Measuring the work your datacenter does

Measuring the work your datacenter does

Summary: New technologies have given us the ability to create a more energy-efficient datacenter that still gets the work done.

TOPICS: Data Centers

While there have been many advances in technology and process that have resulted in the ability to create a datacenter that was more energy efficient, the bottom line has always been that the datacenter still needs to get work done. And a true metric of efficiency needs to be aware of the workload and provide some evaluative method that defines the efficiency of the datacenter in the simple model of comparing what goes in to what comes out.

After five years of work, a global taskforce has released their final set of recommendations. This task force, consisting of representatives from: The Green Grid Association; U.S. Department of Energy’s Save Energy Now and Federal Energy Management Programs (March 2009 – October 2012); U.S. Environmental Protection Agency’s ENERGY STAR Program; European Commission - Joint Research Centre, Data Centres Code of Conduct; Japan’s Ministry of Economy, Trade and Industry; and, Japan’s Green IT Promotion Council focus was on determining directions to improve the key energy efficiency metrics used within datacenters. Their final result is the Data Center Energy Productivity (DCeP) metric which is designed to provide a quantifiable measure of the amount of work a given datacenter is accomplishing.

The DCeP builds on previous work by the task force and the green Grid, and in fact, requires the understanding and usage of earlier established metrics. At this time that means operators will need to have the ability to make use of a series of existing metrics:

  • Power Usage Effectiveness (PUE)
  • Carbon Usage Effectiveness (CUE)
  • Green Energy Coefficient (GEC)
  • Energy Reuse Factor (ERF)

Details on all of these metrics can be found on the Green Grid website.

Combining this information with a detailed knowledge of the workload being sent to the datacenter allows the operator to make use of the DCeP formula:



More details on DCeP and related infrastructure metrics can be found here.

Topic: Data Centers

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.


Log in or register to join the discussion
  • Wow, thanks for the Friday headache ...

    to be more energy-efficient (without all the formulas):
    1. Raise the temperature. Most datacenters are running much colder than they need to based on temps that older hardware needed to operate at. Most new hardware can run fine with temps at 75 degrees Fahrenheit or so (instead of 65 degrees many datacenters run at).
    2. Virtualize anything that you can. Instead of having 20 physical servers running at 10% CPU load, have fewer VM host servers running those same number of virtual servers (you can probably get 10 VMs or more on 1 host unless you are talking serious number crunching). You can also save money on OS purchases by purchasing Windows Server versions that allow one license to cover all Windows VMs on a single VM host.
    3. Make sure remaining CPU rack systems are properly ventilated and have empty slots plugged up so that air flows correctly through rack systems. This allows higher temps mentioned in item 1 and may even allow you to consolidate space out for some unused racks downsized by moving to VMs.
    There are probably a few more ideas, but those above are the low hanging fruit and you don't need a slide rule to calculate them.
  • Virtualize and utilize remote technologies

    Virtualize your data center as much as possible to reduce server costs and free up valuable real estate space. Also, utilize remote technology solutions (like vLeap) to reduce traffic and increase employee productivity.