DSXchange

Hi ,

I have a job where there are two input datasets which act as before and after dataset for a change capture stage.Output of the change capture stage is stored in a dataset.The two input links of change capture stage are sorted and hash partitioned on the key column.

The job is part of a sequence.The input datasets of the job is produced by two predecessor jobs called from the same sequence.When I am running the job through that sequence multiple times.in the job monitor the output link of the change capture stage is showing different counts in different runs,though the input records remains the same.
When I am running the same the job multiple times directly(not calling from sequence)the output record count is always the same for the change capture stage.

Can you please explain why the out put record count of the change capture stage is different for different runs when the job is called from sequence?

Thanks and Regards

Avik Dasgupta

Hi Avik,

Are you sure that running through your sequence does not change the data selected in the two jobs predecessing your CDC-job?

Hi,

Please check if you are getting duplicates on the key columns. This behaviour of change capture stage is often observed when we get duplicates in the two source inputs.

Regards,
Praveen

Guys, please stop quoting everything you are replying to. If you need to do it to make your point, that's fine, but don't just click on Reply with quote out of habit or because you think it's the only option. Use the Reply to topic link instead and save me the work of constantly having to clean up after you... which I'll do later.

Sorry craig,

Maybe I missed something in my browser, but I can't see that link...

(cleared the quote manually here)

It's also at the top of the page (at least in Internet Explorer).