[jira] [Commented] (KAFKA-7088) Kafka streams thread waits infinitely on transaction init
[
https://issues.apache.org/jira/browse/KAFKA-7088?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16769912#comment-16769912
]
Peter Davis commented on KAFKA-7088:
Encountered this bug today with 0.11.0 client. Similar same stack trace and
symptoms. Caused a major issue in production.
In 0.11.0 there is no timeout for this await(). A timeout was added by
KAFKA-6446 in 2.x. KAFKA-6446 also appears to fix the deadlock – instead an
error would be thrown. (Fixed/dup?)
KAFKA-6446 also suggests the root cause – *the broker is either unavailable or
has stopped responding to InitProducerIdRequests.*
**Very interesting error on broker that immediately preceded the outage –
*"failed: this should not happen"* followed by a transaction metadata error.
Eventually, we identified that this particular broker was in a bad state,
restarted it, and the problem immediately resolved. Broker is running CP
4.1.2, which is Apache Kafka 1.1.1-cp1 (not sure what all is in the -cp1
patch). It looks like this particular tx metadata error is thrown when
validating the producer epoch, which means either the broker or producer times
out. This particular broker is on a VM and we've seen issues with
pauses/performance issues in the past that can cause a broker to get into a bad
state after timing out (unrelated but illustrative example: ).
{quote}{{February 15th 2019, 04:32:15.538 xx_broker_xxx.example.com
[2019-02-15 12:32:08,191] ERROR
TransactionMetadata(transactionalId=xx_group_xxx--0_1,
producerId=106000, producerEpoch=427, txnTimeoutMs=6, state=CompleteCommit,
pendingState=Some(Ongoing), topicPartitions=Set(),
txnStartTimestamp=1550233928711, txnLastUpdateTimestamp=1550233927976)'s
transition to TxnTransitMetadata(producerId=106000, producerEpoch=427,
txnTimeoutMs=6, txnState=Ongoing,
topicPartitions=Set(xx_topic_xxx-1), txnStartTimestamp=1550233928001,
txnLastUpdateTimestamp=1550233928001) +*failed: this should not happen*+
(kafka.coordinator.transaction.TransactionMetadata)}}{{February 15th 2019,
04:32:15.538 xx_broker_xxx.example.com [2019-02-15 12:32:08,239] ERROR
[KafkaApi-1] Error when handling request
\{replica_id=103,max_wait_time=500,min_bytes=1,max_bytes=10485760,isolation_level=0,session_id=647691730,epoch=6210667,topics=[{topic=__transaction_state,partitions=[{partition=25,fetch_offset=32613278,log_start_offset=0,max_bytes=10485760}]}],forgetten_topics_data=[]}
(kafka.server.KafkaApis)}}
{{java.lang.IllegalStateException: TransactionalId xx_group_xxx-0_1
*failed transition to state* TxnTransitMetadata(producerId=106000,
producerEpoch=427, txnTimeoutMs=6, txnState=Ongoing,
topicPartitions=Set(xx_topic_xxx-1), txnStartTimestamp=1550233928001,
txnLastUpdateTimestamp=1550233928001) due to unexpected metadata}}
{{ at
kafka.coordinator.transaction.TransactionMetadata.throwStateTransitionFailure(TransactionMetadata.scala:390)}}
{{ at
kafka.coordinator.transaction.TransactionMetadata.*completeTransitionTo(TransactionMetadata.scala:326)*}}
{{ at
kafka.coordinator.transaction.TransactionStateManager$$anonfun$kafka$coordinator$transaction$TransactionStateManager$$updateCacheCallback$1$1.apply$mcV$sp(TransactionStateManager.scala:534)}}
{{ at
kafka.coordinator.transaction.TransactionStateManager$$anonfun$kafka$coordinator$transaction$TransactionStateManager$$updateCacheCallback$1$1.apply(TransactionStateManager.scala:526)}}
{{ at
kafka.coordinator.transaction.TransactionStateManager$$anonfun$kafka$coordinator$transaction$TransactionStateManager$$updateCacheCallback$1$1.apply(TransactionStateManager.scala:526)}}
{{ at kafka.utils.CoreUtils$.inLock(CoreUtils.scala:250)}}
{{ at
kafka.coordinator.transaction.TransactionMetadata.inLock(TransactionMetadata.scala:172)}}
{{ at
kafka.coordinator.transaction.TransactionStateManager.kafka$coordinator$transaction$TransactionStateManager$$updateCacheCallback$1(TransactionStateManager.scala:525)}}
{{ at
kafka.coordinator.transaction.TransactionStateManager$$anonfun$appendTransactionToLog$1$$anonfun$apply$mcV$sp$11.apply(TransactionStateManager.scala:620)}}
{{ at
kafka.coordinator.transaction.TransactionStateManager$$anonfun$appendTransactionToLog$1$$anonfun$apply$mcV$sp$11.apply(TransactionStateManager.scala:620)}}
{{ at kafka.server.DelayedProduce.onComplete(DelayedProduce.scala:129)}}
{{ at kafka.server.DelayedOperation.forceComplete(DelayedOperation.scala:70)}}
{{ at kafka.server.DelayedProduce.tryComplete(DelayedProduce.scala:111)}}
{{ at
kafka.server.DelayedOperation.maybeTryComplete(DelayedOperation.scala:121)}}
{{ at
kafka.server.DelayedOperationPurgatory$Watchers.tryCompleteWatched(DelayedOperation.scala:371)}}
{{ at
kafka.server.DelayedOperationPurgatory.checkAndComplete(DelayedOperation.scala:277)}}
{{ at
kafka.server.ReplicaManager.tryCompleteDelayedProduce(ReplicaManager.scala:289)}}
{{ at
[jira] [Commented] (KAFKA-7088) Kafka streams thread waits infinitely on transaction init
[
https://issues.apache.org/jira/browse/KAFKA-7088?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16523725#comment-16523725
]
Lukasz Gluchowski commented on KAFKA-7088:
--
Sorry but DEBUG is still to verbose. It is impacting resources used by the
application
> Kafka streams thread waits infinitely on transaction init
> -
>
> Key: KAFKA-7088
> URL: https://issues.apache.org/jira/browse/KAFKA-7088
> Project: Kafka
> Issue Type: Bug
> Components: clients
>Affects Versions: 1.0.1
> Environment: Linux 4.14.33-51.37.amzn1.x86_64 #1 SMP Thu May 3
> 20:07:43 UTC 2018
> kafka-streams (client) 1.0.1
> kafka broker 1.1.0
> Java version:
> OpenJDK Runtime Environment (build 1.8.0_171-b10)
> OpenJDK 64-Bit Server VM (build 25.171-b10, mixed mode)
> kakfa config overrides:
> num.stream.threads: 6
> session.timeout.ms: 1
> request.timeout.ms: 11000
> fetch.max.wait.ms: 500
> max.poll.records: 1000
> topic has 24 partitions
>Reporter: Lukasz Gluchowski
>Priority: Major
> Labels: eos
>
> A kafka stream application thread stops processing without any feedback. The
> topic has 24 partitions and I noticed that processing stopped only for some
> partitions. I will describe what happened to partition:10. The application is
> still running (now for about 8 hours) and that thread is hanging there and no
> rebalancing that took place.
> There is no error (we have a custom `Thread.UncaughtExceptionHandler` which
> was not called). I noticed that after couple of minutes stream stopped
> processing (at offset 32606948 where log-end-offset is 33472402).
