DSXchange

The only mechanisms that you can use in a parallel job are two. (1) An interlude before/after subroutine. (2) An interlude transform function routine that can be called from a BASIC Transformer stage (not a parallel Transformer stage). If you create and compile a before/after subroutine successfully...

I suspect that the warning is generated from the metadata. Decimal(11,0) implies a string of 13 characters (including sign and decimal placeholder), so that the metadata generates a warning because your source is smaller than this. Can you try changing the source data type to VarChar(13) either in t...

Read the documentation again. What you are requesting is not how it works (for a number of reasons). You are up for a step of renaming the files.

APT_CONFIG_FILE needs the pathname of the configuration file, not just a parent directory. Conventionally the file name has the ".apt" suffix.

It seems to me that the $APT_CONFIG_FILE environment variable value is not a job parameter in your job design.

First of all I would not regards Optim as a substitute or replacement for DataStage. The tools have tasks that do not overlap (though both may be included in a solution that, for example, requires ETL on masked data). Hashed files can really only be read by DataStage server jobs or by queries execut...

Yes, I believe they're written there when the connector detects a failure in what it's connected to.

Possibly you're not looking in the right place, in much the same way that you're not posting this question in the correct forum.

You have to use sorted data if the aggregation mode is Sort. The Aggregator makes use of the fact that data are sorted by grouping keys to minimize the amount of memory it needs - it only need to keep one key value in memory. Hash mode means that the Aggregator has to keep a table in memory with a r...

I think the failure has to be from the database end.

The answer to your next question is "Remove Duplicates stage preserving Last row in each group and Column Export stage."

skhan2@gmx.us wrote:I have to change 400+ jobs.

So about half a day all up, assuming you can do two per minute (not difficult).

Include in the loop another check for a file which, if it exists, triggers the sequence to shut down gracefully.

ExecSH is not the correct type of routine to call from a Routine activity. ExecSH is a "before/after subroutine" that does not return a value. You need to build a "transform function" type routine. Something like: FUNCTION CallExecSH(Arg1) DEFFUN ExecSH(InputArg, ErrorCode) CALLI...

rkashyap wrote:At IBM Insight, what is the correct forum for giving such suggestions?

There isn't one. That's not what IBM Insight conference is about. Use the standard enhancement request mechanism.

No practical limits in either scenario as far as DataStage is concerned.

But you will need to maintain them, and select parameters from long lists. That might be the governing factor.