Transformer stage - constraint

[vij]

Hi,

I have a job in which i read a source file which has about 100 records and this is passed to a transformer.in the transformer, based on the i/p column value the row has to be moved to the appropriate datasets.

In the above job, i give two constraints on the transformer stage, if i/p column =a, dataset1 and if i/p column =b, dataset2.

the problem is, when a new value for this input column is coming in, for eg., c, then the source is not reading that value at all. wheres the prblm?

the job structure is :

source -> transformer -> target

Thanks

[thebird]

Can you be a little specific on the issue.

As per the constraint that you've given -

i/p column =a, dataset1 and if , i/p column =b, dataset2

Dataset1 will be populated by only those rows where the i/p column =a, and dataset 2 by those where i/p column =b. So there seems to be nothing wrong with the way the job is behaving.
So, if a new value 'c' comes in, it would not go to either dataset1 nor dataset2.

Now if you can also state your requirement clearly, then may be soembody here can help after understanding the issue.

[I_Server_Whale]

The job is behaving correctly according to the constraints defined.

If you want 'c' to go into the output links as well.

Then, you can modify your constraints as:

Code: Select all

i/p column <> b, dataset1

i/p column <> a, dataset2

Whale.

[vij]

thanks for ur reply. i understnd that a new i/p column value will be outputted only if the constraint is specified for that.
my doubt is, if there is some new i/p column value coming in , for eg., V for some 20 records, will datastage read those records are not.

my concern is - in any case, the datastage should read all the records from a file that is defined as a source rite?

pls validate my understanding.

[chulett]

Of course it will always read all of the records, or at least attempt to based on your metadata. They will then 'come into' the Transformer stage but because of the constraint will not.. 'leave'.