Four Tips to Optimize IT Energy Use

Four Tips to Optimize IT Energy Use

A growing number of companies are coming to the realization that they are entering a data center energy crisis.

With data centers nearing full capacity, business growth becomes inhibited as organizations struggle to make necessary capital investments in either heating and cooling equipment, or in added storage.

In fact, a 2006 survey of data centers by power supply vendor Liebert Corp. found that 33 percent of companies actually expect to be out of power and cooling capacity by the end of 2007. What's more, 96 percent expect to be out of capacity by 2011.

Reducing energy use clearly is about more than being environmentally responsible. To ensure their long-term viability, organizations must begin now to find and implement solutions that help decrease power consumption.

The good news? Many of the same tools and practices that have enabled these organizations to reduce IT complexity, streamline operations, and control costs are also highly effective in optimizing energy use, such as data de-duplication, high availability and virtualization, power management, and energy-efficient data center design.

Protecting and Backing up Data
Data de-duplication has emerged as a powerful tool for dramatically reducing storage and bandwidth consumed from disk-based backup.

By eliminating the problem of constantly backing up the same copy of a file again and again, data de-duplication can decrease backup storage consumption by 10 to 50 times compared to traditional tape-based backup methods. Since less data is sent across the infrastructure, data de-duplication also can reduce the bandwidth consumed by traditional network-based backups by up to 500 times.

For many organizations, the hardware savings of data de-duplication can be so significant that it has become a "must-have" in the corporate data center.

Meanwhile, the energy efficiencies associated with data de-duplication are often overlooked. Yet reducing the number of data copies reduces storage capacity requirements and storage power consumption. What's more, once reduced, snapshots and other copies from high performance disks can be moved to lower performance, more energy-efficient disks.

That's not all. By leveraging data de-duplication when backing up remote sites, this technology can provide significant savings in storage capacity and network traffic by not transmitting and storing redundant data, regardless of location. This not only preserves bandwidth-in some cases -- by a factor of 50 -- but also keeps storage capacity requirements to a minimum.

High Availability and Virtualization
To date, many clustering tools have served a specific function. Standby servers, for example, often do not perform productive work in a normal everyday environment; they simply wait for a failure of a primary server, at which point they take on the workload of the main server.

Many clusters are highly localized, usually to a single pair of servers sharing a single disk array and running a single critical enterprise application. In many instances, the primary and redundant infrastructure resides in a single physical location.

Moreover, most clustering tools require identical hardware and operating system environments for implementing a clustering solution. This can be an expensive and restrictive approach since organizations need to ensure platform parity in a cluster.

Investing in the right software, however, can extend the resiliency and usability of virtualization so that it has the ability to address enterprise class applications. For example, advanced clustering technologies combine the ability to act as a traffic director and move applications among servers and storage devices to fine-tune performance, regardless of failure, and in turn reduce hardware, space, and power requirements. Clustering technologies that support heterogeneous and even virtualized environments also keep hardware requirements to a minimum, thereby also decreasing space and power use. These tools enable organizations to add clustering to existing infrastructures without having to purchase additional servers.

At the same time, the degree to which organizations can more efficiently back up, replicate, and host virtualized environments on clustered entities increases the viability of virtualization for enterprise-level applications. This enables IT to extend virtualization to more applications and reduce the number of servers used, along with the power and energy they require.

Managing Power Wisely
In the typical business, computers are used an average of only four hours each business day. Yet, even when they are idle, they continue to use energy.

Experts estimate that approximately 65 percent of energy used by computers and monitors is actually wasted because they often are simply not turned off when employees leave for the evening. Meanwhile, nearly half of monitors are not enabled for power management, thereby missing an opportunity to save money and energy.

However, utilities are now available that enable systems administrators to quickly and easily govern and monitor power management, or sleep, capabilities of all desktop and system monitors throughout the corporate environment. These utilities may be available as free toolkits, and typically enable sleep features that are built into popular operating systems and allow computer monitors to go into a low-power sleep mode when inactive. With these capabilities integrated into their systems management infrastructure, IT can designate monitor shutdown policies at select intervals and administer them from a single web-based management console.

By some estimates, these toolkits can save an organization between $10 and $50 per PC annually, while also helping address environmental responsibility objectives.

Designing Efficient Data Centers
Companies across the globe are realizing that the lack of data center space and appropriate power and cooling facilities is beginning to hinder their deployment of business-critical IT projects. Worse yet, for many, this problem will likely worsen over the next year.

As a result, a growing number of organizations are working with data center design consultants whose concerted project approach to layout can deliver a significant space gain for an existing site while also delivering measurable power and cooling benefits.

Among other activities, these consultants can provide a rapid assessment of how much power is being used to cool unused, redundant, or over-specified systems. They also will evaluate current utilization to help the organization eliminate old or duplicate equipment and get one step ahead in the challenge of becoming more energy efficient.

As costs rise while data centers inch toward full capacity and power and cooling requirements, organizations must begin addressing energy challenges now. The challenges will only likely become more difficult over time.

These tactics and technologies can enable IT to decrease costs, simplify IT operations, reduce data center complexity and optimize energy use.

By decreasing storage consumption through data de-duplication, ensuring high availability through advanced clustering, managing inactive workstations through power management utilities, or maximizing energy efficiencies through intelligent data center design, organizations can better control IT energy usage now and into the future.

Bruce Naegel is senior product manager of Symantec Corp.'s data center management group.