Hello.
I would use an evaluateJson processor to update an attribute rather than parse via a regexp the entire flowfile. Then, I would the routeOnAttribute processor which creates one route per value of the attribute, plus one for the "unmatched" attribute. BTW, you can reach me, by phone or mail or communicator), just search my name in the intranet. Regards. Aurélien DEHAY ________________________________ De : [email protected] <[email protected]> Envoyé : lundi 18 avril 2016 16:51 À : [email protected] Objet : DF design question Hello I have this simple Use Case to implement ( but it's not so clear for me which processors to put in the chain :)) : I have JSON file with records identified by one type property {.......... "type": " smartphone"}, { ......... "type" : "PC" } , { ............ "type": "tablet"} ................. I want to route records based on the "type" property to different sink destinations . Looking to routeText or routeContent procs , seems to be the right direction but I do not see how to route to multiple sinks .... ( 3 in my example ) : I want records of "type": "smartphone" to be route to one sink ( first ElasticSearch processor with index1) , "type": "PC" on another sink ( 2nd ES processor) , and "type": "tablet' to a third ( 3rd ES processor) A kind of demultiplexer to N sinks .... Is it the right design (and processors ) to implement this DF , please? :) Phil
