Page 2 of 3
Question and Answer: Ten Tips for Maximizing Data Virtualization
By Linda Briggs
Global enterprises and government agencies that are implementing data virtualization in IT face a formidable set of challenges -- along with plenty of opportunities. By refining and maximizing the benefits from data virtualization, companies can squeeze more ROI out of their efforts -- for example, by using data virtualization to address challenges such as rogue data marts. IT leaders can also work to ensure that data virtualization efforts pay off by contributing to the corporate bottom line.
Data virtualization has its roots in enterprise information integration (EII) and data federation, and its adoption has accelerated rapidly over the past several years. Given that, what should enterprises and government agencies do to ensure that they get the most from their data virtualization investments? In this interview, Robert Eve, executive vice president of marketing at Composite Software, Inc., offers up ten data virtualization “to dos” for 2010.
TDWI: First, can you provide a quick definition of data virtualization?
Robert Eve: Data virtualization is a form of data integration. It’s used to integrate data from multiple, disparate sources in a logically virtualized manner, rather than a physically consolidated one, for on-demand consumption by end applications. Those applications might include business intelligence and portals.
Another term for data virtualization is data federation. In fact, TDWI recently published a TDWI Checklist Series report on data federation that provides a rich description of popular use cases, underlying capabilities, and more. Wayne Eckerson, director of TDWI Research, researched and wrote the report, which was sponsored by Composite Software.
From your perspective as a data virtualization company, how are enterprises and government agencies currently using the technology?
We saw our business increase by more than 50 percent in 2009 in spite of a sluggish economy, which shows that data virtualization interest and use have both increased significantly as of late. The following three recurring use patterns have become mainstream:
- Data federation: Used for rapid, project-scale data integration using views, data services, caches, virtual data marts, and so forth.
- Data warehouse extension: Used as a way to add data to an existing warehouse to achieve more value from prior warehouse and BI investments.
- Enterprise data sharing: With a focus on the reuse of integrated data across multiple projects, enterprise data sharing is often used in conjunction with standards compliance initiatives and service-oriented architecture (SOA) strategies.
You have a list of 10 virtualization suggestions for 2010. What’s on the list?
The list is intended as a quick set of suggestions to remind companies of ways to make the most of their data virtualization use and experience.
1. Extend existing data warehouses: Data virtualization enables IT to integrate additional data such as external, up-to-the-minute or detail data with existing data warehouse schemas.
2. Eliminate rogue or unauthorized data marts: Virtual data marts can eliminate -- or at least significantly reduce -- the out-of-control, dependent data mart spokes surrounding many of today’s enterprise data warehouse hubs.
3. Think of a 360-degree view: Remember that data virtualization can enable the combination of master data management (MDM) hubs with detailed transaction histories and other related data, all maintained and controlled in dozens of other systems across the extended enterprise.
4. Integrate cloud data sources: Data virtualization integrates on-premise and cloud data without additional replication while conforming to cloud security firewalls and APIs.
5. Conform data to industry standards: Data virtualization can enable rapid development of XML standards-compliant data services to simplify data sharing within and across enterprises and government agencies.
6. Leverage advanced caching options: Although data virtualization is designed to query fresh data on demand, sometimes there is value in temporarily persisting data in a file, database, or in memory. That offers a way for IT to minimize query impact on operational systems, access data from temporarily unavailable sources, and support service-level agreements (SLAs).
7. Cross the relational view and XML data service divide: Beyond relationally oriented data virtualization for BI, data virtualization works equally well with XML- or semi-structured data. In fact, data virtualization is often the only way to combine the two.
8. Deploy a data virtualization integration competency center (ICC): With wider data virtualization use, competency centers focused on integration increase the opportunities for sharing best practice methods, objects, and resources.
9. Move toward an agile enterprise data architecture: Increasingly, IT architects gain data agility by decoupling underlying source data and consuming solution layers using data virtualization middleware.
10. Measure data virtualization success: It’s a good time to calculate the actual return on prior data virtualization investments to make the case for new investments or expansion projects.
Given this list, what are some of the results you’ve seen from companies that have implemented these items?
Results tend to fall into two major categories: business and IT. Business value includes revenue increases, compliance improvements, and cost reductions due to better return on virtualized data assets. Data virtualization delivers this value faster than traditional integration methods. IT results include staff, resource, and total cost of ownership (TCO) savings advantages over traditional integration methods.
What are some general guidelines for companies that want to prioritize potential data virtualization projects?
The best way to think about priorities is based on the data virtualization adoption stage. When IT leaders are just getting started with virtualization, they should prioritize smaller, rapid-return projects to quickly build a foundation of experience, staff, virtual assets, and best practices. Having successfully completed projects or enterprisewide implementations, the priorities are based on business return on investment (ROI).
What does Composite Software bring to this discussion?
Composite Software, Inc. focuses solely on data virtualization. Global organizations faced with disparate, complex data environments, including 10 of the top 20 banks, six of the top 10 pharmaceutical companies, four of the top five energy firms, major media and technology organizations, and government agencies, have chosen Composite’s data virtualization platform to fulfill critical information needs -- faster and with fewer resources. Scaling from project to enterprise, Composite’s middleware enables data federation, data warehouse extension, enterprise data sharing, and real-time and cloud-computing data integration.