DSXchange

The correct and best solution would be loading the data into some temporary stage tables first. Then using ODBC/Oracle stage run a PL/SQL procedure which dumps the data from stage tables to Actual tables. This would be pretty fast. I have worked with 60 millions records using the same concept but wi...

Actually, no - it wasn't a mistake. It was designed to provide the same exact functionality as SQL joins - none of which support a "reject" capability. And at this point there is no such thing as 8.2, my understanding is that the next major release (currently in beta) may be called 8.5 wh...

There are two partition techniques in DataStage. 1) Pipeline Partition 2)Data Partition. There's another level of partition called Component Partition, which is not been supported by DataStage where as its supported in Ab Initio(in addition to the above 2 partition techniques). Pipeline and Data Par...

I disagree. What does a "reject" mean in the context of a join? When I join tables using SQL there are no rejects! ... The "reject" link in JOIN Stage means, the records which don't have the common key will go to reject link. Its the mistake of IBM Software Labs' Tester, for joi...

You could remove duplicates using transformer stage 1) with the help of stage variables 2) by doing unique/stable sort.

Sreenivasulu wrote:Instead of join stage you can use Merge stage. It has reject links.

Regards
Sreeni

The Join Stage in DataStage will have the reject link from 8.2 on wards. Its the IBM Software Labs Tester mistake for Join Stage not having the reject link.

Its my bad, I mean unicode not unique.

Change the Data type to NVARCHAR followed by Unicode. It will work.

[* Above line corrected from unique to Unicode - Andy *]