DSXchange

ray.wurlod

svWeekDay = Oconv(intDate, "DWA")
svMonthDay = Oconv(intDate, "DD")
svFirstSunday = (svWeekDay = "SUNDAY" And svMonthDay <= 7)

ray.wurlod

Otherwise, determine what format DB2 expects dates to be in. Perhaps by selecting some.

ray.wurlod

You're in a very bad situation, then. How would you do it with any other software? Seriously - all you can do is write to a temporary file then, once the job has completed, delete the source file and rename the temporary file. Or use different directories.

ray.wurlod

"Too many columns in record" stems from the fact that the total of column widths (field widths) is less than the actual number of characters in a row (record) in the data. That it worked subsequently suggests that the later set of data had the correct number of characters per line. One cause of this...

ray.wurlod

We have only your assertion that the numbers are "weird". What do you mean by this term? Where's the proof?

Have you tried setting the environment variable that causes player processes to report their row counts?

ray.wurlod

It's known as a "fork join" design.

Code: Select all

----> Copy  --------------> Join ---->
       |                     ^
       |                     |
       +---> Aggregator -----+

ray.wurlod

Why do you claim that you can't group by date in the Aggregator stage?
I don't believe you're right.

Have you tried converting the date to DataStage internal format (integer) and grouping by that?

ray.wurlod

Doesn't always work, of course. Nine women can't have a baby in one month. This is what's known as a sequential process.

ray.wurlod

It's not a bug. It's propagating.

ray.wurlod

Get your head around the fact that you can use Join stage to join the columns back. If "they" say you can't (that is, aren't allowed to), then fight it out with "them".

ray.wurlod

The WHERE clause in the SELECT statement is more efficient, as you end up processing fewer rows and therefore finish faster. If you insist upon filtering within a Transformer stage, then the proper location is within the constraint expression for the output link. The logic is identical.

ray.wurlod

Nothing is automatic. Data types will be converted implicitly if this is feasible, but will generate warnings/errors if not.

ray.wurlod

Do you seek something like a review of existing work?

ray.wurlod

Cleanup Resources is a tool in Director, enabled from Administrator.

Performance monitor is an independent Java application (JobMonApp) that has its own start command, but which is typically setup to start automatically with DataStage.

ray.wurlod

dsjob -jobinfo $prm_project $prm_job

DSXchange

Search found 53125 matches

Advertisement