Jeff Klaus is the GM, Data Center Solutions at Intel Corporation, and Kim Polvsen is VP & GM, Digital Services & Data Center Software at Schneider Electric.
Brooklyn-born songwriter J. Fred Coots was riding the New York City subway in the spring of 1934 and thinking about writing a children’s song when he ran into lyricist Haven Gillespie. What began as a Coots and Gillespie collaboration on the ‘L’ train ended with comedian Eddie Cantor singing “Santa Claus Is Coming to Town” a few months later during his weekly radio show. As the instant megahit’s chorus has it:
You better watch out, you better not cry
Better not pout, I'm telling you why
Santa Claus is coming to town
He's making a list and checking it twice
Going find out who's naughty and nice
And so, data center managers, what kind of data center manager are you? Are you naughty and keeping company with one of the 10 million zombie servers, which, by one estimate, consume energy equivalent to eight large power plants worldwide? Or are you being nice and using DCIM tools to determine what additional capacity you might free up if a zombie server is decommissioned?
Now’s your time to come clean, because we're heading into a new year with new beginnings.
Naughty and Nice Practices Within the Data Center
Naughty: Neglecting to Leverage Real-Time Analytics - In today’s always-on, connected data center environment where unexpected spikes in usage are becoming the new normal, leveraging real-time analytics within the facility is the key to successful operations. Neglecting to do so, puts the data center at a disadvantage and often causes data center managers and their teams to play catch-up. Real-time analytics allow data center operators to work agilely by monitoring data center workloads in real-time and making adjustments as needed, on-demand.
Nice: Introducing New IoT devices Within the Data Center to Create a Holistic View of Infrastructure and Performance - IoT is commonly associated with wearables, refrigerators, connected cars and homes, and all things mobile. However, the introduction of IoT technologies, or those that have an IP address, is also occurring within the data center itself. How? New technology and sensors that help to monitor changes within the data center environment are helping data center managers to obtain a 360-degree, holistic view about how their facility is performing while also giving them insight into where they may need to make adjustments or changes.
Naughty: Not Adopting Automation to Replace Manual Processes - Automation has improved processes for many daily tasks. Water was first brought into homes by fetching buckets of water and advanced into individual automated plumbing systems. Washing cloths required each item to be done by hand, and through technology advancements we now use washing machines to automate the process of washing full loads at ant time. Innovation has a common thread: eliminating a once tedious or manual process with one that does the job more efficiently. So put the Stanley tape measure away and turn to automation.
Automation in the data center allows data center operators to recoup approximately 40 percent of their work week. This is invaluable time that was previously allocated to manual processes such as living in a spreadsheet or physically walking the data center floor with a measuring tape to accomplish capacity planning and forecasting. Still, 45 percent continue to rely on manual processes, wasting time and resources, and leaving them squarely on the naughty list.
Especially Nice: Encouraging Routine or Annual Health Management Assessments of the Data Center - Each year, you pay a visit to your doctor for a routine health check-up. As an ever-changing and developing entity, the data center also requires regular health checks to enable data center managers to stay on the pulse of their facilities, and to maintain business continuity. Preventative measures are critical to avoiding outages and downtime.
Worthy of shiny new toys on Christmas morning, nice data center managers maintain the health of data center hardware by leveraging automated tools that conduct ongoing monitoring, analytics, diagnostics and remediation functions.
Just how important is automating data center health monitoring? According to a study by the Ponemon Institute, the average cost of a single data center outage today is nearly three-quarters of a million dollars. Faced with this scenario, any data center manager expecting a partridge in a pear tree on Christmas morning is grossly misguided.
Decidedly Naughty: Not Using Micro-Level Controls for Individual Servers or Macro-Level Policies for Racks of Servers - Data center managers, we need you to turn your attention to runaway energy consumption, which at 30 billion watts of electricity across the world’s facilities, the equivalent output of 30 nuclear power plants, is enough to power all the households in Italy.
The point is that we can actually do something about this unfortunate circumstance. Rampant energy consumption can be effectively combated with a combination of micro-level controls for individual servers, Power Distribution Units (PDUs), air-flow controllers and cooling units, as well as macro-level controls and policies for racks of servers and entire data centers. These software and technology products can be deployed in less than a week, feature intuitive dashboards and require a short learning curve.
Provided with real-time analytics, data center operators benefit from early detection of thermal spikes and can prepare for unexpected surges in usage by mitigating workloads. They can even reduce a facility’s overall carbon footprint by lowering cooling costs, a major component of data center energy expenditures.
So, do get with the program, naughty data center managers and make resolutions to make next year different.
Opinions expressed in the article above do not necessarily reflect the opinions of Data Center Knowledge and Penton.
Industry Perspectives is a content channel at Data Center Knowledge highlighting thought leadership in the data center arena. See our guidelines and submission process for information on participating. View previously published Industry Perspectives in our Knowledge Library.