Doing Data Virtualization Right to Enable BI Agility
I’ve been participating in the web seminar series hosted by Informatica covering data virtualization (DV) technology. As many of you know I feel strongly about the impact this technology can have on your business and I believe there is a place for it within most companies. I just finished listening to the latest program in the series Doing Data Virtualization Right to Enable Agility and got to hear my friend Wayne Eckerson comment on the history of the technology.
Wayne shared an timeline of how data virtualization has matured over the past 20 years in it he points out that data virtualization first came to the industry under the moniker of Virtual Data Warehousing (VDW) in the early 90’s and was quickly dismissed by the physical data warehouse purists. The technology was interesting but went against the grain of most common data management practices of the time. Data Federation came later and gave way to Enterprise Information Integration (EII) in the 2000’s. The data virtualization technology we see in the market today started to get significant traction over the past five years or so. Why the history lesson? I think its important to point out that DV has been around a good long time and the foundation of this technology is rooted in significant experience coupled with today’s more powerful computing platforms and networks it’s a relevant and important technology especially when it complements existing approaches for managing data.
Rob Myers Manager, BI Architecture/EDW Solution Architect for HeathNow NY was on the program and shared in-depth insights into how they’ve implemented DV technology. Data virtualization has gone somewhat viral at HealthNow they have had success in implementing an SOA style data service for applications, they also leveraged their DV hosted common data model and definitions to enable a stronger Master data Management (MDM) foundation and in the end have addressed rampant data mart spread by delivering the DV access layer.
A critical requirement for HealthNow was the ability to couple Data Federation functionality with traditional data integration features such as data quality, profiling, cleansing and ETL. Addressing data quality issues helped gain acceptance with IT and business shareholders who badly needed a trusted data layer for applications within the company. Having the ability to switch between data federation and ETL has a made the environment more agile and given HealthNow flexibility to serve a wider variety of data requirements.
HealthNow used DV to take control of its data mart sprawl. When they began to implement the technology over 30K data marts were spread across the company. Delivering a unified and trusted data layer enabled them to meet the needs of the users and gain control over an out of control data landscape.
HealthNow’s success is a great example of the flexibility and power a data virtualization solution can provide. Follow this link to replay this last program, Wayne and Rob go into greater detail than I’m able to address here. I’ll be hosting an upcoming session in the series and in the meantime you can follow this link to sign up and listen in.
Previous posts in this series: