Seven Secrets to Data Virtualization Success

Robert Eve is executive vice president of Marketing at Composite Software.

Historically, seven has been a sacred number. Creation took seven days to complete. To the ancient astronomers, there were seven planets. Pythagoreans revered seven because it was the sum of three and four.

Seven Data Virtualization Keys

Consider applying seven secrets practiced by your enterprise counterparts to make your own advanced data virtualization projects and architectures successful.

 

Today's IT teams responsible for data integration may consider applying the following seven secrets practiced by their enterprise counterparts to make their own advanced data virtualization projects and architectures successful.

1. Just Say Yes

Pfizer Inc.'s Portfolio Management team implemented data virtualization to leverage reusable views and extensible caches, thus simplifying the data integration of complex pharmaceutical research data sources. By avoiding the long development cycles of traditional extract-transfer-load (ETL) and physical data consolidation, Pfizer "says yes" to new business requests, typically delivering data for new reports in hours or days instead of weeks or months.

2. Mix Physical and Virtual

The information architecture team at Compassion International designed a mix of physical and virtual data integration in their enterprise-wide Ministry Information Library (MIL). Providing their 150-plus IT team members with greater agility and lower total cost of ownership (TCO), the MIL intelligently combines five physical and virtual data integration solutions including:

• Data virtualization using Composite Software's Composite Information Server
• ETL using Informatica's PowerCenter
• Enterprise Data Warehousing (EDW) using Kalido's Dynamic Information Warehouse
• Master Data Management (MDM) using IBM Initiate's Master Data Management Platform
• Enterprise Service Bus (ESB) using Neuron's Enterprise Service Bus

3. Think Layers

By taking a layered approach, enterprises including Deutsche Bank can decompose complex problems such as huge data volumes in diverse application silos into manageable, yet well-aligned components. Typical layers in any enterprise include:

• Data Consumers Layer-Client applications retrieve data in various formats and known protocols such as SOAP, REST, JDBC and ADO.Net
• Mapping Layer-Business views align with data consumers, for example, formatting into XML from SQL, and accepting parameterized inputs
• Business Layer-Logical views allow consistency and sharing where the business has a standard or canonical way to describe key business entities such as customers and products
• Formatting Layer-Physical data sources are rationalized by name aliasing, value formatting, data type casting, derived column creation, and light data quality
• Physical Layer-This "as-is" metadata layer helps bind the required physical data sources
• Data Sources Layer-These physical data assets include databases, data warehouses, Web services, Excel spreadsheets, packaged applications such as SAP, and more

4. Embrace IT and Data Standards

Dell and the U.S. Army have successfully applied data standards to accelerate their enterprise-wide deployments and reduce TCO. Examples of IT and data standards include:

• Development standards such as SQL, XQuery, Xpath, Java, and MDX
• Web services standards such as WSDL, WS-I Basic Profile, WS-Policy, and more
• Industry data standards such as PIDX and MIMOSA in Energy and ICDL in Intelligence
• Source and consumer API standards such as ODBC, JDBC, ADO.Net, SOAP, JMS, and REST
• Security standards such as LDAP, Active Directory, WS-Security, SAML, and IC-ISM
• System management standards such SNMP

5. Optimize, Optimize, Optimize

Similar to "location" for real estate agents, successful data virtualization initiatives focus on "optimize, optimize, optimize." For example, Wachovia's investment customers with millions of dollars under management have little tolerance for slow queries. Neither do Nielsen's research customers as they analyze TV ratings. With a powerful set of query optimization techniques readily available, data virtualization performance is typically never an issue.

6. Extend Your Integration Competency Center

Integration Competency Centers (ICCs) manage and improve large-scale, ongoing data integration activities. Data virtualization fits easily in ICCs originally developed for ETL, ESB, and other integration technologies. Northern Trust and Qualcomm quickly leveraged data virtualization with their existing governance policies, design and development staffing and best practices, shared services resources, and more to accelerate new development projects, and reduce staffing and hardware costs.

7. Fund Expansion with Savings

Data virtualization adopters including Comag, a joint venture of Hearst Corporation and Conde Nast, often begin with initial smaller project investments to gain rapid ROI - in many cases, up to 200 percent in six months or less. By measuring the savings in time-to-solution, hardware, software, and staffing achieved through this rapid ROI, these users have reversed traditional funding approaches by paying for new data virtualization investments with the already delivered savings.

Looking to overcome complexity in your enterprises' data integration systems? You can achieve success by applying today's enterprise data integration wisdom of "seven."

Robert "Bob" Eve is executive vice president of Marketing at Composite Software. Prior to joining Composite, he held executive-level marketing and business development roles at several other enterprise software companies. At Informatica and Mercury Interactive, he helped penetrate new segments in his role as the vice president of Market Development. Bob ran Marketing and Alliances at Kintana (acquired by Mercury Interactive in 2003) where he defined the IT Governance category. As vice president of Alliances at PeopleSoft, Bob was responsible for more than 300 partners and 100 staff members. Bob has an MS in management from MIT and a BS in business administration with honors from University of California, Berkeley.