First of all, thank you for your response Matt. Sorry I wasn't clear enough in my MergeContent description, but it's as you said; I'm using the Bin-Packing Algorithm merge strategy, and I've set the "Minimum Number of Entries" equal to the amount of ReplaceText processors I have (2 in this case), such that the MergeContent processor will merge a given bin when it meets the configured values for "Minimum Number of Entries" (2) and "group size" (Minimum Group Size is set to 1 B and I haven't set a Maximum Group Size). And with respect to the third point you outlined, I've set the "Max Bin Age" attribute to 10 sec such that if the bin doesn't meet the configured values for "Minimum Number of Entries" and "group size" within 10 sec then it will merge the flowfiles currently allocated to the bin. For more context here's a template of the situation I'm describing, where the second screenshot shows the configurations for the MergeContent processor: For "List1" and "List2" above, as an example, the "Replacement Value" attributes contain the following respectively: abc1 abc2 abc3 and abc2 abc3 abc4 I've since found a way to avoid adding arbitrary max latency and have opted to use a global variable instead that will be incremented after each ReplaceText processor using UpdateAttribute (and connecting UpdateAttribute to MergeContent directly), and I'll set "Minimum Number of Entries" to that global variable. In this way, each time a ReplaceText "stream" is executed, the global variable is incremented signifying that we're expecting one more list, and the MergeContent processor will wait for all the flowfiles required (each containing a different list of course) to merge the bin, assuming of course that I don't assign a value to "Max Bin Age". I believe this approach is more robust than using Max Bin Age/max latency, what do you think? Also, regarding this comment: "Also keep in mind that with a NiFi cluster, each NiFi node is running the MergeContent processor independently of the other nodes and can only merge FlowFiles on the same node. So while an inbound connection queue to the MergeContent processor may show 2 queued FlowFiles, each of those FlowFiles may exist on different nodes and thus would not be merged together in to one FlowFile. " would you suggest then to set the "Execution" of MergeContent to Primary node instead of All nodes? Thank you for your time and I look forward to your response, Chris
... View more
