Preparing for the Future Data Center

The average Joe does not know that when he withdraws money from a local ATM machine, he is initiating a transaction that will update his profile at his bank's main data center. Sometime later, that same transaction will be copied over to another data center to ensure that the bank adequately protects Joe's information, which is a critical bank asset.

The data center houses the most expensive technologies known to the organization. Companies spend tons of money to build data centers that are equipped with the latest in security and built to withstand disasters. This is because the business is dependent on applications that run from the data center. E-mail, data warehouse, billing systems, financial systems, ERP and CRM systems are housed in these near bulletproof rooms.

"Your typical data center is set up to provide a controlled, secure environment for computer equipment. There is a lot of expensive infrastructure set up to enable everything that happens in the production area (computer room) to run smoothly. This includes power systems, environmental controls and security," says Shane Gaskin, manager of the Unisys' Asia Pacific Managed Services Command Center.

The problem today is that the equipment stored in the data center is evolving rapidly. We are seeing faster servers, higher capacity storage systems, intelligent network equipment designed to thwart the latest in hacking best practices. Users are also moving farther away from headquarters taking with them critical applications that require the same quality of service delivered to the branch as that available at HQ.

"We call this real-time infrastructure (RTI) with what drives it is software embedded with business information policy," said Mark Feverston, vice president of enterprise servers and storage solutions at Unisys. "RTI allocates computing resources directly and dynamically to business customers' strategic business processes. The operating environment must be able to share application workloads dynamically and automatically optimize resource usage."

According to Charlie Bess, EDS Fellow, "in the future, the forces of standardization, commoditization and virtualization will drive down the cost of the data center and reduce the time to get a configuration online, providing greater capability through the dynamic assembly of lower cost processors into massive networks of computing capability. Until now data center complexity was managed using the most flexible tool available - people. The data center environment of the future will be far too complex for an individual to comprehend. Industrial techniques that have been applied to process manufacturers will be brought to bear."

Nicholas Gall of the Gartner Group includes innovation as another driving force in the evolution of the data center. "The rate of change of the integrated systems and their interconnections is accelerating," notes Gall. The changes are occurring at the network topology, software configuration, and application integration. We are witnessing a long-term shift in scope from assembly of standard systems based on standard components that comply with standard designs to assembling and configuring standard data centers out of virtual components based on standard designs."

Dr TC Tan, a Distinguished Member of the SYSTIMAX Labs notes that current design parameters for data centers will not change significantly. "A data center is built to provide service, and as such simplicity, flexibility, scalability, availability, reliability, efficiency, security, location and environmental controls will continue to be concerns for data center designers in the future. Bandwidth hogs like virtualized systems will require robust and high performance cabling infrastructure."

EDS Fellow Dr Rene Aerdts believes that, "The future data center will require power than we can only imagine today." Aerdts predicts that some data centers will need to be built near a water source to provide cooling." Today's nuclear reactors of today are water cooled and are thus built near a water reservoir or river.

Present-day remote control technology will also mean that one day we don't need to be physically present at a data center to do routine maintenance. We will have 'lights out' data centers where the only time you need to power on the lights is when you have someone coming over to perform a physical job. Everything else will be done remotely or through robots that are wirelessly connected via the Internet.

The future data center will use historical data to perform real-time assembly of modeled configurations. "These same parameters will feed into autonomous monitors, simplifying the management process into something that people can understand," hints Bess.

In essence, "The data center of the future becomes more of an enterprise value network center, pulling together its resources, adding value to organizations through its flexible and powerful capabilities. Its limits will be on our understanding of how to gain benefit from it. Where the data center is located and who owns the hardware within it will be of less concern than the advantage it provides to the enterprise. Even with all this automation, it will be the people and their understanding of the business that will make the difference."

The data center of the future will be designed for a service-driven organization supporting the business environment. Provisioning of resources will be done in real-time using a combination of pre-determined policies and algorithms that allow for some level of decision making following set parameters -- the precursor of artificial intelligence.

"As a result, automated, virtualized, operational services (i.e., design, assembly, provisioning, monitoring, change management) become essential for managing the complex dynamic relationships among the components," predicts Gall.

"Given that technology is continuously improving coupled with the increasing speed of change in the business environment, predicting what will happen in 10 years' time is a tall order. No one can be certain that some of today's best practices will be inadequate in 5 years' time. Hence, IT organizations must approach planning, implementing, measuring, and improving their high-availability systems as a continuous process," advises Tan.

Well said.