You are viewing a plain text version of this content. The canonical link for it is here.
Posted to jira@kafka.apache.org by "Josep Prat (Jira)" <ji...@apache.org> on 2024/06/19 07:24:00 UTC
[jira] [Commented] (KAFKA-16468) Listener not found error in SendRPCsToBrokersEvent
[ https://issues.apache.org/jira/browse/KAFKA-16468?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17856185#comment-17856185 ]
Josep Prat commented on KAFKA-16468:
------------------------------------
Changing target fix version to 3.9 since this is not a blocker and we are past code freeze
> Listener not found error in SendRPCsToBrokersEvent
> --------------------------------------------------
>
> Key: KAFKA-16468
> URL: https://issues.apache.org/jira/browse/KAFKA-16468
> Project: Kafka
> Issue Type: Bug
> Components: controller, migration
> Reporter: David Arthur
> Priority: Major
> Fix For: 3.9.0
>
>
> During the ZK to KRaft migration, the controller will send RPCs to the ZK brokers using the configured "control.plane.listener.name" or more commonly, the "inter.broker.listener.name". If a ZK broker did not register with this listener, we get a error at the time of sending the first RPC to a broker.
> {code}
> [2024-04-03 09:28:59,043] ERROR Encountered nonFatalFaultHandler fault: Unhandled error in SendRPCsToBrokersEvent (org.apache.kafka.server.fault.MockFaultHandler:44)
> kafka.common.BrokerEndPointNotAvailableException: End point with listener name EXTERNAL not found for broker 0
> at kafka.cluster.Broker.$anonfun$node$1(Broker.scala:94)
> at scala.Option.getOrElse(Option.scala:201)
> at kafka.cluster.Broker.node(Broker.scala:93)
> at kafka.controller.ControllerChannelManager.addNewBroker(ControllerChannelManager.scala:122)
> at kafka.controller.ControllerChannelManager.addBroker(ControllerChannelManager.scala:105)
> at kafka.migration.MigrationPropagator.$anonfun$publishMetadata$2(MigrationPropagator.scala:98)
> at kafka.migration.MigrationPropagator.$anonfun$publishMetadata$2$adapted(MigrationPropagator.scala:98)
> at scala.collection.immutable.Set$Set3.foreach(Set.scala:261)
> at kafka.migration.MigrationPropagator.publishMetadata(MigrationPropagator.scala:98)
> at kafka.migration.MigrationPropagator.sendRPCsToBrokersFromMetadataImage(MigrationPropagator.scala:219)
> at org.apache.kafka.metadata.migration.KRaftMigrationDriver$SendRPCsToBrokersEvent.run(KRaftMigrationDriver.java:777)
> at org.apache.kafka.queue.KafkaEventQueue$EventContext.run(KafkaEventQueue.java:128)
> at org.apache.kafka.queue.KafkaEventQueue$EventHandler.handleEvents(KafkaEventQueue.java:211)
> at org.apache.kafka.queue.KafkaEventQueue$EventHandler.run(KafkaEventQueue.java:182)
> at java.base/java.lang.Thread.run(Thread.java:833)
> {code}
> This is pretty late to be detecting this mis-configuration. By this point, the KRaft controller is the active controller and has already migrated the metadata. Recovery is possible by restarting the brokers with the correct listener names, but we can catch this much sooner in the process.
> When a ZK broker registers with the KRaft controller, we should reject the registration if the expected listener name is not present. This will prevent the migration from starting.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)