[
https://issues.apache.org/jira/browse/CASSANDRA-18200?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17680990#comment-17680990
]
Maciej Sokol edited comment on CASSANDRA-18200 at 1/26/23 1:23 PM:
-------------------------------------------------------------------
[~brandon.williams] From my understanding, prefer_local is only used for the
peers, in this case the node is trying to message itself.
was (Author: JIRAUSER285315):
>From my understanding, prefer_local is only used for the peers, in this case
>the node is trying to message itself.
> Cassandra messaging to self changed behavior
> --------------------------------------------
>
> Key: CASSANDRA-18200
> URL: https://issues.apache.org/jira/browse/CASSANDRA-18200
> Project: Cassandra
> Issue Type: Bug
> Components: Messaging/Internode
> Reporter: Maciej Sokol
> Priority: Normal
> Attachments: patch.txt
>
>
> During testing of Cassandra on AWS, we noticed some behavior changes between
> Cassandra 3.11 and Cassandra 4.0 when it comes to messaging.
> When performing a range query with consistency local_quorum, Cassandra sents
> a request to itself and some peers.
> In case of Cassandra 4.0, it's trying to connect to itself using the
> broadcast_address while in Cassandra 3.11 it's connecting using the local
> address (see
> [https://github.com/apache/cassandra/blob/cassandra-3.11/src/java/org/apache/cassandra/net/OutboundTcpConnectionPool.java#L152].
> This translation seems to be missing in Cassandra 4.X. I think the best place
> to fix it would be here (see attached file):
> [https://github.com/apache/cassandra/blob/cassandra-4.0/src/java/org/apache/cassandra/net/OutboundConnectionSettings.java#L451]
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
