Newsgroups: comp.unix.sysv386
Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!caen!news.cs.indiana.edu!msi.umn.edu!cs.umn.edu!atc!cimcor!mike
From: mike@cimcor.mn.org (Michael Grenier)
Subject: Re: Rebooting Sys V/386
Message-ID: <1991Mar31.125826.25542@cimcor.mn.org>
Organization: Scandia Microsystems
References: <1991Mar29.134732.11457@scuzzy.in-berlin.de>
Date: Sun, 31 Mar 91 12:58:26 GMT

From article <1991Mar29.134732.11457@scuzzy.in-berlin.de>, by src@scuzzy.in-berlin.de (Heiko Blume):
> bill@astph.UUCP (Bill Dripps) writes:
>>One reason we discovered (why do we always do it the hard way?) is that
>>large cache disk adapters do not have enough time to flush disk blocks to
>>the drive before the 'init 6' code jerks the reset line.
> 
> i KNEW using those cache-suckers is risky, that's why i don't use them.
> (altho i think the ones with battery backed up ram won't discard the cache's
> contents on reset.)
> -- 

I'm confused. Why not just modify the rc script on shutdown to add a little
time. The inittab entry on this ESIX box simply kills the processes
and run /etc/unmountall (from /etc/rc0) before calling uadmin to reset
the box. Why not add a sleep statement before the end of /etc/rc0?

     -Mike Grenier
     mike@cimcor.mn.org