Hash Partitioning Question

neroans · Post by **neroans** » Mon Dec 22, 2014 5:28 pm

I have retrieved data from database in sequential execution mode, I sorted the data in sequential execution mode , I pushed the data through a Transformer in sequential execution mode. In the transformer I added a new column ( XYZ ) , In the next step I am trying to create a dataset using XYZ as the key for HASH Partitioning.

I have built all these steps in one job. The partitioning with XYZ is not working. But at the same step, if I use one of the columns that was retrieved from the database as the key for Hash partitioning. It is partitioning fine. Does it mean we should not be using intermediate values as key for partitioning?

qt_ky · Post by **qt_ky** » Mon Dec 22, 2014 8:03 pm

How exactly is it not working? What is the XYZ data type and the values you are using in XYZ?

priyadarshikunal · Post by **priyadarshikunal** » Tue Dec 23, 2014 7:56 am

This question is something like I need to go some where, on a cross road if I take one road i reach my destination and if not, i don't reach there. What is the problem? The problem is you did not read the sign board. Simple mistake but can cause trouble if you are short on gas, cash and time

Are you passing a constant value to that column?

ArndW · Post by **ArndW** » Fri Dec 26, 2014 1:45 am

DataStage has no general problem with using a column created in a transform stage as the hash partitioning key downstream. Could you explain what the column type and contents are and what the partitioning did differently from what you expected?