Hamasaki: With Y2K tests like this, who needs actual failures?

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

Subject:Re: [VAN NUYS] Y2K Test Causes Huge Sewage Spill
Author:cory hamasaki <kiyoinc@ibm.XOUT.net>
  Posting History Post Reply

On Sat, 19 Jun 1999 03:42:04, dtmiller@nevia.net (Dean T. Miller) wrote:
> Ah, but they do.  For example, suppose the Van Nuys system hadn't been
> tested prior to rollover.  Wouldn't it have experienced the same
> malfunction we've seen at rollover (assuming the same input
> conditions)?
Saaaay, check me on this pollys but haven't there been relatively few real tests, Peach Bottom 2, Diane Shield's launch controls, Wendall Ito's Hawaiian Electric and now, Van Nuys?
And the results are doomie.
I'm thinking about the good news from the Wall Street tests where everthing went fine (except for a teeny tiny percentage of easily fixed, not very important transactions.)
Isn't a difference that in the case of Van Nuys, the affect was hard to ignore.  That 99.999 percent of the system worked exactly right.  I'd even venture that there's a chance that the Van Nuys problem might not have surfaced in the real event.
> The light it sheds is that systems can malfunction in strange and
> mysterious ways when operated outside their design parameters -- and
> for too many systems, a date past 12/31/99 is outside.
The catch is, even if most sewage plants roll over correctly, I suspect most will because that industry is certainly abuzz with this minor spill in the news, there will be lots of other industries where 99.999 percent of everything in the plant worked fine, except the one thing that causes it to dump a million gallons of pestacide, halogen gas, whatever, onto the street.
And mull this over, "sewage park" was supposed to be filled with happy millennium partyers, perhaps listening to concerts, picnicing, waiting for midnight fireworks, stretched out on blankets, copping a feel, reaching for another wine cooler.... several hundred yards away, bzzzzz-click, a gate closes, skish-skish-skish, a pump keeps running.
Hey, what's that over there?
Run for it.
Where are the kids?
Yeah so maybe, I *will* take a few days off from cranking code in DeeCee.  Maybe I'll be far from Hill 'n Bill's fun-fest on the mall.
cory hamasaki http://www.kiyoinc.com/current.html

-- a (a@a.a), June 20, 1999


Sewage Park, ready to roll ;^) LA needs a wake-up call; doubt this was enough to get the herd sniffing around much. Hope it prompts other treatment plants --> serious about remediation and testing and getting ready. Man the holes!

-- Ashton & Leska in Cascadia (allaha@earthlink.net), June 20, 1999.

"skish-skish-skish" AAAAAAAHAHAHAHOOO. geeeese, thanks Cory. I needed that. pheeewwwwww Okie-Dokie, I'm good for another 24 hours anyway. Thanks for the laugh a, you firm hunky punk!

-- Will continue (farming@home.com), June 20, 1999.

The guy can type, no doubt there. Apres le deluge, Cory should sit down and write a book, maybe a cross between Nelson Demille and Elmore Leonard. Or maybe not Leonard: Pynchon. Gravity's Donut. Yeah.

-- Spidey (in@jam.com), June 20, 1999.

Yes, the Van Nuys test failed spectacularly, for reasons yet to be determined. It may have been caused by a genuine y2k bug, and it may have been an artifact of a poorly constructed test. We may know soon.

And yes, failures like this can be expected come rollover. Probably quite a few of them. It's hard to imagine this won't happen, given the scope of opportunity for them.

How many such failures will we see? Nobody can know this, but we can try to extrapolate from what we've seen so far. What makes this extrapolation difficult is that spectacular failures get lots of ink. Successful tests are business as usual, unless the media are called in to generate a PR event. Otherwise, we read nothing, because nothing is written. Reported news events are a *long* way from a random sample of what's going on.

-- Flint (flintc@mindspring.com), June 20, 1999.

Flint: you are a *long* way from a random sample of common sense. I suppose the majority are willing to tolerate you, however....just for the fun of it! Hope you've hoarded wet-wipes!

-- Will continue (farming@home.com), June 20, 1999.

I agree with you there Flint.

Poor testing setup, poor test controls, poor "what if" analysis before the test was conducted all contributed.

Problem is, there remain (hundreds, thousands, millions ?) of equally likely (more likely) similar "real" problems that are going to occur.

Damn - I hate being right - told you automated discharge valves at sewer plants are likely to fail back when Aston and Leska first asked the question.

-- Robert A. Cook, PE (Kennesaw, GA) (cook.r@csaatl.com), June 20, 1999.


Let's hope a celebrity failure like this lights a fire under a few responsible butts. At least now they know at least *one* failure mode to examine very carefully.

-- Flint (flintc@mindspring.com), June 20, 1999.

Will Continue:

Anyone who has looked at polling theory, and/or the portion of applied sadistics(statistics) that deals with sampling for QC/prediction will point out that the news media, in covering marquee failures, does NOT use a statistically valid sample base. Nor will a collection of reported failures be representative of the universe of, say, sewage plants. Nor will it NOT be predictive. One SIMPLY can NOT judge from the information presented.

Chuck, who suffered through Engineering Sadistics and then had to go back 3 years later and take a "Management" version.

-- Chuck, a night driver (rienzoo@en.com), June 20, 1999.

In the case of Y2K, would you consider successful tests to be considered "business as usual"?

-- Will continue (farming@home.com), June 20, 1999.


I hope this is a big wake up call for "my" city. However, yesterday a Y2k test was canceled because of fear that the SaMo Bay would be filled with millions of gallons of sewage. This demonstrates a serious lack in confidence of the city officials. I'm not sure how much you know about the area but it would have seriously impacted quite a bit of business along the coastline for weeks if not months...right into summertime. Also, Southern California is a HUGE area with thousands of possible points of failure in everything from water to sewage to electricity...chemical plants, oil refineries...there is NO WAY that we wont suffer serious disruptions.

I think LA might be the first big city in the US to get really serious really quick about warning the public. Let's keep our fingers crossed.

Mike =========================================================

-- Michael Taylor (mtdesign3@aol.com), June 20, 1999.


I was afraid of something like that. I don't think it represents a lack of confidence so much as a lack of political willingness to take the heat if anything should go wrong. Unfortunately, while cases like Van Nuys might cause some others to take their testing more seriously and be more comprehensive and meticulous about it, I fear that in all too many cases, the wrong lesson will be learned -- "testing is risky. Don't take risks. Therefore, don't test." I guess we'll just have to see how irresponsible our elected leaders turn out to be. I'm not very optimistic. They didn't get where they are by sticking their necks out when empty reassurances that everything will be fine sound good and risk nothing (yet).

-- Flint (flintc@mindspring.com), June 20, 1999.

Moderation questions? read the FAQ