Can you confirm if you have the "Topic Auto Creation" disabled: auto.create.topics.enable=false If so, have you created the t1 topic beforehand? Can you do the following: kafka-topics --zookeeper <zkhost>:2181 --list kafka-topics --zookeeper <zkhost>:2181 --describe --topic t1 ... View more

Please confirm, there should be no space between localhost:2181 and /kafka. The zk option should be: --zookeeper localhost:2181/kafka My apologies, it appears I had a typo in my original recommendation. -pd ... View more

There is a regression that was introduced in CDH5.7.0 that throws that stack trace (at INFO level) for every update, or /admin/cores command. The workaround is to set the following in the logging safety valve for the solr service:   log4j.logger.org.apache.solr.servlet.SolrDispatchFilter=WARN This will be resolved in CDH5.7.1.   -pd ... View more

Please define what you mean by "Trigger Flume agent". Are you referring to starting the flume agent, or being able to deliver events to hdfs? ... View more

This error: org.apache.avro.AvroRuntimeException: Excessively large list allocation request detected: 150994944 items! Connection closed.   Is usually caused when something upstream is trying to send non-avro data to the avro source.     In your source config, you are specifying the avro source with the same port as the hdfs namenode port:   tier1.sources.source1.type = avro tier1.sources.source1.bind= 192.168.4.110 tier1.sources.source1.port= 8021 tier1.sources.source1.channels = channel1 tier1.channels.channel1.type= memory tier1.sinks.sink1.type = hdfs tier1.sinks.sink1.channel = channel1 tier1.sinks.sink1.hdfs.path = hdfs://192.168.4.110:8021/user/hadoop/flumelogs/   I believe that will cause issues in your configuration, as the sink will try to connect to the avro source port as it thinks thats the namenode port.  If your namenode port is indeed 8021, then you need to change your avro source port to be something different.   -pd ... View more