[ https://issues.apache.org/jira/browse/CASSANDRA-9706?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14611691#comment-14611691 ]
Benedict commented on CASSANDRA-9706: ------------------------------------- Yep, totally fair. Small 8099 follow up patches aren't a reason to be lazy, especially since I had made mistakes. Pushed with comments, fix and tests. > Precompute ColumnIdentifier comparison > -------------------------------------- > > Key: CASSANDRA-9706 > URL: https://issues.apache.org/jira/browse/CASSANDRA-9706 > Project: Cassandra > Issue Type: Improvement > Components: Core > Reporter: Benedict > Assignee: Benedict > Priority: Minor > Fix For: 3.0 beta 1 > > > Follow up to CASSANDRA-9701. I had hoped to precompute a total order on the > ColumnIdentifier, but decided this would be too risky, with the necessary > periodic rebalancing. So instead, I've hoisted the first 8 bytes of any name > into a long which we can compare to short-circuit all of the expensive work > of ByteBufferUtil.compareUnsigned, making this another very trivial patch (of > debatable necessity to be distinct, but I've already snuck one extra change > in to the previous ticket). -- This message was sent by Atlassian JIRA (v6.3.4#6332)