Imagine you’re upgrading your 40-year-old house with technology to make its functions more efficient. A smart meter system could monitor and manage the home’s energy usage. An automated doorbell would allow you see remotely who’s there. Sensors can cut off the water supply if a pipe breaks, so it won’t cause a costly flood.
Any of these changes would save you money, protect your property or make you and your family more secure. Which, though, is the best investment? Which would give you the most improvement for the expense? The smart energy system, for example, wouldn’t bring as much bang for the buck if the house has old, poorly insulated windows that waste heating and cooling and need replacement.
Federal agency leaders must ask themselves the same questions about upgrades to the aging and expansive infrastructure of their data centers — the houses, so to speak, for the reams of information they keep. Multifaceted technology known as Data Center Infrastructure Management is helping them find the answers.
What Is the DCOI?
In 2010, a federal initiative to streamline data center operations and reduce energy consumption and physical footprint led to the 2014 Federal Information Technology Acquisition Reform Act. The Data Center Optimization Initiative (DCOI) was established two years later to set the requirements for agencies to report on their status and plans for consolidating infrastructure, increasing efficiency, boosting security and reducing costs. The goal: savings of $2.7 billion.
“Right now, some agencies are spending 75 percent of their IT budget on legacy systems. That’s a lot,” says Shawn McCarthy, research director for IDC Government Insights. “By encouraging agencies to combine data centers, combine applications (when possible) and migrate away from older systems, the initiative urges agencies toward great efficiencies in their data centers.”
Congress recently extended the September 2018 deadline for agencies to meet the DCOI requirements to October 2020. A Government Accountability Office report released in May estimated that the 24 agencies targeted by the DCOI will save $1.62 billion by the end of 2018 — just over half the goal — and found that only four of those agencies would meet all optimization requirements by that original deadline.
However, David Powner, the GAO’s outgoing director of IT management issues, says that the DCOI has enabled agencies to discover data centers in their IT environments — even if they are tiny — and, more importantly, to determine the applications running on those servers. In the past, many agencies did not have that information, meaning they were unaware of the security threats that IT equipment and associated apps might be exposed to.
“If you’ve got equipment running out there and you don’t know what’s on it, there’s security vulnerabilities,” he says, noting that DCOI has helped agencies inventory their IT environments and improve their security posture.
What Is DCIM?
Data Center Infrastructure Management lays the path toward greater efficiency. Agency IT officials cannot figure out how to consolidate and optimize without knowing the equipment and data they have, the space and energy demands of that data, the potential security problems and the potential to apply new technology, says Daniel Kent, chief technology officer of U.S. public sector for Cisco Systems’ federal organization, a DCIM service provider to government agencies.
“We have to help them get from the legacy world to this new world,” he says.
DCIM is not a single type of technology. The concept encompasses an array of tools that IT leaders can use to oversee, scrutinize and understand their systems, from power consumption to data storage.
DCIM, as Gartner notes, lets agencies monitor, measure, manage or control the utilization and energy consumption of all IT-related equipment in a data center, like servers, as well as facility infrastructure components, like power distribution units.
“It has a heavy focus on monitoring what is being used in the data center,” McCarthy says. “The end game is to collect some good, hard data from which to make decisions.”
DCIM ideally takes control of a hodgepodge of data collection across the government. From there, agencies can develop strategies to modernize.
“You’re putting tools in the right place to get an end-to-end view of your data center,” says Gary Hix, director of engineering for Hitachi Vantara Federal, which also provides government DCIM services.
A Navy ship has millions of sensors, but without management, that just creates a flow of unrelated information, Hix says. DCIM gathers the data that comes in, correlates it and finds patterns to determine the most relevant improvements. For the Navy, it might show that an easy task like scraping the barnacles off a ship’s surface would reduce drag and improve speed.
“We need to identify the mission benefits of that data,” Hix says.
How Can Agencies Implement DCIM Technology?
An agency’s DCIM strategy may affect how an agency moves to the cloud. Some IT leaders believe that moving all data to the cloud will solve their data center optimization problems.
But those who specialize in DCIM caution that an all-cloud plan isn’t suitable for every agency or every application. IT experts advocate a hybrid cloud approach, as Kent calls it, identifying applications that can easily move to and work best in the cloud and those that are more complicated and better to keep on proprietary systems.
“We believe cloud is not a destination. It’s an operating model,” agrees Steve Septoff, vice president of infrastructure solutions for Dell EMC Federal, another government service provider. Agencies “need to think about workloads. They need to think about applications,” he says.
DCIM shows agencies the space and energy demands of each application, so IT leaders can figure out which would benefit most from moving to the cloud. That analysis is the sweet spot of DCIM.
Agencies with cumbersome systems also can benefit from flexible consumption, Hix says.
If they buy a whole terabyte for more storage, for example, they’ll have to shift data to the new server over time, so a large portion of that storage space goes unused in the interim. With a flexible model, they’d only buy the amount of storage they need when they need it, helping them avoid large costs up front.
“Too often, we try to boil an ocean in IT,” Hix says. “You become paralyzed, almost, just by the size of your infrastructure.”
How Can DCIM Technology Help Agencies Meet DCOI Requirements?
A primary goal of the DCOI is to address the sheer volume of data centers — servers tucked into closets, stacked in rooms never intended for such operations — what Kent calls “the sprawl of compute.”
The DCOI envisions that consolidation would close at last 60 percent of so-called “non-tiered” data centers. “Tiered” data centers encompass a separate physical space with a consistent power supply, dedicated cooling system and backup power generation for prolonged outages.
Data centers use a huge amount of energy to run servers and reduce the heat they emit. To control power consumption, agencies need a good grasp of the hardware in their centers, says Tim Silk, senior manager of systems engineering for Cisco’s U.S. public sector.
“It’s most helpful at helping data center managers and CIOs get the big picture when they have one or more sprawling data centers with a wide range of resources,” McCarthy says.
Agencies with the oldest equipment have the widest gap to traverse to optimization. The Federal Aviation Administration, for example, needs to upgrade its systems, but that takes time, and air travel can’t grind to a halt during the process, Kent says. Moving some data to the cloud could relieve the burden on that system while upgrades take place over time.
The cloud also allows agencies to share services and information — virtual, rather than physical, consolidation — leading to faster deployment and the ability to scale out quickly, Silk says. “It can provide them better management of their data, which is one of the biggest problems we have in government.”
Under DCOI, agencies in most cases cannot expand data center infrastructure. New information, however, continues to pour in from an increasing number of smart devices intended to improve operations.
“There’s data all around us,” Hix says, “and a lot of the time, we aren’t doing anything with it.”