[jira] [Created] (KAFKA-7057) Consumer stop polling

["Moshe Lavi (JIRA)" <ji...@apache.org>]

Moshe Lavi created KAFKA-7057:
---------------------------------

             Summary: Consumer stop polling
                 Key: KAFKA-7057
                 URL: https://issues.apache.org/jira/browse/KAFKA-7057
             Project: Kafka
          Issue Type: Bug
          Components: consumer, controller
    Affects Versions: 0.10.1.1
            Reporter: Moshe Lavi


We build 3 Kafka brokers (0.10.1.1) version using Spring Cloud Stream consumer to poll messages.
We encountered consumer lags alerted and found some consumers were blocked and not polling anymore messages. This requires us to restart the microservice where that consumer resides.

I wonder if this has to do with lack of available threads or to the fact there heartbeat daemon does not exist/work.


*The thread dump shows:*

kafka-coordinator-heartbeat-thread | SiteAgreementItem" #4943 daemon prio=5 os_prio=0 tid=0x00007f3abdd08000 nid=0x83ac waiting for monitor entry [0x00007f3a5dcdb000]

   java.lang.Thread.State: BLOCKED (on object monitor)

                at org.apache.kafka.clients.consumer.internals.ConsumerNetworkClient.disableWakeups(ConsumerNetworkClient.java:409)

                - waiting to lock <*0x00000005df800450*> (a org.apache.kafka.clients.consumer.internals.ConsumerNetworkClient)

                at org.apache.kafka.clients.consumer.internals.ConsumerNetworkClient.pollNoWakeup(ConsumerNetworkClient.java:264)

                at org.apache.kafka.clients.consumer.internals.AbstractCoordinator$HeartbeatThread.run(AbstractCoordinator.java:865)

                - locked <0x00000005df800488> (a org.apache.kafka.clients.consumer.internals.ConsumerCoordinator)

 

-kafka-consumer-1" #4940 prio=5 os_prio=0 tid=0x00007f3a8d433800 nid=0x838e runnable [0x00007f3a5dedd000]

   java.lang.Thread.State: RUNNABLE

                at sun.nio.ch.EPollArrayWrapper.epollWait(Native Method)

                at sun.nio.ch.EPollArrayWrapper.poll(EPollArrayWrapper.java:269)

                at sun.nio.ch.EPollSelectorImpl.doSelect(EPollSelectorImpl.java:79)

                at sun.nio.ch.SelectorImpl.lockAndDoSelect(SelectorImpl.java:86)

                - locked <0x00000005df7705e0> (a sun.nio.ch.Util$2)

                - locked <0x00000005df7705d0> (a java.util.Collections$UnmodifiableSet)

                - locked <0x00000005df7705f0> (a sun.nio.ch.EPollSelectorImpl)

                at sun.nio.ch.SelectorImpl.select(SelectorImpl.java:97)

                at org.apache.kafka.common.network.Selector.select(Selector.java:470)

                at org.apache.kafka.common.network.Selector.poll(Selector.java:286)

                at org.apache.kafka.clients.NetworkClient.poll(NetworkClient.java:260)

                at org.apache.kafka.clients.consumer.internals.ConsumerNetworkClient.poll(ConsumerNetworkClient.java:232)

                - locked <*0x00000005df800450*> (a org.apache.kafka.clients.consumer.internals.ConsumerNetworkClient)

                at org.apache.kafka.clients.consumer.KafkaConsumer.pollOnce(KafkaConsumer.java:1031)

                at org.apache.kafka.clients.consumer.KafkaConsumer.poll(KafkaConsumer.java:979)

                at org.springframework.kafka.listener.KafkaMessageListenerContainer$ListenerConsumer.run(KafkaMessageListenerContainer.java:532)

                at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511)

                at java.util.concurrent.FutureTask.run(FutureTask.java:266)

                at java.lang.Thread.run(Thread.java:745)



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)