[ https://issues.apache.org/jira/browse/CASSANDRA-13263?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16134941#comment-16134941 ]
Sylvain Lebresne commented on CASSANDRA-13263: ---------------------------------------------- Sorry that this kind of felt through the cracks. I'm fine changing that thought the "correct" fix here is to use {{BTree.hashCode}} here, not {{Arrays.hashCode}}. If you change that, I'm +1 on the change, though I'd personally stick to just trunk for this as it's definitively good future-proofing but, I'm relatively confident, doesn't matter with the current code: I don't think that method is called in practice and I see that unlikely to change, at least on 3.X. Definitively won't fight it if it makes you sleep better to have it in 3.0/3.11 though for basically the exact same reasons (and it's a pretty safe change anyway). > Incorrect ComplexColumnData hashCode implementation > --------------------------------------------------- > > Key: CASSANDRA-13263 > URL: https://issues.apache.org/jira/browse/CASSANDRA-13263 > Project: Cassandra > Issue Type: Bug > Reporter: Stefan Podkowinski > Assignee: Stefan Podkowinski > > I went through some of the logs from CASSANDRA-13175 and one of the more > serious issues that we should address seem to be the > {{ComplexColumnData.hashCode()}} implementation. As Objects.hashCode is not > using deep hashing for array arguments, hashed will be based on the identity > instead of the array's content. See patch for simple fix. -- This message was sent by Atlassian JIRA (v6.4.14#64029) --------------------------------------------------------------------- To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org