Where is the Data Center Going? Top Trends Today

Data is evolving into the true universal currency, with organizations of every stripe increasingly reliant on the data center as the linchpin of operations. Regardless of the nature of the endeavor, success more and more frequently depends on reliable, timely access to data and applications and getting the right information to the right people at the right time.

An offshoot of this trend has been tremendous growth in the amount of information and applications data center professionals are being asked to manage. More than a third of data center managers reported databases exceeding a terabyte in size at the end of 2006, and nine of ten expected their operations to grow over the coming year. ¹ And while growth in 2007 IT budgets has been stronger than expected, data center managers are still being asked to do more with less. ²

Following are some of the top trends currently affecting data centers.

1. Huge increases in data volume. It is projected that the world's information base-the raw material for databases-will be doubling in size every 11 hours by 2010, with obvious implications for data centers. ³ Already, the amount of information transmitted globally over the Internet is doubling annually and will continue to do so for at least the next five years, according to market researcher IDC. Databases are growing because bigger databases return greater benefits to their users. Growth is fastest among very large databases, with the biggest database in Winter Corp.'s most recent survey coming in at three times the size of the largest reported two years earlier. ⁴

Many factors are contributing to this growth, including increases in customer transactions, proliferation of new devices, compliance issues, increased systems performance, online business initiatives, and increased replication to secondary or backup sites. ⁵

2. Renewed interest in mainframes. The big draw of today's big iron is straightforward: Modern mainframes can lower costs and complexity in the data center. The modern mainframe approach has the potential to deliver benefits in the areas of reliability, availability, and serviceability, as well as to provide greater degrees of scalability and self-management. ⁶ Compared with clusters of small servers, today's mainframes can offer cost savings, reduced disruption, ease of use, and lowered complexity.

According to a report by Sine Nomine Associates, "Mainframes, such as the IBM System z, take consolidation with virtualization to its logical limit, concentrating dozens, hundreds, or even thousands of servers into a single physical machine." ⁷ Among the top predictions by Gartner analysts for 2007 is a continued boom in mainframe sales because of their mature processes, new open system workloads, cheaper hardware, and a smaller power and cooling footprint. ⁸

3. Blade servers growing in popularity and driving development of new blade-specific power and cooling infrastructure products. The latest generation of blade servers combines up to a fourfold increase in processor density with power savings of 20%-30% over earlier models, all while delivering the flexibility and versatility today's data center managers demand. ⁹ Gartner Dataquest projects that blade shipments will rise to 2.3 million units by 2011, accounting for almost 22% of all server purchases at that time. However, even with the power savings engineered into new models, blade servers are high-demand hardware when it comes to power and cooling. Suppliers have begun introducing modular and adaptive power and cooling systems specifically designed to meet the needs of blade servers.

4. Application server deployments moving from the edge of business to the center of data center core operations. Whether they use the terminology or not, networks moving in this direction are effectively adopting SOA (service oriented architecture) design principles. Application server-centric architectures are being deployed broadly and rapidly in data centers. Solutions based on that architecture are already in place somewhere in the core of up to 80% of data centers, with more deployments to follow. Historically, the emphasis in SOA has been more about the technology than the solution, but the current trend reflects a focus that is shifting squarely to the application and to the management, monitoring, and enhancement of deployed applications. ¹⁰

5. Cost is a primary pain point in data center management. More than half of data center managers count reducing operational costs as being one of their top priorities. However, they consider improving business continuity, scaling capacity to support business growth, and improving service levels to be even more compelling mandates. ¹¹ As a result, many find themselves struggling to improve data center performance under difficult financial constraints. Among the cost-related challenges they must juggle are space, power, cooling, maintenance and amortization of equipment, and personnel.

6. Increasing demand for metrics to measure data center performance. Good management decisions are predicated on the ability to measure performance accurately, and data center managers are gaining access to more resources in this area. Among them are the Uptime Institute's four-tier system to measure data center critical facility reliability and Syska Hennessy Group's Criticality Levels classification system, which incorporates expected availability and reliability between sites that are designed, constructed, commissioned, maintained, and operated at different priority levels. ¹²

A number of professional groups, including the American Society of Heating, Refrigerating and Air-Conditioning Engineers (ASHRAE), have begun publishing universal standards to assist in data center space planning, load-growth estimation and mechanical and electrical system design. Publications such as ASHRAE's Liquid Cooling Guidelines for Datacom Equipment Centers represent the first step toward providing a benchmark for establishing higher standards in data center design. ¹³

Additionally, the EPA Technical Workshop on Energy Efficient Servers and Datacenters in the United States is working on an extensive set of EnergyStar energy use benchmarks for servers and data centers, estimates of cost savings from improved IT equipment and data center energy efficiency, and the potential impact of energy efficiency on product performance and through the supply chain. ¹⁴

7. Dynamically changing nature of virtual environments is driving the need for configuration management databases, and overall management of the virtual environment needs to be improved through better lower-level systems management. Automation of database configuration management was not a big issue in the era of mainframe domination. Proprietary legacy platforms had robust change-management process disciplines, and scale was easily managed manually. Applications in rack-based server systems typically have a dedicated box that can be swapped out if a problem arises. In a virtualization environment, however, multiple applications run on shared infrastructure, meaning even small problems-most of which are caused by change-can have a widespread impact. Configuration management databases help users align management of their IT infrastructure with their business priorities in a virtual environment.

"Dynamically changing nature of virtual environments is driving the need for configuration management databases, and overall management of the virtual environment needs to be improved through better lower-level systems management."

At the systems management level, IBM offers some of the best hardware management tools in the industry, designed to save users time and money by increasing availability, tracking and deploying assets, optimizing performance and enabling remote maintenance. IBM Director, an industry-leading hardware management solution, is an open, integrated suite of software tools that provide a consistent, single point of management and automation.

8. Increased focus on energy issues, including joint hardware vendor/utility rebates for equipment upgrades. Cost is one factor behind this trend, but the bigger concern is the constraint to operations and growth represented by the power consumption of IT equipment. Energy costs already are the second-largest line item associated with data center operations and are predicted to rise to 50% of the overall IT budget within the next few years. Gartner projects that half of IT managers will not have enough power to run their data centers by 2008. ¹⁵

Utilities and hardware/software vendors are teaming up to help end-users reduce electricity bills, often under the auspices of programs such as the Climate Savers Computing Initiative. ¹⁶ CSCI brings together industry, consumers, government, and conservation organizations in an effort to significantly increase the energy efficiency of computers and servers. Some utilities already are teaming up with leading vendors in offering rebates to offset upgrade costs, and many more utility companies and equipment providers are planning to launch similar programs.

9. Increased focus on security issues for both data and applications. Data center managers face a host of challenges on this front, from protecting data and applications from unauthorized access to safeguarding network interactions to securing physical infrastructure and planning for disaster recovery. Among issues of primary concern cited by IT administrators and managers at mid-sized enterprises in a 2006 IDC survey, security-including improving data center availability and disaster recovery-topped the list. ¹⁷

While there is always a certain degree of subjectivity inherent in lists such as this, there is little doubt these trends will play an important role in the ongoing evolution of the data center. As the Greek philosopher Heraclitus put it some 2,500 years ago, "All is flux, nothing stays still."