> Broker itself is not reporting any active consumer in that consumer group and
> the only info I was able to gather was from thread dump:
> {code:java}
> "mp_ads_publisher_pro_madstorage-web-corotos-prod-9db804ae-2a7a-431f-be09-392ab38cd8a2-StreamThread-33"
> #113 prio=5 os_prio=0 tid=0x7fe07c6349b0 nid=0xf7a waiting on condition
> [0x7fe0215d4000]
> java.lang.Thread.State: WAITING (parking)
> at sun.misc.Unsafe.park(Native Method)
> - parking to wait for <0xfec6a2f8> (a
> java.util.concurrent.CountDownLatch$Sync)
> at java.util.concurrent.locks.LockSupport.park(LockSupport.java:175)
> at
> java.util.concurrent.locks.AbstractQueuedSynchronizer.parkAndCheckInterrupt(AbstractQueuedSynchronizer.java:836)
> at
> java.util.concurrent.locks.AbstractQueuedSynchronizer.doAcquireSharedInterruptibly(AbstractQueuedSynchronizer.java:997)
> at
> java.util.concurrent.locks.AbstractQueuedSynchronizer.acquireSharedInterruptibly(AbstractQueuedSynchronizer.java:1304)
> at java.util.concurrent.CountDownLatch.await(CountDownLatch.java:231)
> at
> org.apache.kafka.clients.producer.internals.TransactionalRequestResult.await(TransactionalRequestResult.java:50)
> at
> org.apache.kafka.clients.producer.KafkaProducer.initTransactions(KafkaProducer.java:554)
> at
> org.apache.kafka.streams.processor.internals.StreamTask.(StreamTask.java:151)
> at
> org.apache.kafka.streams.processor.internals.StreamThread$TaskCreator.createTask(StreamThread.java:404)
> at
> org.apache.kafka.streams.processor.internals.StreamThread$TaskCreator.createTask(StreamThread.java:365)
> at
> org.apache.kafka.streams.processor.internals.StreamThread$AbstractTaskCreator.createTasks(StreamThread.java:350)
> at
> org.apache.kafka.streams.processor.internals.TaskManager.addStreamTasks(TaskManager.java:137)
> at
> org.apache.kafka.streams.processor.internals.TaskManager.createTasks(TaskManager.java:88)
> at
> org.apache.kafka.streams.processor.internals.StreamThread$RebalanceListener.onPartitionsAssigned(StreamThread.java:259)
> at
> org.apache.kafka.clients.consumer.internals.ConsumerCoordinator.onJoinComplete(ConsumerCoordinator.java:264)
> at
> org.apache.kafka.clients.consumer.internals.AbstractCoordinator.joinGroupIfNeeded(AbstractCoordinator.java:367)
> at
> org.apache.kafka.clients.consumer.internals.AbstractCoordinator.ensureActiveGroup(AbstractCoordinator.java:316)
> at
> org.apache.kafka.clients.consumer.internals.ConsumerCoordinator.poll(ConsumerCoordinator.java:295)
> at
> org.apache.kafka.clients.consumer.KafkaConsumer.pollOnce(KafkaConsumer.java:1146)
> at
> org.apache.kafka.clients.consumer.KafkaConsumer.poll(KafkaConsumer.java:)
> at
> org.apache.kafka.streams.processor.internals.StreamThread.pollRequests(StreamThread.java:851)
> at
> org.apache.kafka.streams.processor.internals.StreamThread.runOnce(StreamThread.java:808)
> at
> org.apache.kafka.streams.processor.internals.StreamThread.runLoop(StreamThread.java:774)
> at
> org.apache.kafka.streams.processor.internals.StreamThread.run(StreamThread.java:744){code}
>
> I tried restarting application once but the situation repeated. Thread read
> some data, committed offset and
[jira] [Commented] (KAFKA-7088) Kafka streams thread waits infinitely on transaction init
[
https://issues.apache.org/jira/browse/KAFKA-7088?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16522031#comment-16522031
]
Ted Yu commented on KAFKA-7088:
---
Maybe set log level to DEBUG when you have a chance.
Thanks
> Kafka streams thread waits infinitely on transaction init
> -
>
> Key: KAFKA-7088
> URL: https://issues.apache.org/jira/browse/KAFKA-7088
> Project: Kafka
> Issue Type: Bug
> Components: clients
>Affects Versions: 1.0.1
> Environment: Linux 4.14.33-51.37.amzn1.x86_64 #1 SMP Thu May 3
> 20:07:43 UTC 2018
> kafka-streams (client) 1.0.1
> kafka broker 1.1.0
> Java version:
> OpenJDK Runtime Environment (build 1.8.0_171-b10)
> OpenJDK 64-Bit Server VM (build 25.171-b10, mixed mode)
> kakfa config overrides:
> num.stream.threads: 6
> session.timeout.ms: 1
> request.timeout.ms: 11000
> fetch.max.wait.ms: 500
> max.poll.records: 1000
> topic has 24 partitions
>Reporter: Lukasz Gluchowski
>Priority: Major
> Labels: eos
>
> A kafka stream application thread stops processing without any feedback. The
> topic has 24 partitions and I noticed that processing stopped only for some
> partitions. I will describe what happened to partition:10. The application is
> still running (now for about 8 hours) and that thread is hanging there and no
> rebalancing that took place.
> There is no error (we have a custom `Thread.UncaughtExceptionHandler` which
> was not called). I noticed that after couple of minutes stream stopped
> processing (at offset 32606948 where log-end-offset is 33472402).
> Broker itself is not reporting any active consumer in that consumer group and
> the only info I was able to gather was from thread dump:
> {code:java}
> "mp_ads_publisher_pro_madstorage-web-corotos-prod-9db804ae-2a7a-431f-be09-392ab38cd8a2-StreamThread-33"
> #113 prio=5 os_prio=0 tid=0x7fe07c6349b0 nid=0xf7a waiting on condition
> [0x7fe0215d4000]
> java.lang.Thread.State: WAITING (parking)
> at sun.misc.Unsafe.park(Native Method)
> - parking to wait for <0xfec6a2f8> (a
> java.util.concurrent.CountDownLatch$Sync)
> at java.util.concurrent.locks.LockSupport.park(LockSupport.java:175)
> at
> java.util.concurrent.locks.AbstractQueuedSynchronizer.parkAndCheckInterrupt(AbstractQueuedSynchronizer.java:836)
> at
> java.util.concurrent.locks.AbstractQueuedSynchronizer.doAcquireSharedInterruptibly(AbstractQueuedSynchronizer.java:997)
> at
> java.util.concurrent.locks.AbstractQueuedSynchronizer.acquireSharedInterruptibly(AbstractQueuedSynchronizer.java:1304)
> at java.util.concurrent.CountDownLatch.await(CountDownLatch.java:231)
> at
> org.apache.kafka.clients.producer.internals.TransactionalRequestResult.await(TransactionalRequestResult.java:50)
> at
> org.apache.kafka.clients.producer.KafkaProducer.initTransactions(KafkaProducer.java:554)
> at
> org.apache.kafka.streams.processor.internals.StreamTask.(StreamTask.java:151)
> at
> org.apache.kafka.streams.processor.internals.StreamThread$TaskCreator.createTask(StreamThread.java:404)
> at
> org.apache.kafka.streams.processor.internals.StreamThread$TaskCreator.createTask(StreamThread.java:365)
> at
> org.apache.kafka.streams.processor.internals.StreamThread$AbstractTaskCreator.createTasks(StreamThread.java:350)
> at
> org.apache.kafka.streams.processor.internals.TaskManager.addStreamTasks(TaskManager.java:137)
> at
> org.apache.kafka.streams.processor.internals.TaskManager.createTasks(TaskManager.java:88)
> at
> org.apache.kafka.streams.processor.internals.StreamThread$RebalanceListener.onPartitionsAssigned(StreamThread.java:259)
> at
> org.apache.kafka.clients.consumer.internals.ConsumerCoordinator.onJoinComplete(ConsumerCoordinator.java:264)
> at
> org.apache.kafka.clients.consumer.internals.AbstractCoordinator.joinGroupIfNeeded(AbstractCoordinator.java:367)
> at
> org.apache.kafka.clients.consumer.internals.AbstractCoordinator.ensureActiveGroup(AbstractCoordinator.java:316)
> at
> org.apache.kafka.clients.consumer.internals.ConsumerCoordinator.poll(ConsumerCoordinator.java:295)
> at
> org.apache.kafka.clients.consumer.KafkaConsumer.pollOnce(KafkaConsumer.java:1146)
> at
> org.apache.kafka.clients.consumer.KafkaConsumer.poll(KafkaConsumer.java:)
> at
> org.apache.kafka.streams.processor.internals.StreamThread.pollRequests(StreamThread.java:851)
> at
> org.apache.kafka.streams.processor.internals.StreamThread.runOnce(StreamThread.java:808)
> at
> org.apache.kafka.streams.processor.internals.StreamThread.runLoop(StreamThread.java:774)
> at
> org.apache.kafka.streams.processor.internals.StreamThread.run(StreamThread.java:744){code}
>
> I tried restarting application once but the situation repeated. Thread read
> some data, committed offset and stopped processing, leaving that thread in
> wait
[jira] [Commented] (KAFKA-7088) Kafka streams thread waits infinitely on transaction init
[
https://issues.apache.org/jira/browse/KAFKA-7088?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16522004#comment-16522004
]
Lukasz Gluchowski commented on KAFKA-7088:
--
[~yuzhih...@gmail.com] I turned TRACE logging on non production environment to
try it out and it was filling the disk very quickly. Sorry but I can't turn it
on on production.
