Unlike other technology landscapes that hold their shapes with all the consistency of a jellyfish, the storage market takes its sweet time in introducing technologies and trends. This can lead to complacency among data center managers, who might not give a second thought to their storage environments—after all, if everything's working fine, why make changes? However, a closer inspection of the storage industry reveals an ecosystem teeming with change within existing technologies, and much of that change can represent real value to data centers.

A prevailing impetus for change today is the struggling economy, which is forcing data center and IT professionals to consider new methods for handling data. According to recent research from HP, 56% of technology executives say that storage consolidation will be a high priority to be considered for implementation this year. According to the report, investments such as this are expected to result in cost reduction, increased efficiency, and a strong contribution to the company’s strategic goals.

“The architecture of storage hasn’t changed in 50 years in terms of storing full copies of data,” says Chris Gladwin, CEO of Cleversafe (www.cleversafe.com). “However, we are in the midst of rapid change in terms of the type of data we are storing, the access patterns, and the sheer quantities of data.”

Paring Down Storage

Many experts point to cost-cutting as the thrust behind today’s biggest storage trends, and for good reason, explains Carter George, vice president of products for Ocarina Networks (www.ocarinanetworks.com). Because data centers are being forced to do more with less, they’re moving toward storage efficiency and data reduction technologies (such as thin provisioning, unified storage, data reduction, and compression) as potential answers to that challenge. File storage vendors are leading the charge, with George noting that some companies have made storage efficiency and deduplication the cornerstones of their product and marketing campaigns, an effort that is forcing other storage vendors to fall in line.

“With the current economy, customers are looking for backup solutions that will reduce costs and simplify storage operations—deduplication is that solution. Deduplication allows businesses to retain data on disk for longer periods of time, so less time is spent searching for data. It also reduces costs and capacity by reducing duplicate data [and] enables customers to minimize administrative overhead, reduce floor space, and lower energy consumption by decreasing storage capacity requirements,” says Kyle Fitze, director of marketing for HP’s Storage Platforms division (www.hp.com).

According to George, one reason data reduction is a hot topic is the fact that it’s been validated in other parts of the data center. For example, deduplication has become the norm for backups, with disk-based backup targets with built-in deduplication quickly replacing tape methods. But although compression and simple deduplication have become as commonplace as WAN optimization, applying this model to the online storage world is more difficult, he says, because the algorithms need to be focused on different things, and the insertion model is tougher. Further, people aren't apt to replace their storage or storage vendor to obtain data reduction, George says. As such, online storage is just now beginning to see the early stages of data reduction technologies making their way toward its realm.

Keep It Long, Keep It Secure

Despite the down economy, healthcare and government facilities are actually enjoying big growth, says George, which equates to big storage growth. Coinciding with this growth are the demands placed on these facilities by regulations and compliance to store certain files and data types for long periods of time. For this reason, George predicts low-cost archival and storage platforms and data reduction will continue to gain significant traction in these markets.

“There are a couple of other things that tie in,” he says. “One is object-based stores. Things like the cloud—Amazon S3, IBM’s cloud plans, and industry-specific clouds—tend to have object interfaces. You use HTTP GET and PUT to place a file in the object store for long-term storage. It’s a better fit than NFS or CIFS for archival storage, and I think you’ll see uptake for this kind of storage for specific markets like healthcare and medical images.”

Cleversafe’s Gladwin adds that a key trend for health, education, and government institutions will be security, and data centers will be seeking systems that virtualize data to an ultimately secure format in both physical storage and transport.

On The Decline

Although deduplication, storage virtualization, and other technologies could indeed bask in some level of limelight in the coming months and years, other technologies are bound to lose favor in the data center market. One of these, says George, is Fibre Channel block storage. “[These] arrays are disproportionately expensive compared to other offerings. Like the mainframe, they will have a niche in the Fortune 500 database data center, but widespread adoption is going to decrease. . . . The FC crowd is trying to preserve FC by doing FCoE, but it may be too little, too late,” he says.

Similarly, Gladwin says RAID is also headed for a decline, but more for technical reasons. “It mathematically is reaching a breaking point based on 1TB drives. RAID 6, based on parity, cannot recover from more than two simultaneous failures or two failures and a bit-rate error. Typical SATA drives have a published bit rate error of 1014, meaning once every 100,000,000,000,000 bits, there will be a bit that is unrecoverable. Although this failure rate seems insignificant, when reading 100 terabytes, it is nearly certain there will be an unreadable bit, and if this read happens to be during a rebuild, data will be lost,” Gladwin says.

by Christian Perry