DSXchange

It looks like you will have to debug this the hard way. I would remove all columns that do a 1-1 mapping, then recompile and text. Chances are very high the error will still be there. Then start removing single columns, starting with the most complicated derivations first until the error goes away a...

qutesanju wrote:D:\Ascential\DataStage\Projects\proj1\sendmail_SubMasterLoad18608_618_2320.bat

Perhaps that batch file is being called from a before/after job subroutine call. Examine the .bat file to see if that is where your error is coming from.

While you cannot prevent the file creation, you could add an after-job call to UNIX which would delete those files if empty.

In principle you can use the multinstance capability if all of the data sources, interim data, and targets are parameterized and distinct. I would feel uncomfortable doing this, particularly as creating a separate project is simple to do.

Job names must be unique within a project, so that is not possible.

xinhuang66 - My car wouldn't start this morning. I filled it with super unleaded. What is wrong with it? :D


hint - please post your compile error message if you wish qualified assistance.

The generated transform C code is stored in the RT_SC subdirectory with the job number. The easiest way to find the code is to attach to the project directory and then do a "find . -name *{TransformStageName}*.trx".

Are all the tables in the same database? That is probably the most important factor for this question.

Are you connecting through something like Citrix or a similar application?

This is also V8, so the user DataStage uses to create the file might be different from the userid with which you login to DataStage.

It could be that you have reached a file system limit. Attach to your project directory and try entering "mkdir testdir", if successful just "rm testdir" - job 4700 is quite a big number. What OS are you working on?

Does the original job compile or does it have an error?

No such utility exists; this is the major drawback of using a looup fileset and some tests that I did at V7 showed that the performance gains of a lookup fileset versus a dataset were not particularly great. For this reason I don't use them anymore.

This is likely happening when your data contents for a record type is not the same as the PICture data description; perhaps on the last record type in your data.

Check the directory RT_CONFIG4700 in your project directory; does your user have access rights to it?