[ https://issues.apache.org/jira/browse/CASSANDRA-4443?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13423519#comment-13423519 ]
Eric Evans commented on CASSANDRA-4443: --------------------------------------- bq. Eric has some notes up at http://wiki.apache.org/cassandra/VirtualNodes/Balance on implementation but a wiki is a lousy discussion medium so I'll post here: It is, but so is Jira. :) I'm not trying to start a discussion on the wiki, I just wanted to get down the general requirements and brain-storm a bit so that any subsequent discussion (if needed) might be more productive. bq. shuffle should operate on the entire cluster, at least by default. Shuffling node at a time means that for each node i for i in 0..N-1 (where N is the cluster size), i/N of the ranges shuffled will, on average, have been shuffled at least once already. So it's substantially less efficient than shuffling once, then assigning the vnodes out in one cluster-wide pass. Yeah, this is exactly the sort of thing I was trying to capture (and I've added it). > shuffle utility for vnodes > -------------------------- > > Key: CASSANDRA-4443 > URL: https://issues.apache.org/jira/browse/CASSANDRA-4443 > Project: Cassandra > Issue Type: Sub-task > Components: Core > Affects Versions: 1.2 > Reporter: Brandon Williams > Assignee: Eric Evans > Fix For: 1.2 > > > As mentioned on CASSANDRA-4127, for upgrades we need a 'shuffle' command to > split up the contiguous ranges. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira