How feasible is fix-on-failure? (A la Chevron)

greenspun.com : LUSENET : TimeBomb 2000 (Y2000) : One Thread

Chevron's 11/6/1998 SEC 10-Q disclosure admits that they will not be ready for Y2K (as noted on a previous thread). The 10-Q,in its entirety, can be linked to at

Chevron 10-Q

The part of the 10-Q that is of interest to this thread is:

Because of the scope of Chevron's operations, the company believes it is impractical to seek to eliminate all potential Year 2000 problems before they arise. As a result, Chevron expects that its Year 2000 assessment and corrections will include ongoing remedial efforts into the year 2000.


In other words, Chevron is claiming in essence that they will use so-called "fix-on-failure".

It seems to me that fix-on-failure, while it might work under otherwise normal conditions, will be hopeless with Y2K. In normal times, if there is a software failure (for example), the natural thing to ask is "What changed last?", which gives a big clue as to where to look. But with Y2K, what with all the changes due to Y2K code remediation, the answer will be "Nearly everything!".

Also, lets remember that many Y2K problems will not be easily seen failures, but rather just weird problems that defy anyone's ability to grasp from where they might come from within the interdependent layers of complex systems. And this certainly includes embedded systems (e.g., a microchip may not fail, per se, but rather do an improper operation that causes other components to exhibit weird behavior.

-- Jack (jsprat@eld.net), November 11, 1998

Answers

Uh, maybe this would be better formatting for the exerpt from Chevron's 10-Q...

Because of the scope of 
Chevron's operations, the company believes
it is impractical to seek to eliminate all potential Year 2000
problems before they arise. As a result, Chevron expects
that its Year 2000 assessment and corrections will include
ongoing remedial efforts into the year 2000.


-- Jack (jsprat@eld.net), November 11, 1998.

Will Chevron cease to exist?

-- foo@bar (foobar@f.com), November 11, 1998.

Because of the scope of Chevron's operations, the company 
believes
it is impractical to seek to eliminate all potential Year
2000 problems before they arise. As a result, Chevron expects that
its Year 2000 assessment and corrections will include
ongoing remedial efforts into the year 2000.


... 3rd time charming!

-- Jack (jsprat@eld.net), November 11, 1998.

Depends I guess on how long you can afford to fail.

Can you continue to pay the people who are trying to fix the things that have already failed, will fail next, might fail tommorrow, and will fail again because the fix failed the first time....

Sell Chevron. Buy ... horses?

-- Robert A. Cook, P.E. (Kennesaw, GA) (cook.r@csaatl.com), November 11, 1998.


Jack: If you haven't already seen it, make sure you read Ed Yourdon's article on this topic ("The Y2K Fix-On-Failure Strategy") published in Computerworld (May, 1998):

http://www.yourdon. com/articles/9805cw.html

The difference really is whether we're talking about FOF for your date-sensitive screen saver or for your mission-critical, date-sensitive process control / accounting / distribution systems. In my humble opinion, deliberate FOF for the latter rises to the level of criminal negligence. Sorta like playing 'Russian roulette' with the gun pointed at one of your children. Unfortunately, FOF seems to be gaining ground in popularity. ("Great idea John!, It's cheap, easy to implement and things couldn't really be that bad, could they? I've read it's mostly hype anyway. Let's just cross that bridge if and when we come to it. We have much higher priorities at the moment. Plus, it let's us truthfully tell our stockholders we are working on it! Very appealing. Oh yeah, tell the lawyers to put the standard disclaimers in the 10-Q. ")

-Arnie

-- Arnie Rimmer (arnie_rimmer@usa.net), November 11, 1998.



yes, that would explain why there is no hiring boom currently underway for Y2K software people.

-- a (a@a.a), November 11, 1998.

Case in point:

"The U.S. Navy learned the importance of software quality and testing when its Aegis missile cruiser, the USS Yorktown, was put out of commission for several hours due to a software glitch.

"The Yorktown apparently lost control of its propulsion system when an administrator entered "zero" in a data field for the Remote Database Manager program. That action caused the database to overflow and crash all LAN consoles and miniature remote terminal units, according to published reports.

Story at The Enemy Within

(In event of litigation I plead fair use...)

-- Tom Carey (tomcarey@mindspring.com), November 11, 1998.


How feasible? Find a refinery and bring marshmellows 1/2000.

-- R. D..Herring (drherr@erols.com), November 11, 1998.

Feasible to fix the problems? No

Inevitable that many organizations like Chevron will have no choice but FOF for some systems? Yes

Will it work? No

-- Buddy (DC) (buddy@bellatlantic.net), November 11, 1998.


This disclosure by Chevron is to be commended for its candid assessment. Those of us who follow y2k on a day to day basis (for many months) should appreciate this much honesty. We may not like everything that we read but how many companies (or govt. agencies) have provided this much information to date? Chevron will spend between 200 and 300 million dollars this year and next on y2k. Aetna, the largest US health insurer, has stated their 3rd quarter op profits are down due to y2k costs. They have upped their estimate on y2k spending to $195 million for this year and next. Let us hope we see more action and info esp from the utilities.

-- RonBanks (phxbanks@webtv.net), November 11, 1998.


DC Buddy,

Do I detect a slight hint of reality creeping up on you???

-- Nabi Davidson (nabi7@yahoo.com), November 12, 1998.


News flash folks: Chevron is simply telling the truth. Its too big, too complex, and too interdependent for anyone to guarantee compliance regardless how super human the efforts.

For fun, let's suppose your organization was really on top of things and you were 'compliant' as of June of this year. You can sit back,kick your shoes off and pat yourself on the back - right....Wrong.

Why? The follow information was presented at the Wisconsin's Y2K governor's conference and is a matter of public record. ...testing of 4700 PC applications revealed that 28% of those claiming compliance in fact were not compliant ....tests of 500 different BIOS's revealed 93% of pre-1997 were not compliant....

So if you claimed compliance back in June, the odds are that a full quarter of the compliance info you carefully gathered to validate your compliant status is incorrect. And you now get to evaluate your compliance status all over again

There is no current agreed on certification agency or standard for Y2K software or hardware compliance. Definitions of compliance don't agree. (I believe that is how ProveIt 2000 successfully challenged Compaq's compliance claims in England) So the only way to ensure compliance for your organization it to test everything (alone and 'in concert' - we cannot expect these complex system to work flawlessly).

Some electric utilities are making no bones about it, they will not have every computer program in their libraries done. They have classified them by importance -- and they are concentrating on what needs to be done to continue providing service.

I believe they have acknowledged that they plan on doing what has to be done to carry on. The fact of the matter is that a lot of what we do, doens't have to be done for organizations to function.

An honest assessment, the truth, is part of you the information you need to make intelligent decisions.

Good Luck

jh

P.S. Show me a compliant organization, and I'll show an Amish farming community.

-- John Hebert (jhebert@co.waukesha.wi.us), November 12, 1998.


No, Nabi, there just seems to have been a common misconception of my opinion here. I have always believed that Y2K was a very serious problem and that it could wreak havoc. However, I also believe that the collapse of civilization is not inevitable. Awareness is growing exponentially now. Soon, I believe, Y2K will rise to the top of everyone's priority list.

As for Fix-on-failure...it can only work if it is one tactic in an array of Y2K strategic plans. Fix-on-failure is best as part of a well-designed contingency plan.

-- Buddy (DC) (buddy@bellatlantic.net), November 12, 1998.


Moderation questions? read the FAQ