PX creating flat file with duplicate entries

[data_whs_team]

I'm reading an sql table and writing the results to a flat file via a sequential stage. The sql query has 'group by' and an 'order by' statements to ensure unique values in the file.

I'm finding that for every unique row I get back from the sql statement, DataStage is writing one for every node we have defined in our /appl/Ascential/....Configurations/default.apt file. In this case, we have 12 nodes in the default.apt file. Therefore I'm getting 11 MORE sets of data in my file than I want!

The file should look like this:

abc
123
def
456

But I'm getting this:

abc
123
def
456
abc
123
def
456
abc
123
def
456
abc
123
def
456
..... etc.

I can't change the default.apt file.

Any help is greatly appreciated!

Cathy

[kcbland]

Describe the stages in your job. Any chance you set a partitioning to ENTIRE, which sends a full copy of all rows to all nodes, which in turn you write to a sequential file?

[data_whs_team]

We just figured it out. I changed the mode on my Teradata stage from 'parallel' to 'sequential'. Works like a charm now!

TY,
Cathy

[kcbland]

Glad to be of zero help. Next time, state your stages first. I don't see you mentioning Teradata in your original post, just an unmentioned relational stage and a sequential file.

[data_whs_team]

Sorry about the lack of information. I didn't mention the database stage because I didn't think it was relevant. I was focusing on the node count in the default.apt stage.

I'll be more detailed next time!

Cheers,
Cathy

[us1aslam1us]

Cathy,

If you done with this as it appears to be.Please mark this as 'Resolved'