DSXchange

You might try the Stored Procedure stage. Or the Open Command or Close Command options in the Enterprise ODBC stage.

Release all locks held by the PID that had the job locked. Is it still the same message, or a different one, that leads to your "still cannot open the job"?

Data types a strictly enforced in parallel jobs. You would need separate output fields or, indeed, separate output links (not possible with Modify stage) to achieve what you describe.

There is exactly that tool coming in the next ("Hawk") release of DataStage. It generates hyperlinks to the differing parts of the job design, routine code or whatever, so that you can analyze further. But there's nothing in the current version (7.5.1) or earlier. All you can do at the moment is com...

Have you done anything to the uvsh executable (in $DSHOME/bin), such as linking GCI subroutines? You could try renaming uvsh and dssh and re-installing the server. Otherwise do as the message says, and contact your support provider. Core dumps can occur under a number of scenarios, which are describ...

Data elements are cool. I think of them as "luggage tags" for data so that, for example, I can identify one ten digit number as a telephone number and another ten digit number as a part number and yet another ten digit number as a money amount with an implied decimal placeholder. Businesses (or audi...

If it's a date field, the default value must be date of some kind. Maybe use a date that won't occur in data such as 9999-12-31. Otherwise you need to declare the output column to be of type string[max=13] and use a string_from_date() function for the valid dates. It will be tricky in a Modify stage...

Are you actually looking at the job control code for the correct job sequence? We've all done similar things before; gone round and round trying to eliminate an error only to discover that we're looking at the log for the wrong job. Could be something similar here?

There are optional pieces. You can specify truncation or rounding. It's all in the manual.

Modify stage. Use int32_from_decimal() or int64_from_decimal() function.

It's not even a solution for server jobs. You have increased the T30FILE parameter from its default of 200, which means that you can have more dynamic hashed files open. You have disabled automatic sync on close of DataStage files. You have changed the depth of the group latch and record lock tables...

No. You have to persist the highest value, either locally when you run your job (use an Aggregator stage with the Last function or Max function on the key column) or by select max(key) from table; before the next run.

In the Lookup Stage Conditions dialog select the link name from the drop down list captioned "Multiple rows returned from link".

Server jobs are not going away any time soon. They are fully supported in the new ("Hawk") architecture. Sales dudes seem to have confused the fact that there won't be a separate server edition in future with the non-fact about the disappearance of server jobs. Don't be misled. Server jobs are not g...

Pivot preserving @INROWNUM.
Perform lookup, recording found/not found result.
Vertically pivot, recording "And" of results.
Continue if result still equals TRUE.