How Can the Infrastructure Catch Up Quickly?

This is the $64,000 question. The answer is that the infrastructure can't catch up very easily. It takes time and resources. You tell us who in the world of IT has that luxury today? Not too many organizations, but the alternative is to let client/server die a slow death. Everywhere we go management is complaining to anyone who will listen. The poor vendors providing client/server solutions are feeling the brunt of it. The complaints are the same in every part of the world—client/server is not as reliable, available, and serviceable as the mainframe world. We've heard this for almost ten years now. You know what? They're right, and we might also add that on a scale of one to ten, if the mainframe infrastructure was an eight (it had some major flaws), then client/server has to be a two (and that's being generous). Ten years later, and only a two. How pathetic.

You know IBM is sitting back and enjoying this. IBM is telling customers that the mainframe didn't have these RAS issues. And, IBM isn't alone. The Unix vendors, such as Sun, HP and even IBM are bringing to market massively powerful, multi-domain servers which are designed to serve a similar purpose as mainframes, namely to host a variety of applications on the same box. Even so, with n-tier or client/server computing, the fact remains that computing in the next millennium will be more intricate and more network-based which means that the people and process issues are glaring and still with us.

There is hope. Start small. Don't try to take on everything at once. The issues are enormous. Determine what the top three to five problems are within your IT computing environments by doing a high-level assessment, and follow the steps in Chapter 8 on what it takes to build a world-class infrastructure. Then, set up a measurement system to track how well the organization is performing.