[comp.sys.ibm.pc.hardware] 33 MHz 486 system for SCO Unix

L.Chung@ee.surrey.ac.uk (L N Chung) (05/03/91)

We are trying to put together a 486 system for running SCO Unix.
Our basic spec is 33 MHz 486 with 16 MB of ram and a 300 - 600 MB hard disk.
We would like some advice on the choice of disk controller and brand of
hard disk.  Will a Adaptec 1542 or WD 7000 be good enough? How large a disk
cache does SCO Unix perfer? Which Controller is known to have compatibility
problem with SCO Unix?

Also does anybody know of compatibility problem between specific BIOS and 
SCO Unix. Any info on bad experience with 33 MHz 486 motherboard is also
wellcome.

Many Thanks.

-- 
--------------------------------------------------------------------------
L.N. Chung			Dept of Elec. Eng, University of Surrey,
(ees1lc@ee.surrey.ac.uk)	Guildford, Surrey, GU2 5XH. UK.
(l.chung@ee.surrey.ac.uk)	PHONE: +44 483 571281  FAX: +44 483 34139

jim@shograf.com (jim morris) (05/05/91)

From article <1991May3.143520.16386@EE.Surrey.Ac.UK>, by L.Chung@ee.surrey.ac.uk (L N Chung):
> We are trying to put together a 486 system for running SCO Unix.
> ....

I tried getting SCO unix 3.2.2 to run on a Club American 486/33 with NO luck
whatsoever. It wouldn't even Boot!!

I ended up putting the ESI 660Mb HD into an ACMA 386/33, and it has worked
happily ever since!!

Jim

-- 
Jim Morris,	E-Mail: jim@shograf.com    Voice: (415) 903-3887
  _ 
SHO graphics.	Practical PEX

davidsen@sixhub.UUCP (Wm E. Davidsen Jr) (05/07/91)

In article <1991May3.143520.16386@EE.Surrey.Ac.UK> L.Chung@ee.surrey.ac.uk (L N Chung) writes:
| We are trying to put together a 486 system for running SCO Unix.
| Our basic spec is 33 MHz 486 with 16 MB of ram and a 300 - 600 MB hard disk.
| We would like some advice on the choice of disk controller and brand of
| hard disk.  Will a Adaptec 1542 or WD 7000 be good enough? How large a disk
| cache does SCO Unix perfer? Which Controller is known to have compatibility
| problem with SCO Unix?

  If you want the top disk performance get an EISA system and the new
Ultrastor 22 controller. Enable the write back option (be sure you
understand the implications of this) and install their driver. In
various disk benchmarks I really didn't see much difference with the w-b
and driver, but in real usage it made a huge difference, and with only
512k cache installed (but get the 4MB).

  On a very large make the clock time dropped from 23min in "WD 1007
compatible" mode to 15 min in write back plus driver mode. This is
getting close to the 13min of CPU time. Other real applications showed
big gains, too.

  The Adaptek 1542 will show a better throughput for really large
(megabyte+) i/o, but not as good performance as the Ultrastor. The 22
seems much better to use than the earlier (12?) models, which were known
as the "Ultrastrange" by some hardware types at work.
-- 
bill davidsen - davidsen@sixhub.uucp (uunet!crdgw1!sixhub!davidsen)
    sysop *IX BBS and Public Access UNIX
    moderator of comp.binaries.ibm.pc and 80386 mailing list
"Stupidity, like virtue, is its own reward" -me

ees1lc@thorin.ee.surrey.ac.uk (L N Chung) (05/11/91)

Many Thanks to everybody who reply to my query about SCO Unix systems.
I would like to clear up one point:

If SCO Unix does disk cacheing inside its kernel, does it mean it is
pointless in using cacheing controller ? If so, how does SCO Unix
support these controllers, does it just turn the cache off and use its
own?

What is the difference between "Track-at-a-time" controllers and cacheing
ones? Will they do a better job.

Thanks again.

--------------------------------------------------------------------------
L.N. Chung			Dept of Elec. Eng, University of Surrey,
(ees1lc@ee.surrey.ac.uk)	Guildford, Surrey, GU2 5XH. UK.
(l.chung@ee.surrey.ac.uk)	PHONE: +44 483 571281  FAX: +44 483 34139