How can I prevent these errors coming?
When I created 9 partitions on 3 instances and 2 replication-factor, I
didn't have any error. But when I created 36 partitions on 12 instances, I
got these errors.
2014-03-07 18:19:28,208] ERROR Controller 9 epoch 12 initiated state change
of replica 9 for partition [request_trace,3] from NonExistentReplica to
NewReplica failed (state.change.logger)
kafka.common.StateChangeFailedException: Replica 9 for partition
[request_trace,3] cannot be moved to NewReplicastate as it is being
requested to become leader
at
kafka.controller.ReplicaStateMachine.handleStateChange(ReplicaStateMachine.scala:160)
at
kafka.controller.ReplicaStateMachine$$anonfun$handleStateChanges$2.apply(ReplicaStateMachine.scala:96)
at
kafka.controller.ReplicaStateMachine$$anonfun$handleStateChanges$2.apply(ReplicaStateMachine.scala:96)
at
scala.collection.immutable.HashSet$HashSet1.foreach(HashSet.scala:123)
at
scala.collection.immutable.HashSet$HashTrieSet.foreach(HashSet.scala:322)
at
scala.collection.immutable.HashSet$HashTrieSet.foreach(HashSet.scala:322)
at
kafka.controller.ReplicaStateMachine.handleStateChanges(ReplicaStateMachine.scala:96)
at
kafka.controller.KafkaController.onNewPartitionCreation(KafkaController.scala:464)
at
kafka.controller.KafkaController.onNewTopicCreation(KafkaController.scala:452)
at
kafka.controller.PartitionStateMachine$TopicChangeListener$$anonfun$handleChildChange$1.apply$mcV$sp(PartitionStateMachine.scala:421)
at
kafka.controller.PartitionStateMachine$TopicChangeListener$$anonfun$handleChildChange$1.apply(PartitionStateMachine.scala:403)
at
kafka.controller.PartitionStateMachine$TopicChangeListener$$anonfun$handleChildChange$1.apply(PartitionStateMachine.scala:403)
at kafka.utils.Utils$.inLock(Utils.scala:538)
at
kafka.controller.PartitionStateMachine$TopicChangeListener.handleChildChange(PartitionStateMachine.scala:402)
at org.I0Itec.zkclient.ZkClient$7.run(ZkClient.java:576)
at org.I0Itec.zkclient.ZkEventThread.run(ZkEventThread.java:71)
