Virtualization on the plant floor
Preparing your power distribution system
Virtualization places different demands on your infrastructure than traditional architecture.
By increasing the utilization level of servers, virtualization brings potential to deliver incredible savings in terms of server count, footprint, power consumption, and cooling requirements. However, in order to fully reap these benefits without sacrificing electrical reliability, a few important power distribution challenges must be addressed.
For one, overall power consumption will be lower, but it will be of higher variability and concentration. For example, on an un-virtualized platform, the average server CPU runs at only 10%-15% of capacity. With virtualization, that figure increases to about 70%-80%. As CPU utilization increases, so does power consumption per server.
System availability becomes all the more important as servers are pressed to carry these larger workloads. To protect servers, increase the density of enclosure-level power protection and distribution. Enclosure-based power modules are available that can distribute up to 36 kW in only a few U (rack units) of rack space. These cover four to 45 receptacles in an organized manner to meet the needs of a wide range of power densities.
Additionally, modern enclosure-based PDUs (power distribution units) are available that can view the status of each circuit securely from anywhere on the company intranet or the Internet, in order to provide automated alerts of potential trouble.
Increasing power demands
Further up the power system, you’ll want to ensure existing circuits are capable of supporting high-density computing systems. Traditionally, facility managers could plan for about 60 to 100 W of power consumption per U of rack space, so a full rack of equipment averaged 3 to 4 kW. Today’s blade servers have escalated that figure to 600 to 1,000 W per U, which is steadily growing and may soon reach up to 40 kW per rack.
Five or 10 years ago, a typical computer room was designed to feed one 20 A, 208 V circuit to each rack, or less than 3.5 kW per rack. If you now have to support 20 kW of equipment in each rack, it could take up to six of these 20 A circuits. The existing electrical infrastructure will be unable to support this load growth, and could easily run out of circuits or run out of capacity, especially with the growing prevalence of dual- and triple-corded loads.
You can solve this problem by building out a power subdistribution strategy. Instead of running individual cable drops from your large UPS and PDUs to each rack, run higher powered subfeed circuits to an intermediate remote power panel (RPP), power distribution rack (PDR), or rack-mounted power distribution device, and from there to enclosures.
Visibility into demand
Although the power system adjustments suggested will help ensure servers receive adequate power and circuit overload is prevented, it remains highly important to increase your level of visibility into how the movement of applications affects power and vice versa. After all, application demands affect power consumption, and power consumption affects application availability and performance—yet rarely do these two interdependent groups work in unison.
This issue can be addressed with 24x7 power quality metering, monitoring, and management at the branch circuit level, which can be conducted at several points in the data center power distribution system. With this type of universal perspective, you can easily see when irregular conditions threaten processes, and react accordingly to prevent system damage or downtime.
Virtualization enables a facility to meet business objectives with fewer physical servers, which results in dramatic energy and physical footprint savings. However, successful implementation of any virtualization project requires that you prepare for dynamic changes in power demands, higher power densities, and the critical need to have oversight of application movement. Fortunately, proven solutions are available that address these power considerations and allow facilities to extract the full potential of virtualization while ensuring that power demands and business needs are always met.
Jim Tessier is a virtualization product manager for Eaton.
- Virtualization represents a huge advance in industrial network design and implementation.
- This technology is well proven in IT applications and is moving into manufacturing.
- A thorough understanding of the technology and appropriate planning are key to effective implementation.
- Read this story online for additional discussion of each topic.
- Get more information from the companies involved:
|Search the online Automation Integrator Guide|
Case Study Database
Get more exposure for your case study by uploading it to the Control Engineering case study database, where end-users can identify relevant solutions and explore what the experts are doing to effectively implement a variety of technology and productivity related projects.
These case studies provide examples of how knowledgeable solution providers have used technology, processes and people to create effective and successful implementations in real-world situations. Case studies can be completed by filling out a simple online form where you can outline the project title, abstract, and full story in 1500 words or less; upload photos, videos and a logo.
Click here to visit the Case Study Database and upload your case study.