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

Home -> Community -> Usenet -> c.d.o.misc -> Re: 3 table query question

Re: 3 table query question

From: <e_humblot_at_wanadoo.fr>
Date: Wed, 24 Jun 1998 23:39:00 +0200
Message-ID: <6mrs9l$il1$1@platane.wanadoo.fr>


Thank you for your response.

I have just read your problem again and I have noticed you mentioned you used the COST BASE
optimization mode.

This mode runs efficiently only if your table statistics are updated. To do so, for each table :
analyze table my_table estimate statistics;

Or simply use the rull based approach : alter session set optimizer_goal=rule;

Or just put the following hint in your query : select /*+ RULE */ etc...

Please note that in the rule based approach the order of tables in the from clause is important.

To answer your last question, I would say that when you issue a select count(*) statement
you just want to know the number of rows meeting a conditon concerning indexed columns, you don't want to extract the rows themselves. Hence you don't need to access the table data not being part of the index. You would probably get the same result by selecting a field being part of the index.

Salut Received on Wed Jun 24 1998 - 16:39:00 CDT

Original text of this message

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