[
https://issues.apache.org/jira/browse/KAFKA-2738?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14991230#comment-14991230
]
Gwen Shapira commented on KAFKA-2738:
-------------------------------------
The PR above fixes the issue we saw.
Two notes:
1. We keep passing metadata cache references around ReplicaManager. It will be
nice to refactor a bit and initialize ReplicaManager with MetadataCache as a
member, so we can avoid passing references around. I avoided this in this PR to
keep it focused.
2. I kept running into timeouts with controlled shutdown while testing this. It
doesn't seem like the same issue (i.e. they reproduced very inconsistently on
both SSL and PLAINTEXT brokers), but worth keeping in mind.
> Can't set SSL as inter-broker-protocol by rolling restart of brokers
> --------------------------------------------------------------------
>
> Key: KAFKA-2738
> URL: https://issues.apache.org/jira/browse/KAFKA-2738
> Project: Kafka
> Issue Type: Bug
> Reporter: Gwen Shapira
> Assignee: Ben Stopford
> Priority: Blocker
> Fix For: 0.9.0.0
>
>
> Scenario (as carefully documented by [~benstopford]:
> 1. Start 2 or more brokers with listeners on both PLAINTEXT and SSL
> protocols, and PLAINTEXT as security.inter.broker.protocol:
> inter.broker.protocol.version = 0.9.0.X
> security.inter.broker.protocol = PLAINTEXT
> listeners = PLAINTEXT://:9092,SSL://:9093
> 2. Stop one of the brokers and change security.inter.broker.protocol to SSL
> inter.broker.protocol.version = 0.9.0.X
> security.inter.broker.protocol = SSL
> listeners = PLAINTEXT://:9092,SSL://:9093
> 3. Start that broker again.
> You will get replication errors as it will attempt to use SSL on a PLAINTEXT
> port:
> {code}
> WARN ReplicaFetcherThread-0-3, Error in fetch
> kafka.server.ReplicaFetcherThread$FetchRequest@78ca3ba1. Possible cause:
> java.io.IOException: Connection to Node(3, worker4, 9092) failed
> (kafka.server.ReplicaFetcherThread)
> WARN Failed to send SSL Close message
> (org.apache.kafka.common.network.SslTransportLayer)
> java.io.IOException: Broken pipe
> at sun.nio.ch.FileDispatcherImpl.write0(Native Method)
> at sun.nio.ch.SocketDispatcher.write(SocketDispatcher.java:47)
> at sun.nio.ch.IOUtil.writeFromNativeBuffer(IOUtil.java:93)
> at sun.nio.ch.IOUtil.write(IOUtil.java:65)
> at sun.nio.ch.SocketChannelImpl.write(SocketChannelImpl.java:492)
> at
> org.apache.kafka.common.network.SslTransportLayer.flush(SslTransportLayer.java:188)
> at
> org.apache.kafka.common.network.SslTransportLayer.close(SslTransportLayer.java:161)
> at org.apache.kafka.common.network.KafkaChannel.close(KafkaChannel.java:50)
> at org.apache.kafka.common.network.Selector.close(Selector.java:448)
> at org.apache.kafka.common.network.Selector.poll(Selector.java:316)
> at org.apache.kafka.clients.NetworkClient.poll(NetworkClient.java:270)
> at
> kafka.utils.NetworkClientBlockingOps$.recurse$1(NetworkClientBlockingOps.scala:128)
> at
> kafka.utils.NetworkClientBlockingOps$.kafka$utils$NetworkClientBlockingOps$$pollUntilFound$extension(NetworkClientBlockingOps.scala:139)
> at
> kafka.utils.NetworkClientBlockingOps$.kafka$utils$NetworkClientBlockingOps$$pollUntil$extension(NetworkClientBlockingOps.scala:105)
> at
> kafka.utils.NetworkClientBlockingOps$.blockingReady$extension(NetworkClientBlockingOps.scala:58)
> at
> kafka.server.ReplicaFetcherThread.sendRequest(ReplicaFetcherThread.scala:202)
> at kafka.server.ReplicaFetcherThread.fetch(ReplicaFetcherThread.scala:192)
> at kafka.server.ReplicaFetcherThread.fetch(ReplicaFetcherThread.scala:42)
> at
> kafka.server.AbstractFetcherThread.processFetchRequest(AbstractFetcherThread.scala:102)
> at kafka.server.AbstractFetcherThread.doWork(AbstractFetcherThread.scala:93)
> at kafka.utils.ShutdownableThread.run(ShutdownableThread.scala:63)
> {code}
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
