Ah, night shift in the 1970s. Ciggies, hipflasks, ADVENT... and fault-prone disk drives the size of washing machines

Starting from scratch. A really big scratch

An illustration of a mid-century mainframe

Who, Me? Welcome back to Who, Me?, The Register's regular ramble into the dark recesses of readers' memories where we prod consciences with a long, sharp stick.

Today's story, courtesy of "John", takes us back over 40 years to the end of the 1970s – a time of big iron and even bigger server rooms.

John was working in a factory that manufactured terminals for a mainframe vendor. The IT department in which he was a student intern looked after two hunks of PDP-10 iron – one smaller (in mainframe terms) machine used for testing and another heftier beast to deal with the day-to-day production tasks.

Keeping the latter computer ticking over was critical since unexpected downtime meant the 2,000 or so workers in the factory would be left with thumbs twiddling. "This," said John, "was considered a Bad Thing and Never to Be Allowed Upon Pain of Pain."

Lightning

Take DOS, stir in some Netware, add a bit of Windows and... it's ALIIIIVE!

READ MORE

The computers were big old beasts, filling the floorspace of server rooms and requiring disk drives that resembled top-loading washing machines in both appearance and size. "The disk packs themselves were stacks of aluminum platters," recalled John, "that looked like wedding cakes in their smoked plastic covers."

A "string" of around eight drives would be connected to a disk controller cabinet. A mainframe could have one or more controller cabinets. Each of these "washing machines" could hold an eye-popping 176 MB.

"A typical mainframe installation would have rows and rows of 'washing machines'," he explained, "and dedicated people called 'operators' who would mount tapes, switch disk packs and run batch jobs according to 'Run books'.

"Run books were essentially programs followed by humans in order to make things happen."

The smaller test mainframe had only a single controller with six drives. The production machine had three controllers, each with a row of eight RP06 drives attached. It must have been quite a sight.

While different batch jobs (such as payroll or automation) had their own disk packs, it was the night shift that operators looked forward to. After the last tape backup was done, the gang could reach for that special games disk pack and indulge themselves in some ADVENT, CHESS or whatever delight lurked on the most recent DECUS tape.

It is important to note at this point that those drives and disk packs were not sealed, and it was possible for dust or hair to fall into the drive "tub" when packs were changed. This wasn't good since the drive heads "floated" approximately 1/100th of the thickness of a hair above the platter during operation.

Encountering grot would cause a "head crash" and send the heads skittering, borking itself and damaging the platter.

To mitigate this, an air filtration system ran as the platters span up. A minute later, after filtration had done its stuff, the heads would extend and the fun would start.

On the night in question, the senior operator decided that the day was done and games needed to be played. While he popped out for a cigarette and restorative nip from a hipflask (different times), he instructed his junior to pop the games disk pack into the test mainframe.

The junior did so, dropping the pack into drive T (for test) 05. But as it began to spin up, the drive emitted a plaintive beep, showed the FAULT light, and span back down.

"Odd," thought the dutiful and very new operator. He tried putting the pack into drive T04. Same beep, same FAULT light. He tried T03, which also failed.

The junior, keen to ensure playtime was ready for the boss's return, was confused. "The odds of three drives failing at the same time was inconceivable," said John. So it had to be the test mainframe's single disk controller, right?

While you or I might have called a halt at this point, the junior was determined those games would be ready. So he popped the pack into the P12 drive on the production mainframe. Another fault. The same with P11.

Baffled, and thinking that perhaps production might also have a failed disk controller (and keen to demonstrate his ability to both persevere and follow orders), he tried drive P05, which also promptly faulted.

"By the time the lead operator came back from his smoke/tequila break, the junior operator had destroyed the heads in six very expensive disk drives," remembered John.

What had actually happened was the first head crash had sent the heads skittering into the platter, leaving a visible dent that looked like someone had stabbed the poor thing with a screwdriver.

"Each time he moved the pack to a new drive, the heads quickly crashed into the (to them) Himalayan-sized mountain of aluminum, damaging another set of read/write heads and incidentally spraying oxide dust throughout the drive mechanism itself."

With immediate repair not an option, the gang had to put together a functioning storage system out of what was left to avoid a surprise day off for the workforce.

As the student intern, John was tasked (along with the junior operator) with reconfiguring the surviving drives, which involved an overnight crawl through the underfloor cavity, dragging heavy "anaconda" cables to connect the survivors.

It was a long, dirty job and a million miles from the gleaming control panels and flashing lights of the mainframes of fiction. But by 8am it was done and production could get started as though nothing had happened.

Something had happened, of course, and a field engineer was called to deal with the six sets of borked drive heads. Fortunately for all concerned, he was happy to join the conspiracy of silence, having seen a similar thing happen at another facility the previous year.

"They had lost 10 drives to a zealous student operator trying to load a drive full of ASCII pr0n pictures."

As for what happened next? The drive heads were due for replacement in six weeks anyway, so the field engineer chalked up the failures to a faulty AC filter (after taking said filter outside and smacking a shrub with it).

"The junior operator was sworn to secrecy and paid hefty bar tabs for all involved for six months," recalled John, "and the Intern was promised a good grade and evaluation."

Over 40 years on and John still remembers the important lessons learned: "Dust is bad, stop and think, and know when to call for help."

"And most importantly, how to keep a secret."

Ever sacrificed production hardware in your desire for gaming action? Or had your own Die Hard moment in a server room crawlspace? You have? Then share the story via an email to Who, Me?. ®

Sponsored: Detecting cyber attacks as a small to medium business

SUBSCRIBE TO OUR WEEKLY TECH NEWSLETTER


Biting the hand that feeds IT © 1998–2020