DSXchange

CATEGORY is an indexed column in each of the Repository tables in which design-time metadata are stored. That is, because they are in a series of database tables, they are not accessible from UNIX. As you note, it would not be safe to issue an UPDATE statement against the table, because this would n...

How are you handling them on the older version and how are you handling them differently on the new version? In particular, what character maps have you specified?

Then post your question on a UNIX script forum. DSXchange gives DataStage solutions!

How do you propose to capture the SQL code IN the UNIX script? You can use the exit command to return the value - that part's easy.

Heck no. Count back two years. Subtract one from the day if today is February 29th.

That, my friend, is called consulting, and is how some of us make our living. If you would like to post what you have attempted, and why you believe it does not fulfil expectations (with error messages if any exist), then we would certainly comment upon that. Make sure your code is nicely formatted ...

You start by informing us what the target database is, and what its expected date format is. Then you adjust your value accordingly.

It is not possible using the GUI. All you can do is to create the new category, move all items from the old category to the new category (you can use drag-and-drop), then delete the old category.

Put an EXISTS test in the extraction.


Or include a sequentially-executed Transformer stage constrained so that its output link passes only one row. For example:

Create a reject link from your Enterprise database stage. It will automatically contain the columns from its input link plus a column called sqlcode that will contain the SQL error code returned from the database server.

WHAT worked?

Once you've posted the answer to that question, please mark the thread as resolved.

A Copy stage with no output link is an excellent placeholder while you are effecting incremental design. It will consume all the rows sent into it and have nowhere to transfer them, so will simply discard them.

Be more specific about the UPDATE statement. It is in the very nature of DataStage to send a row to be updated for every row that arrives in the stage. The only way to update only one row is to cause only one row to arrive in the stage. To update only one row for each row received into the stage the...

Far better (and automatic) statistics collection is performed in Version 8 plus you can get an all new resource estimator that can be based on the job design alone or on running a sample of data.

At the menu prompt enter each number followed by a question mark; for example 2? - this will give you more information. Which ones to change depends on where your performance problem is. Increasing the reporting interval (so that the stage updates its status less frequently) will always help. Sorted...

Single line batch routine that invokes your UNIX script using DSExecute()