Horror stories
The Man With No Brain
state machine development on a PDP 11 under RSX11. Seems time for the
annual preventitive maintanance came around, and one of the tests is to
ensure the drive can read and write correctly to each and every block of
a disk. The DEC field service tech looked at our rack of disks, saw one
labeled 'Jay's scratch', and decided to use that for a disk. Well, you
know engineers. A disk is a scratch disk until you put something you need
on it, at which time it is the working disk. You also know engineers never
re-name a disk once it has a label on it. Jay comes in the next day,
mounts his disk, and reads out a bunch of E5's. Seems he lost about 3
months of work, only some of which he had listings of. I think the
field service rep also caught hell for doing that to a customer's disk without
asking anyone.
K.Singhal - Systems Design
Well, back a few years ago on a work term in Victoria, B.C., A friend of
mine who was spending his work term at BC Systems told me about what an operator
had done one day. It seems that the machines were normally shut down in the
wee hours of the night and brought back up by the first shift of operators
who came in in the morning. One morning the operators were not to boot up
the system until after some maintenance work was to be done on the drives.
The operator who was on that shift that morning hovever had somehow
forgotten about the scheduled maintenance and tried to boot the system.
It seems he lost his job over the ruined disks.
--
Sherman Lang
Systems Design Engineering
University of Waterloo
Ed Falk
PDP-11/04. that's right, "04". The machine was a quad-width unibus and
the cpu was all low-level ttl. It had a cassette drive, some D/A converters
and one of those big disk drives that you don't often see attached to a
microprocessor. Anyway, the computer-knurds who joined the lab so that they
could play with the computer could never get the idea that you have to
be careful with computer hardware. They were always doing things like
pulling boards out or putting them in when the power was on. At one time
or another, they blew every single board on the bus. The best move they
pulled was when one of them plugged the CPU board, which was only a
dual-width board in the MIDDLE of the bus (not on one side or the other
which would have been ok). As soon as the power was turned on, the
CPU board was blown to kingdom-come. The computer knurds managed to
identify all the bad chips and replace them -- except for the microcode
rom. When they called DEC, they were told that the '04 was out of production
and that there was no way they could get a replacement. What DEC did do
for them was to send an octal dump of the microcode to them. They wound
up in the tech room where I worked with a prom chip and a twelve volt
power supply burning a new chip by hand, bit by bit.
By the way, here's two practical jokes that we thought of but refrained
from pulling (we didn't need to, the computer knurds were doing it for us):
1) go to the big disk drive, remove the pack, and place damp laundry
all around the perimeter.
2) pull out any board, and spray clear Krylon across the fingers. They'd
never have debugged that one.
-ed falk, sun microsystems
Chuq Von Rospach
had an 11/34 with a couple of CDC Hawk 80 meg drives (one of the most
reliable washtubs I've ever worked with). One day, the CS guy was out doing
PM on the drive, and I walked into the machine room to find him smoking
while doing the PM with the drive wide open to the world. He was immediately
escorted out of the machine room and out of the building, and his boss
was called and told that he could send someone else out to finish the PM
and pick up the guy's tools (which we neglected to let him take). I don't
know who was more pissed -- us or his (soon to be ex) boss...
chuq
--
:From catacombs of Castle Tarot: Chuq Von Rospach
ch...@sun.ARPA FidoNet: 125/84
{decwrl,decvax,hplabs,ihnp4,pyramid,seismo,ucbvax}!sun!chuq
Somehow, Toto, I don't think we're in Kansas anymore...
Ross Greenberg
some strange noises. He fiddled with the latch on the cover so it
could spin up to speed with the cover open. He leaned over the spinning
disk to listen.
Then his tie got caught on the *outside* edge of the spinning pack.......
--
------
ross m. greenberg
ihnp4!allegra!phri!sysdes!greenber
[phri rarely makes a guest-account user a spokesperson. Especially not me.]
Jeff Siegal
>[...]. He fiddled with the latch on the cover so it
>could spin up to speed with the cover open. He leaned over the spinning
>disk to listen.
>
>Then his tie got caught on the *outside* edge of the spinning pack.......
