DSXchange

Hi All, I am reading from Oracle Enterprise stage and writing to dataset stage. My question is related to execution mode of the stage. Here i am trying to explain my job design. I have 2 different flows in the same pallet reading from 2 different table

Oracle --> Transformer stage --> Dataset satge (flow 1)
Oracle --> Copy stage --> Dataset satge (flow 2)

Flow 1
It is fetching around 40 million record and the source table is a partition table and the speed is 55000 per sec
Flow 2
It is fetching around 20 million record and the source table is a non-partition table and the speed is 18000 per sec

when i check the oracle stage property, the executation mode is parallel for flow 1 and execuation mode is sequential for flow 2

Now my question is is the mode of execution depends on table partition. Because as i have explain before the number of rows per sec from the first flow is more than second flow.

Appreciate Any help

Yes, that's how the Oracle Enterprise stage works, and indeed how the Oracle Enterprise stage is documented as working.

Thanks a lot Ray..
So you are saying if source table is partition then only the source stage execution mode will be parallel. else it will run in sequential mode.

Is there any way to read the table (not partition) in parallel execution mode, This task i am doing for performance improvement.
Table partition may not be accepted at this point of time..all jobs are in production.

Appreciate any help

hi_manoj wrote:Is there any way to read the table (not partition) in parallel execution mode,

Not if you're using the Oracle Enterprise stage. "Maybe" if you're using an Oracle Connector stage.