> Kafka streams thread waits infinitely on transaction init
> -
>
> Key: KAFKA-7088
> URL: https://issues.apache.org/jira/browse/KAFKA-7088
> Project: Kafka
> Issue Type: Bug
> Components: clients
>Affects Versions: 1.0.1
> Environment: Linux 4.14.33-51.37.amzn1.x86_64 #1 SMP Thu May 3
> 20:07:43 UTC 2018
> kafka-streams (client) 1.0.1
> kafka broker 1.1.0
> Java version:
> OpenJDK Runtime Environment (build 1.8.0_171-b10)
> OpenJDK 64-Bit Server VM (build 25.171-b10, mixed mode)
> kakfa config overrides:
> num.stream.threads: 6
> session.timeout.ms: 1
> request.timeout.ms: 11000
> fetch.max.wait.ms: 500
> max.poll.records: 1000
> topic has 24 partitions
>Reporter: Lukasz Gluchowski
>Priority: Major
> Labels: eos
>
> A kafka stream application thread stops processing without any feedback. The
> topic has 24 partitions and I noticed that processing stopped only for some
> partitions. I will describe what happened to partition:10. The application is
> still running (now for about 8 hours) and that thread is hanging there and no
> rebalancing that took place.
> There is no error (we have a custom `Thread.UncaughtExceptionHandler` which
> was not called). I noticed that after couple of minutes stream stopped
> processing (at offset 32606948 where log-end-offset is 33472402).
> Broker itself is not reporting any active consumer in that consumer group and
> the only info I was able to gather was from thread dump:
> {code:java}
> "mp_ads_publisher_pro_madstorage-web-corotos-prod-9db804ae-2a7a-431f-be09-392ab38cd8a2-StreamThread-33"
> #113 prio=5 os_prio=0 tid=0x7fe07c6349b0 nid=0xf7a waiting on condition
> [0x7fe0215d4000]
> java.lang.Thread.State: WAITING (parking)
> at sun.misc.Unsafe.park(Native Method)
> - parking to wait for <0xfec6a2f8> (a
> java.util.concurrent.CountDownLatch$Sync)
> at java.util.concurrent.locks.LockSupport.park(LockSupport.java:175)
> at
> java.util.concurrent.locks.AbstractQueuedSynchronizer.parkAndCheckInterrupt(AbstractQueuedSynchronizer.java:836)
> at
> java.util.concurrent.locks.AbstractQueuedSynchronizer.doAcquireSharedInterruptibly(AbstractQueuedSynchronizer.java:997)
> at
> java.util.concurrent.locks.AbstractQueuedSynchronizer.acquireSharedInterruptibly(AbstractQueuedSynchronizer.java:1304)
> at java.util.concurrent.CountDownLatch.await(CountDownLatch.java:231)
> at
> org.apache.kafka.clients.producer.internals.TransactionalRequestResult.await(TransactionalRequestResult.java:50)
> at
> org.apache.kafka.clients.producer.KafkaProducer.initTransactions(KafkaProducer.java:554)
> at
> org.apache.kafka.streams.processor.internals.StreamTask.(StreamTask.java:151)
> at
> org.apache.kafka.streams.processor.internals.StreamThread$TaskCreator.createTask(StreamThread.java:404)
> at
> org.apache.kafka.streams.processor.internals.StreamThread$TaskCreator.createTask(StreamThread.java:365)
> at
> org.apache.kafka.streams.processor.internals.StreamThread$AbstractTaskCreator.createTasks(StreamThread.java:350)
> at
> org.apache.kafka.streams.processor.internals.TaskManager.addStreamTasks(TaskManager.java:137)
> at
> org.apache.kafka.streams.processor.internals.TaskManager.createTasks(TaskManager.java:88)
> at
> org.apache.kafka.streams.processor.internals.StreamThread$RebalanceListener.onPartitionsAssigned(StreamThread.java:259)
> at
> org.apache.kafka.clients.consumer.internals.ConsumerCoordinator.onJoinComplete(ConsumerCoordinator.java:264)
> at
> org.apache.kafka.clients.consumer.internals.AbstractCoordinator.joinGroupIfNeeded(AbstractCoordinator.java:367)
> at
> org.apache.kafka.clients.consumer.internals.AbstractCoordinator.ensureActiveGroup(AbstractCoordinator.java:316)
> at
> org.apache.kafka.clients.consumer.internals.ConsumerCoordinator.poll(ConsumerCoordinator.java:295)
> at
> org.apache.kafka.clients.consumer.KafkaConsumer.pollOnce(KafkaConsumer.java:1146)
> at
> org.apache.kafka.clients.consumer.KafkaConsumer.poll(KafkaConsumer.java:)
> at
> org.apache.kafka.streams.processor.internals.StreamThread.pollRequests(StreamThread.java:851)
> at
> org.apache.kafka.streams.processor.internals.StreamThread.runOnce(StreamThread.java:808)
> at
> org.apache.kafka.streams.processor.internals.StreamThread.runLoop(StreamThread.java:774)
> at
> org.apache.kafka.streams.processor.internals.StreamThread.run(StreamThread.java:744){code}
>
> I tried restarting
[jira] [Commented] (KAFKA-7088) Kafka streams thread waits infinitely on transaction init
[
https://issues.apache.org/jira/browse/KAFKA-7088?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16520844#comment-16520844
]
Ted Yu commented on KAFKA-7088:
---
Lukasz:
I was looking at code related to how queued TxnRequestHandler's are processed.
Is it possible for you to turn on DEBUG logging and reproduce this scenario ?
Some logs are at TRACE level:
{code}
if (maybeTerminateRequestWithError(nextRequestHandler)) {
log.trace("Not sending transactional request {} because we are in
an error state",
{code}
But I guess the log file would be chatty once TRACE is turned on.
> Kafka streams thread waits infinitely on transaction init
> -
>
> Key: KAFKA-7088
> URL: https://issues.apache.org/jira/browse/KAFKA-7088
> Project: Kafka
> Issue Type: Bug
> Components: clients
>Affects Versions: 1.0.1
> Environment: Linux 4.14.33-51.37.amzn1.x86_64 #1 SMP Thu May 3
> 20:07:43 UTC 2018
> kafka-streams (client) 1.0.1
> kafka broker 1.1.0
> Java version:
> OpenJDK Runtime Environment (build 1.8.0_171-b10)
> OpenJDK 64-Bit Server VM (build 25.171-b10, mixed mode)
> kakfa config overrides:
> num.stream.threads: 6
> session.timeout.ms: 1
> request.timeout.ms: 11000
> fetch.max.wait.ms: 500
> max.poll.records: 1000
> topic has 24 partitions
>Reporter: Lukasz Gluchowski
>Priority: Major
> Labels: eos
>
> A kafka stream application thread stops processing without any feedback. The
> topic has 24 partitions and I noticed that processing stopped only for some
> partitions. I will describe what happened to partition:10. The application is
> still running (now for about 8 hours) and that thread is hanging there and no
> rebalancing that took place.
> There is no error (we have a custom `Thread.UncaughtExceptionHandler` which
> was not called). I noticed that after couple of minutes stream stopped
> processing (at offset 32606948 where log-end-offset is 33472402).
