The right question
May 14th, 2010 | Published in Uncategorized | 3 Comments
I attended the AITL meeting yesterday. I found it encouraging and frustrating.
What was encouraging? The AITL is made up of intelligent people who are committed to making the University run more effectively and efficiently. They have a clear grasp of the issues we’re facing, and want to do what’s best for the University.
What frustrated me can be illustrated by a sentence from Brad’s “ITS Weekly Update”:
The BSC will use AITL input as it decides whether we “stay status quo with the mainframe” or “move to Open Systems.”
This is wrong on so many levels. We can change the status quo significantly while staying on the mainframe, and the move to Open Systems envisioned by the Mainframe Migration Assessment would preserve the status quo in every dimension except the hardware and operating system. “Status quo on the mainframe” and “move current Adabas/Natural applications to Open Source” are not the only alternatives!
Dennis said at the beginning of the meeting that the purpose of the assessment was to see if the University could save money by moving our existing application portfolio off the mainframe. I think the report answered that question, and the answer is clearly “no.” But I thought when it was first proposed, and I still think now, that “Mainframe vs. Open Systems” is the wrong question. (Personally, I don’t see where there’s that big a difference.) The right question is something like “what should be the University’s long term strategy for administrative information technology?” If we can answer that correctly we will actually have a basis for tactical decisions like whether or how quickly we should migrate off the mainframe. Focusing on the hardware and operating system just distracts us from the real issues.
May 17th, 2010 at 11:03 am (#)
Honestly, “Mainframe vs. Open Systems” isn’t a question at all. I think Dennis was right in framing the original question, and agree that the answer looks like it is “no”, although only the business leadership can really make that call.
If I understand correctly, you are suggesting we should have a long-term strategic vision before considering projects like the migration. While I agree, it does depend on perceived difficulty of the migration — if it could be done in two years, say, it might make sense to do that, save some money, and then consider long-term strategy. I suspect that management was thinking along something like those lines. Of course, many IT folks were skeptical as to feasibility (based on our knowledge of the code-base and architecture), and it seems like our intuitions were well-founded. But that was what the study was about — determining feasibility. I don’t have any intrinsic problem with that.
I _do_ think that it would be wise to reconsider our strategy, because it has been a long time since that strategy was developed, and the world has likely changed. The CASE study did a good job of demonstrating value in our current applications, but it didn’t really address whether we are providing the right IT services or could be doing things in a better way. That said, I think we could have used such a reassessment five years ago, or ten. Like the frog in the proverbial pot, we may sense that things are getting worse but not be impelled to jump.
May 17th, 2010 at 2:23 pm (#)
The way it looks from way down the ladder where I’m at, there are problems with our strategy, but even those are secondary to our biggest problem. That is, that our ability to change anything we do, in accordance with any strategy, is insufficient to the task. Whatever our strategy, however well framed, it will be obsolete in 5 years, and from what I’ve seen 5 years is about the time it takes for an effort to get revved up and going currently.
The end result of this will be, that decisions get made for us, with long periods of status quo interrupted by periodic frenzied and hastily-planned wholesale changes.
May 17th, 2010 at 2:38 pm (#)
I was going to say “very true”, but — being nimble (or not) is a characteristic that would need to be determined by the strategy, no?
I do hope we address strategy but after so many years of seeing it ignored, I’ve become a bit pessimistic.