Worrying about energy costs is intrinsic for most data center owners, operators and designers. And to borrow a bit from Bob Dylan, you don’t need a weatherman to know which way the wind blows.
Energy costs are spiraling, according to recent article in Mission Critical UK. To combat this, analyst Paul Johnson, data center segment lead for ABB, spoke about the role of automation at Data Center World in London. According to Johnson, tackling costs while increasing energy resilience can be achieved by focusing on three elements: Component visibility, intelligent grid connections and elastic critical infrastructure.
Deep Component Visibility
A major premise of Johnson’s talk was the need to make devices smarter and more resilient. Big outages many times start small and have cascading effects. With increased visibility, insights emerge before an event snowballs to an actual outage, thus reducing downtime and business disruption.
Predictive analytics are needed to achieve deep component visibility, and this is where Software Defined Power really shines. SDP’s ability to see into the entire power structure—not just a couple of smart components—delivers better predictive abilities and increased fault avoidance.
“Software Defined Power has the promise to allocate power to the centers with the lowest cost at any given time,” says Carrie Goetz, a highly regarded adviser to data centers worldwide. “This provides significant savings, and also shifts loads around failures, bringing other sites and resources online when and as the need arises.”
Applications Matter Most
Smart devices are important, but at the end of the day, applications matter most. Uptime Institute has spent a lot of time lately focusing on how companies can better manage data center infrastructure to avoid power outages. While this is well intentioned, what’s most important is uptime of mission-critical business applications.
An app doesn’t care if it’s running on an on-premises server or one in the cloud—as long as it’s running. SDP applies this same approach, as the software layer abstracts power distribution, so it can be reallocated across the power hardware infrastructure. Instead of managing independent power distribution platforms, SDP can be used to identify, aggregate and allocate power resources to data center racks, nodes, circuits or application workloads, on demand.
Equally important is the ability to see across different power devices. That’s why VPS embraces a hardware-agnostic approach that seamlessly integrates SDP capabilities into various vendors’ power equipment using Application Programming Interfaces (APIs).
Intelligent Grid Connections
SDP also is ideal at clearing operational obstacles for data center operators managing sites across multiple locations with varying energy sources. Think many locations, a multitude of devices, perhaps even a city with multiple data centers consuming and generating power, especially if renewable energy sources are involved.
As microgrids or multi-grids come into play, so does the need to orchestrate power distribution more easily and effectively, ensuring all devices and energy sources are in sync. SDP leverages analytics to determine the right time to conserve, reallocate or consume energy, similarly to how financial analytics help manage the ebb and flow of the stock market.
“By making the physical plant more aware of what the software is doing, the savings and redundancy that can be built offer major benefits,” adds Goetz. “This also means that data centers can drop power to applications not being used, monitor and control stranded power and just make the IT load a smarter steward of the overall power budget. There is an exciting future here from microgrids to full grids with cloud-enabled applications. Tomorrow’s relationship between facilities and IT will look very different!”
Elastic Critical Infrastructure
After all that has been written about adding redundancy, reliability and resources, why are there still so many outages? Today’s data centers should be built the same way as software: Take into account failure—instead of trying to prevent it.
While this requires a major shift in thinking, it harkens back to the concept that application availability is king. SDP can migrate and manage power across multiple locations to ensure the king never becomes powerless.