DSXchange

DatastageSolutions wrote:Can anyone think of a neat solution to this problem?

No, at least not in a server job.

In the parallel environment you might be able to do something with a set of parallel jobs - push the first row out into a schema file then use that in a subsequent job.

Calling/using nested job sequences to achieve a more modular, and therefore more maintainable, design is a "best practice". While an activity is executing its controller (the job sequence) is effectively sleeping, waiting for the activity to conclude. DLLs on Windows are, by their very nature, share...

Please wrap your job design in Code tags. There is no limit on the number of DB2 stages in a DataStage job, from a technical perspective. From a maintainability perspective, there ought to be. It is apparent, however, that there is a limit on the number of simultaneous connections to your DB2 instan...

what stage type are you using to read the mainframe file? If it's the CFF stage, you can specify to convert the COMP fields into appropriate data types (e.g. integer or decimal). When these are written (exported) to a sequential file, they are converted by the export operator to text.

You can use an I-WAY Connector (in version 8 only).

The Complex Flat File stage can "read data from one of more
complex flat files, including MVS datasets with QSAM and VSAM files." (Chapter 10, Parallel Job Developer's Guide)

Please post the exact error message. What happens if you disable operator combination in the ODBC stage? In the stage downstream of it?

That, I believe, is the theory. I can't speak authoritatively though. But if you ask it to leave the intermediate files in place, you can count these files after the event, and see whether there are four or there is only one.

It's creating a date in %yyyy%mm01 format (the first of the month of the received year and month), as an integer, but with a modified century indicator. Is this, perhaps, intended for a JD Edwards data table?

Do any of the successful file names begin with a numeric character? My thinking is that your FTP tool does not like file names that begin with numeric characters. What actual FTP tool are you using? Have you tried with a different FTP tool?

The only thing I'd query is the trailing slash characters in the pathnames.

Check also that the folders for disk and scratchdisk are writable for all users, and that the file system does not become full when the job is running. Something is preventing a (large) file being created.

There is no limit on the number of stages in a DataStage job from a technical perspective (though from a maintainability perspective there ought to be). There may, however, be a limit on the number of simultaneous connections that can be made to your Oracle instance. Enlist the help of your Oracle D...

Identically partition the two inputs, based on the "join key" field(s). The lookup data will be loaded into memory in every case except a sparse lookup; partitioned data allow the amount of data per node to be minimized. Use as few fields (columns) as possible on the reference input link. Use as few...

It specifies the character used to pad strings, particularly Char(N) data types. Its default value is 0x00. You can change this default in the Administrator. To override the default in a particular job, add the environment variable as a job parameter - use $PROJDEF as the default value for the job p...

Use a custom trigger on the Routine activity. Allow 1 as the returned value.

Use a custom trigger on the Routine activity. Allow 1 as the returned value.