Flow Control
This article will go through the Flow Control Shape of Boomi and its functionalities.
Flow control shape is a very important shape as this allows parallel processing.
To read more about Parallel Processing please check the following blog: Boomi Parallel Processing
Managing Document Flows:
There are three ways Flow control manages the documents.
- No Document Batching: Using this option doesn’t batch the documents and adheres to the Parallel Processing Defined.
If you have 10k documents running in four threads this will send 2500 documents to each thread. (Documents in each thread= Total No of Source documents/No of threads)
- Run Each Document Individually: This option slows integration performance but is unavoidable under some circumstances.
There might be scenarios where the target application can process only one document at a time or you need to do a query connector call that accepts only one document at a time or any other mapping rules in which you want one document to be processed at a time, then this is the only way to do it.
Run Each Document Individually sees the document flow through completion.
- Running Document in Batches: This option is very useful when running a huge data volume.
In Boomi, all the documents flow through each shape at a time together.
If your integration receives 100k records from NetSuite all of them would go to the set properties, then the map, and so on. Thus, even if the first document gets processed it will still wait for the 100000th record to be processed too, and only then move on to the next shape.
In a scenario where you have a lot of shapes in your integration Process with complex functionalities like connector calls or Lookups, and you are dealing with huge volumes it is recommended to run the documents in batches.
For the above scenario, if you decide to run a document in batches of 5000 and select no of threads to four, Boomi will keep sending 5000 documents in the four threads until completion.
Cover Image by fanjianhua on Freepik