> Broker itself is not reporting any active consumer in that consumer group and
> the only info I was able to gather was from thread dump:
> {code:java}
> "mp_ads_publisher_pro_madstorage-web-corotos-prod-9db804ae-2a7a-431f-be09-392ab38cd8a2-StreamThread-33"
> #113 prio=5 os_prio=0 tid=0x7fe07c6349b0 nid=0xf7a waiting on condition
> [0x7fe0215d4000]
> java.lang.Thread.State: WAITING (parking)
> at sun.misc.Unsafe.park(Native Method)
> - parking to wait for <0xfec6a2f8> (a
> java.util.concurrent.CountDownLatch$Sync)
> at java.util.concurrent.locks.LockSupport.park(LockSupport.java:175)
> at
> java.util.concurrent.locks.AbstractQueuedSynchronizer.parkAndCheckInterrupt(AbstractQueuedSynchronizer.java:836)
> at
> java.util.concurrent.locks.AbstractQueuedSynchronizer.doAcquireSharedInterruptibly(AbstractQueuedSynchronizer.java:997)
> at
> java.util.concurrent.locks.AbstractQueuedSynchronizer.acquireSharedInterruptibly(AbstractQueuedSynchronizer.java:1304)
> at java.util.concurrent.CountDownLatch.await(CountDownLatch.java:231)
> at
> org.apache.kafka.clients.producer.internals.TransactionalRequestResult.await(TransactionalRequestResult.java:50)
> at
> org.apache.kafka.clients.producer.KafkaProducer.initTransactions(KafkaProducer.java:554)
> at
> org.apache.kafka.streams.processor.internals.StreamTask.(StreamTask.java:151)
> at
> org.apache.kafka.streams.processor.internals.StreamThread$TaskCreator.createTask(StreamThread.java:404)
> at
> org.apache.kafka.streams.processor.internals.StreamThread$TaskCreator.createTask(StreamThread.java:365)
> at
> org.apache.kafka.streams.processor.internals.StreamThread$AbstractTaskCreator.createTasks(StreamThread.java:350)
> at
> org.apache.kafka.streams.processor.internals.TaskManager.addStreamTasks(TaskManager.java:137)
> at
> org.apache.kafka.streams.processor.internals.TaskManager.createTasks(TaskManager.java:88)
> at
> org.apache.kafka.streams.processor.internals.StreamThread$RebalanceListener.onPartitionsAssigned(StreamThread.java:259)
> at
> org.apache.kafka.clients.consumer.internals.ConsumerCoordinator.onJoinComplete(ConsumerCoordinator.java:264)
> at
> org.apache.kafka.clients.consumer.internals.AbstractCoordinator.joinGroupIfNeeded(AbstractCoordinator.java:367)
> at
> org.apache.kafka.clients.consumer.internals.AbstractCoordinator.ensureActiveGroup(AbstractCoordinator.java:316)
> at
> org.apache.kafka.clients.consumer.internals.ConsumerCoordinator.poll(ConsumerCoordinator.java:295)
> at
> org.apache.kafka.clients.consumer.KafkaConsumer.pollOnce(KafkaConsumer.java:1146)
> at
> org.apache.kafka.clients.consumer.KafkaConsumer.poll(KafkaConsumer.java:)
> at
> org.apache.kafka.streams.processor.internals.StreamThread.pollRequests(StreamThread.java:851)
> at
>
[jira] [Commented] (KAFKA-7088) Kafka streams thread waits infinitely on transaction init
[
https://issues.apache.org/jira/browse/KAFKA-7088?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16520463#comment-16520463
]
Guozhang Wang commented on KAFKA-7088:
--
If you did not override the MAX_BLOCK_MS_CONFIG value, then this call should be
blocked for at most 60 seconds. It is not clear why you observed the thread
hanging on this call forever.
> Kafka streams thread waits infinitely on transaction init
> -
>
> Key: KAFKA-7088
> URL: https://issues.apache.org/jira/browse/KAFKA-7088
> Project: Kafka
> Issue Type: Bug
> Components: clients
>Affects Versions: 1.0.1
> Environment: Linux 4.14.33-51.37.amzn1.x86_64 #1 SMP Thu May 3
> 20:07:43 UTC 2018
> kafka-streams (client) 1.0.1
> kafka broker 1.1.0
> Java version:
> OpenJDK Runtime Environment (build 1.8.0_171-b10)
> OpenJDK 64-Bit Server VM (build 25.171-b10, mixed mode)
> kakfa config overrides:
> num.stream.threads: 6
> session.timeout.ms: 1
> request.timeout.ms: 11000
> fetch.max.wait.ms: 500
> max.poll.records: 1000
> topic has 24 partitions
>Reporter: Lukasz Gluchowski
>Priority: Major
> Labels: eos
>
> A kafka stream application thread stops processing without any feedback. The
> topic has 24 partitions and I noticed that processing stopped only for some
> partitions. I will describe what happened to partition:10. The application is
> still running (now for about 8 hours) and that thread is hanging there and no
> rebalancing that took place.
> There is no error (we have a custom `Thread.UncaughtExceptionHandler` which
> was not called). I noticed that after couple of minutes stream stopped
> processing (at offset 32606948 where log-end-offset is 33472402).
> Broker itself is not reporting any active consumer in that consumer group and
> the only info I was able to gather was from thread dump:
> {code:java}
> "mp_ads_publisher_pro_madstorage-web-corotos-prod-9db804ae-2a7a-431f-be09-392ab38cd8a2-StreamThread-33"
> #113 prio=5 os_prio=0 tid=0x7fe07c6349b0 nid=0xf7a waiting on condition
> [0x7fe0215d4000]
> java.lang.Thread.State: WAITING (parking)
> at sun.misc.Unsafe.park(Native Method)
> - parking to wait for <0xfec6a2f8> (a
> java.util.concurrent.CountDownLatch$Sync)
> at java.util.concurrent.locks.LockSupport.park(LockSupport.java:175)
> at
> java.util.concurrent.locks.AbstractQueuedSynchronizer.parkAndCheckInterrupt(AbstractQueuedSynchronizer.java:836)
> at
> java.util.concurrent.locks.AbstractQueuedSynchronizer.doAcquireSharedInterruptibly(AbstractQueuedSynchronizer.java:997)
> at
> java.util.concurrent.locks.AbstractQueuedSynchronizer.acquireSharedInterruptibly(AbstractQueuedSynchronizer.java:1304)
> at java.util.concurrent.CountDownLatch.await(CountDownLatch.java:231)
> at
> org.apache.kafka.clients.producer.internals.TransactionalRequestResult.await(TransactionalRequestResult.java:50)
> at
> org.apache.kafka.clients.producer.KafkaProducer.initTransactions(KafkaProducer.java:554)
> at
> org.apache.kafka.streams.processor.internals.StreamTask.(StreamTask.java:151)
> at
> org.apache.kafka.streams.processor.internals.StreamThread$TaskCreator.createTask(StreamThread.java:404)
> at
> org.apache.kafka.streams.processor.internals.StreamThread$TaskCreator.createTask(StreamThread.java:365)
> at
> org.apache.kafka.streams.processor.internals.StreamThread$AbstractTaskCreator.createTasks(StreamThread.java:350)
> at
> org.apache.kafka.streams.processor.internals.TaskManager.addStreamTasks(TaskManager.java:137)
> at
> org.apache.kafka.streams.processor.internals.TaskManager.createTasks(TaskManager.java:88)
> at
> org.apache.kafka.streams.processor.internals.StreamThread$RebalanceListener.onPartitionsAssigned(StreamThread.java:259)
> at
> org.apache.kafka.clients.consumer.internals.ConsumerCoordinator.onJoinComplete(ConsumerCoordinator.java:264)
> at
> org.apache.kafka.clients.consumer.internals.AbstractCoordinator.joinGroupIfNeeded(AbstractCoordinator.java:367)
> at
> org.apache.kafka.clients.consumer.internals.AbstractCoordinator.ensureActiveGroup(AbstractCoordinator.java:316)
> at
> org.apache.kafka.clients.consumer.internals.ConsumerCoordinator.poll(ConsumerCoordinator.java:295)
> at
> org.apache.kafka.clients.consumer.KafkaConsumer.pollOnce(KafkaConsumer.java:1146)
> at
> org.apache.kafka.clients.consumer.KafkaConsumer.poll(KafkaConsumer.java:)
> at
> org.apache.kafka.streams.processor.internals.StreamThread.pollRequests(StreamThread.java:851)
> at
> org.apache.kafka.streams.processor.internals.StreamThread.runOnce(StreamThread.java:808)
> at
> org.apache.kafka.streams.processor.internals.StreamThread.runLoop(StreamThread.java:774)
> at
> org.apache.kafka.streams.processor.internals.StreamThread.run(StreamThread.java:744){code}
>
> I tried restarting
[jira] [Commented] (KAFKA-7088) Kafka streams thread waits infinitely on transaction init
[
https://issues.apache.org/jira/browse/KAFKA-7088?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16520396#comment-16520396
]
Lukasz Gluchowski commented on KAFKA-7088:
--
I switched from "exactly once" to "at least once" and the problem went away.
