DSXchange

This style of message usually means that the shared library search path environment variable has not been properly set. The name of this environment variable varies depending on the UNIX; it might be LD_LIBRARY_PATH, SHLIB_PATH, LIBPATH, and so on. Determine where the missing libraries are (using fi...

And the other thing is, this being a sharing place, is that we share our solutions. Can you please report how you solved it?

And the other thing is, this being a sharing place, is that we share our solutions. Can you please report how you solved it? If you'd still like help on the stored procedure from ODBC stage (which CAN be done), please post any error messages that appeared in the job log when you tried this approach.

UniVerse has standard set operations. These are UNION, INTERSECT and DIFFERENCE.

Sorry to disappoint the Oraclers, but MINUS is Oracle only.

There are obviously two difference DIFFERENCE results, but the question ought to return 0 rows if the tables are equivalent.

"(" has special meaning to the shell. What if you "escape" the character (precede it with a backslash), for example ?
This is what you would do in UNIX if you wanted to avoid messy quote characters (for example cd \&PH\&)

Bugs is overrated. Daffy had more personality, while DataStage developers should have most empathy for Wile E Coyote (the business analysts are represented by Acme Corporation).

Can you not use expressions of the kind Sumitgali postulated in version 7.5? That is, use DSGetJobInfo(DSJ.ME, DSJ.JOBINVOCATIONID) in the derivation for the invocation ID?

The easiest way to do this in DataStage is as follows.

In a Transformer stage construct a stage variable, which accumulates the comma-delimited string.
Run the output through a subsequent Aggregator stage, using Last as the aggregate function.

Process this in UNIX, using a utility like sed to change all the line terminator characters to commas.
This leaves a blank final field, you can further process the result if this will be a problem.
It can be done in DataStage, but is easier and faster using the sed utility.

Search the forum for "vertical pivot" (for this is what you are seeking to do). You will find a number of techniques described.

Of course there is, and you do it the same way you'd do it with SQL. Insert the first batch of rows, with suitable default values for those columns to be loaded in the second pass. In the second pass, update the rows with appropriate column values, but only for the columns you need to update. In thi...

How do you "give" the invocation ID? Surely you can simply "give" the same invocation ID to the other jobs in the sequence in exactly the same way.

If it's already in the hashed file, it's a duplicate. Keep writing to the hashed file, so that later duplicates will be detected. To capture only the duplicates, test whether it was found. For example Not(RefInputLink.KeyCol.NOTFOUND) or Not(IsNull(RefInputLink.KeyCol)) The f...

In the job PT05OTDisReceiptRegularLoadControlJob make sure you're trying to process E:\HLLDW\PRETEADW\Inprocess\DataFeeds\OTDISPATCHRECEIPTFEED\251004otinteb_new.csv, not E:\HLLDW\PRETEADW\Input\DataFeeds\OTDISPATCHRECEIPTFEED\251004otinteb_new.csv Feedback from your MOVE (or was it COPY?) command i...

You can generate it with Char(255). Beware, however, that DataStage treats Char(248) through Char(255) in a special way - depending on the context. These characters are what are called dynamic array delimiter characters. If you have NLS enabled, genuine characters in this range will be mapped to the...