Many IT professionals have an alarming lack of confidence in their ability to prevent a stock exchange-style collapse in essential systems.
Readers questioned in this week's Computer Weekly/Harvey Nash Big Question poll were almost unanimous in claiming they had official escalation procedures in place. But unofficially many could not be sure they would prevent an unforeseen technical problem.
Earlier this month, the London Stock Exchange suffered technical difficulties that prevented stockbrokers from trading for nearly eight hours on one of the busiest days of the year. The problem was compounded because escalation procedures that were designed to minimise the effect of an outage were ineffective.
Asked if escalation procedures were adequate, one reader said, "Yes, but I am sure the stock exchange thought the same."
Another said, "Yes, although what I really want to say is yes and no. We have well-established and well-rehearsed procedures for escalating problems but getting user departments involved in those rehearsals is almost impossible."
A third reader said, "Yes we have escalation procedures but in the light of recent events they are to be reviewed to ensure that they really would work."
Many readers said escalation and contingency plans were improved as part of efforts to tackled the year 2000 bug. One asked, "Did the stock exchange treat the Y2K exercise as a one-off and throw such plans away?"
Another said, "As part of our year 2000 work we focused on improving the reliability of our 892 applications. My view is we could respond within a reasonable timescale, as due to our European distributed architecture we do not have the visibility of a centralised trading application."
Few readers were completely confident that the testing of their procedures was sufficient to prevent any embarrassing shortcomings.