DSXchange

It is not possible to use database constraints in an SCD stage therefore I ask again for you to describe exactly what you are doing in the SCD stage.

Grammar Notes In English sentences begin with a capital letter. The present tense plural form of the verb "to be" is "are", not "r". The first person singular pronoun is always capitalized: that is, "I". I believe that what you meant by "prallely" is "in parallel". There is no adverb of which I am ...

You may need to look in the logs of the jobs run by the job sequence as well as in the log of the sequence itself.

It can be done, provided all the preparation (HS.ADMIN etc) has been done and there are no I-types mentioned in the SQL (which will throw "compiled by incompatible compiler" errors). And you must have a UniVerse ODBC driver - it does not install with DataStage, therefore you must have genuine UniVer...

And the question is...?

The word "lock" will probably appear in the answer, once we have been informed of the question.

You're only encouraging them to post in the wrong forum!

You can't, unless you have an ODBC driver for UniVerse. All ODBC references from parallel jobs go through a driver - there's no concept of bypassing the driver on the basis of DBMSTYPE = UNIVERSE. But it remains the case that you can not introduce any blocking operation; that is, a passive stage wit...

Lookup failure is when the lookup fails - that is, the key value being sought does not exist in the reference data set. End of story.

All of the rules are in Chapter 2 of the Parallel Job Developer's Guide You are attempting to violate one of the fundamental principles of parallel job design by introducing a blocking operation - the output of the Hashed File stage can not be opened until its Input has been closed. This is totally ...

Your job design has become corrupted in the Repository. At least one stage record has been lost. You could try re-indexing, but I believe it's worse than that. Hopefully you have a backup (export) from which you can reinstate the job design.

There is no interactive debugger available for parallel jobs (though there is one for server jobs). You can compile the parallel job in trace mode, set on the Execution tab of job properties where you can also set a row limit per node, then run the job and examine a trace of column values handled by...

Not via rollback since you commit in job #2. What you need to do is to include some mechanism for identifying the rows that are loaded (such as a unique run ID or a timestamp) and delete these should that become necessary. If you need to "rollback" updates or deletes, it becomes a little more convol...

What's different about your ID? Do you, for example, have read permission to tnsnames.ora and the other ID does not?

Easy enough. One job to create the files, the other to distribute the data between them using constraints in a Transformer stage and job parameter references. The second job will make use of job parameters, and have as many output links as there are files (probably more output links - as many as the...

Have you set default values for these columns?