DSXchange

I have a very straight forward QS UNDUP match process that takes one input file coming out of a Geomatch and dedupes the results. During the production runs for some reason the result file loses the House number and the house num suffix. So anything that comes in like 123 Main St. ends up being <blank> Main St.

Jobs are constructed very simply :

Input File ---> QS PLugin ---> Output file..

This job is under a sequencer that is under the main sequencer.


Now here's the weird part..

When I run this job QS only it works fine. When I run it from the DS jobs that invokes the QSPlugin it works fine as well..

I checked the parameters in sequencer, that kicks off the DS job and the parameters in the DS job that runs the QS piece, everything is fine and dandy..

I have run out of possible culprits.

Any ideas?

The only thing I can suggest is that you check the datafile definition for the output file, to ensure that the HN column is not being overwritten by some other column. That is, check that the start and length of every field are correct. My suspicion is that you are (accidentally) overwriting the HN field with something that's blank.

ray.wurlod wrote:The only thing I can suggest is that you check the datafile definition for the output file, to ensure that the HN column is not being overwritten by some other column. That is, check that the start and length of every field are correct. My suspicion is that you are (accidentally) overwriting the HN field with something that's blank.

That crossed my mind.. But, then how come it's working on individual runs from both QS and DS ? Needless to say, a case is open with Ascential..

Yeah. Weird. Nothing overwriting it subsequently? Do let us know what "they" have to say.

Nope.. Nothing is overwriting it..
"They" are taking a look at it.. During the process "they" found out that "they" are not able to import an IMF exported from 7.0.1 into 7.5

That will be very handy to know, until "they" fix it. Hopefully that will be soon.