Archives of the TeradataForum

Message Posted: Thu, 03 Oct 2002 @ 11:47:16 GMT

	<Prev	Next>		<<First	<Prev	Next>	Last>>

Here is some DDL I put together to illustrate the problem:

BaseFact
ColA
ColB
ColC
ColD
ColE
ColF
Meas1

MultiDim
ColA
ColA1
ColA2
ColA3
ColB
ColB1
ColB2
ColB3

Create ViewDimA
(ColA,ColA1) As
Select ColA, ColA1,ColA2,ColA3 From MultiDim Group By
ColA,ColA1,ColA2,ColA3;

Create ViewDimB
(ColB,ColB1)
Select ColB,ColB1,ColB2,ColB3 From MultiDim Group By
ColB,ColB1,ColB2,ColB3;

Example SQL
Select    *
From ViewDimA
,    ViewDimB
,    BaseFact
Where     ViewDimA.ColA       =    BaseFact.ColA
and  ViewDimB.ColB       =    BaseFact.ColB
and  ViewDimA.ColA2 =    'Purple'
and  ViewDimB.ColB3 =    'Elephant'

What I would like to happen: Optimizer joins BaseFact to MultiDim (where ColA2 = 'Purple' and ColB3 = 'Elephant') What actually happens: Optimizer joins BaseFact to ViewDimA (where ColA2 = 'Purple') and then joins resulting spool file to ViewDimB (where ColB3 = 'Elephant')

Thanks for the help,

Clay

	<Prev	Next>		<<First	<Prev	Next>	Last>>