What’s the difference between data mesh and data fabric?
Data mesh and data fabric are data analytics and management frameworks that are largely similar and overlapping, but with a few areas of distinction.
The data mesh framework was founded by Zhamak Dehghani (formerly with Thoughtorks consultancy) and is defined by her as a decentralized sociotechnical approach to share, access, and manage analytical data in complex and large environments–within and across environments.
The data fabric framework was developed by Gartner and is defined by Gartner as a design concept that serves as an integrated layer (fabric) of data and connecting processes. A data fabric utilizes continuous analytics over existing, discoverable, and inferenced metadata assets to support the design, deployment, and utilization of integrated and reusable data across all environments. This includes hybrid and multi-cloud platforms.
Both data mesh and data fabric frameworks have a focus on decentralized data ownership and management. Decentralization, in this context, means business groups and data owners manage their own data, or have a shared ownership of data with their corporate IT department, rather than data being centralized and owned through IT.
Data mesh is characterized specifically by four principles: decentralization, self-service access to data, domain ownership of data, and federated computational governance. Data fabric is characterized by broad active metadata, easy access to data, continuous learning from metadata, and automated deployment.
In terms of the areas of distinction, data mesh and followers of this framework tend to place a heavier emphasis on domain-owned data products, not centralized through IT, as the primary mechanism to achieve scale and faster time-to-value from data analytics. While data fabric and Gartner tend to place a heavier emphasis on the importance of global, active metadata as key to enabling decentralized, distributed, data environments to operate at scale and ease.