DSXchange

when there is data growth, adversely affect the execution speed. Lookup can have N+1 input links.

Can you check Stored proc input/output parameter count, parameter type(input/output) etc..

execute command from below directory will help

/etl/InformationServer/Server/PXEngine/bin/

Quick design, ---- remove dup (Col1)----- ---- remove dup (Col2)----- input ---> copy ---- remove dup (Col3)--- transformer- pivot ----> Output ---- remove dup (Col4)----- ---- remove dup (Col5)----- This may help, Pls improve the design.

use functions like NUM to identify . and set it to zero

if you are writing to a file then remove stringtodecimal function.

This may help(Havn't tested it ),

StringToDecimal(Trim(Trim(<string>,"+","A"))),"0","B"))

1. Maybe, But IBM is not providing support for Datastage 7.5 or older versions..

2. Yes, You can use any Datastage job which design to receive IDOc's

Thanks,
Bharath Tipirisetty

pls check there are any index created on the table, What is the remaining table space?

you can also use DSLink3.name[1,10]

In CITRIX virtual machines are mapped under metaframe server, Usually datastage installed in virtual windows server 2003 machine. Login datastge manager, try to export DSX and check for directory which it is pointed to. My guess you can interface Dev windows & Dev unix in putty using FTP. You ca...

Designed mentioned will retain both line 2 and line 5.... As mentined by another user you can implement in transformer also(DS Ver 8.x)

Can you Check max length of that column/ Check any record contains extra delimiters.. like comma in full address column.. etc

have a unix script to just copy source file in before job sob routine, using SED or unix shell script help keep 1 & last line , remove rest. do all your calucations, in after job subroutine again create simple Copy script or ftp script to transfer files....

Introduce alias column like Source_table while selecting data from source, use the same as identifier in reject data.