Plz Help

[hrishy]

Hi


create table test
as select object_id key1,object_name key2,object_id key3
from all_objects


create unique index testidx
on test(key1,key2,key3)

analyze table compute test compute statistics

select distinct(key1),key2
from test


Execution Plan
----------------------------------------------------------
0 SELECT STATEMENT Optimizer=CHOOSE (Cost=306 Card=23004 Bytes
=667116)

1 0 SORT (UNIQUE) (Cost=306 Card=23004 Bytes=667116)
2 1 TABLE ACCESS (FULL) OF 'TEST' (Cost=44 Card=23004 Bytes=
667116)


well the most expensive operation here obviously is

SORT (UNIQUE) investigate wheather you really need that distinc(key1)

regards
Hrishy