From owner-freebsd-hackers  Wed Mar 12 20:15:23 1997
Return-Path: <owner-hackers>
Received: (from root@localhost)
          by freefall.freebsd.org (8.8.5/8.8.5) id UAA02232
          for hackers-outgoing; Wed, 12 Mar 1997 20:15:23 -0800 (PST)
Received: from parkplace.cet.co.jp (parkplace.cet.co.jp [202.32.64.1])
          by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id UAA02220
          for <hackers@freebsd.org>; Wed, 12 Mar 1997 20:15:19 -0800 (PST)
Received: from localhost (michaelh@localhost) by parkplace.cet.co.jp (8.8.5/CET-v2.1) with SMTP id EAA25504; Thu, 13 Mar 1997 04:15:02 GMT
Date: Thu, 13 Mar 1997 13:15:02 +0900 (JST)
From: Michael Hancock <michaelh@cet.co.jp>
To: ccsanady@nyx.pr.mcs.net
cc: FreeBSD Hackers <hackers@freebsd.org>
Subject: Re: Solaris TPC-C benchmarks (with Oracle)
In-Reply-To: <199703130313.WAA21644@jenolan.caipgeneral>
Message-ID: <Pine.SV4.3.95.970313130458.25395A-100000@parkplace.cet.co.jp>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Sender: owner-hackers@freebsd.org
X-Loop: FreeBSD.org
Precedence: bulk

It was a price performance measure.  The actual throughput was quote, "The
performance score of Solaris, measuring transactions per minute, was
equally impressive: 6679.50tpmC throughput."

This was Oracle on an IBM PC Server.

Contrast this with the Compaq/Sybase/UnixWare benchmark in November. "The
TPC-C result of 8,311.43 tpmC was based on a ProLiant 5000 configured with
four 200 MHz Pentium Pro processors each with 512KB cache, 4-GB of memory,
using the SCO's UnixWare operating system version 2.1.1, running Sybase's
SQL Server 11.0.3 database application." 

I wonder which difference make the difference.  Solaris which is
supposedly very advanced with it's kernel threads and slab allocators,
etc doesn't perform as well as UnixWare which I think has a simpler model.

I think UnixWare has a very good AIO implementation, so reads and writes
to raw partitions blaze.

Regards,


Mike Hancock