Sergey Nivens - Fotolia

How a logical data fabric can improve data management

As data continues to grow across different sources, speakers at the Fast Data Strategy virtual summit outline how a logical data fabric approach can help.

With data spread across different applications, databases and both on-premises and cloud environments, pulling it all together is no easy task.

Data virtualization provides one way of helping to connect disparate sources of data, creating a virtual abstraction that enables users to make use of data coming from different locations. But organizations also need an overarching framework above data virtualization to provide enterprise data governance and management. The emerging concept of logical data fabric plays into this key data management requirement.

A logical data fabric is an architecture that uses data virtualization to integrate and help govern data across both on-premises and cloud deployments.

Among the organizations that have embraced a logical data fabric architecture is global financial services firm Prudential Financial Inc., which outlined its approach at the Fast Data Strategy Virtual Summit on April 29, hosted by data virtualization vendor Denodo.

"In today's companies, decisions need to be supported by data at all levels in the organization," said Alberto Pan, CTO of Denodo, during a live-streamed presentation. "This means that now we have more people who need data in order to do their jobs."

Taking a Prudent(ial) approach to logical data fabric

Also, at the virtual summit, Ralph Aloe, director of enterprise information management at Prudential Financial, outlined the challenges his organization faces, that have been eased through the use of a logical data fabric.

Aloe noted that enterprise data governance is a big concern for his organization as data is not always stored in a consistent way.

When he joined the company in 2018, the data architecture was highly fragmented across a complex IT environment, Aloe said. Prudential Financial was also using many extract, transform, load (ETL) tools, to bring in data sets from multiple sources, which was another challenge in terms of both data management and data quality.

The company started out with a proof-of-concept exercise with data virtualization and has since expanded it into a broader logical data fabric architecture. Rather than having users faced with the prospect of trying to figure out where to get data, there is now a unified data process.

"One of the things we came up with was an evolving data fabric that occurred by leveraging data virtualization," Aloe said.

With data virtualization at the core, Prudential Financial could not look at enterprise data governance of all the data sources as part of a unified approach. It was also possible to enable a data catalog, so users within Prudential Financial could more easily identify and access data.

Prudential's Logical Data Fabric
Prudential Financial's logical data fabric architecture unifies multiple data sources and overlays services to improve data operations.
In today's companies, decisions need to be supported by data at all levels in the organization. This means that now we have more people who need data in order to do their jobs.
Alberto PanCTO, Denodo

Logical data fabric as a path to modernization

Beyond helping to enable data governance and data catalog capabilities, by embracing the logical data fabric approach Prudential Financial has been able to accelerate its modernization efforts.

"[Logical data fabric] also kind of helped us leapfrog modernization, where we can leverage virtualization as kind of like an interim component as we modernize back-end systems," Aloe said.

So, for example, if there was a business process that was using Excel spreadsheets, Prudential Financial was able to use data virtualization to help move into something more stable. Also, Aloe noted that the logical data fabric has helped to enable a less complex data architecture that makes data more useful to the business.

"What we're seeing with this data fabric and services built around it is a lot better alignment between our business strategy and our data strategy," Aloe said.

Making the case for logical data fabric

On another track at the conference, David Stodder, senior research director of business intelligence at data science research and education vendor TDWI, said many organizations want to be data-driven, basing many key decisions on data analytics.

During his presentation, Stodder noted that a key part of a successful data-driven model is to remove the delays that come between the steps of getting data from where it originates to where it can actually be used by users and analytics.

With a fast-rising number of organizations using cloud services, Stodder noted that the number of different data silos is also growing, citing TDWI research that shows 51% of organizations surveyed at the start of 2020 reported disconnected data silos as one of their biggest problems.

"Data virtualization provides that logical abstraction layer that shields users from the complexity of knowing how to access the underlying data sources," Stodder said. Data virtualization "leaves the data and so it reduces delays due to data movement," he added.

Over the last several years the concept of data virtualization has been expanded to enable logical data warehouses, where data from different environments is virtually connected into a single data warehouse. The logical data fabric is the next step, and extends data virtualization to integrate, manage and govern enterprise data across a hybrid, multi-cloud environment.

"A logical data fabric is designed to be a flexible and adaptable growing organism that gives us sort of a universal ability to connect things together," Stodder said.

Next Steps

5 principles of a well-designed data architecture

Dig Deeper on Data integration

Business Analytics
Content Management