[ 
https://issues.apache.org/jira/browse/KAFKA-6094?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16213491#comment-16213491
 ] 

Apurva Mehta commented on KAFKA-6094:
-------------------------------------

Hmm. this seems to be an offshoot of the max.in.inflight patch. It appears that 
previously sent batches are expired, causing the producer to bail, since it is 
a fatal error. This is what caused the test failure, from the stack trace. 
Particularly, this is one of the tests which send a decent number of messages 
(10000 in total), so it will span multiple batches.

What is unclear is why there are retries, and why batches should expire. There 
are no broker bounces, and all the brokers are in process. The best I can think 
of is that the jenkins instance was overloaded.

[~junrao] do you know if this was the only test which failed in this run? I 
can't access the results since they were deleted by the time I clicked the link.

> transient failure in testMultipleMarkersOneLeader
> -------------------------------------------------
>
>                 Key: KAFKA-6094
>                 URL: https://issues.apache.org/jira/browse/KAFKA-6094
>             Project: Kafka
>          Issue Type: Bug
>          Components: core
>    Affects Versions: 1.1.0
>            Reporter: Jun Rao
>
> Saw the following transient failure in 
> https://builds.apache.org/job/kafka-pr-jdk7-scala2.11/9005/consoleFull
> kafka.api.TransactionsTest > testMultipleMarkersOneLeader FAILED
>     org.apache.kafka.common.KafkaException: Cannot perform send because at 
> least one previous transactional or idempotent request has failed with errors.
>         at 
> org.apache.kafka.clients.producer.internals.TransactionManager.failIfNotReadyForSend(TransactionManager.java:278)
>         at 
> org.apache.kafka.clients.producer.internals.TransactionManager.maybeAddPartitionToTransaction(TransactionManager.java:263)
>         at 
> org.apache.kafka.clients.producer.KafkaProducer.doSend(KafkaProducer.java:805)
>         at 
> org.apache.kafka.clients.producer.KafkaProducer.send(KafkaProducer.java:761)
>         at 
> org.apache.kafka.clients.producer.KafkaProducer.send(KafkaProducer.java:649)
>         at 
> kafka.api.TransactionsTest$$anonfun$sendTransactionalMessagesWithValueRange$1.apply(TransactionsTest.scala:500)
>         at 
> kafka.api.TransactionsTest$$anonfun$sendTransactionalMessagesWithValueRange$1.apply(TransactionsTest.scala:499)
>         at scala.collection.immutable.Range.foreach(Range.scala:160)
>         at 
> kafka.api.TransactionsTest.sendTransactionalMessagesWithValueRange(TransactionsTest.scala:499)
>         at 
> kafka.api.TransactionsTest.testMultipleMarkersOneLeader(TransactionsTest.scala:475)
>         Caused by:
>         org.apache.kafka.common.errors.TimeoutException: Expiring 213 
> record(s) for largeTopic-8: 10474 ms has passed since batch creation plus 
> linger time



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)

Reply via email to