DCIM Deployment Drives Efficiencies
EP Editorial Staff | September 21, 2017
Data-center-infrastructure-management software helps Sandia National Labs keep up with the growing scale and complexity of its operations.
For more than 60 years, Sandia National Laboratories (sandia.gov, Albuquerque) has delivered essential science and technology to resolve the nation’s most challenging security issues. As a Federally Funded Research and Development Center (FFRDC), Sandia has the vision “to be the nation’s premier science and engineering laboratory for national security and technology innovation.” Efficient operations, including accurate data-center tracking, play a crucial role in making this vision a reality.
In the past, Sandia was using manual calculations and meter readings, combined with field verifications, to track its three data centers (spanning 6,000, 27,000, and 30,000 sq. ft., respectively). As the scale and complexity of operations grew, data-center managers sought an accurate charge-back model to track customer usage and keep tabs on long-term costs.
Dave Martinez, engineering program and project lead with Sandia’s Infrastructure Computing Services, and others in management had, for several years, wanted to acquire a data-center-infrastructure-management (DCIM) tool to improve asset and power tracking, perform trending and predictive analyses, and, ultimately, reduce costs and manual efforts. Sandia’s customers needed more transparency with regard to their particular assets and power usage, and management needed a better cost model for those customers.
With several different tracking systems in place, difficulties associated with fragmented data and system-specific skill sets weren’t unusual. According to Martinez, a single, unified system that the entire staff could use—a system that would also enable knowledge transfer in employee transitions—seemed to be the right direction to take.
To ensure successful adoption, Martinez engaged his staff to obtain buy-in for a DCIM tool. His goal was to empower the team to be self-sufficient by giving them ownership of the solution from the onset. The group came up with the following criteria for the tool:
• quick and easy to deploy
• easy for appropriate personnel to learn and navigate
• capable of showing results in short order to achieve a fast return on investment.
After product demonstrations by eight potential vendors, Martinez and his team narrowed the choices to four. Each finalist was then given the opportunity to meet the key criteria in Sandia’s live environment. After research into all products, Nlyte Software (nlyte.com, San Mateo, CA) was selected.
As the Nlyte tool was deployed, Martinez made sure his team was involved every step of the way. The Nlyte team, in turn, trained Martinez’s team. This vital step not only made the Sandia personnel comfortable with the solution, it helped them gain the knowledge, skills, and independence that led to future success. As hundreds of people began using the tool to varying degrees, a collective knowledge base was formed and became easily transferable.
Gathering the information on the assets and sensors in the data centers took time, but was time well spent as it gave Sandia personnel an accurate picture of everything on the floor. As Martinez noted, “How would we troubleshoot if we didn’t even know where the assets were?” Members of his team soon had visibility to each asset, down to its exact location within a room. They also gained branch circuit monitoring, right down to the rack. This transparency enabled the team to see trending, assets, power usage, and empowered predictive analysis. With these capabilities, the team could finally answer previously unanswerable questions: What would happen if power were lost? How would a failure affect the data center? What would the addition of new equipment do to the surrounding area?
By automating the process, the new DCIM significantly cut manual labor and associated costs. The built-in dashboards and reports in the Nlyte solution eliminated hand calculations and field verification. “Before,” Martinez explained, “we had electric meters on power mains, but we had to go and read them. Now, we just turn on the (Nlyte) screen. We can trust what we see there. We save many, many man hours.”
This kind of information leads to other savings. “We’re not predicting any more,” Martinez said. “We have live studies. We’re turning metrics into dollars and cents. This puts our operational costs more in perspective. The more metrics we have on power and cooling usage, the more we can trim.”
Sandia’s internal customers now have access to their information, and can see their own trending when adding equipment. When adding new customers, Sandia’s management now has an up-to-date cost model for accurate chargebacks.
Sandia began using these Nlyte Enterprise and Nlyte Energy Optimizer features and has plans to use more in the future:
• Data Center Module for viewing and managing all asset information
• Connection Manager for tracing power and data connection with the data center
• Floor Planner for planning and managing physical layout of floors, rooms, and assets within data centers
• Dashboard and Reporting for accessing pre-defined dashboards, as well as the flexibility to create, deploy and maintain user-defined dashboards
• Intelligent Asset Allocation for creating and managing auto-allocation projects with Nlyte’s patented Intelligent Asset Placement.
For the next phase of adoption, Sandia foresees using Nlyte Predict, working to include Nlyte Barcode, and adding Nlyte Integrator at a later stage.
“We’re very pleased about the way everything has developed,” Martinez noted, “including deployment and use of the Nlyte system. It’s beyond what I thought it would be. It has definitely exceeded our expectations.”
Even more important, given its DCIM-readiness using the Nlyte Dashboard and Reporting, Sandia is primed to tackle the upcoming challenges of the DCOI (Data Center Optimization Initiative), a Federal mandate that other U.S. government agencies have been able to address using Nlyte’s DCOI Dashboard. EP
For more information, visit nlyte.com.