% fortune -ae paul murphy

Revisiting the past - and finding the horse dead

I ran into a guy recently who works for the provincial government - and discovered that a client assignment I'd chalked up in the "win" category had actually been a disaster.

The problem looked straight forward - a few years earlier the client had brought in new IT management; the new management had launched a Websphere/DB2 based redevelopment effort aimed at replacing the Windows/Sybase applications in place and, as an incidental preliminary, had IBM replace the then existing SPARC servers with AIX machines.

The whole thing had gone from dumb to dumber with uncontrolled IT cost growth, the then latest failure taking out about a week's worth of patient records, and - well, you can imagine the rest, none of it pretty.

I did what I usually do: counseled them about simplifying their infrastructure while aligning their technology with the skills available; helped get their stuff working again; implemented replication server; talked to IT about replacing paper planning with actual disaster drills; and, got user management started down the path to third party applications. When I left everything seemed fine: years later the people involved still take my calls and they're drifting along the cliff face between budget crunches and minor project successes pretty much the same as everybody else.

So what did I miss?

Pretty much the whole big picture thing. The organization operates as a nominally independent entity within the nationalized health care framework, and I'd known at the time that the money being spent on the Websphere/DB2 stuff came as an ear mark allocation made at the provincial level. What I didn't understand was that the money was being spent at the subsidiary level only because reporting constraints prevented the central authority from directly investing in health care systems redevelopment.

So, bottom line, by "helping" my clients get their own house in order, I helped them cut off a ten year flow of funds with no actual relationship to the operating IT budget at all, and thus robbed them of the opportunity to register a big win on systems at the central authority's expense.

What I should have done, of course, was recommend splitting IT into two groups: operators and skunks - something that's often a good idea and that in this case would have solved the problem while keeping the cash coming. I didn't, another group got the money, and Chris, if you read this: please consider it my apology.


Paul Murphy wrote and published The Unix Guide to Defenestration. Murphy is a 25-year veteran of the I.T. consulting industry, specializing in Unix and Unix-related management issues.