We'd like to be able to provide detailed information to our phone support reps on WHY an individual record failed (which field). We would like to do this without using the audit log (ugly) but haven't have much luck. Has anyone done this?
My first thought was to run the invalid records through a second map that would identify the field in error, but WSTX won't map invalid data.
Mapping invalid fields?
...
Easy! Just send an email off to:
RequestTonsOfSpam@GodDamnThatsATonOfSpam.spam
You should get a response real quick.
RequestTonsOfSpam@GodDamnThatsATonOfSpam.spam
You should get a response real quick.
...
Street corner? No.
That's Willy. And what he was handing out...for a price...would be very diffictult to scan and upload, plus, I'm not sharing any of that.
Jim...Jim is the two year datastage guy that is in my group (development). I'm the guru! And while he is a swell guy, I would recommend meeting the three Russian ladies in the Tech info group. Wow...just, wow. I mean, hey, if you want to meet Jim over the three Russian ladies, there's nothing wrong with that. I'm all for it. Eitherway, the next time you're in New Jersery, we'll all go out and grab a bite. And I'll tell Jim to bring his hand out. I'll tell him to bring all his hand out's.
That's Willy. And what he was handing out...for a price...would be very diffictult to scan and upload, plus, I'm not sharing any of that.
Jim...Jim is the two year datastage guy that is in my group (development). I'm the guru! And while he is a swell guy, I would recommend meeting the three Russian ladies in the Tech info group. Wow...just, wow. I mean, hey, if you want to meet Jim over the three Russian ladies, there's nothing wrong with that. I'm all for it. Eitherway, the next time you're in New Jersery, we'll all go out and grab a bite. And I'll tell Jim to bring his hand out. I'll tell him to bring all his hand out's.
I had to do something like this for a small file ( <10mb ) and a larger file ( >80mb )
For the small one I had a validation map with trace on and another map that read the tace file and summarized issues. This worked fine for most invalid datasets. Problem was wasn't very efficient on large trace files....
The second solution consists of a series of maps.
First map (control map) reads in the data as loops of text. A submap (runmap1) is executed with each loop. The control map records all of the run information about the loop, ID + runmap1 return code + whatever.
The next map (runmap2) that is executed only processes the bad loops (identified by runmap1) at a record/segment level. The control map tracks info from runmap1 failure plus failure information related to the current map, runmap2. ID + runmap1 return code + runmap2 return code + record/segment number + copy of data passed to runmap2.
The last card in the control map reads the data in the card that executed runmap2 looking for issues which it summarises and send to BA.
Not elegant but it works. You might try something like that.
For the small one I had a validation map with trace on and another map that read the tace file and summarized issues. This worked fine for most invalid datasets. Problem was wasn't very efficient on large trace files....
The second solution consists of a series of maps.
First map (control map) reads in the data as loops of text. A submap (runmap1) is executed with each loop. The control map records all of the run information about the loop, ID + runmap1 return code + whatever.
The next map (runmap2) that is executed only processes the bad loops (identified by runmap1) at a record/segment level. The control map tracks info from runmap1 failure plus failure information related to the current map, runmap2. ID + runmap1 return code + runmap2 return code + record/segment number + copy of data passed to runmap2.
The last card in the control map reads the data in the card that executed runmap2 looking for issues which it summarises and send to BA.
Not elegant but it works. You might try something like that.