[jira] [Commented] (KAFKA-6003) Replication Fetcher thread for a partition with no data fails to start
[
https://issues.apache.org/jira/browse/KAFKA-6003?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16188815#comment-16188815
]
Apurva Mehta commented on KAFKA-6003:
-
hi [~schizhov], thanks for filing the JIRA and providing those details.
I think one important detail needs to be clarified: My understanding is that
this replica fetcher thread stopped due to an unrelated error, and when it was
restarted, it would fail again with the {{OutOfOrderSequenceException}}. Is
this correct?
Thanks,
Apurva
> Replication Fetcher thread for a partition with no data fails to start
> --
>
> Key: KAFKA-6003
> URL: https://issues.apache.org/jira/browse/KAFKA-6003
> Project: Kafka
> Issue Type: Bug
> Components: replication
>Affects Versions: 0.11.0.1
>Reporter: Stanislav Chizhov
>Priority: Blocker
> Fix For: 1.0.0, 0.11.0.2
>
>
> If a partition of a topic with idempotent producer has no data on 1 of the
> brokers, but it does exist on others and some of the segments for this
> partition have been already deleted replication thread responsible for this
> partition on the broker which has no data for it fails to start with out of
> order sequence exception:
> {code}
> [2017-10-02 09:44:23,825] ERROR [ReplicaFetcherThread-2-4]: Error due to
> (kafka.server.ReplicaFetcherThread)
> kafka.common.KafkaException: error processing data for partition
> [stage.data.adevents.v2,20] offset 1660336429
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1$$anonfun$apply$2.apply(AbstractFetcherThread.scala:203)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1$$anonfun$apply$2.apply(AbstractFetcherThread.scala:174)
> at scala.Option.foreach(Option.scala:257)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1.apply(AbstractFetcherThread.scala:174)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1.apply(AbstractFetcherThread.scala:171)
> at
> scala.collection.mutable.ResizableArray$class.foreach(ResizableArray.scala:59)
> at scala.collection.mutable.ArrayBuffer.foreach(ArrayBuffer.scala:48)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply$mcV$sp(AbstractFetcherThread.scala:171)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply(AbstractFetcherThread.scala:171)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply(AbstractFetcherThread.scala:171)
> at kafka.utils.CoreUtils$.inLock(CoreUtils.scala:213)
> at
> kafka.server.AbstractFetcherThread.processFetchRequest(AbstractFetcherThread.scala:169)
> at
> kafka.server.AbstractFetcherThread.doWork(AbstractFetcherThread.scala:112)
> at kafka.utils.ShutdownableThread.run(ShutdownableThread.scala:64)
> Caused by: org.apache.kafka.common.errors.OutOfOrderSequenceException:
> Invalid sequence number for new epoch: 0 (request epoch), 154277489 (seq.
> number)
> {code}
> We run kafka 0.11.0.1 and we ran into the situation when 1 of replication
> threads was stopped for few days, while everything else on that broker was
> functional. This is our staging cluster and retention is less than a day, so
> everything for partitions for which replication thread was down was cleaned
> up. At the moment we have a broker which cannot start replication for few
> partitions. I was also able to reproduce in my local test environment.
> Another possible use case when this might cause real pain is disk failure or
> any situation when previously deleting all the data for the partition on a
> broker helped - since it would just fetch all the data from other replicas.
> Now it does not work for topics with idempotent producers. It might also
> affect other not-idempotent topics if those are unlucky to share same
> replication fetcher thread.
> This seems to be caused by this logic:
> https://github.com/apache/kafka/blob/0.11.0.1/core/src/main/scala/kafka/log/ProducerStateManager.scala#L119
> and might be fixed in the scope of
> https://issues.apache.org/jira/browse/KAFKA-5793.
> However any hints on how to get those partition to fully replicated state are
> highly appreciated.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
[jira] [Commented] (KAFKA-6003) Replication Fetcher thread for a partition with no data fails to start
[
https://issues.apache.org/jira/browse/KAFKA-6003?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16188863#comment-16188863
]
Stanislav Chizhov commented on KAFKA-6003:
--
Well in the case we had in staging it was out of order sequence exception for
some other topic/partition. I still need to investigate why exactly did that
happen. It does not seem to matter though how you get to that state with no
data for the partition on one of the brokers.
> Replication Fetcher thread for a partition with no data fails to start
> --
>
> Key: KAFKA-6003
> URL: https://issues.apache.org/jira/browse/KAFKA-6003
> Project: Kafka
> Issue Type: Bug
> Components: replication
>Affects Versions: 0.11.0.1
>Reporter: Stanislav Chizhov
>Priority: Blocker
> Fix For: 1.0.0, 0.11.0.2
>
>
> If a partition of a topic with idempotent producer has no data on 1 of the
> brokers, but it does exist on others and some of the segments for this
> partition have been already deleted replication thread responsible for this
> partition on the broker which has no data for it fails to start with out of
> order sequence exception:
> {code}
> [2017-10-02 09:44:23,825] ERROR [ReplicaFetcherThread-2-4]: Error due to
> (kafka.server.ReplicaFetcherThread)
> kafka.common.KafkaException: error processing data for partition
> [stage.data.adevents.v2,20] offset 1660336429
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1$$anonfun$apply$2.apply(AbstractFetcherThread.scala:203)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1$$anonfun$apply$2.apply(AbstractFetcherThread.scala:174)
> at scala.Option.foreach(Option.scala:257)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1.apply(AbstractFetcherThread.scala:174)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1.apply(AbstractFetcherThread.scala:171)
> at
> scala.collection.mutable.ResizableArray$class.foreach(ResizableArray.scala:59)
> at scala.collection.mutable.ArrayBuffer.foreach(ArrayBuffer.scala:48)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply$mcV$sp(AbstractFetcherThread.scala:171)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply(AbstractFetcherThread.scala:171)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply(AbstractFetcherThread.scala:171)
> at kafka.utils.CoreUtils$.inLock(CoreUtils.scala:213)
> at
> kafka.server.AbstractFetcherThread.processFetchRequest(AbstractFetcherThread.scala:169)
> at
> kafka.server.AbstractFetcherThread.doWork(AbstractFetcherThread.scala:112)
> at kafka.utils.ShutdownableThread.run(ShutdownableThread.scala:64)
> Caused by: org.apache.kafka.common.errors.OutOfOrderSequenceException:
> Invalid sequence number for new epoch: 0 (request epoch), 154277489 (seq.
> number)
> {code}
> We run kafka 0.11.0.1 and we ran into the situation when 1 of replication
> threads was stopped for few days, while everything else on that broker was
> functional. This is our staging cluster and retention is less than a day, so
> everything for partitions for which replication thread was down was cleaned
> up. At the moment we have a broker which cannot start replication for few
> partitions. I was also able to reproduce in my local test environment.
> Another possible use case when this might cause real pain is disk failure or
> any situation when previously deleting all the data for the partition on a
> broker helped - since it would just fetch all the data from other replicas.
> Now it does not work for topics with idempotent producers. It might also
> affect other not-idempotent topics if those are unlucky to share same
> replication fetcher thread.
> This seems to be caused by this logic:
> https://github.com/apache/kafka/blob/0.11.0.1/core/src/main/scala/kafka/log/ProducerStateManager.scala#L119
> and might be fixed in the scope of
> https://issues.apache.org/jira/browse/KAFKA-5793.
> However any hints on how to get those partition to fully replicated state are
> highly appreciated.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
[jira] [Commented] (KAFKA-6003) Replication Fetcher thread for a partition with no data fails to start
[
https://issues.apache.org/jira/browse/KAFKA-6003?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16188909#comment-16188909
]
Apurva Mehta commented on KAFKA-6003:
-
Thanks [~schizhov]. this is indeed a gap in the current logic, on both 1.0.0
and 0.11.0.1. We should be able to handle the loss of producer state on the
replica by simply accepting the leader as the source of truth.
As for a workaround, the only thing I can think of at this point is to rebuild
the replica from scratch :(
> Replication Fetcher thread for a partition with no data fails to start
> --
>
> Key: KAFKA-6003
> URL: https://issues.apache.org/jira/browse/KAFKA-6003
> Project: Kafka
> Issue Type: Bug
> Components: replication
>Affects Versions: 0.11.0.1
>Reporter: Stanislav Chizhov
>Assignee: Apurva Mehta
>Priority: Blocker
> Fix For: 1.0.0, 0.11.0.2
>
>
> If a partition of a topic with idempotent producer has no data on 1 of the
> brokers, but it does exist on others and some of the segments for this
> partition have been already deleted replication thread responsible for this
> partition on the broker which has no data for it fails to start with out of
> order sequence exception:
> {code}
> [2017-10-02 09:44:23,825] ERROR [ReplicaFetcherThread-2-4]: Error due to
> (kafka.server.ReplicaFetcherThread)
> kafka.common.KafkaException: error processing data for partition
> [stage.data.adevents.v2,20] offset 1660336429
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1$$anonfun$apply$2.apply(AbstractFetcherThread.scala:203)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1$$anonfun$apply$2.apply(AbstractFetcherThread.scala:174)
> at scala.Option.foreach(Option.scala:257)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1.apply(AbstractFetcherThread.scala:174)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1.apply(AbstractFetcherThread.scala:171)
> at
> scala.collection.mutable.ResizableArray$class.foreach(ResizableArray.scala:59)
> at scala.collection.mutable.ArrayBuffer.foreach(ArrayBuffer.scala:48)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply$mcV$sp(AbstractFetcherThread.scala:171)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply(AbstractFetcherThread.scala:171)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply(AbstractFetcherThread.scala:171)
> at kafka.utils.CoreUtils$.inLock(CoreUtils.scala:213)
> at
> kafka.server.AbstractFetcherThread.processFetchRequest(AbstractFetcherThread.scala:169)
> at
> kafka.server.AbstractFetcherThread.doWork(AbstractFetcherThread.scala:112)
> at kafka.utils.ShutdownableThread.run(ShutdownableThread.scala:64)
> Caused by: org.apache.kafka.common.errors.OutOfOrderSequenceException:
> Invalid sequence number for new epoch: 0 (request epoch), 154277489 (seq.
> number)
> {code}
> We run kafka 0.11.0.1 and we ran into the situation when 1 of replication
> threads was stopped for few days, while everything else on that broker was
> functional. This is our staging cluster and retention is less than a day, so
> everything for partitions for which replication thread was down was cleaned
> up. At the moment we have a broker which cannot start replication for few
> partitions. I was also able to reproduce in my local test environment.
> Another possible use case when this might cause real pain is disk failure or
> any situation when previously deleting all the data for the partition on a
> broker helped - since it would just fetch all the data from other replicas.
