On Sun, 16 Mar 1997, Michael R. Zucca wrote:

> >On Fri, 14 Mar 1997, Michael R. Zucca wrote:
> >
> >> >I am running 1.2C (March01) and the only difference with GENERIC in my 
> >> >config is I commented out "option M68040" (I have a IIvx). I also 
> >> 
> >> Don't use sbc with the IIvx. I use ncrscsi and it works just fine. It's also
> >> a much faster driver than sbc.
> >
> >This may be a problem. You see, I dont use ncrscsi because it corrupts my 
> >filesystem every 10 minutes. Now that I've switched to sbc it seems "very 
> >problematic" as you say. What am I supposed to use then? I'm starting to 
> >get a bad feeling about this. I cant afford to go buy a new HD right now.
> 
> Hmmm. Well, what drive are you using? I know some people were complaining
> about Quantums but I have a Quantum Fireball 1 Gig which is known to be
> one of the worst drives for this sort of thing and it works just great.
> Perhaps you can get a firware update for the drive either cheap or free from
> the manufacturer? It might be worth looking into.
> 

This is from a while back. I was getting the following error:

  sbc0: parity error!
  sbc0: parity error!
  sbc0: parity error!
  ncr5380_machine: unexpected disconnect.
  sbc0: parity error!

This only seemed to happen under heavy disk load (some compiling and 
large "rm -rf" and a few other things that made it crunch it's heart out).
I've been simply avoiding such situations as a workaround since then.

I've finally contacted DEC (my disk is a DEC RZ26L, 442D, 1001Mb disk) 
and they tell me that I have the right firmware as long as I am not using 
a wide scsi. The only bug they know about is that it may get confused 
when getting SCSI resets too quickly (but this should cause problems only 
at initialisation etc.).

What I'll do is:

   Set up a way to backup my system
   get -current everything
   see if I can reproduce the error

It may take a while though because end of term is upon me at McGill 
University.

> 
> _______________________________________________________________________
>  Michael Zucca - mrz5149@rit.cs.rit.edu - http://www.rit.edu/~mrz5149/
>  "I will choose a path that's clear. I will choose Freewill. "
>   --Rush, Freewill
> _______________________________________________________________________
> 

Thanks for answering, I hadn't though of checking out the firmware.

-----------------------------------------------------------------
"Because user errors often produce unpredictable results, the user
should try to avoid them."

IBM MVS/XA System Programming Library.