cartesian cardinality estimates

From: Richard Kuhler <noone_at_nowhere.com>
Date: Fri, 06 Feb 2004 02:26:31 GMT
Message-ID: <rXCUb.6144$a65.4303@twister.socal.rr.com>

Why would the cardinality estimate for a MERGE JOIN CARTESIAN be lower than the product of the two cardinality estimates used for the cartesian?

MERGE JOIN (CARTESIAN) (Card=1)
   TABLE ACCESS (FULL) OF 'X' (Card=1)
   TABLE ACCESS (FULL) OF 'Y' (Card=1291)

I can see how this is possible if that "Card=1" is a lie and the value is actually much smaller than 1. However, it is my understanding that 1 is the lower bound for any range cardinality calculation (the where condition on table X is COLUMN BETWEEN 'VAL1' and 'VAL2').

What is happening here?

Oracle9i Enterprise Edition Release 9.2.0.2.0 - Production With the Partitioning and Oracle Data Mining options JServer Release 9.2.0.2.0 - Production

--
Richard Kuhler

Received on Thu Feb 05 2004 - 20:26:31 CST