Thanks @SAMSAL for the clarification and the alternatives for looping in Nifi.  We will consider using DuplicateFlowFile processor.  Do you think nifi documentation should be updated to explicitly mention that it is not only for load tests but also can be used in production flows when there is need to clone flow files ? ... View more

Appreciate any help with my above question about DuplicateFlowFile usage.  @SAMSAL    @MattWho    ... View more

@AlokKumar  I am not able to understand the context here.   It would help to get to right solution if you can explain why the inserts into same table should be executed in certain order.   However, Nifi Connection configuration settings have prioritizers.  Please check screenshot below.. If you are able to set attribute "priority" with desired number, "PriorityAttributePrioritizer" can be used.  Please note that these prioritizers operate on node level.  You may have to test these if you have multi node nifi environment.  There is a processor "EnforceOrder" which has more options to enforce execution order.  You may check this processor and find out if it works. https://nifi.apache.org/docs/nifi-docs/components/org.apache.nifi/nifi-standard-nar/nifi2.0-m4/org.apache.nifi.processors.standard.EnforceOrder/index.html   ... View more

@Adyant001  Oracle has auto increment IDENTITY feature as below. Table definition takes care of setting primary key incremental value. No need to send this value in json payload I think all other major RDBMS have similar feature to assign auto increment value. CREATE TABLE roletab (id NUMBER GENERATED BY DEFAULT ON NULL AS IDENTITY (START WITH 1 INCREMENT BY 1 NOCYCLE), role_id NUMBER, role_name VARCHAR(100), PRIMARY KEY (id) ); INSERT INTO roletab (role_id, role_name) VALUES (10, 'Admin'); INSERT INTO roletab (role_id, role_name) VALUES (20, 'Developer'); commit; SELECT * FROM roletab; ID |ROLE_ID|ROLE_NAME| --+-------+---------+ 1    | 10     | Admin | 2   | 20    |Developer| ... View more

@Adyant001    Can you share sample json structure ? If source json contains both parent and child information, you can have two PutDatabaseRecord processors connected in series with right "Data Record Path" mentioned. If child record needs information from parent level, JoltTransformJson processor can be used to create transformed json before passing to PutDatabaseRecord. ... View more