DSXchange

NLS consists of two pieces - the character map (e.g UTF-8) and the locale. The locale is a means for representing different categories of data according to local expectations. The five categories are TIME (dates and times), NUMERIC, CURRENCY, CHARACTER (casing and categories) and COLLATE (sorting ru...

"Release" has a very specific meaning in DataStage; it means "release (a read-only version-stamped copy of the job) for production". I suspect this is not what you mean. If you mean deleting a job, then there is no easy way to do it from a UNIX command prompt. There are many interdependencies in the...

Take an extra output link out of the Transformer stage with the same constraint as your main output. On the new link have one integer column whose value is derived as the constant 1. Run this through an Aggregator stage and into the file that needs to receive the row count.

Hmm. Can you dump the score and see if keyChange occurs there? After all, it's the score, not the design, that gets executed.

No that's not what I said. Please read my post again. I was simply interpreting the underlying cause of your error message. Usually, though, I do prefer to use Char rather than VarChar, as it's better self-documentation. But DataStage itself can cope well with either. It's the Display (or Field Widt...

Or, if that's too long:

Try running DS.CHECKER - it tidies up after incompletely deleted jobs. You must have exclusive access to the project to run this command.

Once you have run DS.CHECKER, if the job is still present, try deleting it again.

"Short input record" indicates that the total widths of your Char data types exceeds the number of characters in the record. Hence the import failures.

A better design is to segregate the inserts from the updates, and run them as separate input links (insert only and update only) into the target stage. The "double-barrelled" update strategy will get slower over time; it has to try the insert then, if there's a failure, determine whether or not it w...

Can you evaluate DateGT( input.CDate_complt, input.CDate_start ) in a stage variable, then run the job in Debugger and see what is being calculated/returned?

Assuming the first lookup on tbl2.col2 returns tbl2.col1 you can use that anywhere downstream to perform a lookup against a hashed file containing tbl3.col1 as its key. You can do this in the same Transformer stage, but only by typing (not with the Expression Editor) - and it will stay red but will ...

Do you have a Sort stage upstream in which a key change column is specified?

In your code you need to assign the result of DSAttachJob to your Handle variable. Handle = DSAttachJob("Header_File", DSJ.ERRWARNING) RealName = DSGetJobInfo(Handle, DSJ.JOBNAME) If RealName = DSJE.BADHANDLE Then Call DSLogFatal("Can not attach Header_File job&quo...

Don't forget that you have to strip the line terminator off.

Use the Data Set Management tool to determine how many rows are actually stored on each partition (node).