Well, what happened!!! 8-)
Jeff Siegal - MIT EECS
Charlie Sorsby
> it to take them out and scrub off the contact fingers with a rubber eraser,
> first. It's amazing how often that works -- the older the equipment, the
> more often it is effective.
It also, as I recall, may void the warranty on boards with gold plated
contact fingers. At the very least I have read service manual that said
"don't do it" because it removes the thin gold plating.
--
The opinions expressed are not necessarily those of my employer,
the government or your favorite deity.
Charlie Sorsby
...!{cmcl2,ihnp4,...}!lanl!crs
c...@lanl.arpa
Tim Bessie
was working as an operator at MIT.
The system they were using had recently been converted to using a new
type of coated fiberglass disk, to replace the old, heavy metal-platter
kind. No problem there. Well, the system they had this 'Emergency Stop'
plug on it that you would pull when an emergency occurred (they assumed
it was for, say, a flood in the machine room). One late evening, a couple
of the operators were sitting around being bored, and decided to see what
would happen when they pulled 'Emergency Stop.' Immediately after pulling
it, they heard a strange sound in the disk cabinet. Looking over, they
saw an arm emerge from the side of the cabinet, on either side of a platter,
and CLAMP down on the platter. Apparently, this wasn't made for use with
fiberglass platters.
They were picking splinters out of the walls for days.
- Tim
PS. Can anyone tell me what kind of system this was? I don't remember.
--
---
Life is like an onion; you keep peeling off the layers,
and sometimes you weep.
- Carl Sandburg
---------------------------------------------------------------
---> Tim Bessie ----- {ucbvax,dual}!unisoft!tim
---> Unisoft Systems; 739 Allston Way; Berkeley, CA 94710
---> (415) 644-1230 TWX II 910 366-2145
Vicarious Oyster
>some strange noises. He fiddled with the latch on the cover so it
>could spin up to speed with the cover open. He leaned over the spinning
>disk to listen.
>
>Then his tie got caught on the *outside* edge of the spinning pack.......
>
unless they go into management; they don't do any real work then, so don't
have to worry about it).
--
- Joel ({allegra,ihnp4,seismo}!uwvax!uwmacc!oyster)
P.S. If my management reads this, I didn't post it...
Henry Spencer
> >...Then his tie got caught on the *outside* edge of the spinning pack...
>
> I *knew* there was a reason computer people don't wear ties ...
If you visit a well-run machine shop, you will note that the machinists do
not wear ties, long sleeves, loose jackets, rings, watches, or gloves while
working in the vicinity of rotating machinery.
--
Henry Spencer @ U of Toronto Zoology
{allegra,ihnp4,linus,decvax}!utzoo!henry
Tom Lowe
we had a Southwest Technical Products UnixFLEX system which was
very similar to a S100 computer. Anyway, we were having intermittent
problems with the thing, and one of the techs there tried cleaning the
card edges. First with an eraser, which didn't work, then with a couple
other things, which didn't work...
Then with ........ *steel wool*!
Well, the card edges were nice and clean -- too bad it shorted out some
of the contacts! We spent hours looking at each of the cards (about a
dozen or so) under stereoscopes looking for the ultra-fine steel wool
particles. Fortunately, nothing got damaged.
Ugggggghhhhhhhhh..............
--
Tom Lowe
uucp: {allegra, ihnp4}!psuvax1!gondor!lowe
Bitnet: LO...@PSUVAXG.BITNET
Carl Greenberg
an ant walked across one of the disk drive heads as it was stepping.. Smeared
the disk, the drive wasn't too good either, and the board was down for several
weeks..
Carl Greenberg
Philip E. OKunewick
>> >...He leaned over the spinning disk to listen.
>> >...Then his tie got caught on the *outside* edge of the spinning pack...
>>
>> I *knew* there was a reason computer people don't wear ties ...
>
>If you visit a well-run machine shop, you will note that the machinists do
>not wear ties, long sleeves, loose jackets, rings, watches, or gloves while
>working in the vicinity of rotating machinery.
