Fractional Truncation Error from Oracle

Post questions here relative to DataStage Server Edition for such areas as Server job design, DS Basic, Routines, Job Sequences, etc.

Moderators: chulett, rschirm, roy

Post Reply
peternolan9
Participant
Posts: 214
Joined: Mon Feb 23, 2004 2:10 am
Location: Dublin, Ireland
Contact:

Fractional Truncation Error from Oracle

Post by peternolan9 »

Hi All,
we have been doing 'day 1 startup' testing where we have been running full volumes of data through DS....one thing we noticed was the /dstage directory where we installed DS is getting used up rapidly....we traced the problem to fractional truncations where we accept a field which is 'number' and move it to number(38) (which means 38,0) and there are SOME numbers on the input tables that have decimal places........we are in a country where the local currency has no decimal places but SOME transactions are in USD....

Problem is, the rows with fractions are spread across the input data. We just didn't see the fractions before because they were buried down the tables quite a bit. We are finding it hard to find the specific columns with the decimal places....

My question is, is there any function in DS that will tell us how many rows or approximate size of a log generated by a particular job? If we can identify the jobs with sizable logs we can then run 10,000 rows and see if we can find the columns with decimal places so we can change them in the target.....

Thanks
Best Regards
Peter Nolan
www.peternolan.com
ray.wurlod
Participant
Posts: 54607
Joined: Wed Oct 23, 2002 10:52 pm
Location: Sydney, Australia
Contact:

Post by ray.wurlod »

That would be the HowLongIsAPieceOfString() function. :lol:

The only thing that can be predicted reliably is the number of entries that will appear in the log file if no warnings are generated. These are:
  • job starting, environment variable settings and job finishing (one each per job)

    active stage starting and active stage finishing (one each per active stage)

    messages from before/after subroutines

    messages from hashed file or table deletion/creation

    messages from passive stages that use SQL
Those are predictable from the job design. The remainder are unpredictable unless you deliberately introduce known errors.
IBM Software Services Group
Any contribution to this forum is my own opinion and does not necessarily reflect any position that IBM may hold.
Post Reply