Path: utzoo!mnetor!uunet!husc6!cmcl2!phri!dasys1!manes
From: manes@dasys1.UUCP (Steve Manes)
Newsgroups: comp.sys.ibm.pc
Subject: Re: XENIX VS Micoport
Message-ID: <2809@dasys1.UUCP>
Date: 2 Feb 88 10:02:11 GMT
References: <16217@watmath.waterloo.edu> <166@bhjat.UUCP> <366@igloo.UUCP> <428@cimcor.UUCP>
Reply-To: manes@dasys1.UUCP (Steve Manes)
Organization: Datamerica Systems, NYC
Lines: 33

In article <428@cimcor.UUCP> mike@cimcor.UUCP (Michael Grenier) writes:
>> 	Igloo, too, has 2 serial ports, a no-name AT clone board and some
>> monster disk drives. It, too, has news running, but ONLY when Bill shuts
>> down EVERY user to do the poll. He's running at 2400 baud now, and going to
>
>
>Strange, my Microport system with the same configuration seems to work
>great. Perhaps, your no name clone does have problems.

I doubt it.  I had the same problem with disk/tty interrupts overflowing
and crashing Microport.  The last time it happened, 'fsck' also destroyed
my root file system on reboot.  This crash occurred with two 1200 baud
users on-line and a 'doscp' to a floppy.  My Microport machine was a
Smartek with a Seagate 4096.

The next week my main UUCP feed, also running Microport V/AT on an AST
machine with a 4096, crashed during a long news feed, destroying all file
systems.

Previous to the big crash (I now run Xenix/386, as does my feed),
"double panics" were a routine event on V/AT.  Virtually everyone I know
who uses/used V/AT with active tty processes has experienced the same
problem, and this includes a couple of genuine Big Blue machines.  The
problem is known and was easily replicated in 2.2... just login with a 9600
baud machine over a null modem and then upload a text file with 'cat >
filename'.  Then fire up 'cu' (providing that you're not already staring at
a screen with Microport's guts spilled all over it).

-- 
+-----
+ Steve Manes            Roxy Recorders, Inc.             NYC
+ decvax!philabs!cmcl2!hombre!magpie!manes       Magpie BBS: 212-420-0527
+ SmartMail:  manes@magpie.MASA.COM