> Kafka streams thread waits infinitely on transaction init
> -
>
> Key: KAFKA-7088
> URL: https://issues.apache.org/jira/browse/KAFKA-7088
> Project: Kafka
> Issue Type: Bug
> Components: clients
>Affects Versions: 1.0.1
> Environment: Linux 4.14.33-51.37.amzn1.x86_64 #1 SMP Thu May 3
> 20:07:43 UTC 2018
> kafka-streams (client) 1.0.1
> kafka broker 1.1.0
> Java version:
> OpenJDK Runtime Environment (build 1.8.0_171-b10)
> OpenJDK 64-Bit Server VM (build 25.171-b10, mixed mode)
> kakfa config overrides:
> num.stream.threads: 6
> session.timeout.ms: 1
> request.timeout.ms: 11000
> fetch.max.wait.ms: 500
> max.poll.records: 1000
> topic has 24 partitions
>Reporter: Lukasz Gluchowski
>Priority: Major
> Labels: eos
>
> A kafka stream application thread stops processing without any feedback. The
> topic has 24 partitions and I noticed that processing stopped only for some
> partitions. I will describe what happened to partition:10. The application is
> still running (now for about 8 hours) and that thread is hanging there and no
> rebalancing that took place.
> There is no error (we have a custom `Thread.UncaughtExceptionHandler` which
> was not called). I noticed that after couple of minutes stream stopped
> processing (at offset 32606948 where log-end-offset is 33472402).
> Broker itself is not reporting any active consumer in that consumer group and
> the only info I was able to gather was from thread dump:
> {code:java}
> "mp_ads_publisher_pro_madstorage-web-corotos-prod-9db804ae-2a7a-431f-be09-392ab38cd8a2-StreamThread-33"
> #113 prio=5 os_prio=0 tid=0x7fe07c6349b0 nid=0xf7a waiting on condition
> [0x7fe0215d4000]
> java.lang.Thread.State: WAITING (parking)
> at sun.misc.Unsafe.park(Native Method)
> - parking to wait for <0xfec6a2f8> (a
> java.util.concurrent.CountDownLatch$Sync)
> at java.util.concurrent.locks.LockSupport.park(LockSupport.java:175)
> at
> java.util.concurrent.locks.AbstractQueuedSynchronizer.parkAndCheckInterrupt(AbstractQueuedSynchronizer.java:836)
> at
> java.util.concurrent.locks.AbstractQueuedSynchronizer.doAcquireSharedInterruptibly(AbstractQueuedSynchronizer.java:997)
> at
> java.util.concurrent.locks.AbstractQueuedSynchronizer.acquireSharedInterruptibly(AbstractQueuedSynchronizer.java:1304)
> at java.util.concurrent.CountDownLatch.await(CountDownLatch.java:231)
> at
> org.apache.kafka.clients.producer.internals.TransactionalRequestResult.await(TransactionalRequestResult.java:50)
> at
> org.apache.kafka.clients.producer.KafkaProducer.initTransactions(KafkaProducer.java:554)
> at
> org.apache.kafka.streams.processor.internals.StreamTask.(StreamTask.java:151)
> at
> org.apache.kafka.streams.processor.internals.StreamThread$TaskCreator.createTask(StreamThread.java:404)
> at
> org.apache.kafka.streams.processor.internals.StreamThread$TaskCreator.createTask(StreamThread.java:365)
> at
> org.apache.kafka.streams.processor.internals.StreamThread$AbstractTaskCreator.createTasks(StreamThread.java:350)
> at
> org.apache.kafka.streams.processor.internals.TaskManager.addStreamTasks(TaskManager.java:137)
> at
> org.apache.kafka.streams.processor.internals.TaskManager.createTasks(TaskManager.java:88)
> at
> org.apache.kafka.streams.processor.internals.StreamThread$RebalanceListener.onPartitionsAssigned(StreamThread.java:259)
> at
> org.apache.kafka.clients.consumer.internals.ConsumerCoordinator.onJoinComplete(ConsumerCoordinator.java:264)
> at
> org.apache.kafka.clients.consumer.internals.AbstractCoordinator.joinGroupIfNeeded(AbstractCoordinator.java:367)
> at
> org.apache.kafka.clients.consumer.internals.AbstractCoordinator.ensureActiveGroup(AbstractCoordinator.java:316)
> at
> org.apache.kafka.clients.consumer.internals.ConsumerCoordinator.poll(ConsumerCoordinator.java:295)
> at
> org.apache.kafka.clients.consumer.KafkaConsumer.pollOnce(KafkaConsumer.java:1146)
> at
> org.apache.kafka.clients.consumer.KafkaConsumer.poll(KafkaConsumer.java:)
> at
> org.apache.kafka.streams.processor.internals.StreamThread.pollRequests(StreamThread.java:851)
> at
> org.apache.kafka.streams.processor.internals.StreamThread.runOnce(StreamThread.java:808)
> at
> org.apache.kafka.streams.processor.internals.StreamThread.runLoop(StreamThread.java:774)
> at
> org.apache.kafka.streams.processor.internals.StreamThread.run(StreamThread.java:744){code}
>
> I tried restarting application once but the situation repeated. Thread read
> some data, committed offset and stopped
[jira] [Commented] (KAFKA-7088) Kafka streams thread waits infinitely on transaction init
[
https://issues.apache.org/jira/browse/KAFKA-7088?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16520094#comment-16520094
]
Ted Yu commented on KAFKA-7088:
---
What's the value for "processing.guarantee" ?
{code}
this.eosEnabled =
StreamsConfig.EXACTLY_ONCE.equals(config.getString(StreamsConfig.PROCESSING_GUARANTEE_CONFIG));
{code}
The call to KafkaProducer.initTransactions is governed by eosEnabled being true.
> Kafka streams thread waits infinitely on transaction init
> -
>
> Key: KAFKA-7088
> URL: https://issues.apache.org/jira/browse/KAFKA-7088
> Project: Kafka
> Issue Type: Bug
> Components: clients
>Affects Versions: 1.0.1
> Environment: Linux 4.14.33-51.37.amzn1.x86_64 #1 SMP Thu May 3
> 20:07:43 UTC 2018
> kafka-streams (client) 1.0.1
> kafka broker 1.1.0
> Java version:
> OpenJDK Runtime Environment (build 1.8.0_171-b10)
> OpenJDK 64-Bit Server VM (build 25.171-b10, mixed mode)
> kakfa config overrides:
> num.stream.threads: 6
> session.timeout.ms: 1
> request.timeout.ms: 11000
> fetch.max.wait.ms: 500
> max.poll.records: 1000
> topic has 24 partitions
>Reporter: Lukasz Gluchowski
>Priority: Major
> Labels: eos
>
> A kafka stream application thread stops processing without any feedback. The
> topic has 24 partitions and I noticed that processing stopped only for some
> partitions. I will describe what happened to partition:10. The application is
> still running (now for about 8 hours) and that thread is hanging there and no
> rebalancing that took place.
> There is no error (we have a custom `Thread.UncaughtExceptionHandler` which
> was not called). I noticed that after couple of minutes stream stopped
> processing (at offset 32606948 where log-end-offset is 33472402).
