mithomas@bsu-cs.UUCP (Michael Thomas Niehaus) (01/08/88)
Here at Ball State University we have recently been having many problems
with parity checks on IBM PC's (running MS-DOS 2.1). In an average day
(15 hours of usage, usually 250 users a day for an average time of one hour
each) we are getting mostly parity check #2's, but there are also some
parity check #1's making appearances. This is occurring in a lab with 30
PC's, each with 640k, a serial port, a parallel port, and CGA graphics card
and monitor. The extra memory and the ports are all on one multifunction card.
The memory was upgraded about 6 months ago by adding more chips to the board
(of course of the kind recommended by the manufacturer, whose name I can't
remember). About three months after that, the trouble started, and continually
increased to where it is now. We have considered static, but have had no
success with recommended solutions. We have tried Isobar boxes. We are
now planning to call in the electrician to check for power line noise and/or
disturbances. If this fails, who knows where we will turn. Anyone out there
have any other suggestions?
Some more info:
The computers almost never get a parity check in the morning. They start
around noon, then gradually increase in frequency until closing time.
The computers all were tested over Christmas break, running IBM Advanced
Diagnostics. In the 24 hour continous loop full test, we came up with
NO errors. Not a one. Everything worked perfectly.
In the lab in the next room, they have 27 IBM PC's with about the same setup,
only most only have a monocrome display/adapter pair. They also use an
AST multifunction board. Could the boards have something to do with it?
They have had only one parity check (that I know of) in the last 3 months.
The computers seem to reset randomly, although certain computers reset more
often than others. No computer has been spared, though.
I have tried everything I can think of. If anyone has any ideas at all, send
me a mail message. I'm open for suggestions. The current situation is not
pleasant. (How do you explain to a beginning user that their last hour of
work is gone and they have to start over?) We have been telling people to save
often, but that is not enough. We need to solve this problem.
Also, recently we have been having disks go bad on the average of 3 per day.
This is also much higher than before.
I hope someone can help.
Sincerely,
Michael Niehaus
Ball State University
Muncie, IN 47306
UUCP: <Backbones>!{uunet,pur-ee,iuvax}!bsu-cs!mithomas
p.s. I hope I didn't offend any English teachers with my typo in the subject
line.oberman@cg-atla.UUCP (Fred Oberman ) (01/15/88)
In article <1795@bsu-cs.UUCP> mithomas@bsu-cs.UUCP (Michael Thomas Niehaus) writes: >Here at Ball State University we have recently been having many problems >with parity checks on IBM PC's (running MS-DOS 2.1). At my last place of employment, we had something similar, though not nearly as drastic. I was told that the cladding metal on the legs of the memory chips was different than the metal used in the chip carriers, and that over time oxides built up which caused random parity errors. Even though these chips were recommended, they may be misbehaving. (who knows they might even be bootleg). As an experiment, try replacing the (newer) memory in one of the machines, and see how it acts during the day. Good luck.