DSXchange

Binary data types such as numbers, dates and timestamps have no internal format. Formatting is applied only when converting from another data type on input or output. So the important question is what is your output (sequential file or database) and what is the datatype defined as in the target. Not...

I'm not sure I understand what you are doing. Have you declared any output columns from the connector? You stated that it hangs for a long time - what message do you see after that time?

Sainath - BINGO, I hadn't thought of that approach and I think that your recommendation will work for us here. Thanks.
p.s. No help from MKS support yet, it seems that the only way to use named pipes is to program them from c.

I realize that this was a deliberate design decision, but it is counterintuitive that a VarChar() column in DS is internally represented as a Char() field for DataSets and thus I maintain that is an anomaly. I gather that this behaviour was different at releases prior to 7.

You are correct, Ray. Just force of habit from ye olde UniVerse days which had no DSExecute() call.

A DataStage job sequence runs as a single process. Parallelility in job execution is achieved by using the asynchronous job run capability of DataStage, i.e. a call to start a job can return to the caller immediately while the job runs in the background. In this case you are not calling jobs but cal...

There is an anomaly with Data Sets where a Varchar(nnn) column will use nnn storage space per row regardless of string contents whereas an unterminated string will use allocate just the number of bytes in the string.

That is a much better explanation, but you are missing the values for employee 200 in the reference stream - you can omit the employee 100 values since you stated that those work correctly. At least by running with 1 node you eliminated partitioning as the source of you problem.

This is not the initial value of the stage variable, but the derivation.

The topic should be "How to handle null posts in DSXchange"

I just read the rest of the program. It won't work at all, there are sevearl problems with it, including that DataStage BASIC has no idea of what commands such as "sqlplus" are... Here is a modified version: JobHandle = DSAttachJob(jobname,DSJ.ERRNONE) Jobstat = DSGetJobInfo(JobHandle, DSJ...

The "IF-THEN-ELSE" construct functions in that if you have the format "IF (condition) THEN (statement)" then only one statement is executed and you don't need an "END", but your format has the "THEN" at the end of the line, which means that all the statements ...

You should do as suggested and use a modify stage or a transform to convert both source columns to the same nullability in order to avoid this problem.

While the DS_AUDIT hashed file contains information on the modification dates and times it doesn't store the compile time, in order to find that when the director log has been purged one would need to look into the appropriate RT_STATUS file - but the file structures for both DS_AUDIT and RT_STATUSn...

What data type does DATE_PROCESSED have in DataStage and what is the COBOL PICture for it?