Re: How to analyze a odd column distribution

After reading Jonathan's response I realized that I missed part of the problem: the inability of the optimizer to do partition pruning based on the to_timestamp predicate. Is FHASTA the partitioning column?

Alfonso León wrote:

>
> SQL> explain plan for
> 2 SELECT * FROM TCUENTA
> 3 WHERE FHASTA =TO_TIMESTAMP('31-DEC-2999','DD-MON-YYYY')
> 4 /
>
> ---------------------------------------------------------------------------------------------------------------
> | Id | Operation | Name |
> Rows | Bytes | Cost (%CPU)| Pstart| Pstop |
> ---------------------------------------------------------------------------------------------------------------
> | 0 | SELECT STATEMENT | |
> 3 | 399 | 5 (20)| | |
> | 1 | PARTITION RANGE SINGLE | |
> | | | KEY | KEY |
> | 2 | TABLE ACCESS BY LOCAL INDEX ROWID| TCUENTA |
> 3 | 399 | 5 (20)| KEY | KEY |
> |* 3 | INDEX RANGE SCAN | ICTAFHASFDESPROSUB |
> 3 | | 4 (25)| KEY | KEY |
> ---------------------------------------------------------------------------------------------------------------
>

-- 
Regards

Wolfgang Breitling
Centrex Consulting Corporation
www.centrexcc.com
--
http://www.freelists.org/webpage/oracle-l