Reading complex flat file in sequential mode

dnat · Post by **dnat** » Thu Dec 13, 2007 1:04 am

Hi,

We have a complex flat file stage to read a cobol file. The default setting to read the file is showing as Parallel mode. But i want datastage to read the file sequentially, as reading in parallel changes the order or the records from the actual file. We are doing some validations with header and trailer records, so i want them to be in the same order as the actual file. Could anyone help. What changes need to be done in the settings?.

Thanks.

ArndW · Post by **ArndW** » Thu Dec 13, 2007 3:09 am

Since you have found the tab in the stage where you can see (and change) in which mode that stage operates I am not sure what you are asking.

aakashahuja · Post by **aakashahuja** » Thu Dec 13, 2007 3:14 am

I can see that the default execution mode (which is disabled as well) of a CFF is Sequential. What settings / options have you used in the stage?

Cheers
Aakash

dnat · Post by **dnat** » Thu Dec 13, 2007 4:57 am

The settings are not editable(disabled). I want to add one more thing here. When i create a new job and pulled the stage, the default setting is shown as sequential. But, i am not sure how it is showing as parallel for few of the jobs which had been created earlier.

Even when it set as sequential, the data is read in parallel, i mean when i try to read the data through datastage, the order is changed(since the records are scattered in all the 4 nodes).

aakashahuja · Post by **aakashahuja** » Thu Dec 13, 2007 5:41 am

Records scattering to 4 nodes will be a function of the type of partitioning used for the next stage (if ur CFF is set to run in Sequential mode).

Secondly, I am sure that some option in ur CFF has rendered it to run in parallel by default. Check the options used / refer to the manual. Your problem lies in ur CFF stage.

Or, to keep things simple, delete the CFF stage, put in another one. :D

Cheers
Aakash

landaghaar · Post by **landaghaar** » Thu Dec 13, 2007 11:59 am

when you view the data from CFF stage, they are read sequencially and in order. This means that your data gets mixed up after reading. so dont change the CFF, but the stages after.