I have encountered a use case where i am fetching rows from tMyslInput and then iterating it one by one to complete a subjob for each of the rows retrieved.
tMysqlInput-----> iterate -------> (job having multiple components such as writing a file, logging it, entry into the database and different processes like this, which is a complete process in itself).
Problem is that since the subjob after iterate link itself takes care of everything, i just want to fork as many subjobs as number of rows fetched from tMysqlInput with different context parameters.
So i tried to do following
tMysqlInput ------>iterate(*n, where n is number of rows fetched)----->(job)
But here what is happening , threads are reading each other context variables hence ending up writing similar context in similar files, same db entry etc..
I want to parrallelize the child job depending on number of rows fetched with threads being in synchronize.
tMysqlInput query lets say, select file_id, input_path , output_path from some table where status='copied';
lets say 4 tuples i got then i want to iterate 4 tuples at the same time. Just execute the child job and let the child job execute on its own.
thanks