Problem with tHashInput -> tFlowToIterate -> tWriteJSONField

One Star

Problem with tHashInput -> tFlowToIterate -> tWriteJSONField

Hi,
as you can see in the screenshots, I created a query, store the result in the hashmap, read it out from there, iterate it parallelized and write a json with it.
But somehow at tWriteJSONField data gets lost. As at tLogRow_3 still all 100 rows arrive, tLogRow2 just gets ~40 (+-5 - it changes randomly each execution).
This happens only if I iterate parallized and the number of iterations are not the same like the number of rows (so 1 (so actually no parallel execution) and 100 parallel executions for 100 rows work in this example any other number parallel executions loses data).
Do you have any suggestions on this, because it looks like a bug to me (something with the java garbage collector or so)?
Regards
Community Manager

Re: Problem with tHashInput -> tFlowToIterate -> tWriteJSONField

Hi
Do you select a column as a group element on tWriteJsonField component? If so, the incoming rows with same value will be grouped as one output row, can you please upload a screenshot of the basic settings of tWriteJsonField?
Shong
----------------------------------------------------------
Talend | Data Agility for Modern Business
One Star

Re: Problem with tHashInput -> tFlowToIterate -> tWriteJSONField

No there is no grouping - and if so, it wouldn't explain that it just happens at parallelizations.
Community Manager

Re: Problem with tHashInput -> tFlowToIterate -> tWriteJSONField

Hi
Thanks for your explanation, it looks like a bug when you set the iteration with parallelizations, I would suggest you to open an issue on our bugtracker with detailed information, such as product version, demo job. Our developer will investigate this issue further and discuss it with you.
Thank you!
Shong
----------------------------------------------------------
Talend | Data Agility for Modern Business