You are viewing a plain text version of this content. The canonical link for it is here.
Posted to jira@kafka.apache.org by "Guozhang Wang (Jira)" <ji...@apache.org> on 2020/04/28 22:24:00 UTC
[jira] [Commented] (KAFKA-9928) Flaky
GlobalKTableEOSIntegrationTest#shouldKStreamGlobalKTableLeftJoin[exactly_once_beta]
[ https://issues.apache.org/jira/browse/KAFKA-9928?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17094902#comment-17094902 ]
Guozhang Wang commented on KAFKA-9928:
--------------------------------------
I found that for the failed run, around the time when the producer of {{produceTopicValues(streamTopic);}} around line 172 is being closed, the following entries are printed (whereas succeeded runs do not have those), cc [~mjsax]:
{code}
[2020-04-28 15:10:58,458] INFO [Consumer clientId=globalTable-eos-test-shouldKStreamGlobalKTableLeftJoin_exactly_once_beta_-0a424027-ab72-4de4-9d83-58989a76b029-StreamThread-1-consumer, groupId=globalTable-eos-test-shouldKStreamGlobalKTableLeftJoin_exactly_once_beta_] Fetch offset 9 is out of range for partition stream-shouldKStreamGlobalKTableLeftJoin_exactly_once_beta_-0, resetting offset (org.apache.kafka.clients.consumer.internals.Fetcher:1261)
[2020-04-28 15:10:58,458] INFO [Consumer clientId=globalTable-eos-test-shouldKStreamGlobalKTableLeftJoin_exactly_once_beta_-0a424027-ab72-4de4-9d83-58989a76b029-StreamThread-1-consumer, groupId=globalTable-eos-test-shouldKStreamGlobalKTableLeftJoin_exactly_once_beta_] Resetting offset for partition stream-shouldKStreamGlobalKTableLeftJoin_exactly_once_beta_-0 to offset 0. (org.apache.kafka.clients.consumer.internals.SubscriptionState:383)
[2020-04-28 15:10:58,459] INFO [Consumer clientId=globalTable-eos-test-shouldKStreamGlobalKTableLeftJoin_exactly_once_beta_-0a424027-ab72-4de4-9d83-58989a76b029-StreamThread-1-consumer, groupId=globalTable-eos-test-shouldKStreamGlobalKTableLeftJoin_exactly_once_beta_] Fetch offset 9 is out of range for partition stream-shouldKStreamGlobalKTableLeftJoin_exactly_once_beta_-0, resetting offset (org.apache.kafka.clients.consumer.internals.Fetcher:1261)
[2020-04-28 15:10:58,460] INFO [Consumer clientId=globalTable-eos-test-shouldKStreamGlobalKTableLeftJoin_exactly_once_beta_-0a424027-ab72-4de4-9d83-58989a76b029-StreamThread-1-consumer, groupId=globalTable-eos-test-shouldKStreamGlobalKTableLeftJoin_exactly_once_beta_] Resetting offset for partition stream-shouldKStreamGlobalKTableLeftJoin_exactly_once_beta_-0 to offset 0. (org.apache.kafka.clients.consumer.internals.SubscriptionState:383)
[2020-04-28 15:10:58,461] INFO [Consumer clientId=globalTable-eos-test-shouldKStreamGlobalKTableLeftJoin_exactly_once_beta_-0a424027-ab72-4de4-9d83-58989a76b029-StreamThread-1-consumer, groupId=globalTable-eos-test-shouldKStreamGlobalKTableLeftJoin_exactly_once_beta_] Fetch offset 9 is out of range for partition stream-shouldKStreamGlobalKTableLeftJoin_exactly_once_beta_-0, resetting offset (org.apache.kafka.clients.consumer.internals.Fetcher:1261)
[2020-04-28 15:10:58,461] INFO [Consumer clientId=globalTable-eos-test-shouldKStreamGlobalKTableLeftJoin_exactly_once_beta_-0a424027-ab72-4de4-9d83-58989a76b029-StreamThread-1-consumer, groupId=globalTable-eos-test-shouldKStreamGlobalKTableLeftJoin_exactly_once_beta_] Resetting offset for partition stream-shouldKStreamGlobalKTableLeftJoin_exactly_once_beta_-0 to offset 0. (org.apache.kafka.clients.consumer.internals.SubscriptionState:383)
[2020-04-28 15:10:58,566] INFO [Producer clientId=globalTable-eos-test-shouldKStreamGlobalKTableLeftJoin_exactly_once_beta_-0a424027-ab72-4de4-9d83-58989a76b029-StreamThread-1-producer, transactionalId=globalTable-eos-test-shouldKStreamGlobalKTableLeftJoin_exactly_once_beta_-0a424027-ab72-4de4-9d83-58989a76b029-1] Discovered group coordinator localhost:54279 (id: 0 rack: null) (org.apache.kafka.clients.producer.internals.TransactionManager:1525)
[2020-04-28 15:11:00,740] INFO [Controller id=0] Processing automatic preferred replica leader election (kafka.controller.KafkaController:66)
{code}
Note that this CLUSTER only have one broker.
> Flaky GlobalKTableEOSIntegrationTest#shouldKStreamGlobalKTableLeftJoin[exactly_once_beta]
> -----------------------------------------------------------------------------------------
>
> Key: KAFKA-9928
> URL: https://issues.apache.org/jira/browse/KAFKA-9928
> Project: Kafka
> Issue Type: Bug
> Components: streams, unit tests
> Reporter: Guozhang Wang
> Assignee: Matthias J. Sax
> Priority: Major
>
> {code}
> Stacktrace
> java.lang.AssertionError: Condition not met within timeout 30000. waiting for final values
> at org.hamcrest.MatcherAssert.assertThat(MatcherAssert.java:26)
> at org.apache.kafka.test.TestUtils.lambda$waitForCondition$5(TestUtils.java:381)
> at org.apache.kafka.test.TestUtils.retryOnExceptionWithTimeout(TestUtils.java:429)
> at org.apache.kafka.test.TestUtils.retryOnExceptionWithTimeout(TestUtils.java:397)
> at org.apache.kafka.test.TestUtils.waitForCondition(TestUtils.java:380)
> at org.apache.kafka.test.TestUtils.waitForCondition(TestUtils.java:368)
> at org.apache.kafka.streams.integration.GlobalKTableEOSIntegrationTest.shouldKStreamGlobalKTableLeftJoin(GlobalKTableEOSIntegrationTest.java:178)
> {code}
> I looked at the below examples:
> https://builds.apache.org/job/kafka-pr-jdk14-scala2.13/149/testReport/junit/org.apache.kafka.streams.integration/GlobalKTableEOSIntegrationTest/shouldKStreamGlobalKTableLeftJoin_exactly_once_beta_/
> https://builds.apache.org/job/kafka-pr-jdk11-scala2.13/6017/testReport/junit/org.apache.kafka.streams.integration/EosIntegrationTest/shouldNotViolateEosIfOneTaskFailsWithState_exactly_once_beta__2/
> https://builds.apache.org/job/kafka-pr-jdk11-scala2.13/6017/testReport/junit/org.apache.kafka.streams.integration/GlobalKTableEOSIntegrationTest/shouldKStreamGlobalKTableLeftJoin_exactly_once_beta__2/
> And also reproduced the flakiness locally after about 180 runs, and the failed one did not have any obvious different traces compared with the successful ones.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)