More or less yes (it transforms sql statements into KafkaStreams jobs). But I simply call a rest endpoint that has as response Transfer-Encoding: chunked and the data comes in as you can check with the linked image. The only thing it's that I have blank rows (but if I debug the code the data is in there) and when I stop the job every blank row gets the correct value.
Il giorno lun 7 ott 2019 alle ore 08:06 Jeff Zhang <zjf...@gmail.com> ha scritto: > I don't know much about KSQL, is each running a streaming job which never > stops ? > > Andrea Santurbano <sant...@gmail.com> 于2019年10月7日周一 上午4:09写道： > > > Hi guys, > > I'm building a KQL interpreter. KSQL is the streaming SQL engine that > > enables real-time data processing against Apache Kafka topics. > > I created an interpreter that leverages the InterpreterOutputStream class > > in order to stream the result from the backend to the frontend, but I get > > the following representation in the frontend: > > > > https://imgur.com/a/J7unk25 > > > > What am I doing wrong? > > > > Following the Interpreter implementation: > > > > > > > https://github.com/conker84/zeppelin/blob/kafka/kafka/src/main/java/org/apache/zeppelin/kafka/KafkaKSQLInterpreter.java > > > > Thanks a lot > > > > Andrea > > > > > -- > Best Regards > > Jeff Zhang >
