"If everything seems under control, you're just not going fast enough."
Mario Andretti—Race Car Driver
Change is a curious word and one that often makes for clichéd slogans, political or otherwise. It’s certainly one of the most overused words of the PowerPoint generation, but the fact is that the need for change is real.
Change happens—whether we like it or not. In the end, staying the same in business as in technology is a surefire way to be left behind.
Transformations take place right before our very eyes without our even noticing it. Look at any major modern city and make a quick survey of the kinds of cars you see on the road. Now compare that view with a flashback of the types of cars roaming around in the seventies. Yes, some cars today are pure vintage, thanks to hobbyists, but a look at the landscape today versus twenty or thirty years ago easily demonstrates that car lines have been updated dynamically in a steady process of sun-setting and replacement. Indeed, the median age for cars in the US is about eight years (creeping up to ten years as people are deferring new car purchases due to the recession).
Look at the same city landscape and observe its buildings. If you are in the downtown area, chances are you will find a number of classic buildings, renovated at great expense if the city has gone through downtown revival; decrepit if the city economic center has shifted to the suburbs. IT systems have become more like the buildings than the cars, and frankly, finding renovated IT systems is rare, most legacy IT systems resemble more a decrepit downtown.
Wouldn’t it be ideal if we were able to get to the point where IT systems can be revamped the way car models get renewed every few years? Well, new technology like the Internet and SOA can help here.
The Internet has been a game changer. Its full impact is still undetermined, but it is certainly forcing a closer look at original IT architectures. Legacy systems were designed around less agile service and maintenance models. Back in the days when it was fashionable to hate IBM (before it was fashionable to hate Microsoft and then to hate Google), centralized mainframe environments closely controlled by the glass-house priesthood forced processes that made it difficult to satisfy user demands and presented serviceability bottlenecks.
The need to overcome existing system deficiencies is not new. In fact, addressing ongoing customer demands has been the lifeline of IT organizations for decades. There has always been a need to transform the original systems, and there has always been the desire to replace the legacy with any available emerging technology. However, legacy technologies were not, and are not, easily replaceable, and the irony is that in every successive generation, the new systems were incorporated, but the old system was never actually retired.
Next week I will analyze in more detail the prototypical data center of today. I believe most large corporations have arrived at a common computing system pattern that must be understood if we are to extricate the system from their current complexity.
For now, don't you agree that the picture above is representative of today’s IT legacy systems?