Data Fabric and Data Mesh continue to sustain legions of hype and debate. Data and analytics leaders are longing for a new roadmap, beyond traditional data management practices, and towards fast and cost-effective solutions. On this quest, many wonder whether Data Fabric and Data Mesh are the same or different; and if it’s one or the other – which is right for their business? The way we see it, the conversation isn’t about what’s better, but that there are merits to both paradigms and both can co-exist together.
Data Fabric modernizes data integration and aids data movement for data that needs to be moved or centralized. Data Mesh is a distributed data pattern carrying many organizational and business process elements that facilitate faster analytics on more data. Both approaches feature the importance of data products, which is what Gartner talks about in their latest research report, Are Data Fabric and Data Mesh the same or different? Professor of Computer Science at the University of Maryland, College Park Daniel Abadi examined the difference between Data Fabric and Data Mesh. Let’s reexamine the difference from the point of view of Gartner.
The difference between Data Fabric and Data Mesh, according to Gartner
“A Data Mesh is a solution architecture for the specific goal of building business-focused data products without preference or specification of the technology involved. A Data Fabric is a technology utilization and implementation design capable of multiple outputs and applied uses.”
Below we go through a few of them:
- Data Fabric (semantic) is promoted by some platform and tool providers as a completely semantic or virtualization solution. Most of these solutions actually include some form of caching of data and included optimization techniques.
- Data Fabric (AI-/machine-learning-augmented), as described by Gartner, begins initially with the accumulation of passive metadata, then progresses into an active metadata scenario. Examples are detailed in Gartner’s research in three case studies, as referenced in the next section.
- Data Fabric (archaic) was initially proposed as a hardware and systems resource allocation model for determining the sizing and assignment of infrastructure components based on processing requirements — whether related to storage, networks, processors, or memory requirements.
- Data Mesh network (dynamically overlaid data management and infrastructure), as proposed by Gartner in 2016, combines the capabilities of communications Mesh networks with an AI-/machine-learning-augmented Data Fabric and a process-based Data Mesh.
- Data Mesh (data-product-based) asserts that authority, governance, and management of data resides in source systems. A Data Mesh further asserts that it can be leveraged to create domain-based and integrated solutions without specifically refactoring the physical data assets in an abstracted and semantic approach supported by use-case-specific processes for use-case context.
- Data Mesh (process-based) is currently represented by the utilization of technologies such as Kubernetes and LinkerD to create a software-enabled, orchestrated process model for relaying data throughout a network of use cases. Such a network includes various performance tuning via resource asset allocation and utilization. Note: This nomenclature predates the usage of “Data Mesh” provided above.
Based on these definitions, it’s apparent that data without context with our current infrastructure is no longer sufficient in a data-driven economy.
Data Fabric versus Data Mesh
- The total cost to deliver either one may ultimately be similar relative to design and deployment. However, the more augmented data management capabilities included in a Data Fabric improve the cost model for ongoing improvement and maintenance.
- Data Mesh and Data Fabric benefit from one another, either adapting to or leveraging best practices.
- Both Data Fabric and Data Mesh materialized from mature data management practices and are based on over 50 years of data management technology advances.
Data Fabric requires deep technological investments as it depends heavily on the capabilities of tools and platforms. Meanwhile, Data Mesh shifts the focus of cost toward services. In the cloud or multi-cloud era, subscription spending is a consideration: services or technology?
Organizations can consider their business needs and act accordingly to budget and data strategy.
Quick Answer: Are Data Fabric and Data Mesh the Same or Different?, 1 November 2021, Ehtisham Zaidi, et. Al.
GARTNER is a registered trademark and service mark of Gartner, Inc. and/or its affiliates in the U.S. and internationally, and is used herein with permission. All rights reserved.