The problem with being a computer tech is that you're supposed to
look like a white collar professional when you arrive at the
customer's site. I ruined more good clothes this way.
When I first started repairing computers, we were required to wear a
tie. Every time I opened a CDC Hawk drive to clean the fixed platter
(We really DID clean them with alcohol & gauze - standard PM) I would
always go through a little ritual of checking my glasses, my watch,
and my tie to make sure they were on tight, and especially make sure
the tie was clear of the drive. One slipped object, and say bye-bye
to a $30 platter and hello to an extra service call.
All the techs were happy when the tie policy was dropped.
---Duck
MKR
This story did not happen to me, and I disremember where I heard it,
so it may not be true, but it's interesting nonetheless, so...
There was a computer system that was experiencing intermittent power
failures that were proving impossible to track down. Every means of recording
device and electrical filter was used, but to no avail. The power failures
always seemed to happen soon after lunch time, but for no apparent reason.
After months of agonizing work, the technician finally figured it out:
The room on the other side of the wall from the computer room was
the men's bathroom. The grounding for the computer room circuits went to
the water pipes that serviced one of the toilets. The building was rather
old, and the toilets were in some need of repair. It seems that when one
sat on the toilet seat, the weight of the sittee would cause the whole
construction to lean forward a bit - not much, but enough to cause the
marginally attached grounding wires to separate from the water pipes as
the pipes bent along with the toilet - voila - the computer re-boots.
I bet that was a hard one to track down!
--
--MKR
"The majority of the stupid is invincible and guaranteed for all time. The
terror of their tyranny, however, is alleviated by their lack of consistency."
- Albert Einstein
David E. Marquardt
This was told to me by a fellow co-worker who worked for another
large main frame manufacture previously.
It seems they delivered a new machine to an overseas site and
during installation every time they applied power to one of the
memory bays they blew every circuit breaker in the computer room.
After reseting the circuits they again applied power to the memory
bay with the same results. Since this was a new machine they
crated it up and shipped it back and got a replacement.
When they got the damaged memory bay back the started to tear it
done to fine the cause of the short. Well what they found was a
small hole about 3/8 in. in diameter going from top to bottom
through some of the memory arrays, which cause a very effective
short. After a lot of research they found the cause, it seems
that after the memory had passed test and evaluation and quality
assurance the bay was crated and put in the warehouse to await
delivery. At some time during its storage an electrician was
hired to do some work and since it was a secure building the
security guard had do go with him. The electrician at one
point said that he had to go back down to his truck to get a
drill and the guard asked why and the electrician said he needed
to drill a hole right here (pointing to a spot on the floor).
The guard then responded by pulling out his sidearm and proceeded
to blow a hole at the appropriate spot which happened to be right
above where the memory bay was being stored.
The last he knew the guard had been reprimanded and re-assigned
to another of the security agency's customers.
Dave Marquardt
sdcrdcf!dem
Gilles Chartrand
had a serious system crash lasting over a day on our AMDAHL 470 V8
(might have been a V6 back then). One day, after the computer was
taken down for PM, there was no file-system left on the disks (3381's I
believe)... What happened was that the computing facility had been
expanded without reguard to the capacity of the air conditioning. The
computer had been left on for some time and in the process, the disks'
internal temperature had far exceeded the specifications. Everything
worked fine, as long as the disks remained "hot". When they were
turned off, they contracted sufficiently to scramble the information on
them... The short term fix was to run with the cabinet doors open
(until adequate air-conditioning was available).
I friend of mine was a system programmer on the above machine. His
current task was to install a new "paging box" (a device containing
vasts amounts of slow ram and looking like a VERY fast drum to
the AMDAHL). He had the computer to himself and, to impose a load on
the paging system, he would start up 10 tasks which would proceed to
get 10 megs of memory and loop around touching the first byte of every
page it had. After finding he a bug, he terminated the tasks and sat
around to figure out a solution. He noticed that his fix took longer
then the usual 3/10ths of a second to compile (after all he had a V8 to
himself!) so he chased down the active processes to see who was
competing with him. It seems he had forgotten to kill one of the
paging processes! The computer had paged in all 10 megs (it was a 16
meg machine) and started to burn CPU like a bat out of hell! All in
all it had consumed something like a $100,000 in "funny money".
