IT manager approved downtime over lunch, but made a meal of it

What happened

It has been reported that in the early 1990s an IT manager — anonymised as "Hugh" — ran a distribution company’s core app off an SCO server driving serial terminals with a single hard disk and nightly tape backups. RAID was new and pricey then, so he took a calculated risk. The gamble seemed to pay off until the lone disk ran out of space and needed replacing. The replacement initially installed cleanly; the server rebooted several times and appeared fine.

But Hugh wanted to tidy some messy cabling and, it has been reported that, got approval to power down "over the lunch hour" to do the job quickly. He finished the rewire within 60 minutes, hit the power button and watched the machine sit there. Panic set in. It has been reported that when he removed the front cover the drive’s LED was flashing a fault code; powering off and on again did nothing. The manufacturer’s support line was allegedly closed for lunch, and he had to sweat through a long 30-minute wait before learning that a small component inside the drive had failed — probably damaged during the initial install and only discovered when the server was powered down for the tidy-up.

Aftermath and takeaways

A replacement drive was ordered, the company was brought back to life, and — notably — the employer accepted the failure as an unforeseeable, rare event and kept Hugh on. The firm then bought a RAID controller and a second drive. Decades later, it has been reported that Hugh moved into independent work and that the distributor remains a customer — a tidy ending, really.

The emotional core here? The gut-sinking minute when everything just won’t boot. Who among us hasn’t tried to squeeze a risky task into a lunch break and paid for optimism? The story is a neat reminder that redundancy, not bravado, wins the long game — and that “it’ll only take an hour” is a famously dangerous phrase in IT.

Sources: The Register