This is probably related to kafka-1382. The root cause is likely ZK session
expiration in the broker. Did you see any?
Thanks,
Jun
On Mon, Jun 9, 2014 at 8:11 PM, Bongyeon Kim <bongyeon....@gmail.com> wrote:
> Hi, team.
>
> I’m using 0.8.1.
> I found some strange log repeatedly on server.log in one of my brokers and
> it keeps logging until now.
>
> server.log
> ======================================================================================
> ...
> [2014-06-09 10:41:47,402] ERROR Conditional update of path
> /brokers/topics/topicTRACE/partitions/1/state with data
> {"controller_epoch":19,"leader":2,"version":1,"leader_epoch":43,"isr":[4,2]}
> and expected version 439 failed due to
> org.apache.zookeeper.KeeperException$BadVersionException: KeeperErrorCode =
> BadVersion for /brokers/topics/topicTRACE/partitions/1/state
> (kafka.utils.ZkUtils$)
> [2014-06-09 10:41:47,402] INFO Partition [topicTRACE,1] on broker 2:
> Cached zkVersion [439] not equal to that in zookeeper, skip updating ISR
> (kafka.cluster.Partition)
> [2014-06-09 10:41:47,402] INFO Partition [topicDEBUG,0] on broker 2:
> Shrinking ISR for partition [topicDEBUG,0] from 1,3,2 to 2
> (kafka.cluster.Partition)
> [2014-06-09 10:41:47,416] ERROR Conditional update of path
> /brokers/topics/topicDEBUG/partitions/0/state with data
> {"controller_epoch":19,"leader":2,"version":1,"leader_epoch":43,"isr":[2]}
> and expected version 1424 failed due to
> org.apache.zookeeper.KeeperException$BadVersionException: KeeperErrorCode =
> BadVersion for /brokers/topics/topicDEBUG/partitions/0/state
> (kafka.utils.ZkUtils$)
> [2014-06-09 10:41:47,432] INFO Partition [topicDEBUG,0] on broker 2:
> Cached zkVersion [1424] not equal to that in zookeeper, skip updating ISR
> (kafka.cluster.Partition)
> [2014-06-09 10:41:47,432] INFO Partition [topicCDR,3] on broker 2:
> Shrinking ISR for partition [topicCDR,3] from 4,1,2 to 2
> (kafka.cluster.Partition)
> [2014-06-09 10:41:47,435] ERROR Conditional update of path
> /brokers/topics/topicCDR/partitions/3/state with data
> {"controller_epoch":19,"leader":2,"version":1,"leader_epoch":46,"isr":[2]}
> and expected version 541 failed due to
> org.apache.zookeeper.KeeperException$BadVersionException: KeeperErrorCode =
> BadVersion for /brokers/topics/topicCDR/partitions/3/state
> (kafka.utils.ZkUtils$)
> [2014-06-09 10:41:47,435] INFO Partition [topicCDR,3] on broker 2: Cached
> zkVersion [541] not equal to that in zookeeper, skip updating ISR
> (kafka.cluster.Partition)
> [2014-06-09 10:41:48,426] INFO Partition [topicTRACE,1] on broker 2:
> Shrinking ISR for partition [topicTRACE,1] from 4,3,2 to 4,2
> (kafka.cluster.Partition)
> ...
>
> =================================================================================================
>
> and found some error and warning in controller.log
>
>
> controller.log
> ======================================================================================
> ...
> [2014-06-09 10:42:03,962] WARN [Controller-3-to-broker-1-send-thread],
> Controller 3 fails to send a request to broker
> id:1,host:c-ccp-tk1-a58,port:9091 (kafka.controller.RequestSendThread)
> java.net.SocketTimeoutException
> at
> sun.nio.ch.SocketAdaptor$SocketInputStream.read(SocketAdaptor.java:229)
> at sun.nio.ch.ChannelInputStream.read(ChannelInputStream.java:103)
> at
> java.nio.channels.Channels$ReadableByteChannelImpl.read(Channels.java:385)
> at kafka.utils.Utils$.read(Utils.scala:375)
> at
> kafka.network.BoundedByteBufferReceive.readFrom(BoundedByteBufferReceive.scala:54)
> at
> kafka.network.Receive$class.readCompletely(Transmission.scala:56)
> at
> kafka.network.BoundedByteBufferReceive.readCompletely(BoundedByteBufferReceive.scala:29)
> at kafka.network.BlockingChannel.receive(BlockingChannel.scala:100)
> at
> kafka.controller.RequestSendThread.doWork(ControllerChannelManager.scala:146)
> at kafka.utils.ShutdownableThread.run(ShutdownableThread.scala:51)
> [2014-06-09 10:42:03,964] ERROR [Controller-3-to-broker-1-send-thread],
> Controller 3 epoch 21 failed to send UpdateMetadata request with
> correlation id 1 to broker id:1,host:c-ccp-tk1-a58,port:9091. Reconnecting
> to broker. (kafka.controller.RequestSendThread)
> java.nio.channels.ClosedChannelException
> at kafka.network.BlockingChannel.send(BlockingChannel.scala:89)
> at
> kafka.controller.RequestSendThread.liftedTree1$1(ControllerChannelManager.scala:132)
> at
> kafka.controller.RequestSendThread.doWork(ControllerChannelManager.scala:131)
> at kafka.utils.ShutdownableThread.run(ShutdownableThread.scala:51)
>
> ...
>
> [2014-06-09 10:42:38,064] WARN [OfflinePartitionLeaderSelector]: No broker
> in ISR is alive for [topicTRACE,0]. Elect leader 3 from live brokers 3.
> There's potential data loss.
> (kafka.controller.OfflinePartitionLeaderSelector)
> ...
>
> =================================================================================================
>
> Why is this happen? Is there any possibilities data loss?
> To normalize my brokers, What I have to do? Do I have to restart this
> broker?
>
>
> Thanks in advance.
>
>
>
