[Search for users] [Overall Top Noters] [List of all Conferences] [Download this site]

Conference sutra::europc

Title:Euro-PC Support Conference
Notice:Stop your whinning and get back to work!
Moderator:SUTRA::BATS
Created:Mon Jun 12 1989
Last Modified:Fri Jun 06 1997
Last Successful Update:Fri Jun 06 1997
Number of topics:3826
Total number of notes:12603

3787.0. "FX, W-NT FDD format crash" by VIVIAN::P_GRIFFIN () Sat Feb 15 1997 09:03

Hi,
I'm not sure if this is a 'real' problem, or a spate of finger trouble. But one
other person has remarked to me about a similar scenario, so I thought I'd offer
mine for comment.

Customer is expecting to roll out hundreds of Celebris FX, once their project
is developed and complete.
They received 15 in November '96, no problems.
They received 21 in January '97, enormous problems.
They are running W-NT v4.0, and get "blue screen" crashes. Over a period of four
weeks they investigated to try to find which ones were suffering (apparantly
at least half), and how often (from twice-a-week to five-times-a-day), and why.
There was a mix of BIOS 1.07 and 1.09 across both batches.
They deduced there was a corrupted Operating System, this was based on fixing
one system by a re-build, and fixing another system by moving the HDD from a
good system.
The systems are bought via a VAR, who adds 16Mb RAM and "sets up" W-NT to some
pre-arranged configuration.

The customer has found a "test". If you START > RUN > format/q a: the blue
screen will appear after 1-9 attempts. The bugcheck is "KMODE - EXCEPTION NOT
HANDLED, Address 8039C9BE has base at 8039B000 - sscntrl.sys".

They had received a further 11 systems in Feb'97 and said the first one they
opened had also failed. Interestingly, these had not been opened by the
intermediate VAR. I opened two, set up W-NT (added 16Mb to one) and they seemed
to be o/k. I also copied a 'good' HDD to a 'bad' system, and did forty formats.

I noticed that there did not seem a great difference in serial numbers
(first batch typical PC64806042, second batch PC70105742, last batch PC70312467.
However, in the ones I opened, first batch MLB=E01, second batch MLB=H01, last
batch MLB=H02+.

The customer has agreed to a plan of identifying failing systems by a 20-pass
of format/q a:, and we will move a 'good master disk' around to regenerate the
HDD on each one (I used HDCP on a floppy to copy D: to C: for this, but would
welcome alternative suggestions. At that time I'll check all the MLB revs.
If the problem all disappears after this, I'll have to attribute it to some
glitch in the FIS or VAR W-NT installation. If the problem continues, I guess
an IPMT will be the next step.

I'm really writing this note to see if it rings any bells, rather than for
assistance, as I can't see why it is only me having this problem, unless my
customers volume of numbers make it more obvious.

Regards Pete
T.RTitleUserPersonal
Name
DateLines
3787.1FX and NT PCBUOA::GKELLEYMon Feb 17 1997 20:0820
    There are a number of things that you can check:
    
    1. Is the memory "Digital" or 3rd party. If 3rd party, remove and
    install Digital (we've seen a lot of problems lately involving 3rd
    party SIMMS and NT problems).
    
    2. Check on the disk for a file called "DIGITAL.FIS". This file is put
    on the HD at manufacture time and contains all the different modules
    necessary for a particular SW load. If the file is present, we can use
    it to verify that all the required pieces are there. If the file is
    missing, it means the VAR installed NT and all bets are off as to the
    stability of the OS.
    
    Check to see if they're using the SISIDE driver. NT 4.0 requires the
    built-in ATAPI.SYS and not the SISIDE driver.
    
    regards,
    
    glen kelley
    pcbu
3787.2Dr Solomon's productivity toolVIVIAN::P_GRIFFINTue Feb 25 1997 03:049
    Thanks for the input Glenn.
    In actual fact, the customer started to build his 'master HDD',
    checking at each stage for its integrity. Guess what? He found the
    problem was caused by installing Dr Solomon's Virus Checker! Further
    investigation revealed it was a 'known problem' (but obviously not
    known by the Windows NT group he had been dealing with for four weeks).
    But I have not been told why it appeared to be 'batch related'.
    Boom, boom, another one bites the dust.
    Regards Pete
3787.3Please update notePCBUOA::GKELLEYTue Feb 25 1997 15:318
    Pete,
    
    Can you update this note on the information about the Dr. Solomom
    problem?
    
    thanks,
    
    glen
3787.4pointerKEIKI::WHITEMIN(2�,FWIW)Thu Feb 27 1997 03:2914
    
    	McAfee also shows similar problems, Microsoft has a patch on
    their ftp site, If the McAfee task manager (Server task) is started
    then whenever you format the floppy the machine, In our case
    a Celebris GL 6200 would crash with IRQL_NOT_LESS_OR_EQUAL in
    ntoskrnl.exe.
    
    	Patch location for Windows-nt
    
    ftp.microsoft.com
    /bussys/winnt/winnt-public/fixes/usa/nt40/hotfixes-postSP2/krnl-fix
    
    						QT141239.txt
    gives description of fix