Data Storage: Maximizing Data Center Power Efficiency: 10 Ways to Do It
Use Smart Grid for Data Collection Across the Environment
Sensors collect power and cooling data from systems and facilities and establish consistent energy benchmarks. This allows IT managers to aggregate information to provide meaningful data on power and cooling efficiency in real time, which can help dynamically manage data center energy usage.
Use Consulting Services to Upgrade
As businesses enter new markets and seek new ways to connect with customers, the demand for new IT services escalates. Experts combine design engineering and architecture with IT expertise while they design and build data centers to meet changing business needs with lower facilities capital and operating costs. To keep up, IT implements flexible infrastructure like blade systems, storage-area networks and virtualization, all matched carefully to business needs.
Use Centralized Management
New-generation data centers are split between environmental, facilities and system/IT management. A comprehensive management system for the complete data center environment allows organizations to take a holistic view of control and of energy use.
Track Power Consumption Data
Track new server installation and provide higher precision, control and automation to power distribution with an automated, energy-aware network between IT systems and technologies.
Use Hot and Cold Aisles to Save Power
Advancements to reduce power consumption include thousands of environmental sensors for gathering data across the facility, and a water-side economizer providing evaporative cooling through a cooling tower, eliminating the use of a power-intensive compressor. It has hot and cold aisles, and every aisle between rows of server racks is bounded with cool-air intakes or hot-air outlets. Air is brought into the cool aisles from underneath and exhausted from the hot aisles overhead to allow for constant air circulation through the racks.
Smart Allocation of Technology Resources
Improve management of IT loads and shift resources to where they are needed. This would include other advanced technologies such as adaptive vent tiles, fine-grain sensing and sophisticated management software. Preliminary research shows an anticipated decrease in power consumption by about 40 percent from these techniques.
Create a Coherent Network
Creating a common, wired-once, virtual I/O network provides a new data center network architecture that is more coherent, flexible and agile. IT can connect thousands of servers to a virtualized, high performance, low latency, multiprotocol network to reduce network latency compared with conventional switches and enhance agility by allowing administrators to change connectivity and bandwidth allocation without disrupting servers.
Check Out Other Data Centers
For example, Hewlett-Packard opened a state-of-the-art research facility in Fort Collins, Colo., where the company is working to advance sustainable data center technologies. It will house about 2,000 server racks used by the R&D business groups for product development. The facility has a 200-watt-per-square-foot cooling capacity. It features overhead power busbars that provide maximum flexibility for the various power types required to run the server racks. Overhead cable trays deliver 10Gb capable network to the equipment. The 30-inch raised floor is used exclusively for cold air plenum.
Institute an IT Sandbox
An IT sandbox is an environmentally separate room within the lab facility, built at 350-watt-per-square-foot cooling capacity. This room also has free air cooling, humidity control, server rack temperature sensors and adaptable vent tiles. HP Labs, for example, will partner with research and development teams to use their equipment in a sandbox to conduct sustainable data center research.
Utilize a Specific Architectural Framework for a Sustainable Data Center
Five principles for integrated design and management that cut across all three infrastructures, and indeed cut across the multiple disciplines that are needed to achieve a sustainable data center:Data Center Scale Lifecycle Design; Flexible and Configurable Building Blocks; Pervasive Sensing; Knowledge Discovery and Visualization; and Autonomous Control.