DSXchange

Change Data Capture (CDC) needs more care to be given. Reading the relevant help document is the first step.

Also make sure:

In the input tab of your Change Capture stage you have used the same partition on the same partition keys. Also your sorting keys are the same and in the same order.

You can use Head stage. You have option to read N rows there. Else use Unix command in filter option of sequential file stage and fetch only first 1000 rows.

You might want to read the FAQ section and learn a bit about the dual engine concept in DataStage. Refer THIS section.

Modularizing and putting comments for log entries is for sure the best way. Option 2.

Again No. Open a parallel job and see if you can see a hashed file option in the palette.

Just change the data type of your Input Column from decimal to integer regardless of what the real input type is and do
in your modify stage.

Is the input column really decimal? You didn't specify any scale in the data type specification nor in the examples/values you posted. If they are expected as date values they shouldn't be expected as decimal at all. That will be: 20070112.00... 20060513.00... If input is coming without sacle values...

That will be
Maveric, Thanks for pointing it out.

If your concern is about 'issues with respect to data stage repository or any affect to the same job' answer is No. Doing it from the director is the safest way (Esply if you are not aware of background processes). Hi, my question is job is opened and my connection lost, when I am try to open again ...

It is a potential bug and I believe there is a fix available from the service provider for this. I had a similar issue in one of my previous projects and IBM responded with a fix. But I found the work around - Passing the parameter from a sequence - elegant and efficient for the requirement you have...

Refer THIS link for a small example to understand the concept

That confirms this is not expected as you have defined the column as VARCHAR(10) and not CHAR(10). Confirm the same on Oracle side.

Minhajuddin wrote: e.g. If I retrieve a column emp_name varchar(10) with value 'John'
The oracle stage is retrieving it as 'John '

Is this expected?

I don't think this is expected You might want to see this on the database side and make sure that data is stored in this way in the source/lookup itself or in the schema it is defined as char and not varchar.