Wireless will keep IT managers awake

It seems an eternal law that, as one disruptive technology is tamed and brought under control, so another is born to perpetuate the anarchy.

Whenever a brave new technology is announced, my heart goes out to the IT managers who have to deliver it. It seems an eternal law that, as one disruptive technology is tamed and brought under control, so another is born to perpetuate the anarchy.

It has always been this way. Small business systems started the trend in the late 1970s, and PCs and client-server architectures started the revolution again a decade later. The next threat to your peace of mind will come from pervasive wireless devices.

You can forget about Wap, but rest assured that wireless devices are going to arrive in droves. When they do, it may be not just the devices that are mobile but also the networks they hook into. The first Manet (mobile ad-hoc network) standard could be with us within a year. So, first find your network, then find your devices and then see what you can do about controlling them. What will that do for the sleep patterns of infrastructure managers?

Control will be needed, not just because it's the only way to sustain service levels, but because auditors will be totting up the amount invested in a few thousand devices and asking where they all are. Privacy regulators will want to know how you control what they are being used for.

When client/server was all the rage a decade ago, I remember sitting with the IT director of Sainsbury's at a Meta Group briefing. As the briefing progressed, he kept repeating to himself: "But how are we expected to control all this and what will it cost?" He finally put the question to the presenters, who replied: "You have to do it because business managers want it."

What on Earth, I thought, were business managers doing, getting their sticky fingers into IT architectures? Why should they bother how their IT services were delivered? Meta was right in saying client/server would happen because business managers wanted it, but it turned out that what the business managers really meant by client/server was just a GUI interface. Windows terminals or thin clients of some sort have since proved a more controllable - and often cheaper - solution.

Soon it will be the turn of wireless to cause the headaches. What's needed is not just the 'any device, any time, anywhere' kind of capability usually associated with mobile wireless, but the ability to exercise control through the same centralised systems used for the rest of the IT junk shop. The ability to do that, even if only on a catch-up basis, is the only thing likely to keep IT managers sane.

Like servers, desktops and the ever-growing number of other boxes, wireless devices need to be on an inventory. They malfunction on occasion, they are subject to change and they experience variable response times. The last thing enterprise managers need are separate inventory, problem, change/configuration and service management systems for wireless devices.

Those who have survived long enough already experienced that problem, when communications and data processing first collided. The commonality of simple network management protocol helped then, and it will with wireless too; but it won't be enough. Console integration is the minimum with, ideally, close integration to all the existing system/service management functions as well.

To really sleep soundly, though, you really could do with all that before the whirlwind hits, not afterwards.