Archives of the TeradataForum

Message Posted: Wed, 01 Mar 2006 @ 10:59:10 GMT

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

Thanks for all your valuable suggestions...

I changed the Datatype from varchar(30) to Char(10) in the filter table to match the definition in the big table.

I have also collected statistics on the columns Data_source_cd on the big table as well as the Filter table.

I made a compression on Data_source_CD for NULLS on Account_event Table.

I havent changed anything else in the view definition .

The reason I am putting a coalesce in the view definition is because When we implemented the warehouse there was no attibute called Data_Source_Cd in Account_Event table. This attribute has been added subsequently as upgrade to the PDM during Basel2 implementation. So most of the 15 million records will have only null in it and only the records from basel2 implemented data will hold value in it .

Since there is a overlap of systems (common systems ) between Basel2 and to the systems implemented prior to it , some of them will be going in as Null and some of them with Values in Data_source_cd and I need to provide the records that are needed only by datamart users (which is null) as well as records common to Basel2 and our datamartusers ( which has value in Data_source_cd) to our business users.

Now the query runs in 6 mins. Since the query against target is only 3 mins ...Is there anything else I could do for improving

Cheers

Suresh

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