Hello:
In the job I do a lookup on the DB2 Dimension table and output it to a Dataset. This is a seperate job.
So I pull the COL_KEY (PK) and COL_ID (FK) into the DataSet.
QUESTION:
Can I inverse the key column and use this Dataset for lookup.....?
So I set the COL_ID to be my Primary key col and COL_KEY the FK
As per my understanding I have to create a seperate dataset and lookup job cause I cannot just flip the key columns in the dataset and use this dataset.
However I am told I can use the old dataset and use a copy stage change the partitioning to Entire on the Input Stage and change the COL_KEY which is the PK to FK and use this same data set. Is this correct...?
Thanks in advance.
MJ
Lookup and DataSet Partitioning
Moderators: chulett, rschirm, roy
Lookup and DataSet Partitioning
Thanks in advance,
So initial lookup job do the following:
Job1
Now to flip the key columns I do the following:
Job2
Some jobs use Dataset_A and some jobs use Dataset_B.
However I am told I can use the Dataset_A with the copystage to flip the key columns.
Job1
Code: Select all
DB2 Table ---> Transformer ----> Dataset_A
Now to flip the key columns I do the following:
Job2
Code: Select all
Dataset_A --->Transform------>Dataset_B
However I am told I can use the Dataset_A with the copystage to flip the key columns.
Code: Select all
Dataset_A ---->Copy_stage---->rest of the job design.
Thanks in advance,
Yes you are right, you can use a copy stage.
But I think one transformer is redundant, you are not doing anything with the dataset A but using a transformer before and after dataset A is created.
Code: Select all
DB2 Table ---> Transformer ----> Dataset_B
|
|
rest of the job design
Last edited by sud on Mon Apr 30, 2007 10:21 am, edited 1 time in total.
It took me fifteen years to discover I had no talent for ETL, but I couldn't give it up because by that time I was too famous.