[
https://issues.apache.org/jira/browse/KAFKA-8154?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16801515#comment-16801515
]
Rajini Sivaram commented on KAFKA-8154:
---------------------------------------
I was trying to recreate this yesterday, but was unable to. Even though it
looks possible to hit this scenario from the code and we should avoid it, with
the default buffer sizes, I am not sure how we would get there. Do you by any
chance have a way to easily recreate this scenario without an extensive set up?
We need a test to merge a fix, but that could be a just a unit test. I am
basically trying to understand how your deployment was hitting this error
frequently, in particular with the Java->Java case with replica fetcher. I was
trying to recreate on my laptop using Oracle JDK version jdk1.8.0_191. I also
looked at buffer sizes on the docker container which we use for system tests
and that uses OpenJDK v8. In both cases, SSL application buffer size is 16916.
But your exception shows a smaller buffer size of 16384. It will be good to
understand how it got there and whether we have other issues around SSL buffer
resize. Thanks.
> Buffer Overflow exceptions between brokers and with clients
> -----------------------------------------------------------
>
> Key: KAFKA-8154
> URL: https://issues.apache.org/jira/browse/KAFKA-8154
> Project: Kafka
> Issue Type: Bug
> Components: clients
> Affects Versions: 2.1.0
> Reporter: Rajesh Nataraja
> Priority: Major
> Attachments: server.properties.txt
>
>
> https://github.com/apache/kafka/pull/6495
> https://github.com/apache/kafka/pull/5785
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
