DSXchange

This is a great example of how you can use stage variables to remember values. A header line is indicated by H1 appearing in the second field. svIsHeaderLine <--- (Field(inLink.TheLine, " ", 2, 1) = "H1") The code is the final field in a header line. If on a non-header line its v...

If you Search DSXchange you will find the answer to this question.

Please try to come up with more meaningful subject lines.

Use a BASIC Transformer stage (or a server job) and, within that, use the Fmt() function - which you can read about in the DataStage BASIC manual or the on-line help - with a "70T" format as the second argument. This will do precisely what you ask. Convert any text marks into field marks o...

Probably.

When's the interview?

Are you allergic to reading manuals or searching fora?

PATH is your command search list. DataStage has searched for some command and been unable to find it because its parent directory is not mentioned in the PATH environment variable. The message about not being able to load orchoracle may be an artifact, or may be another path issue, this time the sha...

In short, not enough fields (= not enough delimiters) in some records.

DataStage parallel jobs expect/require consistent format.

Use server Sequential File stage if you need to handle missing columns. Or read the line as a single string and parse it within your job.

Not those ones. A Merge stage perhaps, but that won't be much different in "performance" (whatever that is) from a Join stage.

There's no good reason any more (since version 7.5) to avoid the Transformer stage. However, what's "best" depends on many variables. Try both, and see what works for your site.

This combination is not possible in version 7.5. It becomes possible with the Super Operator role in version 8.0.

Stop using your old password and use your new one.

Do you have the "return multiple rows from reference link" option selected?

Meanwhile, back on topic, the correct answer (mathematically) is that "division by zero is not defined". In a DataStage parallel Transformer stage it should cause an exception.

Join does not store data anywhere. It relies on the fact that both inputs are sorted to be able to process just one key value at a time.

First time I've encountered COBOL being described as a "liberal art".