[
https://issues.apache.org/jira/browse/CASSANDRA-10816?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15092185#comment-15092185
]
Sylvain Lebresne commented on CASSANDRA-10816:
----------------------------------------------
3.0.2 was a hotfix release that only included CASSANDRA-10822 on top of 3.0.1,
so that patch is definitively not part of it as indicated by the fix-version on
this ticket (it will be part of 3.0.3 however). But I'm not sure what happened
to the changelog, some bad merge added the lines where they shouldn't be
(CASSANDRA-10835 is in the exact same boat). I'll fix the changelog on the 3.0
branch, but 3.0.2 has been release so no point in changing it now so sorry for
the mistake.
> Explicitly handle SSL handshake errors during connect()
> -------------------------------------------------------
>
> Key: CASSANDRA-10816
> URL: https://issues.apache.org/jira/browse/CASSANDRA-10816
> Project: Cassandra
> Issue Type: Bug
> Components: Streaming and Messaging
> Reporter: Stefan Podkowinski
> Assignee: Stefan Podkowinski
> Fix For: 2.2.5, 3.0.3, 3.2
>
>
> Diagnosing internode SSL issues can be a problem as any messaging
> {{IOException}} before this patch is just logged to debug, which is likely
> not enabled in most cases. Logs will just show nothing in case of any
> problems with SSL certs.
> Also the implemented retry semantics in {{OutboundTcpConnection}} will not
> make much sense for SSL handshake errors and will cause unnecessary load for
> both peers.
> The proposed patch will explicitly catch {{SSLHandshakeException}}, log them
> to error and abort connect().
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
