DSXchange

No "magic" transformations occur - only the one(s) you specify. Use a hex editor of some kind to find out what this character is.

One possibility is that you have specified UNIX-style line termination, where the file is acutally using DOS-style.

Yes, but I'd rather tell you not to do it. Why can't the job just end gracefully when the required numbers have been generated?

Unless, of course, you install a full DataStage Engine on every machine. Talk to your account rep - they'd LOVE to hear from you ($$$$$$).

You could try here

Yes. But you will need user-defined SQL, and you will need NOT to pass the key column value from DataStage.

Do you regard a Bourne shell script as an executable file? The analogy is exact.

Partition and sort by the key column, and generate a key change column in the Sort stage. Downstream of that, accumulate your list of non-key columns in stage variables, initializing to null on key change. Downstream of that select the last of each group using a Remove Duplicates stage.

In order for operators to be combined, they must be (a) adjacent, (b) combinable (that is, not have combination disabled), and (c) use the same degree of parallelism. If all these conditions are met, DataStage will most aggressively combine operators to generate the fewest possible number of process...

The Aggregator stage has a Decimal Output property both for the stage (as a default) and for each output field. From memory, I think its default value is "8,2". These numbers are precision and scale respectively.

Moderator: please move to server forum

The answers there will be different.

Preventing the jobs from aborting would be favourite.

What compilation options have you set in your job sequence? Are you using explicit Failure triggers on the Job activities? Is there an Exception Handler in your job sequence?

That format should import OK if your date format string is "%yyyy-%mm-%dd", the year component always has four digits, the month component always has two digits, the day component always has two digits, and all the component values are valid in creating a date.

Yes, DataStage expects constant record length. If FILE AID is too burdensome, could the application create a FILLER to make the shorter record the same length as the longer?

Otherwise you're going to have to have DataStage read and combine the two files.

If it exists for your project then it may be.