> Broker itself is not reporting any active consumer in that consumer group and
> the only info I was able to gather was from thread dump:
> {code:java}
> "mp_ads_publisher_pro_madstorage-web-corotos-prod-9db804ae-2a7a-431f-be09-392ab38cd8a2-StreamThread-33"
> #113 prio=5 os_prio=0 tid=0x7fe07c6349b0 nid=0xf7a waiting on condition
> [0x7fe0215d4000]
> java.lang.Thread.State: WAITING (parking)
> at sun.misc.Unsafe.park(Native Method)
> - parking to wait for <0xfec6a2f8> (a
> java.util.concurrent.CountDownLatch$Sync)
> at java.util.concurrent.locks.LockSupport.park(LockSupport.java:175)
> at
> java.util.concurrent.locks.AbstractQueuedSynchronizer.parkAndCheckInterrupt(AbstractQueuedSynchronizer.java:836)
> at
> java.util.concurrent.locks.AbstractQueuedSynchronizer.doAcquireSharedInterruptibly(AbstractQueuedSynchronizer.java:997)
> at
> java.util.concurrent.locks.AbstractQueuedSynchronizer.acquireSharedInterruptibly(AbstractQueuedSynchronizer.java:1304)
> at java.util.concurrent.CountDownLatch.await(CountDownLatch.java:231)
> at
> org.apache.kafka.clients.producer.internals.TransactionalRequestResult.await(TransactionalRequestResult.java:50)
> at
> org.apache.kafka.clients.producer.KafkaProducer.initTransactions(KafkaProducer.java:554)
> at
> org.apache.kafka.streams.processor.internals.StreamTask.(StreamTask.java:151)
> at
> org.apache.kafka.streams.processor.internals.StreamThread$TaskCreator.createTask(StreamThread.java:404)
> at
> org.apache.kafka.streams.processor.internals.StreamThread$TaskCreator.createTask(StreamThread.java:365)
> at
> org.apache.kafka.streams.processor.internals.StreamThread$AbstractTaskCreator.createTasks(StreamThread.java:350)
> at
> org.apache.kafka.streams.processor.internals.TaskManager.addStreamTasks(TaskManager.java:137)
> at
> org.apache.kafka.streams.processor.internals.TaskManager.createTasks(TaskManager.java:88)
> at
> org.apache.kafka.streams.processor.internals.StreamThread$RebalanceListener.onPartitionsAssigned(StreamThread.java:259)
> at
> org.apache.kafka.clients.consumer.internals.ConsumerCoordinator.onJoinComplete(ConsumerCoordinator.java:264)
> at
> org.apache.kafka.clients.consumer.internals.AbstractCoordinator.joinGroupIfNeeded(AbstractCoordinator.java:367)
> at
> org.apache.kafka.clients.consumer.internals.AbstractCoordinator.ensureActiveGroup(AbstractCoordinator.java:316)
> at
> org.apache.kafka.clients.consumer.internals.ConsumerCoordinator.poll(ConsumerCoordinator.java:295)
> at
> org.apache.kafka.clients.consumer.KafkaConsumer.pollOnce(KafkaConsumer.java:1146)
> at
> org.apache.kafka.clients.consumer.KafkaConsumer.poll(KafkaConsumer.java:)
> at
> org.apache.kafka.streams.processor.internals.StreamThread.pollRequests(StreamThread.java:851)
> at
> org.apache.kafka.streams.processor.internals.StreamThread.runOnce(StreamThread.java:808)
> at
> org.apache.kafka.streams.processor.internals.StreamThread.runLoop(StreamThread.java:774)
> at
>
[jira] [Commented] (KAFKA-7088) Kafka streams thread waits infinitely on transaction init
[
https://issues.apache.org/jira/browse/KAFKA-7088?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16520069#comment-16520069
]
Lukasz Gluchowski commented on KAFKA-7088:
--
[~guozhang] thank you for your response. The producer is created automatically
by kafka streams (that's why I categorized it as stream issue).
I didn't override {{ProducerConfig.MAX_BLOCK_MS_CONFIG}}, property so we rely
on default value (FYI I put all config properties that we explicitly set in the
"Environment" section of this ticket).
Broker itself is healthy and we didn't experienced any network issues when this
error happened, we are running multiple kafka streams applications and vast
majority of the clients have zero lag. If this helps logs from the broker said
that the consumer group is empty.
Note that this situation is repeatable. I restarted the application and the
same situation occurred after ~15 minutes.
{code:java}
kgroups --describe --group mp_ads_publisher_pro_madstorage-web-corotos-prod
Note: This will not show information about old Zookeeper-based consumers.
Consumer group 'mp_ads_publisher_pro_madstorage-web-corotos-prod' has no active
members.
TOPIC PARTITION CURRENT-OFFSET LOG-END-OFFSET LAG CONSUMER-ID HOST CLIENT-ID
madstorage-web-corotos-prod 10 32606948 33472402 865454 - - -
. (other 23 partitions){code}
Let me know if you need more info,
Thanks
> Kafka streams thread waits infinitely on transaction init
> -
>
> Key: KAFKA-7088
> URL: https://issues.apache.org/jira/browse/KAFKA-7088
> Project: Kafka
> Issue Type: Bug
> Components: clients
>Affects Versions: 1.0.1
> Environment: Linux 4.14.33-51.37.amzn1.x86_64 #1 SMP Thu May 3
> 20:07:43 UTC 2018
> kafka-streams (client) 1.0.1
> kafka broker 1.1.0
> Java version:
> OpenJDK Runtime Environment (build 1.8.0_171-b10)
> OpenJDK 64-Bit Server VM (build 25.171-b10, mixed mode)
> kakfa config overrides:
> num.stream.threads: 6
> session.timeout.ms: 1
> request.timeout.ms: 11000
> fetch.max.wait.ms: 500
> max.poll.records: 1000
> topic has 24 partitions
>Reporter: Lukasz Gluchowski
>Priority: Major
> Labels: eos
>
> A kafka stream application thread stops processing without any feedback. The
> topic has 24 partitions and I noticed that processing stopped only for some
> partitions. I will describe what happened to partition:10. The application is
> still running (now for about 8 hours) and that thread is hanging there and no
> rebalancing that took place.
> There is no error (we have a custom `Thread.UncaughtExceptionHandler` which
> was not called). I noticed that after couple of minutes stream stopped
> processing (at offset 32606948 where log-end-offset is 33472402).
> Broker itself is not reporting any active consumer in that consumer group and
> the only info I was able to gather was from thread dump:
> {code:java}
> "mp_ads_publisher_pro_madstorage-web-corotos-prod-9db804ae-2a7a-431f-be09-392ab38cd8a2-StreamThread-33"
> #113 prio=5 os_prio=0 tid=0x7fe07c6349b0 nid=0xf7a waiting on condition
> [0x7fe0215d4000]
> java.lang.Thread.State: WAITING (parking)
> at sun.misc.Unsafe.park(Native Method)
> - parking to wait for <0xfec6a2f8> (a
> java.util.concurrent.CountDownLatch$Sync)
> at java.util.concurrent.locks.LockSupport.park(LockSupport.java:175)
> at
> java.util.concurrent.locks.AbstractQueuedSynchronizer.parkAndCheckInterrupt(AbstractQueuedSynchronizer.java:836)
> at
> java.util.concurrent.locks.AbstractQueuedSynchronizer.doAcquireSharedInterruptibly(AbstractQueuedSynchronizer.java:997)
> at
> java.util.concurrent.locks.AbstractQueuedSynchronizer.acquireSharedInterruptibly(AbstractQueuedSynchronizer.java:1304)
> at java.util.concurrent.CountDownLatch.await(CountDownLatch.java:231)
> at
> org.apache.kafka.clients.producer.internals.TransactionalRequestResult.await(TransactionalRequestResult.java:50)
> at
> org.apache.kafka.clients.producer.KafkaProducer.initTransactions(KafkaProducer.java:554)
> at
> org.apache.kafka.streams.processor.internals.StreamTask.(StreamTask.java:151)
> at
> org.apache.kafka.streams.processor.internals.StreamThread$TaskCreator.createTask(StreamThread.java:404)
> at
> org.apache.kafka.streams.processor.internals.StreamThread$TaskCreator.createTask(StreamThread.java:365)
> at
> org.apache.kafka.streams.processor.internals.StreamThread$AbstractTaskCreator.createTasks(StreamThread.java:350)
> at
> org.apache.kafka.streams.processor.internals.TaskManager.addStreamTasks(TaskManager.java:137)
> at
> org.apache.kafka.streams.processor.internals.TaskManager.createTasks(TaskManager.java:88)
> at
> org.apache.kafka.streams.processor.internals.StreamThread$RebalanceListener.onPartitionsAssigned(StreamThread.java:259)
> at
>
[jira] [Commented] (KAFKA-7088) Kafka streams thread waits infinitely on transaction init
[
https://issues.apache.org/jira/browse/KAFKA-7088?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16519600#comment-16519600
]
Guozhang Wang commented on KAFKA-7088:
--
[~lgluchowski] The blocking code inside producer is:
{code}
if (initTransactionsResult.await(maxBlockTimeMs,
TimeUnit.MILLISECONDS)) {
initTransactionsResult = null;
}
{code}
The {{maxBlockTimeMs}} is defined as {{ProducerConfig.MAX_BLOCK_MS_CONFIG}}, by
default it is 60 seconds. Did you override this config? Also note that under
normal network condition and broker health, producer should wait much less than
60 seconds to get the result from broker. Did you observe any broker
unavailability or network issues during then?
