>>>While I was messing around with Visual Foxpro and SQL Server, I suddenly realize what I think is needed to really make Visual Foxpro a serious tool for
>database product but in fact Foxpro is also a good C/S front-end product. In
>read in different published acticles that MS is in fact working on a lite-version of SQL Server that is easier to maintain by non-technical persons.
people,
perhaps those are the saving graces of ms. i don't wish to douse hopes (they are mine also) but perhaps you might check this out, and it might be technical.
regards,
rostand
rabear@philonline.com
---------- Forwarded message ----------
Return-Path:
Date: Wed, 5 Nov 1997 03:03:31 -0900 (AKST)
From: "Christopher E. Brown"
To: Eric Hoeltzel
Cc: Irfan Akber , linux-admin@vger.rutgers.edu
Subject: Re: SQL Server
In-Reply-To:
Message-ID:
MIME-Version: 1.0
X-Orcpt: rfc822;linux-admin@vger.rutgers.edu
Sender: owner-linux-admin@vger.rutgers.edu
On Tue, 4 Nov 1997, Eric Hoeltzel wrote:
>
> I'm pretty impressed with it. I'd like to do some benchmarking
> against microsoft SQL on my SMP machine.
This spring I was doing development work on a "Management
Reporting System", the proof of concept for the final platform
choice (Linux/MySQL) included setting up a basic system under
NT/MS-SQL, SCO/Oracle, Linux/MSQL, Linux/PostgreSQL, etc...
Well, NT/MS-SQL was horrid, the basic system up for test (some 5
table of ~ 78,000 25 - 40 field records was taking up to 10 minuets for
simple 2 -> 4 table joins. Didn't care for the MS solution to active html
either.
SCO/Oracle worked, but was by no means fast, setup was a pain and
getting everything working right took a while.
Linux/MSQL is easy to setup, stable and fast for single table
selects. Functionality is limited for more complex selects and table
joins to slow (almost as slow as NT/MS-SQL).
Linux/PostgreSQL is reasonably fast, supports complex joins, many
field types, but is still somwhat bug prone. (My experience over the last
14 months)
Linux/MySQL is very fast, supports complex joins/selects, many
field type, very stable. MySQL was faster than any others by a factor of
2 or better ( 2x Postgres, 10x MS-SQL).
The system this was built on is a P133 w/ 32megs, and has been
operating without problems (or maintenance) since May. The final solution
handles several times the # of records, does many complex joins, etc.
My devel workstation (PPro200/256K x 2 SMP, 192MB EDO) runs a
recent version of MySQL, with a total of 12,000,000 records across 23
tables. Worst case query (against a 4.5 million record table) for a
single record against an indexed field is 3 seconds (most < 1 sec).
I am a very very happy camper.
BTW: To make shure this reply fits into linux-admin I do use MySQL for
keeping records of network traffic, DNS records, user info. Pretty much
anything logged or stored over here is in Mysql, dns files for bind are
generated via perl script against the DB, same with network/host traffic
reports.
----------------------------------------------------------------------------
Christopher E. Brown (CB421) +(907) 357-5680
We are Pentium of Borg. Division is futile. You will be approximated.
----------------------------------------------------------------------------
rostand
"le rêve d'étoiles" - (Fr.) the dream of the stars Ü
