[ 
https://issues.apache.org/jira/browse/KAFKA-6634?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16394595#comment-16394595
 ] 

ASF GitHub Bot commented on KAFKA-6634:
---------------------------------------

guozhangwang opened a new pull request #4684: KAFKA-6634: Delay starting new 
transaction in task.initializeTopology
URL: https://github.com/apache/kafka/pull/4684
 
 
   1. As titled, not starting new transaction since during restoration producer 
would have not activity and hence may cause txn expiration.
   1.a. Also delay starting new txn in resuming until initializing topology.
   
   2. Fixed a minor bug, that when resuming process hits a migration exception, 
we should remove that task from the running list if possible.
   
   ### Committer Checklist (excluded from commit message)
   - [ ] Verify design and implementation 
   - [ ] Verify test coverage and CI build status
   - [ ] Verify documentation (including upgrade notes)
   

----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
 
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org


> Delay initiating the txn on producers until initializeTopology with EOS 
> turned on
> ---------------------------------------------------------------------------------
>
>                 Key: KAFKA-6634
>                 URL: https://issues.apache.org/jira/browse/KAFKA-6634
>             Project: Kafka
>          Issue Type: Improvement
>          Components: streams
>            Reporter: Guozhang Wang
>            Assignee: Guozhang Wang
>            Priority: Major
>
> In Streams EOS implementation, the created producers for tasks will initiate 
> a txn immediately after being created in the constructor of `StreamTask`. 
> However, the task may not process any data and hence producer may not send 
> any records for that started txn for a long time because of the restoration 
> process. And with default txn.session.timeout valued at 60 seconds, it means 
> that if the restoration takes more than that amount of time, upon starting 
> the producer will immediately get the error that its producer epoch is 
> already old.
> To fix this, we should consider instantiating the txn only after the 
> restoration phase is done. Although this may have a caveat that if the 
> producer is already fenced, it will not be notified until then, in 
> initializeTopology. But I think this should not be a correctness issue since 
> during the restoration process we do not make any changes to the processing 
> state.



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

Reply via email to