Detail info regarding APT_DUMP_SCORE

[goutam]

My job design is as follows.
SeqFile1 ---> Xform ----> SeqFile2

I have used APT_DUMP_SCORE to see the score.

Here is the score output incase of 1 node config file.
0)main_program: This step has no datasets.
1)It has 1 operator:
2)op0[1p] {(sequential APT_CombinedOperatorController:
3) (Sequential_File_0)
4) (APT_TransformOperatorImplV0S54_test1_Transformer_54 in Transformer_54)
5) (APT_RealFileExportOperator in Sequential_File_59)
) on nodes (
6) node1[op0,p0]
7) It runs 1 process on 1 node.

My questions are:
1)As Ray would has been mentioned in one of his post that LINKs are datasets, Shouldn't there be 2 datasets in the above score?

2)Since the Xform stage runs parallely, I would think that there should be 4 processes in the above score.

If i am running the same job in 2 node config file , i am getting the correct result such as 2 datasets (as there are 2 links) and 4 processes (1 process per sequential stage and 2 processes for Xform stage)

Ray :- If your answer is licensing score, what does this licensing score mean? Please send your post so that every one who had raised this question before colud see is fully. Because we are not premium members.
It's my sincere request.

I would appreciate if anyone including ray could throw some light on this topic.

[DSDexter]

The above score shows that you are running the job on a single node and also the operator is a combined operator of seq.file, transformer and seq. file. So I dont think ther's anything wrong with the above score. It will generate only 1 process in this case.

And also i think that no of links do not govern the no. of process spawned. Its the no. of parallel stages and the no. of nodes that your job is running on.

[mahadev.v]

Number of processes are not governed directly by the number of links or stages on the job canvas. To answer your second question, since you are using only one node config file, Xform is forced to run sequentially. In the first case since all of them are sequential operators, DataStage optimizes it into one process. Whereas in the second case, data has to be imported, portioned, processed, collected and then exported to sequential file. So actually it is 6 process (2 for Xform). Again DataStage optimizes it to 4 processes. Don't know what "Licensing score" is? Hope this helps.

[ray.wurlod]

What you posted is not the licensing score, as yours has a Transformer stage mentioned.

DSDexter's analysis is correct. The score is an optimized variant of the osh that was generated by compiling the job design.