DSXchange

Hi,

Is it possible in Datastage to merge files with different metadata having common key columns?

Thanks

Vinod

vinodhraj wrote:Hi,

Is it possible in Datastage to merge files with different metadata having common key columns?

Thanks

Vinod

Use Merge stage.

Either you can use merger or lookup based on the volume of data. Since your concertn is on metadata, I guess you are on Concatination, if so just use cat file1 file2 > new filename command in any of the unix command execution option, like Execute command activity or Before/After job subroutine.

cat two file. but why you do that? how do you deal with a file with two parts of different meta data?

Thanks.

I am working with datastage server jobs, I can't use merge stage in server since I have to mention different key name for the same column.

Is it possible in unix ?

Somehow I have done this using Set Operators in SQL..

I didn't prefer lookup becoz, there is no relation between these tables though the column name were same.

Moreover SET Operators were better in performance wise.

Thanks

Vinod

If there is no relationship between two table, how can you merge two table without cartision product.
Can you explain your need with some example?

Kumar,

Using Union we can do this i.e if the column name is similar.

Since I am joining with differnent metadata, which ever metadata is mismatching, I will mention them as empty string providing common alias name.

for eg:

table a

name
deptno
sal


table b

name
deptname

select name, deptno,to_char(sal),'' deptname from table a
union all
select name, '' deptno,'' sal,deptname from table b


hope it helps

Vinod

vinodhraj wrote:I am working with datastage server jobs, I can't use merge stage in server since I have to mention different key name for the same column.
Vinod

This is not correct. The column names (when working with sequential files) are totally irrelevant. Adjust the metadata so that you do have a common key column name.

Using transformer, you can create columns with empty strings to match the metadata. Or use script for the same purpose. 'Cat' the two files.

cat is NOT the same as join.

cat is synonymous to union.

Neither is cat synonymous to UNION. A UNION removes duplicates; cat does not. Nor is UNION a join.

Actually i mean UNION ALL. Cat is synonymous to UNION ALL. And yes my statement above doesnt hold true as UNION does remove duplicates. For some reason i left out the 'ALL'.

Learn not to rush, grasshopper. :D

Well caught.