Data Center Optimization

By Mark Gaydos
From the April 2017 Issue

The federal Data Center Optimization Initiative (DCOI) is a mandate from the Office of Management and Budget (OMB) establishing levels of efficiency that federal data centers must meet by the end of FY2018. The unprecedented growth of federal data centers—from 1,000 facilities in 2010 to almost 10,000 in 2014—prompted this mandate by the federal CIO. Under the DCOI mandate, federal agencies are required to:

• develop and report on data center strategies to consolidate inefficient infrastructure;
• optimize existing facilities;
• achieve cost savings and transition to more efficient infrastructures, such as cloud services; and
• install and use Data Center Infrastructure Management (DCIM) software, to replace older, manual, and inefficient ways of monitoring and tracking energy and space usage, assets, and workflow.

One of the most important metrics that will be used to gauge the success of these efforts is Power Usage Effectiveness (PUE). PUE is calculated by dividing Total Data Center Energy Used by Total IT Equipment Energy Used. The DCOI mandate stipulates that existing data centers achieve and maintain a PUE of less than 1.5 while new facilities attain a score of no greater than 1.4, with 1.2 encouraged.

The goals of the DCOI are ambitious, but attainable. The plan calls for saving $1.36 billion between 2016 and 2018, and part of the initiative also involves closing 52% of federal data center inventory. But the savings won’t stop there. Federal data centers will continue to reap the benefits of higher efficiency after the mandate is complete.

Applying DCOI To Any Data Center

Although on differing scales, just about any data center facility can benefit from applying some of the same strategies and metrics that are slated to bring such great savings to the federal government.

Consolidating inefficient infrastructure. First, data center managers must obtain a real-time, accurate inventory of what is in their facility and what work it is performing in order to have a baseline against which to measure the required improvements. This inventory will likely reveal underutilized capacity, whether related to physical space or energy. With this knowledge, a data center operator can improve utilization of servers, floor space, and power.

During consolidation efforts, it is vital to have real-time information about temperature and capacity to prevent overload in any part of the facility. For safe consolidation, it is also prudent to have the ability to run “what if” modeling before making any changes. Operators can test a change that is being considered in a staging environment and determine the effects of the proposed change.

But consolidating assets is just part of optimization. Consolidating whole facilities is the next step. This is achieved by not only using available capacity most efficiently, but also through an accurate inventory of details about assets—not just their locations, but also age and service history.

This capacity tool allows operators to eliminate costly maintenance and service of old equipment and simply retire/replace aging devices. This may require a total tech refresh, which can be done using a DCIM solution—as a guide for tracking equipment from the loading dock to decommissioning (Dock to Decom).

Optimizing existing data centers. Once a data center manager knows where the assets are and how these are operating, making the necessary improvements can boost efficiency and save on energy and cooling. Once it is known where energy dollars are being spent, adjustments can help curb this spending. This may be as simple as rearranging some assets on the data center floor, or adjusting the speed of a fan. Real-time information obtained using a DCIM solution simplifies capacity planning, and enables data center operators to pinpoint where more equipment can be placed safely.

And, with security issues becoming more critical, data center operators must ensure that the DCIM solution they select is fully certified secure by a neutral third party. Failing to do this can expose an organization to unforeseen breaches.

Achieving cost savings. In addition to monitoring the data center in real time, DCIM solutions are now offering next-gen capabilities, which offer additional savings, such as the ability for remote control of assets. This virtual command capability helps save the operator time and travel expense that would be incurred if the changes needed to be made in person and on-site.

As important as real-time information is, it’s also vital to be able to look at historical information about the facility. Evaluating current and historical usage and energy consumption patterns offers insight to inform future planning, and enable more savings.

Just like the federal agencies participating in DCOI, civilian data center managers may also be asked to track or prove cost savings by their C-suite. There are probably several departments or levels involved in any efficiency project, so choosing a DCIM solution with purpose-built modules to track and validate progress to gain specified savings goals is a good start.

Transitioning to more efficient infrastructures. Before moving to the cloud or other shared infrastructures, a data center manager must first gain a clear understanding of the computing power costs in a facility. Only then will they be able to evaluate the cost of on-premise computing, and compare it to outsourcing to the cloud.

The DCIM solution selected should also have a comprehensive portfolio of pre-built connectors to leading virtualization providers to help ease the transition. Pre-built connectors enable facility operators to synchronize information with other systems in a bi-directional fashion. This means operators can pull information out of virtualization systems, for example, and discover out what virtuals are running on which physical servers. In this way, it is known what applications are dependent on specific servers for migration or risk mitigation purposes.

DCIM And DCOI

The Federal DCOI mandate is an ambitious plan designed to gain a handle on data center spending. The methods and tools those facility operators use during implementation are worth learning about and perhaps emulating in any data center environment. DCIM will be an essential part of enabling success now and in the future.

Gaydos is chief marketing officer for Nlyte Software, a data center infrastructure management (DCIM) solution provider for automating data center operations and infrastructure into an enterprise’s IT ecosystem. The company is headquartered in San Mateo, CA.

Do you have a comment? Share your thoughts in the Comments section below or send an e-mail to the Editor at acosgrove@groupc.com