> Now it does not work for topics with idempotent producers. It might also
> affect other not-idempotent topics if those are unlucky to share same
> replication fetcher thread.
> This seems to be caused by this logic:
> https://github.com/apache/kafka/blob/0.11.0.1/core/src/main/scala/kafka/log/ProducerStateManager.scala#L119
> and might be fixed in the scope of
> https://issues.apache.org/jira/browse/KAFKA-5793.
> However any hints on how to get those partition to fully replicated state are
> highly appreciated.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
[jira] [Commented] (KAFKA-6003) Replication Fetcher thread for a partition with no data fails to start
[
https://issues.apache.org/jira/browse/KAFKA-6003?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16188924#comment-16188924
]
Stanislav Chizhov commented on KAFKA-6003:
--
Thanks [~apurva]! By rebuilding "from scratch" - do you mean dropping all the
data in that partition? What about disabling idempotence in the producer and
waiting until all "idempotent" messages are cleaned up - should that work?
> Replication Fetcher thread for a partition with no data fails to start
> --
>
> Key: KAFKA-6003
> URL: https://issues.apache.org/jira/browse/KAFKA-6003
> Project: Kafka
> Issue Type: Bug
> Components: replication
>Affects Versions: 0.11.0.1
>Reporter: Stanislav Chizhov
>Assignee: Apurva Mehta
>Priority: Blocker
> Fix For: 1.0.0, 0.11.0.2
>
>
> If a partition of a topic with idempotent producer has no data on 1 of the
> brokers, but it does exist on others and some of the segments for this
> partition have been already deleted replication thread responsible for this
> partition on the broker which has no data for it fails to start with out of
> order sequence exception:
> {code}
> [2017-10-02 09:44:23,825] ERROR [ReplicaFetcherThread-2-4]: Error due to
> (kafka.server.ReplicaFetcherThread)
> kafka.common.KafkaException: error processing data for partition
> [stage.data.adevents.v2,20] offset 1660336429
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1$$anonfun$apply$2.apply(AbstractFetcherThread.scala:203)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1$$anonfun$apply$2.apply(AbstractFetcherThread.scala:174)
> at scala.Option.foreach(Option.scala:257)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1.apply(AbstractFetcherThread.scala:174)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1.apply(AbstractFetcherThread.scala:171)
> at
> scala.collection.mutable.ResizableArray$class.foreach(ResizableArray.scala:59)
> at scala.collection.mutable.ArrayBuffer.foreach(ArrayBuffer.scala:48)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply$mcV$sp(AbstractFetcherThread.scala:171)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply(AbstractFetcherThread.scala:171)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply(AbstractFetcherThread.scala:171)
> at kafka.utils.CoreUtils$.inLock(CoreUtils.scala:213)
> at
> kafka.server.AbstractFetcherThread.processFetchRequest(AbstractFetcherThread.scala:169)
> at
> kafka.server.AbstractFetcherThread.doWork(AbstractFetcherThread.scala:112)
> at kafka.utils.ShutdownableThread.run(ShutdownableThread.scala:64)
> Caused by: org.apache.kafka.common.errors.OutOfOrderSequenceException:
> Invalid sequence number for new epoch: 0 (request epoch), 154277489 (seq.
> number)
> {code}
> We run kafka 0.11.0.1 and we ran into the situation when 1 of replication
> threads was stopped for few days, while everything else on that broker was
> functional. This is our staging cluster and retention is less than a day, so
> everything for partitions for which replication thread was down was cleaned
> up. At the moment we have a broker which cannot start replication for few
> partitions. I was also able to reproduce in my local test environment.
> Another possible use case when this might cause real pain is disk failure or
> any situation when previously deleting all the data for the partition on a
> broker helped - since it would just fetch all the data from other replicas.
> Now it does not work for topics with idempotent producers. It might also
> affect other not-idempotent topics if those are unlucky to share same
> replication fetcher thread.
> This seems to be caused by this logic:
> https://github.com/apache/kafka/blob/0.11.0.1/core/src/main/scala/kafka/log/ProducerStateManager.scala#L119
> and might be fixed in the scope of
> https://issues.apache.org/jira/browse/KAFKA-5793.
> However any hints on how to get those partition to fully replicated state are
> highly appreciated.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
[jira] [Commented] (KAFKA-6003) Replication Fetcher thread for a partition with no data fails to start
[
https://issues.apache.org/jira/browse/KAFKA-6003?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16188927#comment-16188927
]
Apurva Mehta commented on KAFKA-6003:
-
The answer to both your questions is 'yes'.
1. dropping all data on the replica and making it rereplicate will get the
state back in sync.
2. stopping idempotent writes and waiting for the producer id to be expired
will eventually allow the replica to fetch from the leader again.
> Replication Fetcher thread for a partition with no data fails to start
> --
>
> Key: KAFKA-6003
> URL: https://issues.apache.org/jira/browse/KAFKA-6003
> Project: Kafka
> Issue Type: Bug
> Components: replication
>Affects Versions: 0.11.0.1
>Reporter: Stanislav Chizhov
>Assignee: Apurva Mehta
>Priority: Blocker
> Fix For: 1.0.0, 0.11.0.2
>
>
> If a partition of a topic with idempotent producer has no data on 1 of the
> brokers, but it does exist on others and some of the segments for this
> partition have been already deleted replication thread responsible for this
> partition on the broker which has no data for it fails to start with out of
> order sequence exception:
> {code}
> [2017-10-02 09:44:23,825] ERROR [ReplicaFetcherThread-2-4]: Error due to
> (kafka.server.ReplicaFetcherThread)
> kafka.common.KafkaException: error processing data for partition
> [stage.data.adevents.v2,20] offset 1660336429
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1$$anonfun$apply$2.apply(AbstractFetcherThread.scala:203)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1$$anonfun$apply$2.apply(AbstractFetcherThread.scala:174)
> at scala.Option.foreach(Option.scala:257)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1.apply(AbstractFetcherThread.scala:174)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1.apply(AbstractFetcherThread.scala:171)
> at
> scala.collection.mutable.ResizableArray$class.foreach(ResizableArray.scala:59)
> at scala.collection.mutable.ArrayBuffer.foreach(ArrayBuffer.scala:48)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply$mcV$sp(AbstractFetcherThread.scala:171)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply(AbstractFetcherThread.scala:171)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply(AbstractFetcherThread.scala:171)
> at kafka.utils.CoreUtils$.inLock(CoreUtils.scala:213)
> at
> kafka.server.AbstractFetcherThread.processFetchRequest(AbstractFetcherThread.scala:169)
> at
> kafka.server.AbstractFetcherThread.doWork(AbstractFetcherThread.scala:112)
> at kafka.utils.ShutdownableThread.run(ShutdownableThread.scala:64)
> Caused by: org.apache.kafka.common.errors.OutOfOrderSequenceException:
> Invalid sequence number for new epoch: 0 (request epoch), 154277489 (seq.
> number)
> {code}
> We run kafka 0.11.0.1 and we ran into the situation when 1 of replication
> threads was stopped for few days, while everything else on that broker was
> functional. This is our staging cluster and retention is less than a day, so
> everything for partitions for which replication thread was down was cleaned
> up. At the moment we have a broker which cannot start replication for few
> partitions. I was also able to reproduce in my local test environment.
> Another possible use case when this might cause real pain is disk failure or
> any situation when previously deleting all the data for the partition on a
> broker helped - since it would just fetch all the data from other replicas.
> Now it does not work for topics with idempotent producers. It might also
> affect other not-idempotent topics if those are unlucky to share same
> replication fetcher thread.
> This seems to be caused by this logic:
> https://github.com/apache/kafka/blob/0.11.0.1/core/src/main/scala/kafka/log/ProducerStateManager.scala#L119
> and might be fixed in the scope of
> https://issues.apache.org/jira/browse/KAFKA-5793.
> However any hints on how to get those partition to fully replicated state are
> highly appreciated.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
[jira] [Commented] (KAFKA-6003) Replication Fetcher thread for a partition with no data fails to start
[
https://issues.apache.org/jira/browse/KAFKA-6003?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16188937#comment-16188937
]
Stanislav Chizhov commented on KAFKA-6003:
--
thank you, but that replica does not have any data already and it cant
re-replicate from the leader.
> Replication Fetcher thread for a partition with no data fails to start
> --
>
> Key: KAFKA-6003
> URL: https://issues.apache.org/jira/browse/KAFKA-6003
> Project: Kafka
> Issue Type: Bug
> Components: replication
>Affects Versions: 0.11.0.1
>Reporter: Stanislav Chizhov
>Assignee: Apurva Mehta
>Priority: Blocker
> Fix For: 1.0.0, 0.11.0.2
>
>
> If a partition of a topic with idempotent producer has no data on 1 of the
> brokers, but it does exist on others and some of the segments for this
> partition have been already deleted replication thread responsible for this
> partition on the broker which has no data for it fails to start with out of
> order sequence exception:
> {code}
> [2017-10-02 09:44:23,825] ERROR [ReplicaFetcherThread-2-4]: Error due to
> (kafka.server.ReplicaFetcherThread)
> kafka.common.KafkaException: error processing data for partition
> [stage.data.adevents.v2,20] offset 1660336429
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1$$anonfun$apply$2.apply(AbstractFetcherThread.scala:203)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1$$anonfun$apply$2.apply(AbstractFetcherThread.scala:174)
> at scala.Option.foreach(Option.scala:257)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1.apply(AbstractFetcherThread.scala:174)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1.apply(AbstractFetcherThread.scala:171)
> at
> scala.collection.mutable.ResizableArray$class.foreach(ResizableArray.scala:59)
> at scala.collection.mutable.ArrayBuffer.foreach(ArrayBuffer.scala:48)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply$mcV$sp(AbstractFetcherThread.scala:171)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply(AbstractFetcherThread.scala:171)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply(AbstractFetcherThread.scala:171)
> at kafka.utils.CoreUtils$.inLock(CoreUtils.scala:213)
> at
> kafka.server.AbstractFetcherThread.processFetchRequest(AbstractFetcherThread.scala:169)
> at
> kafka.server.AbstractFetcherThread.doWork(AbstractFetcherThread.scala:112)
> at kafka.utils.ShutdownableThread.run(ShutdownableThread.scala:64)
> Caused by: org.apache.kafka.common.errors.OutOfOrderSequenceException:
> Invalid sequence number for new epoch: 0 (request epoch), 154277489 (seq.
> number)
> {code}
> We run kafka 0.11.0.1 and we ran into the situation when 1 of replication
> threads was stopped for few days, while everything else on that broker was
> functional. This is our staging cluster and retention is less than a day, so
> everything for partitions for which replication thread was down was cleaned
> up. At the moment we have a broker which cannot start replication for few
> partitions. I was also able to reproduce in my local test environment.
> Another possible use case when this might cause real pain is disk failure or
> any situation when previously deleting all the data for the partition on a
> broker helped - since it would just fetch all the data from other replicas.
> Now it does not work for topics with idempotent producers. It might also
> affect other not-idempotent topics if those are unlucky to share same
> replication fetcher thread.
> This seems to be caused by this logic:
> https://github.com/apache/kafka/blob/0.11.0.1/core/src/main/scala/kafka/log/ProducerStateManager.scala#L119
> and might be fixed in the scope of
> https://issues.apache.org/jira/browse/KAFKA-5793.
> However any hints on how to get those partition to fully replicated state are
> highly appreciated.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
[jira] [Commented] (KAFKA-6003) Replication Fetcher thread for a partition with no data fails to start
[
https://issues.apache.org/jira/browse/KAFKA-6003?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16188959#comment-16188959
]
Apurva Mehta commented on KAFKA-6003:
-
Sorry, I meant that need to decommission the old replica and create a new
replica by rereplicating from the leader (through the reassignment tool). That
will bring the state of the new replica in sync with the leader. The patch I am
preparing will essentially make the replica accept the leader's state without
validation when it has no local state of its own.
> Replication Fetcher thread for a partition with no data fails to start
> --
>
> Key: KAFKA-6003
> URL: https://issues.apache.org/jira/browse/KAFKA-6003
> Project: Kafka
> Issue Type: Bug
> Components: replication
>Affects Versions: 0.11.0.1
>Reporter: Stanislav Chizhov
>Assignee: Apurva Mehta
>Priority: Blocker
> Fix For: 1.0.0, 0.11.0.2
>
>
> If a partition of a topic with idempotent producer has no data on 1 of the
> brokers, but it does exist on others and some of the segments for this
> partition have been already deleted replication thread responsible for this
> partition on the broker which has no data for it fails to start with out of
> order sequence exception:
> {code}
> [2017-10-02 09:44:23,825] ERROR [ReplicaFetcherThread-2-4]: Error due to
> (kafka.server.ReplicaFetcherThread)
> kafka.common.KafkaException: error processing data for partition
> [stage.data.adevents.v2,20] offset 1660336429
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1$$anonfun$apply$2.apply(AbstractFetcherThread.scala:203)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1$$anonfun$apply$2.apply(AbstractFetcherThread.scala:174)
> at scala.Option.foreach(Option.scala:257)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1.apply(AbstractFetcherThread.scala:174)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1.apply(AbstractFetcherThread.scala:171)
> at
> scala.collection.mutable.ResizableArray$class.foreach(ResizableArray.scala:59)
> at scala.collection.mutable.ArrayBuffer.foreach(ArrayBuffer.scala:48)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply$mcV$sp(AbstractFetcherThread.scala:171)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply(AbstractFetcherThread.scala:171)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply(AbstractFetcherThread.scala:171)
> at kafka.utils.CoreUtils$.inLock(CoreUtils.scala:213)
> at
> kafka.server.AbstractFetcherThread.processFetchRequest(AbstractFetcherThread.scala:169)
> at
> kafka.server.AbstractFetcherThread.doWork(AbstractFetcherThread.scala:112)
> at kafka.utils.ShutdownableThread.run(ShutdownableThread.scala:64)
> Caused by: org.apache.kafka.common.errors.OutOfOrderSequenceException:
> Invalid sequence number for new epoch: 0 (request epoch), 154277489 (seq.
> number)
> {code}
> We run kafka 0.11.0.1 and we ran into the situation when 1 of replication
> threads was stopped for few days, while everything else on that broker was
> functional. This is our staging cluster and retention is less than a day, so
> everything for partitions for which replication thread was down was cleaned
> up. At the moment we have a broker which cannot start replication for few
> partitions. I was also able to reproduce in my local test environment.
> Another possible use case when this might cause real pain is disk failure or
> any situation when previously deleting all the data for the partition on a
> broker helped - since it would just fetch all the data from other replicas.
> Now it does not work for topics with idempotent producers. It might also
> affect other not-idempotent topics if those are unlucky to share same
> replication fetcher thread.
> This seems to be caused by this logic:
> https://github.com/apache/kafka/blob/0.11.0.1/core/src/main/scala/kafka/log/ProducerStateManager.scala#L119
> and might be fixed in the scope of
> https://issues.apache.org/jira/browse/KAFKA-5793.
> However any hints on how to get those partition to fully replicated state are
> highly appreciated.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
[jira] [Commented] (KAFKA-6003) Replication Fetcher thread for a partition with no data fails to start
[
https://issues.apache.org/jira/browse/KAFKA-6003?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16188989#comment-16188989
]
ASF GitHub Bot commented on KAFKA-6003:
---
GitHub user apurvam opened a pull request:
https://github.com/apache/kafka/pull/4004
KAFKA-6003: Accept appends on replicas unconditionally when local producer
state doesn't exist
Without this patch, if the replica's log was somehow truncated before
the leader's it is possible for the replica fetcher thread to
continuously through an OutOfOrderSequenceException because the
incoming sequence would be non-zero and there is no local state.
This patch changes the behavior so that the replica state is updated to
the leader's state if there was no local state for the producer at the
time of the append.
You can merge this pull request into a Git repository by running:
$ git pull https://github.com/apurvam/kafka
KAFKA-6003-handle-unknown-producer-on-replica
Alternatively you can review and apply these changes as the patch at:
https://github.com/apache/kafka/pull/4004.patch
To close this pull request, make a commit to your master/trunk branch
with (at least) the following in the commit message:
This closes #4004
commit 341a0d2ba3ec0716f1830860869bf773f1bf8d85
Author: Apurva Mehta
Date: 2017-10-02T22:41:19Z
KAFKA-6003: Accept appends on replicas unconditionally when local
producer state doesn't exist.
Without this patch, if the replica's log was somehow truncated before
the leader's it is possible for the replica fetcher thread to
continuously through an OutOfOrderSequenceException because the
incoming sequence would be non-zero and there is no local state.
This patch changes the behavior so that the replica state is updated to
the leader's state if there was no local state for the producer at the
time of the append.
> Replication Fetcher thread for a partition with no data fails to start
> --
>
> Key: KAFKA-6003
> URL: https://issues.apache.org/jira/browse/KAFKA-6003
> Project: Kafka
> Issue Type: Bug
> Components: replication
>Affects Versions: 0.11.0.1
>Reporter: Stanislav Chizhov
>Assignee: Apurva Mehta
>Priority: Blocker
> Fix For: 1.0.0, 0.11.0.2
>
>
> If a partition of a topic with idempotent producer has no data on 1 of the
> brokers, but it does exist on others and some of the segments for this
> partition have been already deleted replication thread responsible for this
> partition on the broker which has no data for it fails to start with out of
> order sequence exception:
> {code}
> [2017-10-02 09:44:23,825] ERROR [ReplicaFetcherThread-2-4]: Error due to
> (kafka.server.ReplicaFetcherThread)
> kafka.common.KafkaException: error processing data for partition
> [stage.data.adevents.v2,20] offset 1660336429
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1$$anonfun$apply$2.apply(AbstractFetcherThread.scala:203)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1$$anonfun$apply$2.apply(AbstractFetcherThread.scala:174)
> at scala.Option.foreach(Option.scala:257)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1.apply(AbstractFetcherThread.scala:174)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1.apply(AbstractFetcherThread.scala:171)
> at
> scala.collection.mutable.ResizableArray$class.foreach(ResizableArray.scala:59)
> at scala.collection.mutable.ArrayBuffer.foreach(ArrayBuffer.scala:48)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply$mcV$sp(AbstractFetcherThread.scala:171)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply(AbstractFetcherThread.scala:171)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply(AbstractFetcherThread.scala:171)
> at kafka.utils.CoreUtils$.inLock(CoreUtils.scala:213)
> at
> kafka.server.AbstractFetcherThread.processFetchRequest(AbstractFetcherThread.scala:169)
> at
> kafka.server.AbstractFetcherThread.doWork(AbstractFetcherThread.scala:112)
> at kafka.utils.ShutdownableThread.run(ShutdownableThread.scala:64)
> Caused by: org.apache.kafka.common.errors.OutOfOrderSequenceException:
> Invalid sequence number for new epoch: 0 (request epoch), 154277489 (seq.
> number)
> {code}
> We run kafka 0.11.0.1 and we ran into the situation when 1 of replication
> threads was stopped for few days, while everything else on that broker was
[jira] [Commented] (KAFKA-6003) Replication Fetcher thread for a partition with no data fails to start
[
https://issues.apache.org/jira/browse/KAFKA-6003?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16189012#comment-16189012
]
Apurva Mehta commented on KAFKA-6003:
-
Actually, [~schizhov], my recommendation will not work. Even rebuilding a
replica from scratch will hit this same bug, because the leader's sequences
most likely will not start from zero. So the first append on the replica during
reassignment will also fail.
> Replication Fetcher thread for a partition with no data fails to start
> --
>
> Key: KAFKA-6003
> URL: https://issues.apache.org/jira/browse/KAFKA-6003
> Project: Kafka
> Issue Type: Bug
> Components: replication
>Affects Versions: 0.11.0.1
>Reporter: Stanislav Chizhov
>Assignee: Apurva Mehta
>Priority: Blocker
> Fix For: 1.0.0, 0.11.0.2
>
>
> If a partition of a topic with idempotent producer has no data on 1 of the
> brokers, but it does exist on others and some of the segments for this
> partition have been already deleted replication thread responsible for this
> partition on the broker which has no data for it fails to start with out of
> order sequence exception:
> {code}
> [2017-10-02 09:44:23,825] ERROR [ReplicaFetcherThread-2-4]: Error due to
> (kafka.server.ReplicaFetcherThread)
> kafka.common.KafkaException: error processing data for partition
> [stage.data.adevents.v2,20] offset 1660336429
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1$$anonfun$apply$2.apply(AbstractFetcherThread.scala:203)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1$$anonfun$apply$2.apply(AbstractFetcherThread.scala:174)
> at scala.Option.foreach(Option.scala:257)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1.apply(AbstractFetcherThread.scala:174)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1.apply(AbstractFetcherThread.scala:171)
> at
> scala.collection.mutable.ResizableArray$class.foreach(ResizableArray.scala:59)
> at scala.collection.mutable.ArrayBuffer.foreach(ArrayBuffer.scala:48)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply$mcV$sp(AbstractFetcherThread.scala:171)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply(AbstractFetcherThread.scala:171)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply(AbstractFetcherThread.scala:171)
> at kafka.utils.CoreUtils$.inLock(CoreUtils.scala:213)
> at
> kafka.server.AbstractFetcherThread.processFetchRequest(AbstractFetcherThread.scala:169)
> at
> kafka.server.AbstractFetcherThread.doWork(AbstractFetcherThread.scala:112)
> at kafka.utils.ShutdownableThread.run(ShutdownableThread.scala:64)
> Caused by: org.apache.kafka.common.errors.OutOfOrderSequenceException:
> Invalid sequence number for new epoch: 0 (request epoch), 154277489 (seq.
> number)
> {code}
> We run kafka 0.11.0.1 and we ran into the situation when 1 of replication
> threads was stopped for few days, while everything else on that broker was
> functional. This is our staging cluster and retention is less than a day, so
> everything for partitions for which replication thread was down was cleaned
> up. At the moment we have a broker which cannot start replication for few
> partitions. I was also able to reproduce in my local test environment.
> Another possible use case when this might cause real pain is disk failure or
> any situation when previously deleting all the data for the partition on a
> broker helped - since it would just fetch all the data from other replicas.
> Now it does not work for topics with idempotent producers. It might also
> affect other not-idempotent topics if those are unlucky to share same
> replication fetcher thread.
> This seems to be caused by this logic:
> https://github.com/apache/kafka/blob/0.11.0.1/core/src/main/scala/kafka/log/ProducerStateManager.scala#L119
> and might be fixed in the scope of
> https://issues.apache.org/jira/browse/KAFKA-5793.
> However any hints on how to get those partition to fully replicated state are
> highly appreciated.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
[jira] [Commented] (KAFKA-6003) Replication Fetcher thread for a partition with no data fails to start
[
https://issues.apache.org/jira/browse/KAFKA-6003?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16189014#comment-16189014
]
Stanislav Chizhov commented on KAFKA-6003:
--
Aha! Got it. Will try reassigning then. Used to think though that after
partition reassignment new broker would be exactly in the same position as the
current one and thus fail replicating as well.
> Replication Fetcher thread for a partition with no data fails to start
> --
>
> Key: KAFKA-6003
> URL: https://issues.apache.org/jira/browse/KAFKA-6003
> Project: Kafka
> Issue Type: Bug
> Components: replication
>Affects Versions: 0.11.0.1
>Reporter: Stanislav Chizhov
>Assignee: Apurva Mehta
>Priority: Blocker
> Fix For: 1.0.0, 0.11.0.2
>
>
> If a partition of a topic with idempotent producer has no data on 1 of the
> brokers, but it does exist on others and some of the segments for this
> partition have been already deleted replication thread responsible for this
> partition on the broker which has no data for it fails to start with out of
> order sequence exception:
> {code}
> [2017-10-02 09:44:23,825] ERROR [ReplicaFetcherThread-2-4]: Error due to
> (kafka.server.ReplicaFetcherThread)
> kafka.common.KafkaException: error processing data for partition
> [stage.data.adevents.v2,20] offset 1660336429
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1$$anonfun$apply$2.apply(AbstractFetcherThread.scala:203)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1$$anonfun$apply$2.apply(AbstractFetcherThread.scala:174)
> at scala.Option.foreach(Option.scala:257)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1.apply(AbstractFetcherThread.scala:174)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1.apply(AbstractFetcherThread.scala:171)
> at
> scala.collection.mutable.ResizableArray$class.foreach(ResizableArray.scala:59)
> at scala.collection.mutable.ArrayBuffer.foreach(ArrayBuffer.scala:48)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply$mcV$sp(AbstractFetcherThread.scala:171)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply(AbstractFetcherThread.scala:171)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply(AbstractFetcherThread.scala:171)
> at kafka.utils.CoreUtils$.inLock(CoreUtils.scala:213)
> at
> kafka.server.AbstractFetcherThread.processFetchRequest(AbstractFetcherThread.scala:169)
> at
> kafka.server.AbstractFetcherThread.doWork(AbstractFetcherThread.scala:112)
> at kafka.utils.ShutdownableThread.run(ShutdownableThread.scala:64)
> Caused by: org.apache.kafka.common.errors.OutOfOrderSequenceException:
> Invalid sequence number for new epoch: 0 (request epoch), 154277489 (seq.
> number)
> {code}
> We run kafka 0.11.0.1 and we ran into the situation when 1 of replication
> threads was stopped for few days, while everything else on that broker was
> functional. This is our staging cluster and retention is less than a day, so
> everything for partitions for which replication thread was down was cleaned
> up. At the moment we have a broker which cannot start replication for few
> partitions. I was also able to reproduce in my local test environment.
> Another possible use case when this might cause real pain is disk failure or
> any situation when previously deleting all the data for the partition on a
> broker helped - since it would just fetch all the data from other replicas.
> Now it does not work for topics with idempotent producers. It might also
> affect other not-idempotent topics if those are unlucky to share same
> replication fetcher thread.
> This seems to be caused by this logic:
> https://github.com/apache/kafka/blob/0.11.0.1/core/src/main/scala/kafka/log/ProducerStateManager.scala#L119
> and might be fixed in the scope of
> https://issues.apache.org/jira/browse/KAFKA-5793.
> However any hints on how to get those partition to fully replicated state are
> highly appreciated.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
[jira] [Commented] (KAFKA-6003) Replication Fetcher thread for a partition with no data fails to start
[
https://issues.apache.org/jira/browse/KAFKA-6003?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16189022#comment-16189022
]
Stanislav Chizhov commented on KAFKA-6003:
--
OK. Then I will not try reassigning :) .
> Replication Fetcher thread for a partition with no data fails to start
> --
>
> Key: KAFKA-6003
> URL: https://issues.apache.org/jira/browse/KAFKA-6003
> Project: Kafka
> Issue Type: Bug
> Components: replication
>Affects Versions: 0.11.0.1
>Reporter: Stanislav Chizhov
>Assignee: Apurva Mehta
>Priority: Blocker
> Fix For: 1.0.0, 0.11.0.2
>
>
> If a partition of a topic with idempotent producer has no data on 1 of the
> brokers, but it does exist on others and some of the segments for this
> partition have been already deleted replication thread responsible for this
> partition on the broker which has no data for it fails to start with out of
> order sequence exception:
> {code}
> [2017-10-02 09:44:23,825] ERROR [ReplicaFetcherThread-2-4]: Error due to
> (kafka.server.ReplicaFetcherThread)
> kafka.common.KafkaException: error processing data for partition
> [stage.data.adevents.v2,20] offset 1660336429
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1$$anonfun$apply$2.apply(AbstractFetcherThread.scala:203)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1$$anonfun$apply$2.apply(AbstractFetcherThread.scala:174)
> at scala.Option.foreach(Option.scala:257)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1.apply(AbstractFetcherThread.scala:174)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1.apply(AbstractFetcherThread.scala:171)
> at
> scala.collection.mutable.ResizableArray$class.foreach(ResizableArray.scala:59)
> at scala.collection.mutable.ArrayBuffer.foreach(ArrayBuffer.scala:48)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply$mcV$sp(AbstractFetcherThread.scala:171)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply(AbstractFetcherThread.scala:171)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply(AbstractFetcherThread.scala:171)
> at kafka.utils.CoreUtils$.inLock(CoreUtils.scala:213)
> at
> kafka.server.AbstractFetcherThread.processFetchRequest(AbstractFetcherThread.scala:169)
> at
> kafka.server.AbstractFetcherThread.doWork(AbstractFetcherThread.scala:112)
> at kafka.utils.ShutdownableThread.run(ShutdownableThread.scala:64)
> Caused by: org.apache.kafka.common.errors.OutOfOrderSequenceException:
> Invalid sequence number for new epoch: 0 (request epoch), 154277489 (seq.
> number)
> {code}
> We run kafka 0.11.0.1 and we ran into the situation when 1 of replication
> threads was stopped for few days, while everything else on that broker was
> functional. This is our staging cluster and retention is less than a day, so
> everything for partitions for which replication thread was down was cleaned
> up. At the moment we have a broker which cannot start replication for few
> partitions. I was also able to reproduce in my local test environment.
> Another possible use case when this might cause real pain is disk failure or
> any situation when previously deleting all the data for the partition on a
> broker helped - since it would just fetch all the data from other replicas.
> Now it does not work for topics with idempotent producers. It might also
> affect other not-idempotent topics if those are unlucky to share same
> replication fetcher thread.
> This seems to be caused by this logic:
> https://github.com/apache/kafka/blob/0.11.0.1/core/src/main/scala/kafka/log/ProducerStateManager.scala#L119
> and might be fixed in the scope of
> https://issues.apache.org/jira/browse/KAFKA-5793.
> However any hints on how to get those partition to fully replicated state are
> highly appreciated.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
[jira] [Commented] (KAFKA-6003) Replication Fetcher thread for a partition with no data fails to start
[
https://issues.apache.org/jira/browse/KAFKA-6003?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16189026#comment-16189026
]
Stanislav Chizhov commented on KAFKA-6003:
--
Are there any expectation of when 0.11.0.2 is coming out?
> Replication Fetcher thread for a partition with no data fails to start
> --
>
> Key: KAFKA-6003
> URL: https://issues.apache.org/jira/browse/KAFKA-6003
> Project: Kafka
> Issue Type: Bug
> Components: replication
>Affects Versions: 0.11.0.1
>Reporter: Stanislav Chizhov
>Assignee: Apurva Mehta
>Priority: Blocker
> Fix For: 1.0.0, 0.11.0.2
>
>
> If a partition of a topic with idempotent producer has no data on 1 of the
> brokers, but it does exist on others and some of the segments for this
> partition have been already deleted replication thread responsible for this
> partition on the broker which has no data for it fails to start with out of
> order sequence exception:
> {code}
> [2017-10-02 09:44:23,825] ERROR [ReplicaFetcherThread-2-4]: Error due to
> (kafka.server.ReplicaFetcherThread)
> kafka.common.KafkaException: error processing data for partition
> [stage.data.adevents.v2,20] offset 1660336429
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1$$anonfun$apply$2.apply(AbstractFetcherThread.scala:203)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1$$anonfun$apply$2.apply(AbstractFetcherThread.scala:174)
> at scala.Option.foreach(Option.scala:257)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1.apply(AbstractFetcherThread.scala:174)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1.apply(AbstractFetcherThread.scala:171)
> at
> scala.collection.mutable.ResizableArray$class.foreach(ResizableArray.scala:59)
> at scala.collection.mutable.ArrayBuffer.foreach(ArrayBuffer.scala:48)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply$mcV$sp(AbstractFetcherThread.scala:171)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply(AbstractFetcherThread.scala:171)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply(AbstractFetcherThread.scala:171)
> at kafka.utils.CoreUtils$.inLock(CoreUtils.scala:213)
> at
> kafka.server.AbstractFetcherThread.processFetchRequest(AbstractFetcherThread.scala:169)
> at
> kafka.server.AbstractFetcherThread.doWork(AbstractFetcherThread.scala:112)
> at kafka.utils.ShutdownableThread.run(ShutdownableThread.scala:64)
> Caused by: org.apache.kafka.common.errors.OutOfOrderSequenceException:
> Invalid sequence number for new epoch: 0 (request epoch), 154277489 (seq.
> number)
> {code}
> We run kafka 0.11.0.1 and we ran into the situation when 1 of replication
> threads was stopped for few days, while everything else on that broker was
> functional. This is our staging cluster and retention is less than a day, so
> everything for partitions for which replication thread was down was cleaned
> up. At the moment we have a broker which cannot start replication for few
> partitions. I was also able to reproduce in my local test environment.
> Another possible use case when this might cause real pain is disk failure or
> any situation when previously deleting all the data for the partition on a
> broker helped - since it would just fetch all the data from other replicas.
> Now it does not work for topics with idempotent producers. It might also
> affect other not-idempotent topics if those are unlucky to share same
> replication fetcher thread.
> This seems to be caused by this logic:
> https://github.com/apache/kafka/blob/0.11.0.1/core/src/main/scala/kafka/log/ProducerStateManager.scala#L119
> and might be fixed in the scope of
> https://issues.apache.org/jira/browse/KAFKA-5793.
> However any hints on how to get those partition to fully replicated state are
> highly appreciated.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
[jira] [Commented] (KAFKA-6003) Replication Fetcher thread for a partition with no data fails to start
[
https://issues.apache.org/jira/browse/KAFKA-6003?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16189082#comment-16189082
]
Apurva Mehta commented on KAFKA-6003:
-
Unfortunately, I don't think there is an ETA for 0.11.0.2 at this time.
0.11.0.1 was just released a couple of weeks ago.
> Replication Fetcher thread for a partition with no data fails to start
> --
>
> Key: KAFKA-6003
> URL: https://issues.apache.org/jira/browse/KAFKA-6003
> Project: Kafka
> Issue Type: Bug
> Components: replication
>Affects Versions: 0.11.0.1
>Reporter: Stanislav Chizhov
>Assignee: Apurva Mehta
>Priority: Blocker
> Fix For: 1.0.0, 0.11.0.2
>
>
> If a partition of a topic with idempotent producer has no data on 1 of the
> brokers, but it does exist on others and some of the segments for this
> partition have been already deleted replication thread responsible for this
> partition on the broker which has no data for it fails to start with out of
> order sequence exception:
> {code}
> [2017-10-02 09:44:23,825] ERROR [ReplicaFetcherThread-2-4]: Error due to
> (kafka.server.ReplicaFetcherThread)
> kafka.common.KafkaException: error processing data for partition
> [stage.data.adevents.v2,20] offset 1660336429
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1$$anonfun$apply$2.apply(AbstractFetcherThread.scala:203)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1$$anonfun$apply$2.apply(AbstractFetcherThread.scala:174)
> at scala.Option.foreach(Option.scala:257)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1.apply(AbstractFetcherThread.scala:174)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1.apply(AbstractFetcherThread.scala:171)
> at
> scala.collection.mutable.ResizableArray$class.foreach(ResizableArray.scala:59)
> at scala.collection.mutable.ArrayBuffer.foreach(ArrayBuffer.scala:48)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply$mcV$sp(AbstractFetcherThread.scala:171)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply(AbstractFetcherThread.scala:171)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply(AbstractFetcherThread.scala:171)
> at kafka.utils.CoreUtils$.inLock(CoreUtils.scala:213)
> at
> kafka.server.AbstractFetcherThread.processFetchRequest(AbstractFetcherThread.scala:169)
> at
> kafka.server.AbstractFetcherThread.doWork(AbstractFetcherThread.scala:112)
> at kafka.utils.ShutdownableThread.run(ShutdownableThread.scala:64)
> Caused by: org.apache.kafka.common.errors.OutOfOrderSequenceException:
> Invalid sequence number for new epoch: 0 (request epoch), 154277489 (seq.
> number)
> {code}
> We run kafka 0.11.0.1 and we ran into the situation when 1 of replication
> threads was stopped for few days, while everything else on that broker was
> functional. This is our staging cluster and retention is less than a day, so
> everything for partitions for which replication thread was down was cleaned
> up. At the moment we have a broker which cannot start replication for few
> partitions. I was also able to reproduce in my local test environment.
> Another possible use case when this might cause real pain is disk failure or
> any situation when previously deleting all the data for the partition on a
> broker helped - since it would just fetch all the data from other replicas.
> Now it does not work for topics with idempotent producers. It might also
> affect other not-idempotent topics if those are unlucky to share same
> replication fetcher thread.
> This seems to be caused by this logic:
> https://github.com/apache/kafka/blob/0.11.0.1/core/src/main/scala/kafka/log/ProducerStateManager.scala#L119
> and might be fixed in the scope of
> https://issues.apache.org/jira/browse/KAFKA-5793.
> However any hints on how to get those partition to fully replicated state are
> highly appreciated.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
[jira] [Commented] (KAFKA-6003) Replication Fetcher thread for a partition with no data fails to start
[
https://issues.apache.org/jira/browse/KAFKA-6003?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16189731#comment-16189731
]
Stanislav Chizhov commented on KAFKA-6003:
--
So to summarise the impact:
In case there is topic with finite retention and idempotent producer there is
very high risk of not being able to:
- replace a dead broker
- scale up cluster
- reassign partitions (BTW I did try it and it failed in the same way)
If the above is true it renders idempotent producer feature pretty much
unusable.
Please correct me if I am wrong.
> Replication Fetcher thread for a partition with no data fails to start
> --
>
> Key: KAFKA-6003
> URL: https://issues.apache.org/jira/browse/KAFKA-6003
> Project: Kafka
> Issue Type: Bug
> Components: replication
>Affects Versions: 0.11.0.1
>Reporter: Stanislav Chizhov
>Assignee: Apurva Mehta
>Priority: Blocker
> Fix For: 1.0.0, 0.11.0.2
>
>
> If a partition of a topic with idempotent producer has no data on 1 of the
> brokers, but it does exist on others and some of the segments for this
> partition have been already deleted replication thread responsible for this
> partition on the broker which has no data for it fails to start with out of
> order sequence exception:
> {code}
> [2017-10-02 09:44:23,825] ERROR [ReplicaFetcherThread-2-4]: Error due to
> (kafka.server.ReplicaFetcherThread)
> kafka.common.KafkaException: error processing data for partition
> [stage.data.adevents.v2,20] offset 1660336429
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1$$anonfun$apply$2.apply(AbstractFetcherThread.scala:203)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1$$anonfun$apply$2.apply(AbstractFetcherThread.scala:174)
> at scala.Option.foreach(Option.scala:257)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1.apply(AbstractFetcherThread.scala:174)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1.apply(AbstractFetcherThread.scala:171)
> at
> scala.collection.mutable.ResizableArray$class.foreach(ResizableArray.scala:59)
> at scala.collection.mutable.ArrayBuffer.foreach(ArrayBuffer.scala:48)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply$mcV$sp(AbstractFetcherThread.scala:171)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply(AbstractFetcherThread.scala:171)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply(AbstractFetcherThread.scala:171)
> at kafka.utils.CoreUtils$.inLock(CoreUtils.scala:213)
> at
> kafka.server.AbstractFetcherThread.processFetchRequest(AbstractFetcherThread.scala:169)
> at
> kafka.server.AbstractFetcherThread.doWork(AbstractFetcherThread.scala:112)
> at kafka.utils.ShutdownableThread.run(ShutdownableThread.scala:64)
> Caused by: org.apache.kafka.common.errors.OutOfOrderSequenceException:
> Invalid sequence number for new epoch: 0 (request epoch), 154277489 (seq.
> number)
> {code}
> We run kafka 0.11.0.1 and we ran into the situation when 1 of replication
> threads was stopped for few days, while everything else on that broker was
> functional. This is our staging cluster and retention is less than a day, so
> everything for partitions for which replication thread was down was cleaned
> up. At the moment we have a broker which cannot start replication for few
> partitions. I was also able to reproduce in my local test environment.
> Another possible use case when this might cause real pain is disk failure or
> any situation when previously deleting all the data for the partition on a
> broker helped - since it would just fetch all the data from other replicas.
> Now it does not work for topics with idempotent producers. It might also
> affect other not-idempotent topics if those are unlucky to share same
> replication fetcher thread.
> This seems to be caused by this logic:
> https://github.com/apache/kafka/blob/0.11.0.1/core/src/main/scala/kafka/log/ProducerStateManager.scala#L119
> and might be fixed in the scope of
> https://issues.apache.org/jira/browse/KAFKA-5793.
> However any hints on how to get those partition to fully replicated state are
> highly appreciated.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
[jira] [Commented] (KAFKA-6003) Replication Fetcher thread for a partition with no data fails to start
[
https://issues.apache.org/jira/browse/KAFKA-6003?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16189937#comment-16189937
]
Apurva Mehta commented on KAFKA-6003:
-
Unfortunately your summary is correct :(
We have tests which actually do cluster expansion / replica movement, but
unfortunately the workload in those tests doesn't include idempotent messages,
so we didn't catch this.
We will update the tests to use the idempotent producer so that we can catch
any further bugs in these scenarios.
For now, we can start the conversation about the 0.11.0.2 release on the
mailing list, and see what the community thinks of a quick release just to fix
this issue.
> Replication Fetcher thread for a partition with no data fails to start
> --
>
> Key: KAFKA-6003
> URL: https://issues.apache.org/jira/browse/KAFKA-6003
> Project: Kafka
> Issue Type: Bug
> Components: replication
>Affects Versions: 0.11.0.1
>Reporter: Stanislav Chizhov
>Assignee: Apurva Mehta
>Priority: Blocker
> Fix For: 1.0.0, 0.11.0.2
>
>
> If a partition of a topic with idempotent producer has no data on 1 of the
> brokers, but it does exist on others and some of the segments for this
> partition have been already deleted replication thread responsible for this
> partition on the broker which has no data for it fails to start with out of
> order sequence exception:
> {code}
> [2017-10-02 09:44:23,825] ERROR [ReplicaFetcherThread-2-4]: Error due to
> (kafka.server.ReplicaFetcherThread)
> kafka.common.KafkaException: error processing data for partition
> [stage.data.adevents.v2,20] offset 1660336429
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1$$anonfun$apply$2.apply(AbstractFetcherThread.scala:203)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1$$anonfun$apply$2.apply(AbstractFetcherThread.scala:174)
> at scala.Option.foreach(Option.scala:257)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1.apply(AbstractFetcherThread.scala:174)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1.apply(AbstractFetcherThread.scala:171)
> at
> scala.collection.mutable.ResizableArray$class.foreach(ResizableArray.scala:59)
> at scala.collection.mutable.ArrayBuffer.foreach(ArrayBuffer.scala:48)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply$mcV$sp(AbstractFetcherThread.scala:171)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply(AbstractFetcherThread.scala:171)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply(AbstractFetcherThread.scala:171)
> at kafka.utils.CoreUtils$.inLock(CoreUtils.scala:213)
> at
> kafka.server.AbstractFetcherThread.processFetchRequest(AbstractFetcherThread.scala:169)
> at
> kafka.server.AbstractFetcherThread.doWork(AbstractFetcherThread.scala:112)
> at kafka.utils.ShutdownableThread.run(ShutdownableThread.scala:64)
> Caused by: org.apache.kafka.common.errors.OutOfOrderSequenceException:
> Invalid sequence number for new epoch: 0 (request epoch), 154277489 (seq.
> number)
> {code}
> We run kafka 0.11.0.1 and we ran into the situation when 1 of replication
> threads was stopped for few days, while everything else on that broker was
> functional. This is our staging cluster and retention is less than a day, so
> everything for partitions for which replication thread was down was cleaned
> up. At the moment we have a broker which cannot start replication for few
> partitions. I was also able to reproduce in my local test environment.
> Another possible use case when this might cause real pain is disk failure or
> any situation when previously deleting all the data for the partition on a
> broker helped - since it would just fetch all the data from other replicas.
> Now it does not work for topics with idempotent producers. It might also
> affect other not-idempotent topics if those are unlucky to share same
> replication fetcher thread.
> This seems to be caused by this logic:
> https://github.com/apache/kafka/blob/0.11.0.1/core/src/main/scala/kafka/log/ProducerStateManager.scala#L119
> and might be fixed in the scope of
> https://issues.apache.org/jira/browse/KAFKA-5793.
> However any hints on how to get those partition to fully replicated state are
> highly appreciated.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
[jira] [Commented] (KAFKA-6003) Replication Fetcher thread for a partition with no data fails to start
[
https://issues.apache.org/jira/browse/KAFKA-6003?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16192028#comment-16192028
]
ASF GitHub Bot commented on KAFKA-6003:
---
GitHub user apurvam opened a pull request:
https://github.com/apache/kafka/pull/4020
KAFKA-6003: Accept appends on replicas and when rebuilding the log
unconditionally
This is a port of #4004 for the 0.11.0 branch.
With this patch so that we _only_ validate appends which originate
from the client. In general, once the append is validated and written to
the leader the first time, revalidating it is undesirable since we can't
do anything if validation fails, and also because it is hard to maintain
the correct assumptions during validation, leading to spurious
validation failures.
For example, when we have compacted topics, it is possible for batches
to be compacted on the follower but not on the leader. This case would
also lead to an OutOfOrderSequencException during replication. The same
applies to when we rebuild state from compacted topics: we would get
gaps in the sequence numbers, causing the OutOfOrderSequence.
You can merge this pull request into a Git repository by running:
$ git pull https://github.com/apurvam/kafka
KAKFA-6003-0.11.0-handle-unknown-producer-on-replica
Alternatively you can review and apply these changes as the patch at:
https://github.com/apache/kafka/pull/4020.patch
To close this pull request, make a commit to your master/trunk branch
with (at least) the following in the commit message:
This closes #4020
commit 0a6a0213c091c8e6b6a9c5ce7655b7e0d06c9db0
Author: Apurva Mehta
Date: 2017-10-04T20:42:17Z
KAFKA-6003: Accept appends on replicas and when rebuilding state from
the log unconditionally.
With this patch so that we _only_ validate appends which originate
from the client. In general, once the append is validated and written to
the leader the first time, revalidating it is undesirable since we can't
do anything if validation fails, and also because it is hard to maintain
the correct assumptions during validation, leading to spurious
validation failures.
For example, when we have compacted topics, it is possible for batches
to be compacted on the follower but not on the leader. This case would
also lead to an OutOfOrderSequencException during replication. The same
applies to when we rebuild state from compacted topics: we would get
gaps in the sequence numbers, causing the OutOfOrderSequence.
> Replication Fetcher thread for a partition with no data fails to start
> --
>
> Key: KAFKA-6003
> URL: https://issues.apache.org/jira/browse/KAFKA-6003
> Project: Kafka
> Issue Type: Bug
> Components: replication
>Affects Versions: 0.11.0.1
>Reporter: Stanislav Chizhov
>Assignee: Apurva Mehta
>Priority: Blocker
> Fix For: 1.0.0, 0.11.0.2
>
>
> If a partition of a topic with idempotent producer has no data on 1 of the
> brokers, but it does exist on others and some of the segments for this
> partition have been already deleted replication thread responsible for this
> partition on the broker which has no data for it fails to start with out of
> order sequence exception:
> {code}
> [2017-10-02 09:44:23,825] ERROR [ReplicaFetcherThread-2-4]: Error due to
> (kafka.server.ReplicaFetcherThread)
> kafka.common.KafkaException: error processing data for partition
> [stage.data.adevents.v2,20] offset 1660336429
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1$$anonfun$apply$2.apply(AbstractFetcherThread.scala:203)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1$$anonfun$apply$2.apply(AbstractFetcherThread.scala:174)
> at scala.Option.foreach(Option.scala:257)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1.apply(AbstractFetcherThread.scala:174)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1.apply(AbstractFetcherThread.scala:171)
> at
> scala.collection.mutable.ResizableArray$class.foreach(ResizableArray.scala:59)
> at scala.collection.mutable.ArrayBuffer.foreach(ArrayBuffer.scala:48)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply$mcV$sp(AbstractFetcherThread.scala:171)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply(AbstractFetcherThread.scala:171)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply(AbstractFetcherThread.scala:1
[jira] [Commented] (KAFKA-6003) Replication Fetcher thread for a partition with no data fails to start
[
https://issues.apache.org/jira/browse/KAFKA-6003?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16192484#comment-16192484
]
ASF GitHub Bot commented on KAFKA-6003:
---
Github user asfgit closed the pull request at:
https://github.com/apache/kafka/pull/4004
> Replication Fetcher thread for a partition with no data fails to start
> --
>
> Key: KAFKA-6003
> URL: https://issues.apache.org/jira/browse/KAFKA-6003
> Project: Kafka
> Issue Type: Bug
> Components: replication
>Affects Versions: 0.11.0.1
>Reporter: Stanislav Chizhov
>Assignee: Apurva Mehta
>Priority: Blocker
> Fix For: 1.0.0
>
>
> If a partition of a topic with idempotent producer has no data on 1 of the
> brokers, but it does exist on others and some of the segments for this
> partition have been already deleted replication thread responsible for this
> partition on the broker which has no data for it fails to start with out of
> order sequence exception:
> {code}
> [2017-10-02 09:44:23,825] ERROR [ReplicaFetcherThread-2-4]: Error due to
> (kafka.server.ReplicaFetcherThread)
> kafka.common.KafkaException: error processing data for partition
> [stage.data.adevents.v2,20] offset 1660336429
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1$$anonfun$apply$2.apply(AbstractFetcherThread.scala:203)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1$$anonfun$apply$2.apply(AbstractFetcherThread.scala:174)
> at scala.Option.foreach(Option.scala:257)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1.apply(AbstractFetcherThread.scala:174)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1.apply(AbstractFetcherThread.scala:171)
> at
> scala.collection.mutable.ResizableArray$class.foreach(ResizableArray.scala:59)
> at scala.collection.mutable.ArrayBuffer.foreach(ArrayBuffer.scala:48)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply$mcV$sp(AbstractFetcherThread.scala:171)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply(AbstractFetcherThread.scala:171)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply(AbstractFetcherThread.scala:171)
> at kafka.utils.CoreUtils$.inLock(CoreUtils.scala:213)
> at
> kafka.server.AbstractFetcherThread.processFetchRequest(AbstractFetcherThread.scala:169)
> at
> kafka.server.AbstractFetcherThread.doWork(AbstractFetcherThread.scala:112)
> at kafka.utils.ShutdownableThread.run(ShutdownableThread.scala:64)
> Caused by: org.apache.kafka.common.errors.OutOfOrderSequenceException:
> Invalid sequence number for new epoch: 0 (request epoch), 154277489 (seq.
> number)
> {code}
> We run kafka 0.11.0.1 and we ran into the situation when 1 of replication
> threads was stopped for few days, while everything else on that broker was
> functional. This is our staging cluster and retention is less than a day, so
> everything for partitions for which replication thread was down was cleaned
> up. At the moment we have a broker which cannot start replication for few
> partitions. I was also able to reproduce in my local test environment.
> Another possible use case when this might cause real pain is disk failure or
> any situation when previously deleting all the data for the partition on a
> broker helped - since it would just fetch all the data from other replicas.
> Now it does not work for topics with idempotent producers. It might also
> affect other not-idempotent topics if those are unlucky to share same
> replication fetcher thread.
> This seems to be caused by this logic:
> https://github.com/apache/kafka/blob/0.11.0.1/core/src/main/scala/kafka/log/ProducerStateManager.scala#L119
> and might be fixed in the scope of
> https://issues.apache.org/jira/browse/KAFKA-5793.
> However any hints on how to get those partition to fully replicated state are
> highly appreciated.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
[jira] [Commented] (KAFKA-6003) Replication Fetcher thread for a partition with no data fails to start
[
https://issues.apache.org/jira/browse/KAFKA-6003?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16193721#comment-16193721
]
Apurva Mehta commented on KAFKA-6003:
-
Added a system test which reproduces this scenario:
https://github.com/apache/kafka/pull/4029
> Replication Fetcher thread for a partition with no data fails to start
> --
>
> Key: KAFKA-6003
> URL: https://issues.apache.org/jira/browse/KAFKA-6003
> Project: Kafka
> Issue Type: Bug
> Components: replication
>Affects Versions: 0.11.0.1
>Reporter: Stanislav Chizhov
>Assignee: Apurva Mehta
>Priority: Blocker
> Fix For: 1.0.0
>
>
> If a partition of a topic with idempotent producer has no data on 1 of the
> brokers, but it does exist on others and some of the segments for this
> partition have been already deleted replication thread responsible for this
> partition on the broker which has no data for it fails to start with out of
> order sequence exception:
> {code}
> [2017-10-02 09:44:23,825] ERROR [ReplicaFetcherThread-2-4]: Error due to
> (kafka.server.ReplicaFetcherThread)
> kafka.common.KafkaException: error processing data for partition
> [stage.data.adevents.v2,20] offset 1660336429
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1$$anonfun$apply$2.apply(AbstractFetcherThread.scala:203)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1$$anonfun$apply$2.apply(AbstractFetcherThread.scala:174)
> at scala.Option.foreach(Option.scala:257)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1.apply(AbstractFetcherThread.scala:174)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1.apply(AbstractFetcherThread.scala:171)
> at
> scala.collection.mutable.ResizableArray$class.foreach(ResizableArray.scala:59)
> at scala.collection.mutable.ArrayBuffer.foreach(ArrayBuffer.scala:48)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply$mcV$sp(AbstractFetcherThread.scala:171)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply(AbstractFetcherThread.scala:171)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply(AbstractFetcherThread.scala:171)
> at kafka.utils.CoreUtils$.inLock(CoreUtils.scala:213)
> at
> kafka.server.AbstractFetcherThread.processFetchRequest(AbstractFetcherThread.scala:169)
> at
> kafka.server.AbstractFetcherThread.doWork(AbstractFetcherThread.scala:112)
> at kafka.utils.ShutdownableThread.run(ShutdownableThread.scala:64)
> Caused by: org.apache.kafka.common.errors.OutOfOrderSequenceException:
> Invalid sequence number for new epoch: 0 (request epoch), 154277489 (seq.
> number)
> {code}
> We run kafka 0.11.0.1 and we ran into the situation when 1 of replication
> threads was stopped for few days, while everything else on that broker was
> functional. This is our staging cluster and retention is less than a day, so
> everything for partitions for which replication thread was down was cleaned
> up. At the moment we have a broker which cannot start replication for few
> partitions. I was also able to reproduce in my local test environment.
> Another possible use case when this might cause real pain is disk failure or
> any situation when previously deleting all the data for the partition on a
> broker helped - since it would just fetch all the data from other replicas.
> Now it does not work for topics with idempotent producers. It might also
> affect other not-idempotent topics if those are unlucky to share same
> replication fetcher thread.
> This seems to be caused by this logic:
> https://github.com/apache/kafka/blob/0.11.0.1/core/src/main/scala/kafka/log/ProducerStateManager.scala#L119
> and might be fixed in the scope of
> https://issues.apache.org/jira/browse/KAFKA-5793.
> However any hints on how to get those partition to fully replicated state are
> highly appreciated.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
[jira] [Commented] (KAFKA-6003) Replication Fetcher thread for a partition with no data fails to start
[
https://issues.apache.org/jira/browse/KAFKA-6003?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16194347#comment-16194347
]
Stanislav Chizhov commented on KAFKA-6003:
--
Hi [~apurva]]. So now there are no plans to have that fixed in 0.11.0.2 as far
as I can see from the fix version of this ticket. Can u point me to a related
discussion thread somewhere or shed some light on this here?
Thank you.
> Replication Fetcher thread for a partition with no data fails to start
> --
>
> Key: KAFKA-6003
> URL: https://issues.apache.org/jira/browse/KAFKA-6003
> Project: Kafka
> Issue Type: Bug
> Components: replication
>Affects Versions: 0.11.0.1
>Reporter: Stanislav Chizhov
>Assignee: Apurva Mehta
>Priority: Blocker
> Fix For: 1.0.0
>
>
> If a partition of a topic with idempotent producer has no data on 1 of the
> brokers, but it does exist on others and some of the segments for this
> partition have been already deleted replication thread responsible for this
> partition on the broker which has no data for it fails to start with out of
> order sequence exception:
> {code}
> [2017-10-02 09:44:23,825] ERROR [ReplicaFetcherThread-2-4]: Error due to
> (kafka.server.ReplicaFetcherThread)
> kafka.common.KafkaException: error processing data for partition
> [stage.data.adevents.v2,20] offset 1660336429
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1$$anonfun$apply$2.apply(AbstractFetcherThread.scala:203)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1$$anonfun$apply$2.apply(AbstractFetcherThread.scala:174)
> at scala.Option.foreach(Option.scala:257)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1.apply(AbstractFetcherThread.scala:174)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1.apply(AbstractFetcherThread.scala:171)
> at
> scala.collection.mutable.ResizableArray$class.foreach(ResizableArray.scala:59)
> at scala.collection.mutable.ArrayBuffer.foreach(ArrayBuffer.scala:48)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply$mcV$sp(AbstractFetcherThread.scala:171)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply(AbstractFetcherThread.scala:171)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply(AbstractFetcherThread.scala:171)
> at kafka.utils.CoreUtils$.inLock(CoreUtils.scala:213)
> at
> kafka.server.AbstractFetcherThread.processFetchRequest(AbstractFetcherThread.scala:169)
> at
> kafka.server.AbstractFetcherThread.doWork(AbstractFetcherThread.scala:112)
> at kafka.utils.ShutdownableThread.run(ShutdownableThread.scala:64)
> Caused by: org.apache.kafka.common.errors.OutOfOrderSequenceException:
> Invalid sequence number for new epoch: 0 (request epoch), 154277489 (seq.
> number)
> {code}
> We run kafka 0.11.0.1 and we ran into the situation when 1 of replication
> threads was stopped for few days, while everything else on that broker was
> functional. This is our staging cluster and retention is less than a day, so
> everything for partitions for which replication thread was down was cleaned
> up. At the moment we have a broker which cannot start replication for few
> partitions. I was also able to reproduce in my local test environment.
> Another possible use case when this might cause real pain is disk failure or
> any situation when previously deleting all the data for the partition on a
> broker helped - since it would just fetch all the data from other replicas.
> Now it does not work for topics with idempotent producers. It might also
> affect other not-idempotent topics if those are unlucky to share same
> replication fetcher thread.
> This seems to be caused by this logic:
> https://github.com/apache/kafka/blob/0.11.0.1/core/src/main/scala/kafka/log/ProducerStateManager.scala#L119
> and might be fixed in the scope of
> https://issues.apache.org/jira/browse/KAFKA-5793.
> However any hints on how to get those partition to fully replicated state are
> highly appreciated.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
[jira] [Commented] (KAFKA-6003) Replication Fetcher thread for a partition with no data fails to start
[
https://issues.apache.org/jira/browse/KAFKA-6003?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16194382#comment-16194382
]
Ismael Juma commented on KAFKA-6003:
There is a PR open for the 0.11.0 branch:
https://github.com/apache/kafka/pull/4020
> Replication Fetcher thread for a partition with no data fails to start
> --
>
> Key: KAFKA-6003
> URL: https://issues.apache.org/jira/browse/KAFKA-6003
> Project: Kafka
> Issue Type: Bug
> Components: replication
>Affects Versions: 0.11.0.1
>Reporter: Stanislav Chizhov
>Assignee: Apurva Mehta
>Priority: Blocker
> Fix For: 1.0.0
>
>
> If a partition of a topic with idempotent producer has no data on 1 of the
> brokers, but it does exist on others and some of the segments for this
> partition have been already deleted replication thread responsible for this
> partition on the broker which has no data for it fails to start with out of
> order sequence exception:
> {code}
> [2017-10-02 09:44:23,825] ERROR [ReplicaFetcherThread-2-4]: Error due to
> (kafka.server.ReplicaFetcherThread)
> kafka.common.KafkaException: error processing data for partition
> [stage.data.adevents.v2,20] offset 1660336429
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1$$anonfun$apply$2.apply(AbstractFetcherThread.scala:203)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1$$anonfun$apply$2.apply(AbstractFetcherThread.scala:174)
> at scala.Option.foreach(Option.scala:257)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1.apply(AbstractFetcherThread.scala:174)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1.apply(AbstractFetcherThread.scala:171)
> at
> scala.collection.mutable.ResizableArray$class.foreach(ResizableArray.scala:59)
> at scala.collection.mutable.ArrayBuffer.foreach(ArrayBuffer.scala:48)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply$mcV$sp(AbstractFetcherThread.scala:171)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply(AbstractFetcherThread.scala:171)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply(AbstractFetcherThread.scala:171)
> at kafka.utils.CoreUtils$.inLock(CoreUtils.scala:213)
> at
> kafka.server.AbstractFetcherThread.processFetchRequest(AbstractFetcherThread.scala:169)
> at
> kafka.server.AbstractFetcherThread.doWork(AbstractFetcherThread.scala:112)
> at kafka.utils.ShutdownableThread.run(ShutdownableThread.scala:64)
> Caused by: org.apache.kafka.common.errors.OutOfOrderSequenceException:
> Invalid sequence number for new epoch: 0 (request epoch), 154277489 (seq.
> number)
> {code}
> We run kafka 0.11.0.1 and we ran into the situation when 1 of replication
> threads was stopped for few days, while everything else on that broker was
> functional. This is our staging cluster and retention is less than a day, so
> everything for partitions for which replication thread was down was cleaned
> up. At the moment we have a broker which cannot start replication for few
> partitions. I was also able to reproduce in my local test environment.
> Another possible use case when this might cause real pain is disk failure or
> any situation when previously deleting all the data for the partition on a
> broker helped - since it would just fetch all the data from other replicas.
> Now it does not work for topics with idempotent producers. It might also
> affect other not-idempotent topics if those are unlucky to share same
> replication fetcher thread.
> This seems to be caused by this logic:
> https://github.com/apache/kafka/blob/0.11.0.1/core/src/main/scala/kafka/log/ProducerStateManager.scala#L119
> and might be fixed in the scope of
> https://issues.apache.org/jira/browse/KAFKA-5793.
> However any hints on how to get those partition to fully replicated state are
> highly appreciated.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
[jira] [Commented] (KAFKA-6003) Replication Fetcher thread for a partition with no data fails to start
[
https://issues.apache.org/jira/browse/KAFKA-6003?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16194469#comment-16194469
]
Stanislav Chizhov commented on KAFKA-6003:
--
OK, I see. Thank you. Just wondering: if there is a plan to fix it for 0.11.0.2
shoudn't this ticket has 0.11.0.2 in its "Fix Versions" list?
