[ https://issues.apache.org/jira/browse/CASSANDRA-3919?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13627124#comment-13627124 ]
Jonathan Ellis commented on CASSANDRA-3919: ------------------------------------------- Why do we only call discardDropped if it has re-added columns? Why is the dropped time the last part of the cell name? Isn't it the value as well? That seems odd. Nit: would prefer to store the values natively as micros rather than fix it up on load into CFMD. Nit 2: {{!cf.metadata().getDroppedColumns().isEmpty()}} could move into isDropped. Leaning towards "we should probably not put this into 1.2 this late in the release cycle," is it going to kill people to wait for 2.0? > Dropping a column should do more than just remove the definition > ---------------------------------------------------------------- > > Key: CASSANDRA-3919 > URL: https://issues.apache.org/jira/browse/CASSANDRA-3919 > Project: Cassandra > Issue Type: Sub-task > Components: Core > Reporter: Jonathan Ellis > Assignee: Aleksey Yeschenko > Labels: compaction, cql > Fix For: 1.2.5 > > > Dropping a column should: > - immediately make it unavailable for {{SELECT}}, including {{SELECT *}} > - eventually (i.e., post-compaction) reclaim the space formerly used by that > column -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira