RE: select count(*)

Hi!

I would like add that following things are same in both the databases.

init.ora parameters

optimizer parameters

data is also similar

but when I looked at explain plans they are different.

My question is basically how oracle treats a query with predicate as " 0 = 1".

Thanks and Regards,

Sarma

---

From: "Richard J. Goulet" <rgoulet@kanbay.com>
Reply-To: rgoulet@kanbay.com
To: <oracle-l@freelists.org>
Subject: RE: select count(*)
Date: Fri, 6 Oct 2006 11:32:22 -0400

Sarma,

 

    Before you run the statement in each instance try setting "set autotrace on".  That should give you your answer.

 

 
Dick Goulet, Senior Oracle DBA

45 Bartlett St  Marlborough, Ma 01752, USA
Tel.: 508.573.1978 |Fax:  508.229.2019 | Cell:508.742.5795

RGoulet@kanbay.com
: POWERING TRANSFORMATION

 

---

From: oracle-l-bounce@freelists.org [mailto:oracle-l-bounce@freelists.org] On Behalf Of Sarma Aryasomayajula
Sent: Friday, October 06, 2006 11:10 AM
To: oracle-l@freelists.org
Subject: select count(*)

Hi!

 

I'm running the below query in PROD(10g R1) and ACPT(10g R2). the query returns in fraction of a second in PROD and, If I run the same query in ACPT, it's taking 30+ mins.

 

In Prod

S5UVAD@powls01 > set timin on
S5UVAD@powls01 > select count(*) from wlcbs_master.rpt_08_vew where 0=1;

 

  COUNT(*)
----------
         0

 

Elapsed: 00:00:00.07
S5UVAD@powls01 > sho user

 

In ACPT:

 

N7OTHA@AOWLS01 > set timin on
N7OTHA@AOWLS01 > select count(*) from rpt_08_vew where 0=1;

 

  COUNT(*)
----------
         0

 

Elapsed: 00:31:02.27

1. If I apply condition "where 0 = 1" on a single table, does oracle reads
entire table then applies this condition or otherwise since this is negative
condition does it apply without reading the entire table?

2. If the same condition is applied on a view which is join of 3 big
tables(paritioned)? How does oracle executes the query?

Can any one explain the above scenarious.

Regards,

Sarma

-- http://www.freelists.org/webpage/oracle-l
<< klogo.gif >>