Data centers

Time to Unplug Comatose Servers (If You Can Find Them)


As corporate and cloud data center workloads continue to grow, demand for associated servers remains on the rise. Indeed, worldwide server shipments grew 13 percent in Q1 2015, according to Gartner. Now for the irony: Instead of selling your customers more servers -- or buying servers for your own data center -- take a close look at existing footprints.

"Wasteful, or comatose, servers hide in plain sight in even the most sophisticated IT organizations," according to Uptime Institute. "These servers, abandoned by application owners and users but still racked and running, represent a triple threat in terms of energy waste—squandering power at the plug, wasting data center facility capacity, and incurring software licensing and hardware maintenance costs."

Deadly Hardware

More specifically, Uptime Institute estimates that 15 percent to 20 percent of servers in data centers are obsolete, outdated, or unused. The problem: Many data center owners and co-location residents don't perform annual audits to determine which servers are:

  • Over-utilized and in need of more capacity;
  • Under-utilized and ready to take on more workloads; or
  • Antiquated and unable to fulfill modern compute requirements.

Of course, IT monitoring and management tools -- coupled with analytics and reporting software -- can rapidly and easily reveal these shortcomings to IT leaders. But sometimes, the day-to-day grind -- "keeping the lights on" -- takes priority over long-term data center planning.

Over the past decade, many businesses pursued data center consolidation and server consolidation projects -- virtualizing more applications onto fewer servers, and condensing hardware footprints in the process. But the server sprawl problem often transformed into a virtualization sprawl issue -- with many virtual machines running unchecked and unmonitored.

If physical and virtual server workloads aren't carefully managed, data center power costs can skyrocket. That's bad news for customers -- but a potential opportunity for cloud providers and/or channel partners that design and manage data centers.

Establish Best Practices

Indeed, Uptime Institute sees nine areas where partners or corporate IT can get data center sprawl and power consumption under control. They include:

  1. A formal reporting relationship between IT and data center facilities management with a chargeback model that takes into account procurement and operations/maintenance costs
  2. Key performance indicators (KPIs) and mandated reporting for power, water, and carbon utilization
  3. A culture of continuous improvement with incentives and recognition for staff efforts
  4. Cost modeling of efficiency improvements for presentation to senior management
  5. Optimization of resource efficiency through ongoing management and operations
  6. Computer room management: rigorous airflow management and no bypass airflow
  7. Testing, documenting, and improving IT hardware utilization
  8. IT asset management: consolidating, decommissioning, and recycling obsolete hardware
  9. Managing software and hardware life cycles from procurement to disposal

Translation: Before you run out and buy more servers, get existing data center footprints under control.

Joe Panettieri

Joe Panettieri is co-founder & editorial director of MSSP Alert and ChannelE2E, the two leading news & analysis sites for managed service providers in the cybersecurity market.