<I think it's important to keep in mind what is the most likely result when a Y2K test fails: the test is aborted, and the testers simply return to the old system.>
No, that's not the most likely result of a Y2K test failure. Actually that's the best thing that most could hope for.
The most likely scenario, is that wrong info is generated. Or, wrong actions occur, because of wrong information being generated. BUT, you don't find out until down the road, instead of immediately.
Oxford Health Care computer problems are a good example. It took almost a year before they realized wrong financial information was being generated. Stock kept on rising because of great earnings. In actuality, for almost a year - they were losing money. Stock tanked big time, when it finally came to light. Patients weren't hurt. Stockholders were.
Finding out later, after incorrect information is generated and built upon, and calculations made based on that incorrect data ... makes a fix much more difficult, and harder to find.
Plus incorrect decisions could have been made, based on wrong information during the interim.
The biggest problem I saw with CBOT and MCI fiasco, was that CBOT had no back-up phone carrier (like some ATM systems) in the event of network failure. That still amazes me.
Fortunately, in this case, it was pretty much a mechnical failure. [As far as we know.] Data couldn't be transmitted.
If incorrect financial data were being transmitted ... now that's another ... and potentially worse, problem.
Cheryl |
