SCO 3.2.x problem during unmount()

[Horst Laumer]

Hello, out there in the networld !

I got a tiny little problem at work, where I deal in a little
Ethernet-LAN consisting of 5 386-boxes running SCO-OS'es.

The problem is, that during normal shutdown / init 0 /etc/mnttab
gets crashed. On screen, there appears the message

	bad hertz value
	using nn from <sys/param.h>

unfortunately, this message is also inserted at a random position
into /etc/mnttab, thus causing the system not to boot properly.

On another 386-box, running SCO's 3.2.0 as standalone-system, this
problem also occured, so it cannot be a problem of the add-on
packages (TCP/IP Runtime&Development, INGRES Runtime, partly INGRES Dev.).

Sometimes the error occurs not only during shutdown, but at runtime, thus
causing df e.a. to produce silly output.

To my astonishment the problem is mostly on an 16MHz-Box with FPU running
ODT and on our server, an 33MHz-Box without FPU running 3.2.1. Our stations
are quite normal, since 3.2.1 was installed; these are 25MHz without FPU.

Has anybody else experienced this problem ?
Couls anybody help me, since its foolish to restore /etc/mnttab from a
backup every 3 to 5 days?

EMail welcome, I will resume on the net.

Thanks ahead.
       HL

-- 
============================================================================
Horst Laumer, Kantstrasse 107, D-1000 Berlin 12 ! Bang-Adress: A Relict Of
Domain: hotte at sunrise.in-berlin.de              ! Pre-Routing-Times, When
Bang:   ...unido!fub!geminix!sunrise!hotte      ! Computers Were Unmanagable