TrendPoint offers four-prong attack to greening your datacenter

Devising an energy budget and measuring datacenter efficiency among key strategies

Datacenter operators struggling to cut energy costs and to reduce their carbon emissions in anticipation of environmental regulations need an action plan, a point made this week by TrendPoint Systems. A provider of energy management solutions, the company unveiled a four-prong plan of attack to tackle the aforementioned issues. Although the plan doubles as a pitch for the company's product line, there's certainly sound advice to be found within.

1. Set an energy budget. "In the same way that companies set and manage travel and other line-item budgets, datacenters need to be able to set energy and carbon budgets that can be broken down among users, departments and sites," suggests TrendPoint. "Companies can then set budgets and give each user access to manage their individual energy and carbon usage against assigned metrics. Colocation facilities, in particular, need to be able to provide each customer with the ability to manage their own energy and carbon usage and to provide a system to bill back customers appropriately."

One company embracing this strategy is Microsoft. The company has devised an internal datacenter monitoring program called Scry. The system gathers all sorts of data on energy usage, temperature, carbon emissions, and more from all of Microsoft's datacenters. It also ties in to the company's asset management, ticketing, and CMDB (configuration management database) systems. The company uses the system to charge business units for the specific datacenter resources they use.

According to Microsoft, the approach has compelled employees to find ways to reduce energy consumption through techniques such as writing trimmer code.

[For more about Microsoft's Scry system, please read "Microsoft developers trim code to cut costs internally."]

2. Virtualize servers. "Companies can reap instant savings by consolidating underutilized data servers onto virtual machines on a physical server," according to TrendPoint. Indeed, InfoWorld has seen example after example of organizations reaping significant savings in energy consumption as well as floor space thanks to virtualization.

Virtualize with care, however, TrendPoint cautions: The company reports that it "has seen that virtualized servers generate significantly more heat vis-a-vis the under-utilized machines and, therefore, need careful attention with their cooling management. ... Just as smog recycles and builds up in an atmospheric thermal inversion, a hot server can create a thermal inversion that causes cool air to be trapped and recycled, wasting cooling resources. 'Server thermal inversions' can reduce potential energy savings and lead to costly heat-related downtime."

[To learn more about other virtualization pitfalls, please read "Virtualization's dirty little secrets."]

3. Equalize heat and cooling balance. "Datacenters waste enormous amounts of energy by over-cooling the majority of their data cabinets. This is because CRAC units and control systems make macro-cooling decisions," TrendPoint explains. "By matching cooling resources (from floor vents or liquid cooling units) to the actual needs of each individual cabinet, datacenters can realize significant savings on cooling energy use of 25 percent or more."

[To learn more about granular approaches to cooling, please read The cool new look in datacenter design as well as "Chillin' at the HP datacenter."]

According to the company, "further savings can be achieved by balancing heat loads on an intra-cabinet basis. By grouping servers in a minimum of two and preferably three zones within a cabinet and moving towards equalized heat loads between the zones, 'server thermal inversions' are avoided and users can reap additional cooling savings of 10 to 30 percent."

4. Manage to the metrics . "As datacenters add, move, and change servers and equipment, managers need to continually monitor and manage IT equipment heat generation and cooling," TrendPoint recommends.

That's where metrics such as The Green Grid's Power Usage Effectiveness (PUE) come into play: Tracking datacenter efficiency on a continual basis can help you assess whether changes are having a positive or detrimental impact on your overall energy consumption. (PUE compares the total amount of power a facility uses with how much of that power actually gets applied to computing instead of cooling, power conversion, and other support functions.)

Figuring out just how to measure PUE on an ongoing basis is the tricky part, but vendors such as TrendPoint, SynapSense, EDSA, and others offer tools to perform the task.

Ted Samson is a senior analyst at InfoWorld where he writes a blog called Sustainable IT, tracking green technology trends. Subscribe to his free weekly Green Tech newsletter.


Copyright © 2008 IDG Communications, Inc.

How to choose a low-code development platform