Many tech decision makers are tagging new power, cooling and space efficiency strategies such as data tiering and archiving as keys to dramatic, immediate and practical improvements in data center server performance and storage capacity.

Terri McClure, an analyst with Milford, Mass.-based Enterprise Strategy Group, says technological advances—particularly when it comes to collaboration and virtualization—dramatically optimize power, cooling and space efficiency while delivering optimal levels of server capacity, input/output and services.

After server consolidation, storage consolidation holds the biggest potential for saving energy in the data center, according to McClure. Techniques such as data tiering and archiving let users drive up sagging storage utilization rates and drive out aging subsystems that consume excessive amounts of energy, she says.

Determining the exact amount of those sagging rates, says McClure, is a solid start toward increasing the positive effect storage utilization rates have on storage consolidation strategies.

For instance, Lawrence Schwartz, a senior Symmetrix product-marketing manager with Hopkinton, Mass.-based EMC, agrees with a McClure estimate that 30 percent to 50 percent of disk capacity is typically unused.

Most enterprise firms, he says, are aiming for 60 percent to 70 percent utilization, which can typically be achieved with an appropriate combination of storage resource management tools. In some cases, says Schwartz, the best choice of disk drive has the potential to boost utilization to 90 percent or more.

Data tiering is widely viewed as one of the most important storage consolidation strategies partly because it offers many options. From McClure’s perspective, the complexity of data requires that it be managed and stored according to the tier in which it is bundled.

Data tiers are broken into different categories, including:

• Inactive data that is stored offsite and rarely, if ever, accessed (Tier 4);

• Persistent inactive data that is rarely accessed, but must be available within seconds or minutes (Tier 3);

• Persistent active online data, which is no longer appropriate for “big iron” storage systems (Tier 2); and

• Dynamic active online data, which is constantly changing (Tier 1).

McClure suggests that users focus the most money, energy and time on the dynamic, active Tier-1 storage.

Schwartz describes a broader vision of data tiering that starts with RAID configurations and flash drives on the higher end, and works down to information lifecycle planning and data classification on the lower end.

McClure describes archiving, meanwhile, as a popular storage consolidation technique that can be even more beneficial than efficient power supplies or spinning down drives. This is because the biggest aid to consolidation comes from reducing the amount of stored data; thereby reducing the overall number of drives that need to be plugged in.

“This can be done by leveraging thin provisioning, space-saving snapshots, and archiving data off Tier-1 storage onto lower-cost, denser disks, then eventually to tape or optical,” says McClure. “There are significant savings to be had in floor space, power and cooling by migrating data to nearline and offline storage.”

Bruce Hoard is a freelance writer based in Bangor, Maine.

Related content

Storage vendors are setting the SAN table for SMBs

File virtualization is worth a second look