> Kafka streams thread waits infinitely on transaction init
> -
>
> Key: KAFKA-7088
> URL: https://issues.apache.org/jira/browse/KAFKA-7088
> Project: Kafka
> Issue Type: Bug
> Components: clients
>Affects Versions: 1.0.1
> Environment: Linux 4.14.33-51.37.amzn1.x86_64 #1 SMP Thu May 3
> 20:07:43 UTC 2018
> kafka-streams (client) 1.0.1
> kafka broker 1.1.0
> Java version:
> OpenJDK Runtime Environment (build 1.8.0_171-b10)
> OpenJDK 64-Bit Server VM (build 25.171-b10, mixed mode)
> kakfa config overrides:
> num.stream.threads: 6
> session.timeout.ms: 1
> request.timeout.ms: 11000
> fetch.max.wait.ms: 500
> max.poll.records: 1000
> topic has 24 partitions
>Reporter: Lukasz Gluchowski
>Priority: Major
> Labels: eos
>
> A kafka stream application thread stops processing without any feedback. The
> topic has 24 partitions and I noticed that processing stopped only for some
> partitions. I will describe what happened to partition:10. The application is
> still running (now for about 8 hours) and that thread is hanging there and no
> rebalancing that took place.
> There is no error (we have a custom `Thread.UncaughtExceptionHandler` which
> was not called). I noticed that after couple of minutes stream stopped
> processing (at offset 32606948 where log-end-offset is 33472402).
> Broker itself is not reporting any active consumer in that consumer group and
> the only info I was able to gather was from thread dump:
> {code:java}
> "mp_ads_publisher_pro_madstorage-web-corotos-prod-9db804ae-2a7a-431f-be09-392ab38cd8a2-StreamThread-33"
> #113 prio=5 os_prio=0 tid=0x7fe07c6349b0 nid=0xf7a waiting on condition
> [0x7fe0215d4000]
> java.lang.Thread.State: WAITING (parking)
> at sun.misc.Unsafe.park(Native Method)
> - parking to wait for <0xfec6a2f8> (a
> java.util.concurrent.CountDownLatch$Sync)
> at java.util.concurrent.locks.LockSupport.park(LockSupport.java:175)
> at
> java.util.concurrent.locks.AbstractQueuedSynchronizer.parkAndCheckInterrupt(AbstractQueuedSynchronizer.java:836)
> at
> java.util.concurrent.locks.AbstractQueuedSynchronizer.doAcquireSharedInterruptibly(AbstractQueuedSynchronizer.java:997)
> at
> java.util.concurrent.locks.AbstractQueuedSynchronizer.acquireSharedInterruptibly(AbstractQueuedSynchronizer.java:1304)
> at java.util.concurrent.CountDownLatch.await(CountDownLatch.java:231)
> at
> org.apache.kafka.clients.producer.internals.TransactionalRequestResult.await(TransactionalRequestResult.java:50)
> at
> org.apache.kafka.clients.producer.KafkaProducer.initTransactions(KafkaProducer.java:554)
> at
> org.apache.kafka.streams.processor.internals.StreamTask.(StreamTask.java:151)
> at
> org.apache.kafka.streams.processor.internals.StreamThread$TaskCreator.createTask(StreamThread.java:404)
> at
> org.apache.kafka.streams.processor.internals.StreamThread$TaskCreator.createTask(StreamThread.java:365)
> at
> org.apache.kafka.streams.processor.internals.StreamThread$AbstractTaskCreator.createTasks(StreamThread.java:350)
> at
> org.apache.kafka.streams.processor.internals.TaskManager.addStreamTasks(TaskManager.java:137)
> at
> org.apache.kafka.streams.processor.internals.TaskManager.createTasks(TaskManager.java:88)
> at
> org.apache.kafka.streams.processor.internals.StreamThread$RebalanceListener.onPartitionsAssigned(StreamThread.java:259)
> at
> org.apache.kafka.clients.consumer.internals.ConsumerCoordinator.onJoinComplete(ConsumerCoordinator.java:264)
> at
> org.apache.kafka.clients.consumer.internals.AbstractCoordinator.joinGroupIfNeeded(AbstractCoordinator.java:367)
> at
> org.apache.kafka.clients.consumer.internals.AbstractCoordinator.ensureActiveGroup(AbstractCoordinator.java:316)
> at
> org.apache.kafka.clients.consumer.internals.ConsumerCoordinator.poll(ConsumerCoordinator.java:295)
> at
> org.apache.kafka.clients.consumer.KafkaConsumer.pollOnce(KafkaConsumer.java:1146)
> at
> org.apache.kafka.clients.consumer.KafkaConsumer.poll(KafkaConsumer.java:)
> at
>
[jira] [Commented] (KAFKA-7088) Kafka streams thread waits infinitely on transaction init
[
https://issues.apache.org/jira/browse/KAFKA-7088?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16519572#comment-16519572
]
Matthias J. Sax commented on KAFKA-7088:
According to the stack trace it hangs in the producer. Thus, it seems to be
client issues, not a Kafka Streams issues. \cc [~hachikuji]
> Kafka streams thread waits infinitely on transaction init
> -
>
> Key: KAFKA-7088
> URL: https://issues.apache.org/jira/browse/KAFKA-7088
> Project: Kafka
> Issue Type: Bug
> Components: clients
>Affects Versions: 1.0.1
> Environment: Linux 4.14.33-51.37.amzn1.x86_64 #1 SMP Thu May 3
> 20:07:43 UTC 2018
> kafka-streams (client) 1.0.1
> kafka broker 1.1.0
> Java version:
> OpenJDK Runtime Environment (build 1.8.0_171-b10)
> OpenJDK 64-Bit Server VM (build 25.171-b10, mixed mode)
> kakfa config overrides:
> num.stream.threads: 6
> session.timeout.ms: 1
> request.timeout.ms: 11000
> fetch.max.wait.ms: 500
> max.poll.records: 1000
> topic has 24 partitions
>Reporter: Lukasz Gluchowski
>Priority: Major
>
> A kafka stream application thread stops processing without any feedback. The
> topic has 24 partitions and I noticed that processing stopped only for some
> partitions. I will describe what happened to partition:10. The application is
> still running (now for about 8 hours) and that thread is hanging there and no
> rebalancing that took place.
> There is no error (we have a custom `Thread.UncaughtExceptionHandler` which
> was not called). I noticed that after couple of minutes stream stopped
> processing (at offset 32606948 where log-end-offset is 33472402).
