Is there any down side of not providing the lengths for the Varchar fields through out the entire job flow?
We have created few jobs and removed the lengths for the Varchar fields from all the way source (Sequentail files) to the target (Datasets).
We removed the lengths for the Varchar fields even in the Transformers, Joins.. etc within the job flow.
Trying to understand , if there there any down side of doing this.
Will this approach have any impact with DataStage 8.1?
Thanks in Advance.
Lengths for Varchar Fields
Moderators: chulett, rschirm, roy
Lengths for Varchar Fields
Suren V
No significant downside and no difference at Version 8.1 to the former version.
You will have to ensure that you don't write a VarChar value that is longer than the target (e.g. an Oracle table) can contain, but using unterminated VarChar columns within DataStage jobs is not a problem - in fact it is an advantage in several cases such as when writing to Datasets (storage space is minimized).
You will have to ensure that you don't write a VarChar value that is longer than the target (e.g. an Oracle table) can contain, but using unterminated VarChar columns within DataStage jobs is not a problem - in fact it is an advantage in several cases such as when writing to Datasets (storage space is minimized).
I think you might get a mild warning on some reports that the column length doesn't match the database setting. But I'm not at a site with Metadata WorkBench installed, so I can't confirm.
I seem to remember getting that warning in the director logs as well. It is easy enough to get rid of, just put a length on the column when it goes to / from the database.
I seem to remember getting that warning in the director logs as well. It is easy enough to get rid of, just put a length on the column when it goes to / from the database.