The quickstart tutorial from imply 1.3 works fine and was able to run the wikiticker kafka tutorial without issues. However, as soon as i try with some real data with a modified version of supervisor.json from kafka tutotial - i am seeing issues.
The Issue is I can see the supervisor task in the Indexer coordinator console and it says it is finished SUCCESSFULLY but there are no segments or data that has been processed. Also, the task finishes too soon .I tried with a very small sample like 20 rows and still it seems to have issue . It does connect to kafka and seek to the right offset but nothing happens after that. One notable error is zookeeper has an error with NoNode .( I am able to load the same data using traditional batch load fine, however i really want to take advantage of this new type of load as the data load has significant delays/stragglers in collection process which touches older segments. I don’t want to bother with tinkering realtime loads/window configuration or intervals/segments on the delta batch load configuration)
INDEX TASK LOG var/druid/task/index_kafka_realfeed_ce4f41348970e46_plffpncd/log