AnsweredAssumed Answered

Hiveserver2 leak of memory

Question asked by rvelfre on Mar 17, 2016
Latest reply on Mar 24, 2016 by mufeed

Hello !


I'm facing an issue after migrating from MapR 3 to MapR 5.

Our hiveserver2 (that we updated from 0.13-1409 to 0.13-1510) throws a java heap space error every 5 hours so I assume there is a memory leak somewhere...


In the following log :  /opt/mapr/hive/hive-0.13/logs/mapr/hive.log

The same ERROR message is throwing many times/sec and says :


2016-03-17 17:35:03,823 ERROR [HiveServer2-Handler-Pool: Thread-16]: server.TThreadPoolServer ( - Error occurred during processing of message.
java.lang.RuntimeException: org.apache.thrift.transport.TTransportException

at org.apache.hadoop.hive.thrift.HadoopThriftAuthBridge$HiveSaslServerTransportFactory.getTransport(
at org.apache.thrift.server.TThreadPoolServer$
at java.util.concurrent.ThreadPoolExecutor.runWorker(
at java.util.concurrent.ThreadPoolExecutor$
Caused by: org.apache.thrift.transport.TTransportException
at org.apache.thrift.transport.TTransport.readAll(
at org.apache.hadoop.hive.thrift.HadoopThriftAuthBridge$HiveSaslServerTransportFactory$1.receiveSaslMessage(
at org.apache.thrift.transport.TSaslServerTransport.handleSaslStartMessage(
at org.apache.hadoop.hive.thrift.HadoopThriftAuthBridge$HiveSaslServerTransportFactory.getTransport(
... 4 more


It seems to be related to the SALS authentication. For you information, we’re running a non-secure cluster.

We use default hive.server2.authentication properties in hive-site.xml (which is SASL mode).


Has anyone facing this issue ?


Thanks for your help !