Y2K: DOD leaving nothing to chance during leap year transition periodgreenspun.com : LUSENET : Grassroots Information Coordination Center (GICC) : One Thread
Y2K: Looking Ahead, Looking Back by Paul Stone
Washington - January 21, 2000
- After more than a year of intense preparation -- as well as media hype and occasional hysterics -- for a millennium bug that, in the end, barely took a nibble out of key information systems throughout the world, it might be hard to do little more than yawn as the next Y2K deadline draws near. - But DoD -- just like it did during the Year 2000 rollover -- is leaving nothing to chance for the leap year transition period at the end of February.
Bill Curtis, principal director for DoD's Y2K repair effort during the past two years, said the Pentagon's Y2K cell will be fully operational from the end of February through the first few days of March to monitor and fix any glitches that may occur during the period.
The Y2K leap year problem results from the fact software developers long ago failed to add the extra day into their programs. Normally, "century years" -- 1800, 1900 -- are not leap years. However, when a century is divisible by 400, as well as by 100 -- such as is the case with the year 2000 -- it is a leap year. Therefore, unless systems have been repaired, it's likely they will not recognize the leap year. Curtis said Y2K tests conducted during the past 18 months uncovered almost as many glitches caused by the leap year problem as they did from the Year 2000 rollover. But he remains confident that the end of February will just as much a non-event for DoD as Jan. 1.
"I don't believe we're going to see a lot happen over the leap year," Curtis said. "We had things show up during our [Y2K] testing that focused people on the issue. I suspect the period will go very much like the first of January did. The key thing is, if there are problems, our first team will be right there watching. And if anything needs to be fixed, we'll fix it."
Except for a satellite-based intelligence system that experienced a Y2K failure and was inoperable for a few hours, DoD experienced only a few minor Y2K failures on Jan. 1, and they did not affect military operations or readiness.
Although the leap year problem is seen as the last major Y2K battle, Curtis said DoD will watching throughout 2000 to catch problems that may pop up down the road.
"We have to stay vigilant about our data bases," he said. "Very few routines in computer code are actually executed in any given time. Some occur every week, some every month, some quarterly, and some are yearly updates. So we will likely experience minor problems at various points throughout the year." While getting through 2000 without major Y2K problems is a key goal, the Pentagon will be dealing with Y2K leftovers for years to come. Curtis explained that many of the system repairs made during the past two years were only temporary fixes, involving a technique called "windowing."
Windowing is a programming technique that enables software to recognize four-digit year fields instead of just two-digit fields, and thus allowing information technology users to temporarily avoid the Y2K problem. For example, a typical windowing fix would reconfigure software so that years entered as 00-29 are assumed to represent 2000 through 2029, and years entered as 30-99 represent 1930 through 1999. Essentially, this delays the need for a permanent Y2K fix until the year 2029. The technique has been widely used in both government and private industry.
The technique -- at least in the short run, saved the Pentagon time, money and ensured continued reliability of systems into the millennium," Curtis said. Some systems were "windowed" for a five-year period, while others go out as many as 50 years.
"Had we tried to do permanent fixes on everyone's systems during the past few years there would have been a huge information technology traffic jam," Curtis said. "Now we can go back and make the necessary modifications in a coordinated way without impacting interaction between one system and another."
DoD spent approximately $3.6 billion during its four-year Y2K repair and testing effort -- an investment of time and money Curtis said will pay DoD big dividends for years to come.
Looking back on his role as one of DoD's key Y2K managers, he said Curtis said the military has learned valuable lessons that will help the department manage information technology in the future.
During the course of the Y2K challenge, he said leaders at all levels came to appreciate the military's dependency on information technology.
"We fixed a lot of infrastructure and an awful lot of computer code got cleaned up," Curtis said. "We've gone into the year 2000 with a much better set of systems than we had before and a far better system for maintaining them."
Other benefits he cited included: A clear understanding of what systems are systems are vulnerable to computer hackers and how to better protect them in the future. Development of models to manage and track the use of information technology throughout DoD. Better working relationships with both federal agencies and foreign nations -- all of which DoD worked closely with to ensure Y2K did not impact either U.S. or overseas installations.
"It was a tremendous effort and we've all learned a great deal from the experience," he said. I owe a great deal of thanks to those who led the way, from our top leaders on down to those who were fixing the problems in systems throughout DoD. They're the real heroes of Y2K."
Link to story:
-- Carl Jenkins (Somewherepress@aol.com), January 24, 2000