By Richard Muirhead, CEO and Founder, Tideway Systems
Going green is a red hot issue these days, particularly in the data center which is a big consumer of energy. Rising energy costs, combined with the increasing number of servers, cooling equipment and related infrastructure being deployed, are creating new pressures on corporate bottom lines in financial services organizations and in many other industry sectors.According to a recent study by Stanford University's Jonathan Koomey, the energy consumed by data center equipment and infrastructure more than doubled in the U.S. and worldwide between 2000 and 2005. In the U.S., the electricity bill totalled $2.7 billion in 2005 and rose to $7.2 billion worldwide in the same year. Put another way, our own estimates indicate that a single, moderate-sized server in the data center of a large financial institution has about the same carbon footprint as a gas-guzzling family SUV that gets 15 miles to the gallon.
There are many issues to consider other than power consumption, however. These include the broader environmental concerns of limiting carbon and greenhouse gas emissions, using materials from renewable resources, recycling materials and reusing heat from data centers.
A Wider View of Your IT Assets
To take the first steps in creating a greener data center, financial institutions need to get a better handle on their IT assets. Some two and a half percent of IT assets continue consuming electricity even though the business thinks the equipment has been disposed of. This lack of overall visibility can lead to inefficient use of current assets and wasted investments in new equipment.
Financial services companies are dramatically reducing their IT carbon footprint, energy consumption and overhead through data center optimization. This is typically a manually intensive and risky process, especially since technologies such as virtualisation make it difficult to understand the impact of removing real-world pieces of hardware. New technologies, however, such as application dependency mapping, now make it possible to automate more of the process, and critically show the definitive business and technical impact of removing any one piece of hardware. The following is a top-level view of the typical phases of a data center optimization strategy:
Diligence: A complete IT inventory of servers, storage and applications, including identifying inter-system and application dependencies and relationships, is central to any sustained data center optimization effort. When done manually, this intensive and highly-detailed activity can take a great deal of time to complete to attain the required granularity and data quality. Data gathered is employed in subsequent planning, costing and risk-mitigation activities, and used as the basis of a configuration database that will help manage change during the migration itself.
Understand: Once a sustainable method of collecting and maintaining accurate inventory and dependency data is established, an organization can look beyond the data to reach a deeper understanding of how technology underpins its business operations. In this phase, the focus is on addressing questions that correlate technology and application assets with their business owners and stakeholders, and on mapping the interdependent relationships.
Design and plan: Armed with a full description of the IT and business asset landscape, the project team can concentrate on finding optimization opportunities, designing the desired end-goals, and planning the process for getting there.
Implement: In the implementation phase you need to obtain a precise picture of the state of the IT environment from day to day. This allows project progress to be monitored, provides a continuously updated view of what changed incorrectly in the environment when incidents occur, and assists with necessary re-planning. If the project has been planned effectively, then the implementation itself will be a controlled, predictable exercise.
As you delve into your data center optimization project, it is important to consider ways to create a continuous, repeatable and adaptable program. The newest application dependency mapping tools automate the process of understanding IT assets and their interdependencies, including the ramifications of planned changes, or services impacted by component problems. According to Gartner, a service dependency map is a "prerequisite for achieving success" with a host of IT-related issues, including service-level agreements and availability, systems configuration and changes, and performance management."
Greener Data Center Benefits the Bottom Line
As the appetite for information grows, so does the infrastructure required to support it. And while technology vendors are developing more energy-efficient products, businesses can do their part in reducing the impact the data center has on the environment.
Organizations are implementing a number of environmentally friendly data center strategies including reducing the number of servers through server virtualization, switching off unused hardware, deploying more efficient cooling and power supply products and replacing slow, energy-hogging servers. A few are even taking advantage of new financial incentives from electric power companies by reducing energy usage during peak months. Green data centers that operate with greater energy efficiency make sound financial sense, and ultimately improve the environment we live in.
Richard Muirhead is CEO and founder of Tideway Systems. Tideway helps companies gain transparency of their IT estates and realize the significant "green" aspects of data center optimization.