DSXchange

Hi There,

We are using a CFF stage to read one of our input file

We are able to view the all columns properly but when we view it in datastage each record is shown as 4 times

If i have 2 records in the input file, it shows 8 records when i view the data from Datastage, when i tried to load the file to Dataset it loads 8 records but the performance statistics shows only 2 records


Any ideas please

Regards

Is your datastage configuration running with 4 nodes, with the partitioning (dataset properties, input tab, partitioning subtab) set to ENTIRE ?

One possible scenario: you have an array of 4 elements (e.g. OCCURS 4 TIMES) that you've told the CFF stage to normalize: 2 records * 4 elements = 8 normalized records.

Mike

Hi Mike,

Please let me know where can i set the array size

Regards

You don't.

The array is defined in the COBOL FD that you presumably imported. You decide at design time whether to normalize or flatten the array based on your transformation requirements.

Mike

Hi Mike,

We selected the option of As-Is while importing

Should we make changes to this option.

As per the design we have to get rid of these duplicates before processing

Please advise

Many thanks for your support

Regards

It's been a while, but I don't think you decide when importing the COBOL FD.

It would be an option when you load the column metadata into the CFF stage.

Mike

That's certainly true in 7.5.2.

I'm an old COBOL dog, so I don't hesitate to edit the copybook before importing it, deciding on arrays and redefines myself. When I deal with mainframe data, it's always going to or from some other platform (usually Unix) where COBOL formatting is not very useful.

Hi All,

We got this working by selecting the Flatten Arrays while selecting the columns.

We are getting source files from the Packed Decimal (Comp 3) and EBCDIC format

But when we select the character set as EBCDIC, we are not able to view the data, but we can view the data (Partially-some columns) if we select the character set as ASCII

When we select the character set as EBCDIC, we are getting the below error as

Input buffer overrun at field "FIELD NAME"

Please advise if i have to bring any settings in

Regards

The fact that you can partially view the data when you select ascii likely means that the file was transferred in ascii mode... all of your packed decimals will be destroyed.

There was a recent post, I think, where James and Franklin provided some excellent information regarding ftp of mainframe files.

Mike