IT Myth 2: Eighty percent of corporate data resides on mainframes

It’s past time to retire the myth that mainframes, those impenetrable-looking boxes understood by only a few IT magicians, still store 80 percent of all corporate data.

Free IT resource TechNet: More ways to know it, share it, and keep it running. Sponsored by Microsoft Free IT resource Attend the SOA Executive Forum: Breaking SOA Bottlenecks SOAExecForum.com/may2007 Sponsored by InfoWorld DOWNLOAD PDF Click here to download InfoWorld's special report Myths of IT

Since their introduction in the 1950s, mainframes have largely been the unchallenged gatekeeper for all mission-critical corporate data. IBM became Big Blue, the color of their early mainframes, by popularizing mainframes with the company’s hardware and operating systems -- and eventually its line of applications -- and then gained an iron grip on the entire market for decades.

But IBM’s early monopoly of the mainframe market came under attack in the 1970s and 1980s. With the arrival of the first minicomputers and then microcomputers, which both held the promise of distributing centralized data closer to users doing the work, Fortune 1000 companies started demanding less reliance on mainframes.

Even with the desktop revolution, the notion that mainframes held at least 80 percent of all corporate data remained intact through the mid-1990s in the minds of many.

But the birth of the Internet and the resulting flood of unstructured corporate data, such as e-mails, Web pages, Microsoft Word documents, and various technologies to manage and store this digital data has led many to conclude that the stranglehold mainframes have held on corporate data has been slipping.

“In dealing with some of our clients, it is almost shocking to see some large organizations’ financials being managed in a couple of Excel spreadsheets. Plus with all the blogs, instant messages, e-mails that do not pass through a mainframe, the amount of data now residing on mainframes is now probably in the neighborhood of 40 to 50 percent,” says Stephen O’Grady, senior analyst at RedMonk.

Reinforcing this growing trend, there is already an impressive amount of mission-critical financial data being generated, shared, and managed out of the sight of mainframes, says Dana Gardner, senior analyst at The Yankee Group. “Some corporate users now have Spreadmarts -- big, honking flat files in spreadsheets used to manage many business processes and in a really decentralized way,” according to Gardner.

And with the aggressive promotion the past few years of dozens of integration strategies that threaten to tear down the technology borders between mainframes and distributed platforms, some question the relevance of where data resides.

Click for larger view.

“Does it even matter anymore [where corporate data resides] is the more relevant question. I think it has less meaning today than it did a few years ago. In fact, the more you hold onto that old axiom, the more you point out it is a proprietary and isolated environment. I wouldn’t think anyone would want to continue promoting that idea,” says Steve Josselyn, research director of the global enterprise server solutions program at IDC.

Another trend eating away at the mainframe’s dominance is the rise of SANs and NAS appliances. Although many such environments have direct pipelines into mainframes where data can be shared back and forth, the inclination of more and more corporate users is to plant data on SANs and NAS devices.

“Increasingly, the type of computer becomes irrelevant with the local-area storage networks and the increasingly sophisticated storage that has come into play,” says Hadley Reynolds, research director at Delphi Group.

(Return to special report)