DSXchange

The key in SDKSequences obviously got out of synchronization with the key of the Oracle table, or perhaps you are not using the correct sequence name in your KeyMgtGetNextKey() call.

Method 1 - write to hashed file with all columns as key then read from that hashed file.

Method 2 - sort data then set up a pair of stage variables for each column, one to detect changes and one, executed after the first, to "remember" the value from the previous row.

Please advise where this limit is documented. The solution is to build a string (initialize a stage variable) containing all 31 characters and convert that to "". The example below abbreviates the stage variable initialization expression. svStripChars <== Char(1):Char(2):...:Char(...

Status code 3 may mean that the job could not be reset. Check the job log.

I've encountered this error when writing to Oracle with NLS enabled. It seems to be a bug in sqlldr, because INSERT handles the same value quite happily. In my case it was a 94 character string, and the column data type was VARCHAR2(100). It was suggested that the data type declaration should have b...

Maybe because no-one else cares enough to harrass the vendor about it, mainly for the reasons in Vincent's post. As a sometime contractor, I know that I sign some fairly heavy duty non-disclosure agreements, and am happy to do so, thus even that I can access my clients' data is not a problem. What's...

Yes/no/maybe. If you export the job executables as well as the designs, then they should run OK. If you don't, then you will need to compile them first, perhaps using the Multi Job Compile tool (under Tools menu). I always like to compile jobs after moving them, because it alerts me if any required ...

Ok, so stop using it and "do it yourself".

Disable operator combination to learn which stage is throwing the error. This may provide more accurately-targeted diagnosis.

You can - physically, of course - move the project folder. But it won't work until and unless all the system tables that describe objects in it and all the internal pointers in those objects have been updated/repaired.

Create, export and import looks after all of that for you.

Change your job design to implement a DecimalToDecimal() conversion and make the metadata in your job design match what's in the tables.

Everyone knows that a DataStage server can support an infinite number of simultaneously executing jobs!

An "all terms" search for SMP, MPP and configuration yielded a number of hits, including this one

It's looking for the "resource" table which contains the text associated with message IDs. Check that DSHOME or APT_RESPATH is set.