DSXchange

Hi All,
I have a question on join stage.
I have 6000 records in my source and I'm doing left outer join with one table, after the join I'm getting 6039 but I should get only 6000 records.
Can some please explain me why it is happening?

Thanks In Advance
Eze_ke

How many records are in table you are joining with?

anntaylor,
there are 1195 records in the table.

Thanks
Eze_ke

No there is no bug in Join stage. What I think that you must have some duplicate key in the right table. Please check it and I hope you get the answer.

eze_ke wrote:Hi All,
I have a question on join stage.
I have 6000 records in my source and I'm doing left outer join with one table, after the join I'm getting 6039 but I should get only 6000 records.
Can some please explain me why it is happening?

Thanks In Advance
Eze_ke

If u do a Left outer join with tables then it will take all matching rows from the left table and also null values from the right table i think this might happening

Hi,

Are you doing some kind of partitioning in the input?

Are you using the configuration file with multiple nodes for running this job?

--Rich

Most probably girija would be right. Duplicates would cause more number of row in the output than expected.

obviously u will be getting excess records I dont think so any Bug in join stage.u are using left outer join the link from left all rows u will be getting it. and the matching records from the right link.just see through the link ordering.

arasan wrote:obviously u will be getting excess records I dont think so any Bug in join stage.u are using left outer join the link from left all rows u will be getting it. and the matching records from the right link.just see through the link ordering.

Even in this case, the number of output cannot exceed the number of records in the master link. Unless it has dulicates or improperly partitioned.