls. The Denodo Platform can also implement dynamic access controls, such as connecting user to their location at the time of their query. Because all queries come through the Denodo Platform, it is also in the privileged position of being able to make powerful, artificial intelligence (AI)-based recommendations based on usage patterns, such as recommending additional data sets or even potential joins.

Data Governance for Modern Data Architectures

With its logical data management capabilities, the Denodo Platform functions as a complete system of engagement from metadata to actual data, across a wide range of sources in a data mesh, a data fabric, or both at once. We have resources for more information on this topic; please reach out if you have any questions.

Suresh Chandrasekaran