[
https://issues.apache.org/jira/browse/CASSANDRA-16677?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17377298#comment-17377298
]
Berenguer Blasi commented on CASSANDRA-16677:
---------------------------------------------
There are a few test cases that can fail in this test. Idk if the all have a
common root cause. But so far after diving among all the async, SSL and some
netty code I can't pin down yet who is closing that connection. There are so
many asyncs, threads, loose optimistic concurrent assumptions, etc... I don't
see anything obvious in the netty code either but I don't know it thoroughly.
I have been focusing on the outbound closing logic. But now I am focusing on
the inbound side of things where a well placed Thread.sleep() _seems_ to fix it
(still to be confirmed). I am afraid I have only 3 options:
- Download netty sources, add debug and continue the hunt. Worth it?
- Add a Thread.sleep()
- Mark as flaky
I need to think a bit more about this...
> Fix flaky ConnectionTest
> ------------------------
>
> Key: CASSANDRA-16677
> URL: https://issues.apache.org/jira/browse/CASSANDRA-16677
> Project: Cassandra
> Issue Type: Bug
> Components: Messaging/Internode
> Reporter: Brandon Williams
> Assignee: Berenguer Blasi
> Priority: Normal
> Fix For: 4.0, 4.0-rc
>
>
> https://ci-cassandra.apache.org/job/Cassandra-devbranch/785/testReport/junit/org.apache.cassandra.net/ConnectionTest/testMessageDeliveryOnReconnect_compression/
> https://app.circleci.com/pipelines/github/adelapena/cassandra/460/workflows/cf7dcec6-612c-45d1-8471-623bde481dca/jobs/4069
> https://app.circleci.com/pipelines/github/adelapena/cassandra/460/workflows/b750cd38-0263-4b5e-9bb8-a8be98214378/jobs/4065
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