Gilles
...!ihnp4!alberta!myrias!ggc
Steve Fritzinger
been defending horror stories in net.rumor by saying there all folk lore,
sorry.
Back in the summer of '84 I was setting up a PC lab at my school. We were
converting an old chem. lab, and of course had to make some major
modifications, including installing air conditioning to handle the heat.
After setting up about 50 Dec Pro 350s we had the normal break in trouble
but soon everything settled down and ran fine until about mid October.
I came in one Saturday morning to open the lab and found it a little warm,
but didn't think anything about it. After cramming close to 100 freshmen
into the lab to work on their homework, the temperature reached close to 90
and 3/4s of the machines were down with random hardware errors. Seems that
building services had decided on Friday afternoon that it was time to turn
off the air conditioner, and fire up the heat for the winter. They had of
course locked the door behind them, and we had rivited all the windows shut
that summer to prevent theft. The whole lab was down until late Monday when
we finally convinced building services that we would need our air conditioner
all winter.
The cause of the second one was a little more difficult to find. Recently
one of our customers was having trouble with a group of terminals getting
periodic line noise, sometimes to the point of locking up the comm processor.
After finding nothing wrong in the hard or software a team of crack
support people went to site. There they found a bunch of RS232 lines almost
600 ft. long that ran through an elevator shaft. Every time the elevator
came by with it's big electric motor on top the RS232 line would pick up the
RF noise like any good antena and drive the comm board insane.
--
Steve Fritzinger
CCI-OSG Reston,Va.
seismo!rlgvax!jsf
"Most guys spend their entire life with one woman, right?"
"Right."
"I've got three under 'Q'."
Hank Buurman
>
> This story did not happen to me, and I disremember where I heard it,
>so it may not be true, but it's interesting nonetheless, so...
I'm curious. Why did you use the (non)word "disremember"? Wouldn't it
have been easier to type "forget"? I believe this is an example of
Richard Nixon's contributions to the english language, i.e. disremember=
forget, "misspoke myself"=lied, etc. I'm not flaming you, I'm just
puzzled as to why these examples of "nixonese" are entering common
useage. Or are they? Does anybody else remember any other examples?
Hank Buurman Tektronix Inc. tektronix!tekla!hankb
===============================================================
"I've had people run out on me before, but never when I was
so...charming." Rick Deckard, LAPD, Bladerunner Squad, B265354
===============================================================
Curtis Jackson
facility (which then consisted of on one DEC-10 for staff, students, *and*
research!!) by adding two IBM 4341's (one for staff in day, research at
night, one for students and day research). There was no operator on duty
from Friday evening (5pm) 'til Sunday noon. When the operator came into
the computer room this particular Sunday, the temperature was hovering
above 100F. It seems that the Physical Plant people got a bit slack and
decided they didn't need to clean those nasty leaves off the air conditioning
input ducts.... The DEC never did boot again properly, the IBM had two
badly damaged disk drives but seemed unaffected otherwise -- one of the
few good things I've ever said about IBM!!
As for the +1:
I was working feverishly on my compiler for my compiler construction course,
so I came in Saturday night and the damned thing wouldn't work. It
transitioned to weird states, did all sorts of strange stuff. I spent
over 10 hours debugging the godawful thing on our powerful PDP-11/34 :-)
before I was finally reduced to the program:
main()
{
int i;
i=5;
printf("%d\n",i);
}
The answer printed was a six-digit number.... I finally called our system
administrator (it was early Sunday morning by this time) and he admitted
that he had been fooling with our V6/V7 halfbreed C compiler and...
I made him replace the C compiler with the old (correct) version before
I killed him.
Eric, if you're out there anywhere, I forgive you ;-)
--
The MAD Programmer -- 919-228-3313 (Cornet 291)
alias: Curtis Jackson ...![ ihnp4 ulysses cbosgd allegra ]!burl!rcj
...![ ihnp4 cbosgd decvax watmath ]!clyde!rcj