DSXchange

OpenLink is another provider of ODBC drivers. There may even be one available from Unisys.

If your data types are genuinely date or timestamp, then the format is irrelevant; internally they are stored in binary format. If the data type is string of some kind, then there are conversion functions such as TimestampToString() and StringToTimestamp() that you can use with appropriate format st...

Welcome aboard.

DSXchange has a Search facility that can often yield the required answer faster than posting a question that has been answered in the past.

What happens if you use Entire as the partitioning algorithm on the reference input link? I'm asking as a diagnostic question, not necessarily with a view to recommending this as the solution.

There was no such bug. The problem would appear to have been between the keyboard and that job's developer's chair.

This is a known bug in DataStage, if you explicity specify a partition in Transformer Stage it throws this Warning message. You can add this warning to an Existing Message handler or new Message handler in your project and demote it to Informational. It's not a bug, as I've explained before. It's a...

Can you open the file RT_BP985/JOB.150667382.DT.1485838458 and post a few lines either side of line #49 ?

Assuming that the real names have more than one character:

Care to share the URL? A search at ibm.com didn't find it.

You can also "escape" the dollar sign. \$APT_CONFIG_FILE

This and "hard quotes" are standard shell scripting techniques. Perhaps some learning is required.

jobmon_ident is simply the identifier used by the job monitor - it is not the job monitor itself. If the stage is not there, it is not possible to monitor its output. You need to adopt a new mindset. What gets executed is not necessarily exactly what you designed, though it is functionally equivalent.

What's different at the times when it works and the times when it fails? What else is happening in SAP then? The message "waiting for BW to indicate..." suggests that the problem is on that side rather than in DataStage.

RowSplitter and RowMerger stages are available at least back to version 7.5, maybe even to 7.0.

You need a "rules dispatcher" routine and individual routines that check each rule, returning a true/false value. Use "indirect call" (see DataStage BASIC manual) to achieve late binding in the rules dispatcher; your job passes the rule name "Rule42" and the rule routine has a name that is, or inclu...

You can call a stored procedure that returns a record set (if, indeed, this is what you're talking about) using a connection stage, such as an ODBC stage.