DSXchange

Split the job so that rows are written to a staging area (text files) by one job, then loaded (server job?) from those. That way you have preserved the data to be loaded even in cases where you need to rollback. You can monitor the load in your load job so that, if you want to restart from a known p...

Check out the prices of ODBC drivers from Data Direct then post back if you still want to persist with this approach.

We're thinking about adding an Australia New Zealand user group to DSXchange. If you are based in Australia or New Zealand and we can't determine this from your profile, please respond to this post, indicating in which city you are working. You never know, it may lead to beer!

Better than that, why not read all about it in the help? The IMS components give you the ability to access IMS databases and viewsets and to import table definitions from IMS definitions. You get an IMS stage type once IMS is licensed. Open designer help (C:\Program Files\Ascential\DataStage\dsdesig...

Or does it mean that DB Connect v8 can not send correct commands to DB2/390 v7 (version mismatch), or does it mean that DB2/390 and DB2/UDB are different beasts?

Check the Format tab of your Sequential File stage. If the "null representation" field is empty, "" ought to be translated to null. Also check the Columns grid in case the null representation has been overridden on a per-column basis - you will need to scroll right to find this setting. Otherwise, i...

Your subject was "unable to export", not import. Re-indexing the Repository database makes the export utility more likely to be able to find all the components. To import a successfully exported project into a different project ought not to require any re-indexing, unless that project also had a cor...

Yes, but you will need to install/license the IMS components of DataStage. Contact your vendor for pricing.

Not if you want/require int64. Even with decimal_from_decimal, if the target is a smaller size than the source, DataStage will generate a warning.

In star schema terminology, do the snowflakes first, then the dimensions, and finally the fact tables.

The best Oracle to Oracle performance is usually to be had from not using DataStage. Every layer of software in the picture adds resource consumption. However, if the transformation rules are sufficiently complex, the argument for using DataStage becomes more compelling - there is a limit to what ki...

If the view is defined with WHERE Table1.ID = Table2.ID then you don't need that particular WHERE clause in a SELECT from the view - indeed, it would generate an error.

Additional Suggestions: Disable any screen saver on the Windows server. Verify that the firewall and/or anti-virus software is not imposing too great an impact on startup times for executables. Verify that the Windows server is not being used as either a primary or backup domain controller. Monitor ...

Only load into the hashed file those columns - and indeed those rows - that will actually be needed in the job. This will help in a number of ways. First, you are not doing unnecessary work moving these data into the hashed file. Second, you are keeping the average record size in the hashed file sma...

You could also use ExecSH (or your own routine) as a before-job subroutine to echo the header into the file, append detail records in the job and use ExecSH (or your own routine) as an after-job subroutine to append the trailer using echo and the >> redirection operator.