Network storage's annus horribilis
Analyst Ashok Kumar attributes an abysmal year in network storage to the narrow pursuit of self-interest of warring industry factions who only succeeded in fostering more customer confusion.
Wannabe storage networking companies dropped off the face of the earth and many established companies saw their profits shrivel and margins sink. Only a year ago, we were being told by industry executives that storage was not a discretionary purchase because there was an insatiable appetite among information technology users for storage capacity.
Bah, Humbug! The fact of the matter is that technology advancements were minimal, with the increase of Fibre Channel from 1Gb to 2Gb as the crowning achievement.
The network storage industry is dysfunctional--behaving more like warring tribal leaders than an industry trying to move to the front of the IT value chain. Brocade Communications Systems, the Fibre Channel industry's biggest player, prefers to keep its technology proprietary for as long as possible, and in the process it's stunting industry growth. EMC, the biggest fish in the storage tank, pulls no punches when attacking competitors and former employees without regard to the negative impact those actions may have on the image of the industry.
As long as network storage companies actively derail their own railroad, the market will continue to be confused and industry growth will be disappointing, just as it was in 2001.
The industry entered 2001 short on talent to educate the market. At the end of the year, after the layoffs and contraction, the total head count in the industry is probably less than it was at the beginning. That does not sound much like the next great boom industry. Considering the enormous work that needs to be done before the broad base of buyers trusts the technology, 2001 was a lost cause. But numbers alone won't do the job; the industry needs to agree on some common themes and language to further its cause. Instead, the market is presented with cacophonous hucksterism doled out in the name of education.
A prime example of the industry's spastic messaging is storage virtualization--the "hot-dogger" of the storage industry for 2001. While the appeal of storage virtualization worked well enough to allow companies to raise money and attract trade-press attention, it failed to produce a big boost in the sale of storage products. The fundamental problem with virtualization is that customers do not need a new, more expensive, less familiar way of doing something they can already do with familiar products such as RAID (redundant array of independent--or inexpensive--disks) controllers and volume management software.
The hyping of virtualization was raised to such a fevered pitch that storage networking companies tried to cast almost every technology innovation in 2001 as some type of virtualization. Older technologies, such as HSM (hierarchical storage management), were also repositioned as virtualization, even though they have already been sold for decades.
The primary problem with virtualization is that it is a technology looking for a solution, rather than a solution to customer problems. There is a very real need to find ways to manage storage more effectively, but believing virtualization is a solution is akin to believing the wheel is a solution to mass transportation needs. Virtualization can be a valuable component of powerful storage management, but it is much more of a technique than an end in itself.
Instead of virtualization, the storage networking industry needs comprehensive management solutions that unlock the potential of storage networks. Legacy storage and network management applications fail to even meet minimum requirements because they don't integrate the management of networking, system and storage functions in a way customers need.
What's on tap for 2002?
Among the most promising new storage networking developments to look forward to in 2002 are TCP protocol offload technologies, storage provisioning, dynamically scaling NAS (network attached storage) and DAFS (Direct Access File System, a new file-access protocol designed to take advantage of standard memory).
TCP protocol offload processing will allow Ethernet/IP networks to be used instead of Fibre Channel. The main benefit here is the enormous Ethernet/IP talent pool with the accompanying expertise that can be applied to solving customer problems. Storage provisioning is an old telecommunications concept that's being applied to the specifics of storage networks. It allows customers to use and redeploy storage network resources in an efficient, automated fashion.
Dynamically scaling NAS will allow network attached storage, with its superior data management capabilities, to be used for larger data center environments than it is now.
Finally, DAFS will enable NAS products to be used for mission-critical transaction processing environments. In general, NAS will move from being a departmental type of solution into the data center and give customers more architectural options than they have today.