DSXchange

You must use the backups/exports you did before upgrading.

Uninstall 8.1 completely, install 7.5.1 then import from your exports (or restore from backups).

It's obvious you're hitting a precision limit somewhere. I trust that you did uvregen and restart DataStage after changing EXACTNUMERIC ?

Create a job parameter called, say, jpRunDate. Assign its value from a job sequence that generates the appropriate value, whether from a function, a macro, or whatever. Use a reference to this parameter instead of current date in your WHERE clause.

That would depend entirely on how you specify "current date" and how the database server handles that. For example, if you specify it as "job start date" then there's no way it can carry over.

Try floor() for rounding down. Explore all the Transformer stage functions in Appendix A of Parallel Job Developer's Guide manual.

Why do you "have to describe" them? Metadata for the Repository tables is incomplete. This is, and always has been, a deliberate decision by the manufacturer. So, while LIST DICT tablename will show you the extant metadata, there is usually far more in the record structure than is shown in these re...

Worst case - every row has the same key value. Every row from the left input will therefore match every row from the right output. You'll get 100 rows output.

It would have been useful had you posted the error message generated rather than "it was not working". Regular SQL rules apply even in DataStage/SQL. The routinename is a string (VarChar) data type, and so the value in the WHERE clause needs to be surrounded with single quote characters. UPDATE DS_R...

Date data types don't have a format - they are binary. The format picture in TimestampToDate must be that of the timestamp.

The default date format (for display) can be overridden in the date column's extended properties.

There are different kinds of routine, including transformation function (callable from Routine Activity and server jobs), before/after subroutine (callable from job properties of server and parallel jobs and active stages in server jobs) and parallel routines (callable from Transformer stages in par...

The version 8 DataStage Essentials course code is DX444 (DX434 is for version 7.5).

Then the only remedy is to edit the table definition after you've imported it. Add a comment in the Description field to alert other developers that you've had to do this, and why.

Are you talking about a Transformer stage here? Without that information it is really difficult to proffer any advice at all. For example you may be better served by Column Auto-Match than by Propagate to reinstate mappings across a Transformer stage.

Use some of the environment variables like APT_STARTUP_STATUS that let you track what's going on.

Read the remainder of my post, which deals with suppressing the warning.

Premium membership costs less than 30c (Rs12) per day. These funds are totally devoted to assisting with the hosting and bandwidth costs incurred by DSXchange. It is a worthwhile investment.