[
https://issues.apache.org/jira/browse/CASSANDRA-16148?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17205090#comment-17205090
]
Jordan West commented on CASSANDRA-16148:
-----------------------------------------
[~dcapwell] good catch re: volatile. I had that locally but forgot to commit.
Thank you. Regarding the hot path, it certainly adds more work during mixed
mode. Can you say more about what you mean by "avoid caching when the version
isn't known". If memoization is involved won't we be "caching" the value? Or is
there a short circuit I am not aware of? Otherwise, I think we need to opt for
correctness over performance here.
Will remove the extraneous comment
> Test failures caused by merging CASSANDRA-15833
> -----------------------------------------------
>
> Key: CASSANDRA-16148
> URL: https://issues.apache.org/jira/browse/CASSANDRA-16148
> Project: Cassandra
> Issue Type: Bug
> Components: Cluster/Gossip
> Reporter: Jordan West
> Assignee: Jordan West
> Priority: Normal
>
> Three issues were caused by merging CASSANDRA-15833:
> 1. `GossiperTest#testHaveAnyVersion3Nodes` was failing on trunk:
> https://app.circleci.com/pipelines/github/jrwest/cassandra/53/workflows/95f9f401-1ef8-4b8d-9c64-3703d9669d95/jobs/771
> 2. python dtest ReadRepairTest#test_atomic_writes[blocking] was failing
> 3. In-jvm dtests being worked on as part of CASSANDRA-15977 uncovered an
> issue with how CASSANDRA-15833 changes interacted with in-jvm dtests running
> without {{Feature.GOSSIP}}
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org
For additional commands, e-mail: commits-h...@cassandra.apache.org
