Virtualization Magazine on Ulitzer

Large enterprises and government agencies are drowning in data. IT teams deploy a myriad of data warehouse-centric solutions - BI, predictive analytics, data and content mining, portals and dashboards - to harness and deliver data for intelligent decision-making.

Yet, large enterprises are also expected to act like start-ups: nimble, agile and flexible to adapt to ever-changing market conditions. Impossible? By examining the best practices of their successful peers and adapting these to their own enterprises, data teams and enterprise architects can contribute to their corporate initiatives.

The intersection of data virtualization and enterprise data warehouses represents corporate best practices for delivering the rich data assets available in the enterprise data warehouse with the myriad sources of data now available outside the data warehouse. In Part One of this two-part series, I examine ways to use data virtualization to improve data warehouse effectiveness. By extending the warehouse schema to include additional data, data virtualization delivers greater business value from existing warehouse and non-warehouse data assets. Part Two will target improving data warehouse efficiency by showing four best practices where data virtualization, used alongside data warehouses, save time and money.

1. Data Warehouse Extension
Despite proliferating data silos accommodating exponentially growing data volumes, valuable data still remains outside the warehouse. Enterprise IT teams responsible for providing users with complete data in support of revenue, cost and risk management goals typically require the following combinations of data from both the data warehouse plus external sources:

• Historical data as well as current data from transactional systems or operational data stores.
• Summarized data and drill-down detail from transactional systems or operational data stores.
• Proprietary data and external data from outside sources including cloud computing

Data virtualization federates data-warehouse contents with additional information sources. These complementary views are conducive to adding current data to historical warehouse data, detailed data to summarized warehouse data, and external data to internal warehouse data.

In Figure 1, data virtualization middleware hosts new views that integrate additional RDBMS and web service data sources as well as extend existing sources such as packaged applications.

Industry Examples
A global energy company uses data virtualization to federate its real-time data about crew, equipment and production status from its wells with its SAP maintenance management system of historical surface, subsurface and business data.  This information enables the optimal deployment of repair crews and equipment across 10,000-plus production oil wells, resulting in faster repairs for greater uptime and more profitability.

To analyze the effectiveness of its pharmaceutical sales and marketing programs, a life sciences leader uses data virtualization to federate externally sourced competitor sales data from an industry data services provider with internal prescription sales data from its sales data warehouse. This total market view provides the sales team with the intelligence required for more effective sales and marketing programs that increase revenues.

2. MDM Hub Extension
Master Data Management (MDM) hubs maintain and control critical master data attributes, but not the detailed transaction histories and other related data maintained and controlled in dozens of other systems across the extended enterprise. Data virtualization acts as the foreign key to quickly and easily federate master data within an MDM hub with additional transactional and historical data, delivering a comprehensive view of customer, product and employee information.

In the integration pattern shown in Figure 2, the data virtualization middleware hosts new complementary views that integrate additional RDBMS and Web service data sources as well as extend existing sources, such as packaged applications.

Industry Examples
A European mobile phone operator, known for its customer service leadership, maximizes revenue per customer and customer service levels by using data virtualization to combine customer service data from its customer reporting data warehouse, billing data from its financial systems, and call and configuration data from its operational support systems to deliver a 360o view of customers to customer service representatives. Faster issue resolution and more productive up-sell programs reduce churn, increase customer satisfaction, and ultimately contribute to revenues.

A global investment bank federates its HR master data with myriad internal benefits and compensation systems as well as external payroll services to provide its employees with a 360o view of their total compensation through a self-service employee benefits portal. Securely exposing this information improves retention and lessens HR staff workload.

3. Data Warehouse Federation
A fundamental reason enterprises implement data warehouses is to overcome the disparity of transactional and analytical system silos typical in today's large enterprise and government agencies. However, the single "enterprise" data warehouse remains elusive. Instead, multiple data warehouses and data marts have been developed and deployed, perpetuating, rather than overcoming, data silo proliferation.

Optimizing business performance requires data from across these various warehouses and marts. The effort of physically combining multiple marts and warehouses into a complete enterprise-wide data warehouse is simply too costly and time-consuming.

Data virtualization federates multiple physical warehouses. Two examples include combining data from sales and financial data warehouses, or combining two sales data warehouses after a corporate merger or acquisition. This approach creates an integrated view by using abstraction to rationalize the different schema designs, and thereby achieves a logical consolidation of multiple warehouses.

In the federation pattern shown in Figure 3, the data virtualization middleware hosts federated warehouse views that logically integrate both data warehouses.

Industry Examples
To enable more flexible customer self-service reporting and meet SEC compliance reporting mandates, a leading prime brokerage federates equity, fixed income and other investment positions and trades information from siloed trading data warehouses into a data virtualization layer. The net result is higher customer satisfaction and lower reporting costs.

One of the world's leading pharmaceutical companies uses data virtualization to enable research scientists to access and analyze data from research, clinical trial, FDA submission and other data warehouses. Scientists use this data to accelerate time-to-market for new compounds and drugs, thereby increasing revenues in an otherwise lengthy and costly development process.

4. Data Warehouse in Enterprise Architectures
Increasingly enterprises are seeking unified ways to integrate warehouse and other data in an enterprise-wide information architecture. According to Cambridge, Mass.-based industry analyst firm Forrester Research, "New architectural approaches such as Information-as-a-Service (IaaS) have emerged to provide flexible, real-time, service-oriented data integration and data-quality capabilities that support both structured data and unstructured content, delivering a true information integration platform." (1)

Data virtualization, which is included in IaaS, integrates data warehouses into an unified enterprise information architecture, as shown in Figure 4. IT teams use data virtualization middleware to form an enterprise data virtualization layer that is home to a consistent and complete logical schema covering multiple consolidated and virtual sources. When designing the enterprise information architecture, developers use data virtualization design tools to develop semantic abstractions in the form of web services or relational views. At runtime, end user-level applications, reports or mash-ups are created to call web data services, on demand, to query, federate, abstract and deliver the requested data to information consumers.

Industry Examples
To meet the information needs of its diverse group of technical and business users, an energy company has deployed enterprise information architecture to deliver disparate warehouse and operational data from more than 12 refineries located around the world. As a result, the company's refinery yields are increasing, and it proactively maintains its equipment, reducing downtime. In addition, the company complies with a myriad of regulations more consistently and at a reduced cost.

Several government agencies are using data virtualization to create a common information layer that spans agency information databases and enables intelligence analysts to better control threats. Agencies involved include the Drug Enforcement Administration (DEA) and the Immigration and Naturalization Service (INS). The common information layer delivers access to passenger, crew and manifest data from a U.S. Coast Guard port arrivals data warehouse, for example.

Conclusion
As data sources proliferate, including many web-based and cloud computing sources outside traditional enterprise data warehouses, enterprises and government agencies are deploying solutions that represent the intersection of enterprise data warehouses and data virtualization. These hybrid solutions deliver the most comprehensive information to decision-makers and positively impact the bottom line. Additional results include extended life for existing information system investments, greater agility for adding new BI and other analytic technologies, and less disruption from corporate activities including mergers and acquisitions.

Resources
1. 2009 Update: Evaluating Integration Alternatives, Scenario-Based Guidance For Choosing Products That Provide Application, Process, And Data Integration Features, Ken Vollmer Copyright (r) 2009, Forrester Research, Inc.