IT Management

IT Sense: Innovative Mainframe Computing

When I was offered an opportunity a few weeks ago from CA to brief me on the latest release of its mainframe database product, IDMS—a database system that I knew in a former life as Cullinet—I took the meeting mainly out of a sense of nostalgia. I had much experience with Cullinet in the early ’90s, when the technology was being used as the foundation of the U.S. Secret Service’s post stander systems, used to coordinate the myriad details of protection services for U.S. politicians.

Hearing Greg Beedy, senior principle product manager for CA IDMS, so enthusiastically discuss the cool features of the latest release, I realized I had fallen prey to the very mindset that had been casting shadows on the future of the mainframe in the minds of many IT planners for the last few years. I was seeing the mainframe and mainframe databases as “old” technology that must inevitably succumb to something newer and better. Innovation and mainframe tech rarely found themselves in the same sentence in the trade press or in discussions with business clients. Upon doing some research to refresh my knowledge of the history of IDMS, it dawned on me that some important lessons can be learned about business and IT innovation.

What made Cullinet cool in its day was the fact its database was organized as files mapped to areas and subdivided into pages that directly corresponded to blocks on a disk. Originally, it was less a relational database engine than a networked one. Its design closely mapped to COBOL coding conventions, with relational capabilities added over time as RDMS popularity increased.

What really made Cullinet, now CA-IDMS, of interest to me was that its roots were firmly planted in business requirements, and not in the plans of a tech company seeking to finesse competitors for market share. The underlying technology was the brainchild of Dr. Charles Bachman at General Electric, who called it an Integrated Data Store. It was subsequently rewritten into a new portable language by IT folks within the Chemical Products Division of B. F. Goodrich, who wanted to enable it to run on several target machines deployed at the company to serve various business objectives. Over time, it found its way onto IBM mainframes as well as Digital Equipment Corp. boxes, but its development was done within a business for a business.

Somewhere along the way, B. F. Goodrich bosses decided the company wasn’t in the software products business. That opened the doors for John Cullinane to buy the rights to market the product. By the ’90s, Cullinet could boast more than 2,500 installations. Many of these, deployed as early as the ’70s, are still in operation today.

CA acquired Cullinet in September 1989, for $300 million. When they began adding SQL capabilities from a relational database technology, the once-heated debates waned over whether Cullinet/IDMS was really a relational database system or a network database with a relational veneer.

CA announced Release 17 of IDMS a few weeks ago, delivering another round of architectural improvements that continue to meet customer requests, including much anticipated support for those “chameleon” processor modules that IBM has been selling for a couple of years: the zIIPs and zAAPs. IDMS R17 enables workload to be offloaded to these auxiliary processors and customers to see significant improvements in throughput and total cost of ownership, while mitigating the need to add expensive processing capacity to meet growing workload requirements.

This column isn’t a platform for pitching vendor wares, of course. Rather, what I want to emphasize here is a side issue: innovative mainframe computing.

Ingenuity and innovation have been the hallmark of U.S. business and a key to its success. In these challenging times, IT needs to support business innovation more than ever before. When it does harness technology to advance business goals, great things are possible—as the story of IDMS illustrates.

Sometimes, we think business technology innovation comes from outside the business. A mythology has been created around the Silicon Valley engineer, working in quiet isolation in his garage/workshop, figuring out technical solutions to problems that businesses don’t know they have. The way I see it, really innovative concepts—the foundations of great technology—have been born in the business world itself and driven by tangible business requirements.

The computing platform, whether a mainframe, x86 server or something else, provides the canvas for IT innovation; the application software technologies are the brushes and the paints. At the end of the day, the IT practitioner is an “artist” who is driven by a business muse named Necessity to create technology that supports and advances business objectives.