Our six-step program to network computing success Forget technology, the real challenge in implementing client/server computing is people.

December  1995

Deconstructing the mainframe and distributing the servers is all well and good, but the key to rightsizing is the people. The authors outline some ways to build unity within the IT department and implement the transformations that lead to a successful client/server installation. (1,000 words)

Mail this article to a friend

When we first started evangelizing the how-to's of rightsizing, downsizing, and capsizing back in 1993, we spoke about our #1 priority, a client/server production acceptance process. It was, and still is the most critical process, the key link in implementing the proper infrastructure to support today's enterprise. We'll update you on the latest revisions to this process in future columns.

After visiting with thousands of companies around the world, we've found the biggest problem large firms face when moving from host-based to network-based computing is not technology, it's people!

In previous columns we outlined how to retrain mainframe and Unix people, but education is only a small piece of the puzzle. The biggest piece is dealing with the diverse cultures within the organization. You must break down the walls not only between PC, Unix, and mainframe people, but between these groups and (horrors!) users.

It's the big guy's fault
To a) get this out of our systems, and b) place the blame elsewhere, let's point the finger at who's responsible for underestimating the difficulty in adopting network-based computing. It's the CIO's fault. (If you are a CIO, feel free to change the "I" in "CIO" to an "E.")

You know the scene: Well-dressed, lightly perfumed executives shmoozing in ivory-tower boardrooms about mergers and acquisitions, company politics, and reorganizations, punctuated by occasional pointed questions why computers cost so darn much, and when the board will start to see the ROI on all of the client/server spending. (Life as a CIO can't be all fun.) Executives aren't stupid. Some just don't have a clue what it takes to implement this crazy new enterprise, or don't care to listen to a savvy CIO's warnings and explanations.

Lead by the CIO, the IT team must refresh the corporate memory in regards to how difficult it was in years past to exchange one mainframe operating system for another. It took months of planning and testing to move the organization from, essentially, one tractor-trailer rig to another. How quickly people forget the three decades it took to establish the procedures that established the foundation for a secure and reliable central data center.

Executives and users are demanding and impatient. Spurred by competition and oftentimes glowing accounts in trade magazines of client/server bliss, executives and users fail to recognize that not only is a safe change tough under ideal circumstances, it's very difficult when the organization is adopting a computing system with double or perhaps triple the number of variables. If moving from one mainframe OS to another is like changing truck brands, then switching from centralized, mainframe-style computing to a network-based system is like trading trucks for airplanes.

Both executives and users need to understand that today's networked computing paradigm implies wide-ranging organizational changes far beyond Microsoft's plan for selling Windows95 to the hoi polli and Windows NT to the more demanding user. By change, we mean establishing a new, fast-moving, flexible organization where information is available in a timely manner to those who need to make decisions rapidly.

Isolate, right? Wrong!
In our experience, the first reaction IT departments have when contemplating a network-based installation is to isolate the legacy stuff in a cocoon. To salvage morale, they will bring over some of the legacy gang to work on the new fun stuff. A nice gesture, appease the troops right? Oh, what a mistake that is!

You need to mesh your entire organization. Never, ever separate legacy (usually mainframe) and client/server. Don't even refer to part of your organization as the mainframe group and the other as the client/server group. That's when this barrier, thicker than the Great Wall of China, comes up. We see it in just about every company around the world. Yet the CIO usually has no idea it's there and that it's a virus slowing client/server implementation. No one will talk about it, because they're afraid it will jeopardize their careers. But how can you successfully implement such a huge undertaking without everyone moving in the same direction, working as a single unit?

Tear down the walls
An opportunity has arisen as IT organizations move forward with client/server computing: an opportunity to make a change in the way IT provides services to its users.

In the 1970's the mavens of mainframe-based corporate Information Technologies (IT) hoarded computing resources and dictated computing practices and disciplines to their customers. Even the computing architecture of host-terminal technologies showed an ambivalence, if not downright disdain toward anyone outside the glass walls of the data center.

In the 1980's, users got their revenge. By hook or by crook and rarely under the auspices of corporate IT, they sneaked in PCs with an assortment of personal-productivity applications and joined the power-on-the-desktop revolution. "Who needs IT?" they shouted! Those renegade PCs revolutionized business but made a mess of corporate computing, which IT is called to clean up.

Now in the 1990s, the opportunity is there for IT to be a factor again, but not a dictator. IT needs to win its customers back. As companies shift paradigms to distributed computing it will be the perfect time to drastically improve customer satisfaction. This will not be easy to do because of the damage created over the past few decades by host-based computing. It will be extremely important for them to pick a partner that has first-hand experience in dealing with these issues.

To borrow a phrase or two from the self-help industry, IT needs a six-step program to help bring itself and its enterprise not only to health, but in fighting trim to compete in the modern global economy. Here are the steps we advise IT organizations to follow as they transition to the New Enterprise:

1. IT needs a clear understanding of how to organize and better support the new enterprise. It needs a relevant mission, a roadmap to a more efficient enterprise.

2. IT needs unity to fulfill its mission statement. All IT people (mainframe, Unix, and PC) must work together as one team.

3. Before it can change other parts of the organization, IT itself must change the way it works. It must demonstrate the ROI of client/server computing, reengineering, or whatever catch-phrase you choose, with real-life, undeniable evidence.

4. The IT staff needs to learn and use techniques found in sales and marketing. IT needs to sell itself.

5. IT needs to develop skills in identifying processes and bureaucratic procedures ripe for streamlining.

6. Hardest of all, IT needs backing from an organization's highest levels to help when encountering sticky political or bureaucratic resistance to change.

The introduction of client/server means an opportunity for change and by following this program, that change means breaking down the walls between IT and its users, between mainframe and Unix and PC people. You must take the time and implement the proper infrastructure for the New Enterprise. It all starts with people!

Resources

• A list of other Unix Enterprise
  /sunworldonline/common/swol-backissues-columns.html#unix

If you have technical problems with this magazine, contact webmaster@sunworld.com

URL: http://www.sunworld.com/swol-12-1995/swol-12-unix.html
Last modified: