[
https://issues.apache.org/jira/browse/CASSANDRA-13084?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15795841#comment-15795841
]
Jason Brown commented on CASSANDRA-13084:
-----------------------------------------
As this is not really a bug fix (a perf improvement), I'll consider this for
3.X and trunk. On the whole, lgtm, and I've created branches with the patch to
run on cassci:
||3.X||trunk||
|[branch|https://github.com/jasobrown/cassandra/tree/13084-3.X]|[branch|https://github.com/jasobrown/cassandra/tree/13084-trunk]|
|[dtest|http://cassci.datastax.com/view/Dev/view/jasobrown/job/jasobrown-13084-3.X-dtest/]|[dtest|http://cassci.datastax.com/view/Dev/view/jasobrown/job/jasobrown-13084-trunk-dtest/]|
|[testall|http://cassci.datastax.com/view/Dev/view/jasobrown/job/jasobrown-13084-3.X-testall/]|[testall|http://cassci.datastax.com/view/Dev/view/jasobrown/job/jasobrown-13084-trunk-testall/]|
I'm everything passes, I'll commit
> Avoid byte-array copy when key cache is disabled
> ------------------------------------------------
>
> Key: CASSANDRA-13084
> URL: https://issues.apache.org/jira/browse/CASSANDRA-13084
> Project: Cassandra
> Issue Type: Bug
> Reporter: Marcus Olsson
> Assignee: Marcus Olsson
> Priority: Trivial
> Attachments: CASSANDRA-13084-3.0.patch, CASSANDRA-13084-3.X.patch
>
>
> While profiling a cluster with the key cache disabled some unnecessary heap
> pressure was observed. It seems that even though the key cache is disabled we
> still create the *KeyCacheKey* object which uses _ByteBufferUtil.getArray()_
> to copy the keys byte array.
> I don't believe it's a big issue but it's still unnecessary when not using
> the key cache.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
