Increasing efficiency is a perennial mandate for IT professionals everywhere, but what that means to the organization has undergone a subtle change. Over the past few years there has been an increase in focus on the energy efficiencies of the IT infrastructure, and more specifically the Data Center. This shift is driven both by the economics of energy efficiency as well as a heightened awareness of the environmental impact that inefficient energy consumption can have on the business and the community at large.

 

As the IT organization strives to eke efficiencies out of their IT infrastructure, however, they are also being asked to support more with that infrastructure than ever before. New services and applications are exploding. Applications, such as video on demand (VOD), Web conferencing, Voice over IP (VoIP), and applications we haven’t even thought of yet, offer the promise of enhancing our lives and improving our communication.

 

Yet, to deliver a satisfying experience to users and achieve the potential productivity gains companies are expecting requires an underlying high-performance infrastructure. It requires an infrastructure with the performance, reliability and security levels capable of supporting the demands of today’s applications and the dynamic, rich-media applications of the future.

 

IT Managers know that the importance of this flexible, scalable infrastructure will only increase, as our dependence grows on these technologies, which make both our work and personal lives a little easier and more enjoyable and keep us connected to the people and information we need. But IT organizations are struggling with how to increase the performance, capacity, flexibility, etc. of their IT infrastructure, while mindful of energy consumption constraints.

 

IT departments are not only facing rising energy costs, they are also struggling with meeting the basic power needs of their growing IT infrastructure, due to already strained power grids and limits inherent in the physical capabilities of their facilities to support the growing power demands. The reality is that IT departments cannot simply add devices in order to add capacity or functionality. IT organizations are finding they are unable to draw the power or afford the costly facility upgrades that are required to support an unending proliferation of devices. Instead, they are looking at replacing this old practice with new ways to get additional efficiencies and functionality from every device they have on the network to support the exponential growth of application and service demands on the infrastructure.

 

As IT Managers look at ways to quickly meet the business requirements of the organization, within the bounds of responsible and cost-effective energy consumption, the first project on the list is generally based in the Data Center. When the Environmental Protection Agency recently pointed out that Data Center consumption is roughly 1.5% of total U.S. electricity consumption, it probably didn’t come as a surprise to many in IT. It also isn’t a shock that the Data Center represents a great opportunity for IT to reduce energy consumption and operational costs to realize overall efficiency gains. The EPA identified “the potential to save up to $4 billion in annual electricity costs through more energy efficient equipment and operations, and the broad implementation of best management practices,” in a press release, EPA Reports Significant Energy Efficiency Opportunities for U.S. Servers and Data Centers, dated 8/3/07.

 

Many IT organizations start tackling the Data Center by ensuring the general layout is as efficient as possible, following a hot-aisle, cold-aisle layout, installing efficient power and air conditioning systems, ensuring tighter controls over the administration of the facility, including the ability to monitor and adjust energy allocations and temperature, etc. We, at Juniper, were able to realize immediate energy efficiencies by improving the layout of our data centers — in our East coast headquarters, in Westford, Mass., we reduced the facility’s carbon footprint in 2006 by 23% (from 7,714.81 metric tons of carbon emissions in 2005 to 5928.93 in 2006).

 

The next phase, which can be done in parallel, involves the IT department taking a good hard look at the leverage they are getting from each and every device on their network. Many companies are challenging their vendors to deliver more energy efficient solutions. Logically, IT departments followed the “power” trail, which led them to first focus on computing and server technologies that typically make up the majority of a company’s energy consumption.

 

Server and computing vendors have been answering the call, delivering increasingly efficient solutions that help companies more effectively manage their energy consumption.

 

It is also no coincidence that technologies, such as virtualization that maximize device utilization, have taken off. These technologies enable a single physical server to house multiple, independent operating systems and applications that are logically separated. IT organizations are leveraging virtualization in their Data Center to achieve better data consumption and overall server utilization, ultimately enabling server consolidation that results in power, cooling, space and operational management efficiencies. We implemented and recently completed a virtualization project for our Data Centers that enabled us to retire 325 servers. This represented a reduction of 146,250 Watts consumed per hour, 3,510,000 Watts per day, 1,277,640,000 Watts per year — the equivalent to what 1,252 U.S. residents would consume in a month. We expect the return on the investment (capital and software licenses) will be between 12–14 months.

 

But once an IT Manager has started the journey of deploying energy efficient servers and maximizing their utility with virtualization software, what’s next? The next wave of energy efficiency will be found in the other infrastructure that makes up the Data Center—within the storage, routers, switches, security, etc. industries.

 

We are already seeing movement in this direction, with new core routers announced last year that are 30% more efficient, in terms of power and cooling, than competitive products on the market. This is the start of a trend that is only going to grow, due, in part, to the maturation and availability of more energy efficient technologies, as well as a strong commitment by leading vendors to identify new and innovative ways to create energy efficiencies within their solutions — from power supplies to custom silicon development. The key is for vendors to find ways to deliver the functionality customers need to support the onslaught of services and applications that rely on the network infrastructure, within a more efficient form factor that helps them reduce energy consumption — without sacrificing one for the other.

 

In addition, we are seeing a consolidation of functionality into high-performing, predictable platforms that can be deployed in the Data Center. IT Managers may have been hesitant in the past to deploy multi-function devices in parts of their network, such as the Data Center, that necessitate a high-level of reliability and performance. The potential risk was that as functionality was turned on, the overall performance and reliability of the device would degrade.

 

For example, some security solutions, due to the way they are architected — with multiple technologies on a single device, but with very little true integration — or the way they handle traffic under load — simply passing it on rather than inspecting it — causes IT managers to adopt a policy of single device, single function. Now, however, there are tested, multi-function solutions, such as Unified Threat Management devices, that can achieve the performance, reliability and security needed to operate in the Data Center. The benefit of being able to deploy a single unit instead of multiple devices enables IT organizations to realize energy, cooling and space gains as well as general operational efficiencies.

 

Another aspect of a high-performance network infrastructure that IT organizations should evaluate as they look to drive efficiencies into their operations is the timing of implementing change. It is important to protect the investments that have been made in the IT infrastructure and look for ways to make the most of each resource to achieve efficient ongoing operations. Ideally, the solution can be redeployed or the hardware upgraded to minimize waste. At Juniper, there is a strong commitment to the reuse of hardware; for example, with interchangeable line cards that can be deployed across product families, or by offering an upgrade path whenever possible to maximize the life and utility of a platform. These are all parts of the equation that should play a role in achieving the efficiency mandate.

 

There is a real shift taking place, driven by an increasing awareness of the benefits, both cost and environmental, that an energy efficient IT strategy can have on the business. As IT organizations look at how to create a trusted and responsive environment that will support the onslaught of new experience-rich services and applications that users demand, they will find that managing both their productivity and energy efficiencies will be critical to their success.

 

-----