> Broker itself is not reporting any active consumer in that consumer group and
> the only info I was able to gather was from thread dump:
> {code:java}
> "mp_ads_publisher_pro_madstorage-web-corotos-prod-9db804ae-2a7a-431f-be09-392ab38cd8a2-StreamThread-33"
> #113 prio=5 os_prio=0 tid=0x7fe07c6349b0 nid=0xf7a waiting on condition
> [0x7fe0215d4000]
> java.lang.Thread.State: WAITING (parking)
> at sun.misc.Unsafe.park(Native Method)
> - parking to wait for <0xfec6a2f8> (a
> java.util.concurrent.CountDownLatch$Sync)
> at java.util.concurrent.locks.LockSupport.park(LockSupport.java:175)
> at
> java.util.concurrent.locks.AbstractQueuedSynchronizer.parkAndCheckInterrupt(AbstractQueuedSynchronizer.java:836)
> at
> java.util.concurrent.locks.AbstractQueuedSynchronizer.doAcquireSharedInterruptibly(AbstractQueuedSynchronizer.java:997)
> at
> java.util.concurrent.locks.AbstractQueuedSynchronizer.acquireSharedInterruptibly(AbstractQueuedSynchronizer.java:1304)
> at java.util.concurrent.CountDownLatch.await(CountDownLatch.java:231)
> at
> org.apache.kafka.clients.producer.internals.TransactionalRequestResult.await(TransactionalRequestResult.java:50)
> at
> org.apache.kafka.clients.producer.KafkaProducer.initTransactions(KafkaProducer.java:554)
> at
> org.apache.kafka.streams.processor.internals.StreamTask.(StreamTask.java:151)
> at
> org.apache.kafka.streams.processor.internals.StreamThread$TaskCreator.createTask(StreamThread.java:404)
> at
> org.apache.kafka.streams.processor.internals.StreamThread$TaskCreator.createTask(StreamThread.java:365)
> at
> org.apache.kafka.streams.processor.internals.StreamThread$AbstractTaskCreator.createTasks(StreamThread.java:350)
> at
> org.apache.kafka.streams.processor.internals.TaskManager.addStreamTasks(TaskManager.java:137)
> at
> org.apache.kafka.streams.processor.internals.TaskManager.createTasks(TaskManager.java:88)
> at
> org.apache.kafka.streams.processor.internals.StreamThread$RebalanceListener.onPartitionsAssigned(StreamThread.java:259)
> at
> org.apache.kafka.clients.consumer.internals.ConsumerCoordinator.onJoinComplete(ConsumerCoordinator.java:264)
> at
> org.apache.kafka.clients.consumer.internals.AbstractCoordinator.joinGroupIfNeeded(AbstractCoordinator.java:367)
> at
> org.apache.kafka.clients.consumer.internals.AbstractCoordinator.ensureActiveGroup(AbstractCoordinator.java:316)
> at
> org.apache.kafka.clients.consumer.internals.ConsumerCoordinator.poll(ConsumerCoordinator.java:295)
> at
> org.apache.kafka.clients.consumer.KafkaConsumer.pollOnce(KafkaConsumer.java:1146)
> at
> org.apache.kafka.clients.consumer.KafkaConsumer.poll(KafkaConsumer.java:)
> at
> org.apache.kafka.streams.processor.internals.StreamThread.pollRequests(StreamThread.java:851)
> at
> org.apache.kafka.streams.processor.internals.StreamThread.runOnce(StreamThread.java:808)
> at
> org.apache.kafka.streams.processor.internals.StreamThread.runLoop(StreamThread.java:774)
> at
> org.apache.kafka.streams.processor.internals.StreamThread.run(StreamThread.java:744){code}
>
> I tried restarting application once but the situation repeated. Thread read
> some data,
[jira] [Commented] (KAFKA-7088) Kafka streams thread waits infinitely on transaction init
[
https://issues.apache.org/jira/browse/KAFKA-7088?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16519542#comment-16519542
]
Ted Yu commented on KAFKA-7088:
---
Is it possible to reproduce this behavior using 1.1.1 RC ?
There have been some changes in related code since 1.0.1
Thanks
> Kafka streams thread waits infinitely on transaction init
> -
>
> Key: KAFKA-7088
> URL: https://issues.apache.org/jira/browse/KAFKA-7088
> Project: Kafka
> Issue Type: Bug
> Components: streams
>Affects Versions: 1.0.1
> Environment: Linux 4.14.33-51.37.amzn1.x86_64 #1 SMP Thu May 3
> 20:07:43 UTC 2018
> kafka-streams (client) 1.0.1
> kafka broker 1.1.0
> Java version:
> OpenJDK Runtime Environment (build 1.8.0_171-b10)
> OpenJDK 64-Bit Server VM (build 25.171-b10, mixed mode)
> kakfa config overrides:
> num.stream.threads: 6
> session.timeout.ms: 1
> request.timeout.ms: 11000
> fetch.max.wait.ms: 500
> max.poll.records: 1000
> topic has 24 partitions
>Reporter: Lukasz Gluchowski
>Priority: Major
>
> A kafka stream application thread stops processing without any feedback. The
> topic has 24 partitions and I noticed that processing stopped only for some
> partitions. I will describe what happened to partition:10. The application is
> still running (now for about 8 hours) and that thread is hanging there and no
> rebalancing that took place.
> There is no error (we have a custom `Thread.UncaughtExceptionHandler` which
> was not called). I noticed that after couple of minutes stream stopped
> processing (at offset 32606948 where log-end-offset is 33472402).
> Broker itself is not reporting any active consumer in that consumer group and
> the only info I was able to gather was from thread dump:
> {code:java}
> "mp_ads_publisher_pro_madstorage-web-corotos-prod-9db804ae-2a7a-431f-be09-392ab38cd8a2-StreamThread-33"
> #113 prio=5 os_prio=0 tid=0x7fe07c6349b0 nid=0xf7a waiting on condition
> [0x7fe0215d4000]
> java.lang.Thread.State: WAITING (parking)
> at sun.misc.Unsafe.park(Native Method)
> - parking to wait for <0xfec6a2f8> (a
> java.util.concurrent.CountDownLatch$Sync)
> at java.util.concurrent.locks.LockSupport.park(LockSupport.java:175)
> at
> java.util.concurrent.locks.AbstractQueuedSynchronizer.parkAndCheckInterrupt(AbstractQueuedSynchronizer.java:836)
> at
> java.util.concurrent.locks.AbstractQueuedSynchronizer.doAcquireSharedInterruptibly(AbstractQueuedSynchronizer.java:997)
> at
> java.util.concurrent.locks.AbstractQueuedSynchronizer.acquireSharedInterruptibly(AbstractQueuedSynchronizer.java:1304)
> at java.util.concurrent.CountDownLatch.await(CountDownLatch.java:231)
> at
> org.apache.kafka.clients.producer.internals.TransactionalRequestResult.await(TransactionalRequestResult.java:50)
> at
> org.apache.kafka.clients.producer.KafkaProducer.initTransactions(KafkaProducer.java:554)
> at
> org.apache.kafka.streams.processor.internals.StreamTask.(StreamTask.java:151)
> at
> org.apache.kafka.streams.processor.internals.StreamThread$TaskCreator.createTask(StreamThread.java:404)
> at
> org.apache.kafka.streams.processor.internals.StreamThread$TaskCreator.createTask(StreamThread.java:365)
> at
> org.apache.kafka.streams.processor.internals.StreamThread$AbstractTaskCreator.createTasks(StreamThread.java:350)
> at
> org.apache.kafka.streams.processor.internals.TaskManager.addStreamTasks(TaskManager.java:137)
> at
> org.apache.kafka.streams.processor.internals.TaskManager.createTasks(TaskManager.java:88)
> at
> org.apache.kafka.streams.processor.internals.StreamThread$RebalanceListener.onPartitionsAssigned(StreamThread.java:259)
> at
> org.apache.kafka.clients.consumer.internals.ConsumerCoordinator.onJoinComplete(ConsumerCoordinator.java:264)
> at
> org.apache.kafka.clients.consumer.internals.AbstractCoordinator.joinGroupIfNeeded(AbstractCoordinator.java:367)
> at
> org.apache.kafka.clients.consumer.internals.AbstractCoordinator.ensureActiveGroup(AbstractCoordinator.java:316)
> at
> org.apache.kafka.clients.consumer.internals.ConsumerCoordinator.poll(ConsumerCoordinator.java:295)
> at
> org.apache.kafka.clients.consumer.KafkaConsumer.pollOnce(KafkaConsumer.java:1146)
> at
> org.apache.kafka.clients.consumer.KafkaConsumer.poll(KafkaConsumer.java:)
> at
> org.apache.kafka.streams.processor.internals.StreamThread.pollRequests(StreamThread.java:851)
> at
> org.apache.kafka.streams.processor.internals.StreamThread.runOnce(StreamThread.java:808)
> at
> org.apache.kafka.streams.processor.internals.StreamThread.runLoop(StreamThread.java:774)
> at
> org.apache.kafka.streams.processor.internals.StreamThread.run(StreamThread.java:744){code}
>
> I tried restarting application once but the situation repeated. Thread read
> some data, committed offset and stopped
