Page 1 of 1

COLUMN IMPORT (schema file used)

Posted: Mon Mar 14, 2011 1:26 pm
by vdr123
I have a job which is trying to read some datasets(DS) and writting to a table.
Data Flow is: (RCP enabled, using schema file in CE/CI)
DS --> COLUMN EXPORT(CE) --> TRANSFORM --> COLUMN IMPORT(CI) --> DB

The datasets are created by a different flow(say JOB#A).

The dataset has a timestamp as one of the column. The job#A is writting "*******" (Star's) into this column.

When I try to read it with the abouve data flow, it complains on the format of the timestamp, which is reading Star's.

But this same dataset can be read by a job with a dataset with RCP enabled.

The data flow above fails in Column Import(CI), compalining about the timestamp format.

Any thoughts!!!

Posted: Mon Mar 14, 2011 1:41 pm
by ray.wurlod
Check the timestamp format picture used in each job.

Posted: Wed Mar 16, 2011 8:13 am
by vdr123
I guess you are referring to "Timestamp string" under the "Defaults-TAB" of Job Properties?

This has the regualr timestamp: %yyyy-%mm-%dd %hh:%nn:%ss

I tried to change this, but didnt help.

Which APT variable has the default values for PX for INT or TIMESTAMP.

example:
In a lookup stage, if the lookup is not found, it returns a zero(0) if the field is NOT NULL as default.

In this case, for timestamp, its returning a "*" (star) as default after the join fails, as the timestamp is NOT NULL.

Thoughts!