Link

From Computer World for Educational-Research Purposes Only

Y2K's nastiest work

By Ed Yourdon 08/16/99 Y2K contingency planning is hard work, and your business users will need all the help they can get. It's not hard just because it forces people to confront unpleasant scenarios that they would prefer to ignore. It's also hard because the people doing the planning are often the same ones who were involved in the year 2000 remediation. It's hard because people think of Y2K failures as all-or-nothing scenarios. And it's hard because people assume that Y2K problems won't occur until midnight on Dec. 31.

One difficulty involves the people who typically lead the contingency planning effort. Suppose the question is raised: What happens if the billing system is down? The Y2K team member assigned to the finance department is likely to shout, "That can't happen! We've tested it! It's compliant!"

The contingency planning group should assume that all systems could fail; but that's a difficult mind-set for someone who has spent the past year doing everything possible to prevent the system from failing. If the entire team has this kind of mental block, the contingency plan is likely to focus only on the external systems (for example, utilities) over which the team has no direct control. Make sure your team has at least one or two members who were not involved in the remediation effort and who tend to have a pessimistic view of IT systems. Auditors, quality-assurance specialists and security experts are good candidates.

Contingency planning is also difficult because planners sometimes assume that all failures are permanent. But that's rarely true. In some cases, the failure can be repaired in moments (for example, by rebooting a desktop PC); in many cases, it can be repaired within two to three days. Of course, there is the possibility of a monthlong or yearlong disruption, and the contingency planners need to address those scenarios, too.

But the first question is: "Would anyone in the organization even notice if the XYZ system was unavailable for a few minutes?" Chances are they'll shrug. Then you can ask, "How serious would the impact be if the XYZ system was down for a few days? Would we go bankrupt?" In many cases they've already experienced outages of two to three days and have already found ways of coping with the problem. It's only when you reach durations of a few weeks or a few months that users become visibly nervous about the prospects of keeping their business processes operating in some acceptable fashion.

Finally, remember that some contingency-planning scenarios may actually occur before Jan. 1. That means that contingency plans have to be finished substantially before that date and such things as war rooms need to be in place. Y2K project teams are already familiar with some of the "trigger" dates that could cause such problems  for example, the Aug. 22 rollover date for the Global Positioning Satellite system. But consider also the impact of anticipatory policy decisions on the part of business and industry. The U.S. government has announced, for example, that on Oct. 1 it will publish its list of "risky" countries. That may lead to pre-Y2K restrictions on air travel to those nations. Similarly, contingency planners should assume that civil unrest might occur during the final few days of the year. What will you do, for example, if your corporate headquarters is inaccessible because the police have outlawed vehicular traffic into and out of the region? Don't think it can happen? Ask the Broadway theater owners in New York. They may already have been ordered to close their theaters for New Year's Eve.

None of these difficulties is insurmountable, but they do require some pragmatic, "out-of-the-box" thinking. Start now, because it will be too late to do so on Jan. 1.

Yourdon heads the year 2000 service at Cutter Consortium in Arlington, Mass. Contact him at www.yourdon.com.

-- Mike Lang (webflier@erols.com), August 17, 1999

Answers

The U.S. government has announced, for example, that on Oct. 1 it will publish its list of "risky" countries.

I wonder if part of these contingencies is a plan for "flight to quality" (there is an excellent article at http://www.csis.org/html/y2kw4.html)

This has already begun and it will only snowball. If the government is publishing this information October 1 regarding "risky" countries then my guess would be that the flight to quality would become intense as we move toward the end of the year.

Can a company that is not yet ready plan a contingency to keep it's customers, investors, etc. from seeking more stable alternatives? Even if the entity is on track to meet the deadline yet isn't there yet this could become a major problem for them.

Come to think of it...wouldn't it be a great strategy to continually say the US is far ahead of the rest of the world so that foreign investments and businesses would feel comfortable either continuing operations here, bringing investments here, etc?

Mike

=========================================================

-- Michael Taylor (mtdesign3@aol.com), August 17, 1999.

Thanks Ed. Good advice.

Diane

-- Diane J. Squire (sacredspaces@yahoo.com), August 17, 1999.

I wonder if Ed knows the amount of info out on the web on contingency planning, from how to documents, to actual plans to forums and web sites. I also wonder who the audience is for this article. If it's an IT organization as a possible customer for Ed's consulting, then I find it very obtuse and offensive. When Ed puts up these employee scenarios of shouting, "That can't happen! We've tested it! It's compliant!", I find I'm no longer interested in what he has to sell.

FWIW, our billing department didn't say this. They not only had their contingency plans in place but already tried various aspects to see that it would work.

Sometimes I wonder what Ed's thinking.

-- Maria (anon@ymous.com), August 17, 1999.

Maria,

The audience for Computerworld's editorial columns is IT managers; that's who I assume my "audience" is.

As for the members of your billing department, I'm delighted that they're ready and that they've tested their contingency plans. They are definitely in the minority: I've just finished analyzing a survey of Y2K projects, and found that while roughly 80% have begun their continency planning efforts, only 10% have finished.

The example of the emotional reaction from the billing department is a real one, from a recent consulting assignment of mine. I think the point is a valid one: the people who have been doing Y2K remediation have often developed a mind-set that makes them less effective for contingency planning than someone from the disaster-recovery or security area who has paranoid assumptions about EVERYTHING breaking.

Ed

-- Ed Yourdon (HumptyDumptyY2K@yourdon.com), August 17, 1999.