I was on the CNN website and noticed this little gem:

http://cnn.com/TECH/computing/9903/30y2k.critical.ap/

It turns out that, in order to give possitive results for mission critical testing, they have decided to shorten the mission critical list. Rather scary!

-- Anonymous, March 31, 1999

Answers

Sean, to see an example of redefining critical systems in a utility report, read the comments to the thread "Conectiv's Recent SEC 10Q". While Conectiv is still reporting percentages based on their original critical systems, they have redefined a subset as "essential" systems and work has been prioritized to those.

-- Anonymous, March 31, 1999

Sean,

The referenced link is not working, but let me share a (blind) thought on this.

I assume that it is a fact that "they have decided to shorten the mission critical list".

What I challenge is the stated reason "in order to give positive results for mission critical testing". My guess is that someone made this part up, because that is the type of "spin" they wanted to give this message.

Let me give you a much more common sense, and actually observed reason why this factual event might have occurred.

Everyone likes to think their job is critically important, maybe because no one wants to be considered indispensable. As a result, every department has called their most important systems as mission critical. And maybe for that department, it is mission critical. But when the department at a electric utility that produces the internal newsletter calls their software mission critical, someone will eventually catch up with that and say no, lets work on the plant systems first.

It is my guess that this is all that has occurred, but the hype around Y2K has turned this into a conspiracy plot.

bob

-- Anonymous, March 31, 1999

I think Bob's observation is correct. In my experience, system criticality is determined initially in a decentralized fashion using a survey. Each respondent is asked to make a subjective determination of criticality, usually along a three-point scale. For the reason Bob stated, these initial criticality rankings tend to be inflated.

Later, the initial rankings are re-evaluated centrally and (sometimes) more objective criteria are applied, such as potential revenue loss, life safety impact, regulatory requirements, or effect on the firm's reputation. This process results in a shorter list of "truly mission critical" systems.

The re-evaluation process is usually subect to review by a body of senior managers constituted as a "risk management committee" or "y2k steering committee". They approve the final list of "truly mission critical" systems and any further prioritization within that list.

The shrinkage in the number of mission-critical systems being reported by almost every organization is largely the result of this process. It is not an attempt to hide real problems.

-- Anonymous, April 01, 1999

So, Peter and Bob are saying that the process of identifying which systems are "mission-critical" continues, lo, even until today (4/1/99)! Very interesting!

Perhaps the earlier reports should have referred to these systems as "candidate critical systems," because top management had not yet reviewed the list! But, interestingly, top management signed off on the earlier reports to Congress that included the higher numbers of "mission-critical" systems. Hmmmm. I don't think the bureaucrats can have it both ways!

-- Anonymous, April 01, 1999

"It is not an attempt to hide real problems." (Peter Nicholson)

Given the real-world constraints of estimated quarterly earnings, stock price, managerial reluctance to look bad, it seems unlikely that "real problems" will be broadcast to the world. Best that can be hoped is that any "real problems" are being found and successfully addressed.

-- Anonymous, April 01, 1999