DSXchange

Not really. Because you never NEED to use them. However a common use is when obtaining the result from an Execute Command activity in a job sequence. The command that is executed has one or more line terminators, which are "translated" into field marks by DataStage (to achieve the same result irresp...

Welcome aboard. What you are trying to achieve is called a "vertical pivot". Search DSXchange to learn various techniques for implementing vertical pivot. Yours is a slight variation in that, having accomplished the vertical pivot, you then require to concatenate the columns. But that's no big deal ...

Welcome aboard.

When you import the table definition into DataStage, what data type is recorded in the table definition?

All target tables' rows have some indication of what run it was that inserted/updated them. By this means even committed changes can be undone if that is deemed necessary. The indication can be a unique "run batch" identifier and/or a timestamp. Committed deletions are trickier - you need to build a...

Like what? Arnd's descriptions are reasonably complete. They are system variables that contain particular byte values which are important in the DataStage Respository, UniData and UniVerse database products, primarily as dynamic array delimiters.

The reason for rejection is NOT stored in any DataStage repository table.

First line is column headings?

You may need to suppress reading of this row if such is the case.

There is a little-known feature in the DataStage Director that allows you to navigate from one job to another without leaving Log View. This is typically used when job control - such as that executed by a job sequence or by a server job containing job control code or by a Batch job - invokes another...

What is the value of the NLSMODE parameter in your uvconfig file?

What backup/export do you have?

NLS is enabled, and default map selected for a project, using the Administrator client NLS button once that project has been selected.

Should your output be 2008-08-31 ?

Who knows? Define "performance". Nothing to prevent you using more than one Lookup stage if monitoring shows that one Lookup stage with many reference inputs is a bottleneck in the job design.

Welcome aboard. Oconv() is not available in parallel jobs.

Use the IsValid() function.