Page 1 of 1

Merge Stage functionality

Posted: Wed Dec 08, 2010 1:47 pm
by GowthamSen
Hi,


I created a job which merges data from 1 Master file and 1 update file using Merge stage.


here is the data as follows:

Master file content:

mKey,Col1

1,A

2,C

3,D


Upadate1 fiel content:

1,P2

3,t2

3,u2

2,Q2

2,R2


The output file gives output as:

1,A,P2

2,C,Q2

2,C,R2

3,D,t2

3,D,u2


This works fine.


But when I add another update file as follows:

Update2 file content:

1,P

2,Q

4,S


The output file gives output as:

1,A,P2,P

2,C,Q2,Q

3,D, -------- This record is coming as I used Unmatched records: KEEP.


But, Here I am confused, why the merge stage is behaving differently with respect to 1 update file and 2 update files.


I am assuming that, even with 2 update files, the output should contain all duplicate records from update files.


Please let me know, if I am missing anything.

Posted: Wed Dec 08, 2010 7:10 pm
by ray.wurlod
It is a documented requirement that, when there are more than one Update inputs, all Update inputs must be de-duplicated. This does not appear to be the case in your example.

Posted: Thu Dec 09, 2010 10:58 am
by GowthamSen
Thank you Ray.

Previously I was worried that, some thing I am missing in stage properties.

So, now its clear to me that, in case of multiple update links, there shouldn't be duplicates.