With Y2K looming this is not good news:

http://www.ncpa.org/pd/budget/pd091499a.html

Analysts report that the well-being of a major sector of the U.S. economy hinges on an aging and overworked computer located in Newington, Va., a quiet Washington, D.C., suburb. The computer's assigned task is to count imports coming into the U.S. But the system is 15 years old and with imports soaring it stands near collapse, experts warn.

If the computer breaks down for more than a few hours, goods from abroad will start stacking up at the U.S. border. U.S. companies that rely on just-in-time deliveries might have to shut down production. Economists say the ripple effect could cost the U.S. billions of dollars.

When the computer -- known as ACS for Automated Commercial System -- was first brought on line, U.S. imports stood at $332 billion annually, but are close to $1 trillion now.

The first sign of trouble came in September 1998, causing a six hour shut down that threw the U.S. Customs Service three weeks behind schedule.

After a second glitch a little more than two weeks later, the computer muddled along for a two-day period.

A new system would cost around $1.5 billion -- but the White House, Customs and Congress have reportedly dragged their feet on spending the money.

Several months ago, Customs ran a simulated computer shutdown test in the ports of Charleston, S.C., and Savannah, Ga. Instead of using the computer, officials counted using pens and paper. The result was that backlogs exploded after just six hours.

After a theoretical 30-day shut down, cargoes would have been delayed by as long as eight days in port. Manufacturers would have been deprived of vitally important parts and agriculture imports would have rotted.

Source: Daniel J. Murphy, "A Computer-Age Trade Meltdown?" Investor's Business Daily, September 14, 1999.

For more on Commerce Dept. http://www.ncpa.org/pd/budget/budget-7.html

-- helium (heliumavid@yahoo.com), September 15, 1999

Answers

Fixing this just requires regulation to be waived. Not even the U.S. government is stupid enough to strange the economy because imports can't be recorded through customs properly.

There are going to be real failures to worry about. This kind of failure is nothing.

-- You Know... (notme@nothere.com), September 15, 1999.

The big picture....A real good memory and hundreds of articles like this,make me a GI

-- Eric michael (bizarr2@hotmail.com), September 15, 1999.

Quoting from the reply above:

<< Not even the U.S. government is stupid enough to strange (sic) the economy because imports can't be recorded through customs properly.

There are going to be real failures to worry about. This kind of failure is nothing. >>

__

Wrong, dead wrong. The typical federal bureacrats are exactly just stubborn enough to strangle the econmy by (1) not allowing imports to clear customs (ie, get pay tariff's and taxes) without using this computer; (2) not consider this computer a "critical mission" and so leave it unremediated (no word in the story about what kind of computer or whether it (or its databases) are y2k compliant!) (3) "rule-bidden" so far as to attempt to do things by hand.

However, if this computer is as described, they may need to "go by hand" - which obviously failed - for weeks, not hours or days.

Another reason to follow imports and JIT supplies not just during the 1st week, nor even the first 4 weeks, but rather across a long transition period before youi can begin to feel optimisitc.

-- Robert A. Cook, PE (Marietta, GA) (cook.r@csaatl.com), September 15, 1999.

Something wrong with the math here:

The first sign of trouble came in September 1998, causing a six hour shut down that threw the U.S. Customs Service three weeks behind schedule.

Several months ago, Customs ran a simulated computer shutdown test in the ports of Charleston, S.C., and Savannah, Ga. Instead of using the computer, officials counted using pens and paper. The result was that backlogs exploded after just six hours.

Note... no mention of Y2K. So it is probably just a coincidence that the Sept. 1998 "glitch" happened in a key deadline month for Y2K readiness. And just a coincidence that "several months ago" they decided to test manual work arounds and the computers were out for the same period of time.. six hours.

Then the funny math:

After a theoretical 30-day shut down, cargoes would have been delayed by as long as eight days in port.

A real 6 hour outage caused a 3 week backlog.

By my math a 30-day shutdown (based on only a 8 hour work day.. is that reasonable for customs?) would = an 840 day backlog (over 2 years).

-- Linda (lwmb@psln.com), September 15, 1999.

Linda, you are using your thinking cap again. Now stop that! Trust the written word. Trust the author, the government. Think nice thoughts and be a nice compliant person.

Fine job, sis. Keep it up. Only think 'exponentially' not 'arithmaticly'. Those ships will also be delayed on return routes, which will stack up warehouses in exporting ports, which will shut down production there, which will throw the exporter nations into a financial tail spin, etc. It ain't so much to be so pretty, that.

-- ..- (dit@dot.dash), September 15, 1999.

Well, it may blow up for some other reason, but apparently they have completed Y2K remediation about a year ago (10/98):

http://www.customs.ustreas.gov/about/y2k/mcs.htm

But man, their web site is screwed up. Lots of outdated links.

-- Lewis (aslanshow@yahoo.com), September 15, 1999.

Well then, apparently the Sept 98 "manual" testing was part of their y2k tests - if they are claiming as of Oct 98 everything was complete.

But certainly, "manual backup methods" couldn't do the job. So, lets hope the "primary", the one, the only other system works next January.

-- Robert A. Cook, PE (Marietta, GA) (cook.r@csaatl.com), September 15, 1999.