Join behaves strangely

rachitha · Post by **rachitha** » Fri Jun 25, 2010 12:12 am

As DS1 and DS2 is left outer join. The column coming from DS2 i.e Key_Col is giving null value. Though matching data exists, Key_Col is populating with NULL value which is wrong.

Thanks,
Rachitha

rachitha · Post by **rachitha** » Fri Jun 25, 2010 12:14 am

Even I am facing the same problem in Test Environment. Only change is my job uses Inner join. Though data is matching.. join is not capturing data and target dataset is not created.

Kindly post ur suggestions

Thanks,
Rachitha

ray.wurlod · Post by **ray.wurlod** » Fri Jun 25, 2010 2:17 am

Is Key_Col one of the join keys? Are the data partitioned on both inputs using a key-based partitioning algorithm that names the join keys? Sorted?

rachitha · Post by **rachitha** » Fri Jun 25, 2010 5:21 am

No Ray, Key_Col is not a join key field.
Applied hash partition based on key column, giving sorting, case sensitive, perform nulls first.

Still same problem exist..

AM not understanding y it is working in dev. environment but not in TEST env. what might be the reasons

Will it be a environment issue.. if yes what kind of issue it can be..

Should we check the number of nodes given in config file???

Thanks,

chulett · Post by **chulett** » Fri Jun 25, 2010 6:03 am

Of course, you need to check everything that could change from environment to environment. That would include the config file used, actual data, source/target metadata, etc etc.

mjgmc · Post by **mjgmc** » Thu Jul 22, 2010 2:48 am

ray.wurlod wrote:Is Key_Col one of the join keys? Are the data partitioned on both inputs using a key-based partitioning algorithm that names the join keys? Sorted? ...

Sorry for this late reply.

Yes, both links are hash/sort by the Key_Col