DSXchange

Welcome aboard. :D

Try IBM or Google for starters.

In the subroutine itself. On return, the job will detect the non-zero ErrorCode value and abort.

Please post the entire error message. You may need to double-click on the log entry to get it.

And lose the habit of using kill on any DataStage process. It only ever leads to messy cleanup problems. For example, how do you know you correctly identified all child, sibling and parent processes?

Craft your job designs so that they don't hang in the first place.

From your data listing I would recommend trying with "|" as the final delimiter, unless you have 14 fields defined in the record layout.

Please verify that you have correctly specified the delimiter character as "|" and either 14 fields or 13 fields and a final delimiter of "|".

Set ErrorCode (the second argument to the subroutine) to a non-zero value. This has the effect of preventing further execution by aborting the job.

I continue to disagree that you "have to abort" in any situation, but the above is the appropriate mechanism in a before/after subroutine.

Define an extra output link from your transformer. This link carries one column, call it RowCount, derived as @INROWNUM. The new link leads to an Aggregator stage that applies the Last function (Max would do just as well, but Last is slightly more efficient) to the RowCount column. Then to another T...

Welcome aboard. :D I have worked at many sites, and seen many different formats. There does not appear to be any standard. Take a look at the (free) ETLStats tools, which you can find from either of the two links from Kim Duke's signature block. These will generate the information you require. After...

It's extremely unusual that &PH& is empty but for these two files. Do not delete .uvnolsmap (since it specifies the NLS map to be used for file names in that directory). If &PH& is otherwise empty you can delete .Type1, which is an indicator to DataStage how to treat this directory i...

Unless, of course, one were to write one's own server Funnel stage...

The value " 4 " is reported. This suggests that the field has surrounding space characters, and perchance its data type is a numeric data type, such as Integer of some kind. You need to fix either the data or the metadata.

Memory sharing is a technique only applicable to the Entire partitioning method on SMP systems; it enables less use of memory for the Data Set in question. If you don't like the warning, enable a message handler to demote it to Informational. You can not eliminate the warning; it is something about ...

Change the requirement. Anything you do within the job has to be done for every row. While you could update the table with the value of @OUTROWNUM for every row processed, it's unnecessary work and will slow things down. Then you'd complain about performance. Do it afterwards, when you only have to ...

You might also have been able to enclose the reserved names in double quotes in your SQL.