The Transformation of Oracle IT’s Data Center Practices
Oracle IT went from being a siloed, functionally oriented organization to a key entity in supporting the transformation of Oracle as a serial acquirer.
by Nardo van der Rijst, June 2010
In 2006, Oracle IT was faced with significant physical and power limitations. Oracle’s data centers were going to run out of space by the end of the year. The proliferation of servers, storage, and communications hardware resulted in significant capital expense. With Oracle’s merger-and-acquisition strategy in full force, a perfect storm of data-center capacity, new acquisitions, growth in the On Demand business, and technology challenged the ability to scale Oracle’s growth.
How IT Changed its Data Center Practices Oracle IT has three main customer segments for its IT services: On Demand, internal operations and development, and Oracle University. Among the internal data-center services, IT manages large, single-instance applications for CRM, financials, human resources and business analytics. Currently, IT operates two main data centers to support worldwide service delivery — one in Austin, Texas, and one in Colorado Springs, Colorado. Both data centers are in active-active connection, with the Colorado Springs data center (RMDC) serving as the backup site for Austin (ADC).
Playing such an important role in accomplishing Oracle’s strategic vision, IT needed to develop a systematic, multi-year transformation plan to establish a true utility computing services delivery organization. This implies a set of shared services that can be delivered on demand. In this avatar, IT runs as a business within a business, providing a customer-centric approach leveraging frameworks such as Information Technology Infrastructure Library (ITIL). The resulting services are delivered for the lowest cost possible, with the highest efficiency, while improving service levels at the same time. In its transformation, IT needed to implement the following key steps:
Standardize hardware, infrastructure, and operations
Virtualize across the technology stack
The end result is complete services automation by leveraging a utility computing model.
Consolidation Consolidation was a necessary first step in the road towards IT as a service, to reap immediate benefits from less complexity in the overall infrastructure. Major benefits were achieved by consolidating all enterprise applications that supported key operational processes at Oracle, such as procure to pay and hire to retire, into one environment supported on one platform. A total of 51 application instances were consolidated to one instance. The consolidated infrastructure would allow newly acquired companies to be absorbed quickly.
Standardization IT adopted ITIL to standardize the processes in the data center, and applied a similar logic to the platforms needed to deliver those services. IT services management (ITSM) was used to focus on real customer needs that added value.
Virtualization Virtualization of all assets — servers and storage — meant a further abstraction away from specific vendor platforms, applications and skill sets. IT was an early adopter of storage virtualization tools and techniques as a way to aggressively increase storage capacity while reducing the physical footprint. Virtualization was also applied at the server level to allow multiple OS images to run on one physical server. As an added benefit of standardizing on Oracle Database, all of its features to support virtualization at the database level could be exploited to create a shared database services environment. Server virtualization has resulted in a utilization that is close to 70 percent, up from seven percent as measured at the beginning of the initiative.
Utility Computing Oracle IT has created a physical environment in its data centers that has achieved a high level of abstraction for platforms, server, storage and applications. Skill sets are now generic, and as a result available staff can be deployed on a wide variety of tasks following standardized processes which, in many cases, are completely automated. All of these accomplishments are key ingredients to working toward a utility computing model to serve internal as well as external customers.
With several elements in place for the next phase in IT’s master plan — to establish utility computing — IT is focused on the self-service provisioning of environments for all of its customers. Both internal customers such as developers and Oracle University, as well as current On Demand customers, use self-service capabilities to kick off automatic provisioning processes.
Conclusion IT went from being a siloed, functionally oriented organization to a key entity in supporting the transformation of Oracle as a serial acquirer. Scalability of IT processes, once seen as a real bottleneck for integrating companies, is now a strong suit. IT managed to implement a solid foundation for utility computing through the adoption of the ITSM framework, and standardizing on Oracle technologies in the data center has provided the automation needed to be cost efficient and effective at providing IT services.
Nardo van der Rijst is director of Industry Strategy and Insight, and a member of Oracle Insight Global Data Center Practice at Oracle.