DSXchange

Not DataStaging today, so I can't check. First step would be to check whether it's documented as behaving the same way, or whether different control codes are used for familiar (from server jobs) functionality.

I also suspect you're not running version 7.5x2 This is the only pre-8.0 version that supports running of parallel jobs on Windows platforms. It is not supported on Windows XP, but can be gotten to run there. You may also have difficulty compiling a parallel job that includes a parallel Transformer ...

Now that you've copied the relevant section, there's no need to read the Parallel Job Developer's Guide !

The command is dsexport, which has a /JOB=name option. Search the documentation or the forum for full detail.

Note that exports are initiated from the client machine, and the export file resides on the client machine.

Get a better requirement. There is no sane reason to restrict this design to a single job. The Sequential File stage can generate the actual file name read into an additional output column. You could use this to filter the rows, but you must already have in place in the job an output Sequential File...

No. That's not required. (That is, I don't have that in place, and it's working fine for me.)

Premium membership is only a few cents per day, is definitely worth it, and helps to pay for the bandwidth costs of DSXchange.

As well as havingAPT_ORCHHOME set and PATH including $APT_ORCHHOME/bin you must have APT_CONFIG_FILE set (pointing to the same configuration file that was used to create the Data Set), and your shared library search path must include $APT_ORCHHOME/lib.

The job type is marked as server edition and it's posted in the server forum. On that basis, let's offer a server solution. It can't conveniently be done in the job itself. In the job sequence that controls the job use a Nested Condition activity to determine whether the file name (job parameter?) i...

No spaces around the comma.

If there is a Trillium plug-in for version 7.5 then it is supported. IBM does not cease support of part of a current product version. Challenge the consultant to provide a statement in writing from IBM that it is no longer supported. Plug-in stages are written to a consistent API; chances are that y...

There is no more information in the error message than the VALSORT module ran out of memory, or tried to access a memory address that its process did not own. It could be a difference between versions. Or the amount of data could be sufficiently larger to trigger this problem. If there were any othe...

The job status is available on file unit #2 (stderr). If you specify the -jobstatus option, the job status of dsjob is the exit status of the job itself, on file unit #1 (stdout). It is documented thus.

Now I'm confused. All of the following must be in SHLIB_PATH, and before any database libraries:
$DSHOME/lib
$APT_ORCHHOME/lib
branded_odbc/lib

As far as I can determine, there is no branded_odbc within the $APT_ORCHHOME directory.

DSSetJobLimit() is used from the DataStage API, not from dsjob.

DSJ.LIMITWARN is a DataStage constant used as an argument to DSSetJobLimit() and which specifies that it is the warnings limit (not the rows limit) that is being set.

Research the on-line help for DSSetJobLimit.

The second field is "", which you can safely discard.