Anyone working in IT could be forgiven for suffering a little green-fatigue. IT vendors are eager to throw "greenwash" around, highlighting the power-saving features of their existing technologies.

However, this is not necessarily all bad. There are many existing technologies that have a greater power-saving potential than some of the dedicated "green" solutions designed for that purpose.

Companies have to avoid the trap of rushing out to buy expensive, new energy-saving technologies that only marginally reduce the cost of bad practice. Focusing on incremental improvements in the energy efficiency of new equipment, on its own, is only easing the symptoms and not addressing the illness. Companies, particularly medium-sized companies, need to focus on one or two "green storage initiatives" that can deliver maximum immediate benefit.

In a recent report entitled "The Greening of the Data Center," the Taneja Group highlighted this point, explaining how power consumption in the data center is a function of data volumes, the service levels surrounding how that data is processed, stored and backed-up, and finally the technical efficiency with which those service levels are delivered.

Storage service levels are certainly one area where a tightening up of practice has as many savings to offer as the implementation of many dedicated "green" technologies. Up until now, the relatively low cost of storage has lead to complacency in this area. In many cases, for simplicity of deployment, the IT department has made all storage, as well as back-up and recovery, as fast as possible. Storage systems have been designed to the highest common denominator to deliver the quality of service required by the company's most business-critical systems and users. With the storage system purchased and in place, all company data has then been given this top-level treatment.

However, speed, resilience and high levels of hardware redundancy now come with a new cost, not only the hardware cost. Not even the direct cost of consuming more, ever-higher priced power or the indirect costs to the environment. The fact that power shortages in many urban areas are causing competition for data center capacity is pushing up costs even further. Suddenly, the question has to be asked as to whether all data needs to be delivered with the same quality of service.

For certain, all business storage needs to be safe and reliable. But whether it all needs to be on the same power-hungry FC SAN, or needs the same high-speed/low-volume disk, or needs the same RAID configuration, or whether the back-up and restore targets times all need to be the same, is more questionable.

But while large enterprises can invest in several physical tiers of storage, this is not a viable option for most medium businesses.

Moreover, the dispersal of storage onto multiple systems will, in itself, increase energy consumption, as well as increase complexity and administrative overheads.

For the medium business, heterogeneous virtualization provides a solution by opening up choice to the administrator. The ability to connect multiple heterogeneous arrays to a storage system enables users to tier storage on the same system. By combining this with effective virtualization, administrators have the ability to create multiple virtual storage systems, with different qualities of service, behind one single storage system.

A high-performing power-hungry SAS or Fiber Channel disk can be allocated to mission-critical systems, while less critical applications and file storage can be placed on the less power-hungry SATA disk. RAID levels can be set differently on different arrays or software RAID can be applied by the storage system, allowing different RAID levels on the same bunch of disks. Backups can be tiered separately from the first-line storage of an iSCSI/IP SAN or FC SAN, allowing for the incorporation of a low-powered SATA disk to be the remote backup storage pool.

With all storage virtualized behind a single storage system, the administrator has the broadest possible number of permutations of which physical storage resources to use for first-, second- and, where necessary, third-line storage. Backup systems and dependencies can be tailored to the needs of each storage host. In essence, a new virtual system can be built with its own service levels for each new application or file system.

Equally important, with all storage behind a single virtualization engine, storage volumes can be moved between physical resources without disrupting the application. For the administrator, different tiers of storage are not separate physical storage islands. For example, spare tier-one capacity can be allocated to tier-two applications until it is needed by a tier-one host. This alleviates the build-up of separate pools of redundant capacity at each stage of the tiering process, reducing overall levels of redundant capacity, which amount to wasted energy and capital outlay.

The real benefit of heterogeneous virtualization is more than reducing power consumption and hardware costs, the real benefit is choice.

Companies are not tied to a specific vendor or grade of disk by their initial system purchase. Existing storage volumes can be moved between resources as required and new arrays can be purchased according to the company's real current need. This freedom of choice also allows companies to take full advantage of new, faster or more efficient disk technologies from any vendor as they arise, without waiting for their incumbent supplier to catch up.