can SAN make network HISTORY?

You may remember the old Yogi Berra quote, "It's deja vu all over again." If you were to look at network storage today, you would swear you had seen it all before. And you have ... in 1984.

Back then, the networking industry was driven by the high cost of storage devices. Novell, for instance, began as a PC vendor. (Honest.) It got into networking software because no one would pay $100/MB for its 10MB workstations. By sharing these pricey drives, however, Novell found it could convince users it was all worth it.

Lots of other vendors copied the approach, but with a proprietary edge. If you bought software from A, it wouldn't work on hardware from B. Then Novell stepped in and made NetWare work with everyone's hardware--a strategy that leveraged the entire installed base of PC networks at the time. The network industry exploded. It stopped being a niche player and entered the mainstream of business MIS.

Jump forward to 1999. What do we find? It costs so much to implement Windows NT that we now purchase software to share the server's processing power among many dumb terminals. Otherwise, we wouldn't buy into this massive processing unit and storage facility.

The newest contending shared storage strategy, Storage Area Networks (SANs), may take the storage industry from commodity backwater to the sexy forefront. Every vendor is rushing SAN-enabled devices to market. SANs are written up in the press as the best way to enhance network performance.

But now a proprietary problem is back. One vendor's SAN technology promises, but doesn't match up with another's. The current SANs also lack the specialized gateways to provide file and record locking on each device. Current OS options, such as NetWare, NT, and UNIX have no idea what to do with the SAN. SAN is great if you buy pretty much everything from the same vendor. But its lack of proven interoperability threatens to slow adoption, as history amply demonstrates.

unify!

The next two to three years could be the best ever for network optical storage. It could explode, just like business networking did after 1984. But unless the parochial nature of the industry changes, it will surely stagnate.

Today's network optical storage industry seems content to pursue modest goals in meeting the demands of the high-end user in various niche markets. Unit sales in the four figures are adequate to maintain the organization, so why pursue anything else? Jukebox vendors wait on drive manufacturers to demonstrate value to consumers before new drives appear in their systems. Drive vendors have their hands too full maintaining margins to take leadership roles.

So who will lead as Novell did, vanquishing every other competitor by leveraging all PC users' current hardware investments and thus demonstrating to average businesses how networking technology can save them money? Fewer printers, fewer software licenses, and fewer device conflicts all add up to fewer headaches.

SAN technology sits in the same position as networking in 1984--waiting for a savvy SAN vendor to come along and make the concept useful. The time is ripe for a new NetWare to step up and provide the full level of SAN services at a price that departments and small to medium businesses can afford.

simplify!

But what about history's other lessons? If Novell knew the secret to network pre-eminence, why did it ultimately cede its title to Microsoft? Because Microsoft reduced the buying equation. Corporate buyers no longer had to be computer engineers (as, arguably, NetWare adopters did) to assemble a network system. In theory, Microsoft saved users from multiple buying decisions attendant to using NetWare: Which desktop operating system to use? Which network OS? Which application suite? All could be decided immediately with Microsoft, and the pre-installed office suite triumphed.

The same is true of optical storage today. There are simply too many buying decisions for hardware and software. Today's jukebox management software scene has begun consolidating software either as part of a vertical package, such as a digital content management application or as a seamlessly integrated part of the OS itself. Even at the high-end level where optical storage shines, demand remains for one less buying decision.

Who of the various vendors will stop probing and become the "NetWare" of storage? For example, where's the vendor with the simple buying equation for this product: the bundled "Y2K Protection Package" consisting of a DVD-RAM jukebox, an NAS thin server, and pre-installed network backup software? It could be plugged into the network, initialized, and then told to begin a continuous backup of network files.

And where's the vendor who will take its optical storage management capability and incorporate it with both SAN and Sun's new JINI technology? Here you would have what I would call "instantaneous network"--the Internet of local networks. Intelligent storage devices (with JINI) could be part of the same view as devices controlled on the SAN. Both high- and low-end users would prefer one solution that integrates many pieces than having multiple solutions for the pieces.

One-step solutions are ready-made for today's network storage market. As Plasmon's president Rob McPherson puts it, "This isn't about competing technologies. So the question is, who provides not just the right tool, but the right storage solution?"

NETWORK OBSERVER columnist David Doering [dave@techvoice.com], an EMedia Professional contributing editor, is also senior analyst with TechVoice Inc., an Orem, Utah-based consultancy.

Comments? Email us at letters@onlineinc.com, or check the masthead for other ways to contact us.

COPYRIGHT 1999 Online, Inc.
COPYRIGHT 2000 Gale Group