Message-Id: <4.2.0.58.20000728204043.00c24d20@mail.diligence.com>
Date: Fri, 28 Jul 2000 20:49:53 -0600
To: pgsql-interfaces@postgresql.org
From: Tim Uckun <tim@diligence.com>
Subject: Access and Postgres question.
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"; format=flowed

I am testing postgres 7.02 and ms-sql server as back ends to ms access. I 
created identical databases containing three tables with about 1.2 , 1.6 
and 3 million records each. The machine that is running SQL server is a 
dual pIII zeon machine with raid 0 and 512 megs of ram. The postgres 
machine is a dual pIII 500 with 256 megs of ram no raid but nice fast scsi 
drive running red hat. Given this scenario I do expect postgres to be 
slower then ms sql server. What I find very odd is that in some tasks it 
actually faster (very slightly) like when doing simple select * from table 
where primarykey=x. But when I do select * from very_big_table (I know it's 
not such a good idea) it is insanely slow. On one test the table with 1.2 
million records too 12 seconds in postgres and less then a second in sql 
server!. I suspect that this has something to do with the ODBC driver 
because I am having a hard time believing that postgres is that slow. I 
suspect that the MS-SQL server ODBC driver is not actually pulling up ALL 
the records but just a few at a time (and perhaps spinning off another 
thread to fetch the rest). Can anybody confirm this?
----------------------------------------------
              Tim Uckun
       Mobile Intelligence Unit.
----------------------------------------------
    "There are some who call me TIM?"
----------------------------------------------