[jira] [Created] (KAFKA-4268) huge server.logs during the error frequently happen "Message format version for partition 200 not found"

["Peyton Peng (JIRA)" <ji...@apache.org>]

Peyton Peng created KAFKA-4268:
----------------------------------

             Summary: huge server.logs during the error frequently happen "Message format version for partition 200 not found" 
                 Key: KAFKA-4268
                 URL: https://issues.apache.org/jira/browse/KAFKA-4268
             Project: Kafka
          Issue Type: Bug
          Components: core
    Affects Versions: 0.10.0.1
         Environment: Ubuntu 14.04.2 LTS
            Reporter: Peyton Peng


We have three broker running on three different ubuntu servers, there are about more than 200 topics served for different type of data. Each topic have 3 partitions. While I added a new consumer with the group "wlas" on two different machine(centos), which is running within flume, to subscribe all above topics in one consumer, after a while, we got the exception as below, the exception throws in server.log, and loops frequently, which cause the disk space leak.

I doubt if the issue related with "one consumer subscribe so many topics", and running on two machine.
(even got this exception, the topic data consume logic is running well)

-------------- exception segment:
[2016-10-07 08:49:27,982] INFO [GroupCoordinator 1]: Assignment received from leader for group wlas for generation 10 (kafka.coordinator.GroupCoordinator)
[2016-10-07 08:49:27,982] ERROR [KafkaApi-1] Error when handling request {group_id=wlas,generation_id=10,member_id=flume-1011-6ec8e841-9fa8-4014-97b8-211b2ee2465e,group_assignment=[{member_id=flume-1011-6ec8e841-9fa8-4014-97b8-211b2ee2465e,member_assignment=java.nio.HeapByteBuffer[pos=0 lim=2301 cap=5023]},{member_id=flume-1009-7f6eacb5-8885-4131-8286-3848ed17b2a9,member_assignment=java.nio.HeapByteBuffer[pos=0 lim=2669 cap=2669]}]} (kafka.server.KafkaApis)
java.lang.IllegalArgumentException: Message format version for partition 200 not found
   at kafka.coordinator.GroupMetadataManager$$anonfun$14.apply(GroupMetadataManager.scala:633)
   at kafka.coordinator.GroupMetadataManager$$anonfun$14.apply(GroupMetadataManager.scala:633)
   at scala.Option.getOrElse(Option.scala:121)
   at kafka.coordinator.GroupMetadataManager.kafka$coordinator$GroupMetadataManager$$getMessageFormatVersionAndTimestamp(GroupMetadataManager.scala:632)
   at kafka.coordinator.GroupMetadataManager.prepareStoreGroup(GroupMetadataManager.scala:171)
   at kafka.coordinator.GroupCoordinator.doSyncGroup(GroupCoordinator.scala:276)
   at kafka.coordinator.GroupCoordinator.handleSyncGroup(GroupCoordinator.scala:240)
   at kafka.server.KafkaApis.handleSyncGroupRequest(KafkaApis.scala:933)
   at kafka.server.KafkaApis.handle(KafkaApis.scala:90)
   at kafka.server.KafkaRequestHandler.run(KafkaRequestHandler.scala:60)
   at java.lang.Thread.run(Thread.java:745) 



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)