DSXchange

And even then you may need to get rid of the translated line terminator characters.

Most easily in a separate job, run after the first, that checks the link row count in the first job and generates the zeroes row conditionally.

What is the precise value returned by the UNIX command (including all line terminator characters)?
How are you loading that value into your job parameter? Are you, for example, trimming any line terminator and/or field mark characters?

Accepted. If you had premium membership you would see that I posted a suggested solution function. It's a worthwhile investment, and helps to fund the hosting and bandwidth costs needed to keep DSXchange alive.

Probably from /etc/profile or your .profile script (in your home directory), possibly from the DataStage dsenv script.

After only two hours?!! Some of us have to work for a living. We don't do "urgent" here. If you need urgent help, sign up with your support provider for premium service and learn the true cost of urgent. Put in some diagnostic calls to DSLogInfo() or UtilityMessageToLog(), for example to report the ...

Are you certain that job #3122 was started from the project in question? Could it have been started from a different project?

Nor do you have version 8.1 of DataStage. It is scheduled for release in the second quarter of 2008.

This is why I took the trouble to offer a Transformer stage solution.

I would have thought as you have. Inspection of the generated job control code - especially the part where the job sequence waits for multiple jobs to finish - might answer the question formally. As a workaround, you might put a command such as SLEEP 2 (via ExecUV) in to each job as an after-job sub...

In the metadata on the Columns grid of the Modify stage input define ROW_ID as Integer, not BigInt. If you are defining it using a schema file, specify uint32, not uint64, in the record schema.

The approach you suggest in your subject is the correct one.

You can write the after job subroutine yourself or hire someone else to write it.

Please post the current version of your code, along with the results of any diagnostic attempts you may have made (for example capturing the actual command that was sent).

It seems suspicious that a deleted job is still running - DataStage ordinarily would not permit deletion of a running job. Are you certain that job #3122 was started from the project in question? Does the process that is the running job hold any locks in the DataStage repository (use the LIST.READU ...

A fully general solution using this approach would require an unwieldy expression involving ten Index() functions, one for each numeric character, and a lot of Or operators. You would probably be better off either using QualityStage (as mentioned) which has inbuilt name and address parsing capabilit...

Try creating a routine with at least one input argument. DataStage expects to pass at least one value to a routine and get something back.