Join growth on distinct Account ID

[jiegao]

Hi,

Currently we are doing a left join (hash partition on Account ID) on two Oracle tables on Account ID. The left table is a simple select statement with more than 1 unique Account ID, whereas the right statement has the 'Distinct' keyword and only selecting Account ID's from the table.

My questions is: Why do the link statistics show join growth on the output of the join? Clearly, if there is a distinct right account id the join output should be the same as the left input.

I replicated the query in Oracle and it gives me the results which I think are correct which leads me to believe something is happening in Datastage that I don't understand.

Thanks in advance!

[jiegao]

I think I know what is happening here. This is caused by the "Enable partitioned read" on the data source. DISTINCT and "Enable partitioned read" should not be used at the same time.