> Replication Fetcher thread for a partition with no data fails to start
> --
>
> Key: KAFKA-6003
> URL: https://issues.apache.org/jira/browse/KAFKA-6003
> Project: Kafka
> Issue Type: Bug
> Components: replication
>Affects Versions: 0.11.0.1
>Reporter: Stanislav Chizhov
>Assignee: Apurva Mehta
>Priority: Blocker
> Fix For: 1.0.0
>
>
> If a partition of a topic with idempotent producer has no data on 1 of the
> brokers, but it does exist on others and some of the segments for this
> partition have been already deleted replication thread responsible for this
> partition on the broker which has no data for it fails to start with out of
> order sequence exception:
> {code}
> [2017-10-02 09:44:23,825] ERROR [ReplicaFetcherThread-2-4]: Error due to
> (kafka.server.ReplicaFetcherThread)
> kafka.common.KafkaException: error processing data for partition
> [stage.data.adevents.v2,20] offset 1660336429
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1$$anonfun$apply$2.apply(AbstractFetcherThread.scala:203)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1$$anonfun$apply$2.apply(AbstractFetcherThread.scala:174)
> at scala.Option.foreach(Option.scala:257)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1.apply(AbstractFetcherThread.scala:174)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1.apply(AbstractFetcherThread.scala:171)
> at
> scala.collection.mutable.ResizableArray$class.foreach(ResizableArray.scala:59)
> at scala.collection.mutable.ArrayBuffer.foreach(ArrayBuffer.scala:48)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply$mcV$sp(AbstractFetcherThread.scala:171)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply(AbstractFetcherThread.scala:171)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply(AbstractFetcherThread.scala:171)
> at kafka.utils.CoreUtils$.inLock(CoreUtils.scala:213)
> at
> kafka.server.AbstractFetcherThread.processFetchRequest(AbstractFetcherThread.scala:169)
> at
> kafka.server.AbstractFetcherThread.doWork(AbstractFetcherThread.scala:112)
> at kafka.utils.ShutdownableThread.run(ShutdownableThread.scala:64)
> Caused by: org.apache.kafka.common.errors.OutOfOrderSequenceException:
> Invalid sequence number for new epoch: 0 (request epoch), 154277489 (seq.
> number)
> {code}
> We run kafka 0.11.0.1 and we ran into the situation when 1 of replication
> threads was stopped for few days, while everything else on that broker was
> functional. This is our staging cluster and retention is less than a day, so
> everything for partitions for which replication thread was down was cleaned
> up. At the moment we have a broker which cannot start replication for few
> partitions. I was also able to reproduce in my local test environment.
> Another possible use case when this might cause real pain is disk failure or
> any situation when previously deleting all the data for the partition on a
> broker helped - since it would just fetch all the data from other replicas.
> Now it does not work for topics with idempotent producers. It might also
> affect other not-idempotent topics if those are unlucky to share same
> replication fetcher thread.
> This seems to be caused by this logic:
> https://github.com/apache/kafka/blob/0.11.0.1/core/src/main/scala/kafka/log/ProducerStateManager.scala#L119
> and might be fixed in the scope of
> https://issues.apache.org/jira/browse/KAFKA-5793.
> However any hints on how to get those partition to fully replicated state are
> highly appreciated.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
[jira] [Commented] (KAFKA-6003) Replication Fetcher thread for a partition with no data fails to start
[
https://issues.apache.org/jira/browse/KAFKA-6003?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16194868#comment-16194868
]
Apurva Mehta commented on KAFKA-6003:
-
Hi [~schizhov] : I update the fix version to include 0.11.0.2. We should
absolutely fix this in an 0.11.0.2 release. The only open question is _when_
that release will happen: we don't have a date yet. I think it should happen
sooner rather than later, personally.
> Replication Fetcher thread for a partition with no data fails to start
> --
>
> Key: KAFKA-6003
> URL: https://issues.apache.org/jira/browse/KAFKA-6003
> Project: Kafka
> Issue Type: Bug
> Components: replication
>Affects Versions: 0.11.0.1
>Reporter: Stanislav Chizhov
>Assignee: Apurva Mehta
>Priority: Blocker
> Fix For: 1.0.0, 0.11.0.2
>
>
> If a partition of a topic with idempotent producer has no data on 1 of the
> brokers, but it does exist on others and some of the segments for this
> partition have been already deleted replication thread responsible for this
> partition on the broker which has no data for it fails to start with out of
> order sequence exception:
> {code}
> [2017-10-02 09:44:23,825] ERROR [ReplicaFetcherThread-2-4]: Error due to
> (kafka.server.ReplicaFetcherThread)
> kafka.common.KafkaException: error processing data for partition
> [stage.data.adevents.v2,20] offset 1660336429
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1$$anonfun$apply$2.apply(AbstractFetcherThread.scala:203)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1$$anonfun$apply$2.apply(AbstractFetcherThread.scala:174)
> at scala.Option.foreach(Option.scala:257)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1.apply(AbstractFetcherThread.scala:174)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1.apply(AbstractFetcherThread.scala:171)
> at
> scala.collection.mutable.ResizableArray$class.foreach(ResizableArray.scala:59)
> at scala.collection.mutable.ArrayBuffer.foreach(ArrayBuffer.scala:48)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply$mcV$sp(AbstractFetcherThread.scala:171)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply(AbstractFetcherThread.scala:171)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply(AbstractFetcherThread.scala:171)
> at kafka.utils.CoreUtils$.inLock(CoreUtils.scala:213)
> at
> kafka.server.AbstractFetcherThread.processFetchRequest(AbstractFetcherThread.scala:169)
> at
> kafka.server.AbstractFetcherThread.doWork(AbstractFetcherThread.scala:112)
> at kafka.utils.ShutdownableThread.run(ShutdownableThread.scala:64)
> Caused by: org.apache.kafka.common.errors.OutOfOrderSequenceException:
> Invalid sequence number for new epoch: 0 (request epoch), 154277489 (seq.
> number)
> {code}
> We run kafka 0.11.0.1 and we ran into the situation when 1 of replication
> threads was stopped for few days, while everything else on that broker was
> functional. This is our staging cluster and retention is less than a day, so
> everything for partitions for which replication thread was down was cleaned
> up. At the moment we have a broker which cannot start replication for few
> partitions. I was also able to reproduce in my local test environment.
> Another possible use case when this might cause real pain is disk failure or
> any situation when previously deleting all the data for the partition on a
> broker helped - since it would just fetch all the data from other replicas.
> Now it does not work for topics with idempotent producers. It might also
> affect other not-idempotent topics if those are unlucky to share same
> replication fetcher thread.
> This seems to be caused by this logic:
> https://github.com/apache/kafka/blob/0.11.0.1/core/src/main/scala/kafka/log/ProducerStateManager.scala#L119
> and might be fixed in the scope of
> https://issues.apache.org/jira/browse/KAFKA-5793.
> However any hints on how to get those partition to fully replicated state are
> highly appreciated.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
[jira] [Commented] (KAFKA-6003) Replication Fetcher thread for a partition with no data fails to start
[
https://issues.apache.org/jira/browse/KAFKA-6003?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16197376#comment-16197376
]
ASF GitHub Bot commented on KAFKA-6003:
---
Github user apurvam closed the pull request at:
https://github.com/apache/kafka/pull/4020
> Replication Fetcher thread for a partition with no data fails to start
> --
>
> Key: KAFKA-6003
> URL: https://issues.apache.org/jira/browse/KAFKA-6003
> Project: Kafka
> Issue Type: Bug
> Components: replication
>Affects Versions: 0.11.0.1
>Reporter: Stanislav Chizhov
>Assignee: Apurva Mehta
>Priority: Blocker
> Fix For: 1.0.0, 0.11.0.2
>
>
> If a partition of a topic with idempotent producer has no data on 1 of the
> brokers, but it does exist on others and some of the segments for this
> partition have been already deleted replication thread responsible for this
> partition on the broker which has no data for it fails to start with out of
> order sequence exception:
> {code}
> [2017-10-02 09:44:23,825] ERROR [ReplicaFetcherThread-2-4]: Error due to
> (kafka.server.ReplicaFetcherThread)
> kafka.common.KafkaException: error processing data for partition
> [stage.data.adevents.v2,20] offset 1660336429
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1$$anonfun$apply$2.apply(AbstractFetcherThread.scala:203)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1$$anonfun$apply$2.apply(AbstractFetcherThread.scala:174)
> at scala.Option.foreach(Option.scala:257)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1.apply(AbstractFetcherThread.scala:174)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2$$anonfun$apply$mcV$sp$1.apply(AbstractFetcherThread.scala:171)
> at
> scala.collection.mutable.ResizableArray$class.foreach(ResizableArray.scala:59)
> at scala.collection.mutable.ArrayBuffer.foreach(ArrayBuffer.scala:48)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply$mcV$sp(AbstractFetcherThread.scala:171)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply(AbstractFetcherThread.scala:171)
> at
> kafka.server.AbstractFetcherThread$$anonfun$processFetchRequest$2.apply(AbstractFetcherThread.scala:171)
> at kafka.utils.CoreUtils$.inLock(CoreUtils.scala:213)
> at
> kafka.server.AbstractFetcherThread.processFetchRequest(AbstractFetcherThread.scala:169)
> at
> kafka.server.AbstractFetcherThread.doWork(AbstractFetcherThread.scala:112)
> at kafka.utils.ShutdownableThread.run(ShutdownableThread.scala:64)
> Caused by: org.apache.kafka.common.errors.OutOfOrderSequenceException:
> Invalid sequence number for new epoch: 0 (request epoch), 154277489 (seq.
> number)
> {code}
> We run kafka 0.11.0.1 and we ran into the situation when 1 of replication
> threads was stopped for few days, while everything else on that broker was
> functional. This is our staging cluster and retention is less than a day, so
> everything for partitions for which replication thread was down was cleaned
> up. At the moment we have a broker which cannot start replication for few
> partitions. I was also able to reproduce in my local test environment.
> Another possible use case when this might cause real pain is disk failure or
> any situation when previously deleting all the data for the partition on a
> broker helped - since it would just fetch all the data from other replicas.
> Now it does not work for topics with idempotent producers. It might also
> affect other not-idempotent topics if those are unlucky to share same
> replication fetcher thread.
> This seems to be caused by this logic:
> https://github.com/apache/kafka/blob/0.11.0.1/core/src/main/scala/kafka/log/ProducerStateManager.scala#L119
> and might be fixed in the scope of
> https://issues.apache.org/jira/browse/KAFKA-5793.
> However any hints on how to get those partition to fully replicated state are
> highly appreciated.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
