DSXchange

Thanks all.
Actually there was a problem with the data type.Datastage can not take Int 16 .It should be bigint 16.

There is not much difference even if I sort the data, so please let me know what to do

I have the following query in DB2 ------------------------------------------------------------------------------------- COLUMN NAME DATA TYPE LENGTH -------------------------------------------------------------------------------------- TSS_CUST_ID InTEGER 8 RPS_ACCT_ID_14 InTEGER 16 LMT_VAL DECIMAL ...

Thanks wnogalski.
I will try it out.

Aparna

wnogalski wrote:If You'll try to join everything on the DB side and the performance is poor see which join makes the query run so long, then try to perform this join using DS, and so on.

Thanks ray.Lookup on table will be costly when the data is huge.Also every time it will hit the table and connection will be open.There will be posiibility of deadlock.So can't have lookup on table. Can try the second option. Also will it take more time to process the SQL query having so many joins?...

I want to join ten DB2 tables and put the data into a sequential file.There will be millions of records.Is it better to write a sql query in a DB2 stage
or can i extract the data into hash files and the have a join on these files?
Performace wise which one will be better?

Aparna