Oracle FAQ Your Portal to the Oracle Knowledge Grid
HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US
 

Home -> Community -> Usenet -> c.d.o.server -> Performance difference between 2 machines

Performance difference between 2 machines

From: Etunimi Sukunimi <etunimi.sukunimi_at_sonera.com>
Date: Mon, 20 Jun 2005 06:16:02 GMT
Message-ID: <Catte.26$%h7.18@read3.inet.fi>

Does anyone have any info/advice on the following:

Our customer has a Data Warehouse using Oracle 9.2.0.6 on an HP_UX 9000/800 machine, os version 11.00, 4GB of memory, 2 processors.

Database files are stored on an HP VA7410 virtual disk array.

The same database tables are stored on our test machine, which is basically a desktop machine, 2.8GHz P4 with 1GB of memory, an S-ATA disk tied to a RAID card (no RAID configured, though), running Red Hat Linux 8.

The problem is that a simple "select count(1) from TABLE_A" takes about 250 seconds on the customers machine, and about 280 seconds on ours - even though we have much more data in our table (190M vs. 113M).

The execution plan is similar on both machines, and trace indicates that both use about the same amount of CPU time.

The only difference I can see is this:

Customers machine:

   Event waited on                   Times   Max. Wait  Total Waited

------------------------------- Waited ---------- ------------
SQL*Net message to client 3 0.00 0.00 SQL*Net message from client 3 0.00 0.00 db file scattered read 13916 0.34 197.26 db file sequential read 1 0.00 0.00 Our machine: Event waited on Times Max. Wait Total Waited
------------------------------- Waited ---------- ------------
SQL*Net message to client 2 0.00 0.00 db file sequential read 3 0.03 0.07 db file scattered read 47719 0.10 25.86 SQL*Net message from client 2 0.00 0.00

(our database is using block size 8192, customers 16384)

Can we draw any conclusions from this data? For example that their disk I/O is simply slower?

The customers machine should give much better performance then our pitiful test server, shouldn't it?

This simple test is not really the problem, but having a build process take 2,5 hours on the customers machine and 0,5 hours on ours, is...

G: Mikael Received on Mon Jun 20 2005 - 01:16:02 CDT

Original text of this message

HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US