I installed cassandra spark-hadoop cluster on 3 ec2 nodes. Yesterday, I was able to start the spark thrift server on node0, and actually executed a simple sql statement in beeline. Today, after a schema change, I restarted the thrift server, now I get a SparkException java.lang.IllegalArgumentException: ip-172-30-4-140 at org.apache.hadoop.hive.cassandra.cql3.input.HiveCqlInputFormat.getRecordReader(HiveCqlInputFormat.java:212)
the ip-172-30-4-140 is simply the private ip of that node
I tried running the same sequence from the other two cassandra nodes, and for those, the sql statement gets stuck and never returns.
What is this error? any one knows?