Generic CDC job that can be used across any file

Bhargav86 · Post by **Bhargav86** » Tue Dec 30, 2014 6:18 am

Hi,

Could you please help me anyone how to impliment generic CDC logic for any number of fields and any number of keys.

Thanks,
Bhargav

priyadarshikunal · Post by **priyadarshikunal** » Tue Dec 30, 2014 7:43 am

That will be a bit tough to explain, I have done it in past though. Why don't you explain what you are trying to achieve and what all information about file metadata you will have to design such a job.

priyadarshikunal · Post by **priyadarshikunal** » Tue Dec 30, 2014 7:51 am

In a nutshell, you will need following information to get started.

whether the file is delimited or fixed width. if delimited which all fields are to be considered as keys and which all to be used for comparison. if fixed width start and end position of key and data fields.

You will need a little bit of looping here to concatenate all key fields together in one variable and data field in another one as they are not fixed. So you have to be on DataStage 8.5 or higher. if you need you can join back the result to get actual records and mark them with change codes.

priyadarshikunal · Post by **priyadarshikunal** » Tue Dec 30, 2014 7:54 am

if on earlier version you need to have enough stage variables to capture the max number of key and data fields. In fixed length you can optimize it by multiple continuous fields defined as one.