[ https://issues.apache.org/jira/browse/CASSANDRA-3297?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Jonathan Ellis updated CASSANDRA-3297: -------------------------------------- Attachment: 3297.txt The primary fix here is this: {noformat} + // flush the CF being truncated before forcing the new segment + forceBlockingFlush(); {noformat} Without this, forcing a new segment doesn't help us if the CF-to-truncate was dirty, since its last memtable will be in the new, non-deletable (since it is the last) segment. The rest of the patch does three things: - removes redundant code from RMTruncateTest - fixes CL.resetUnsafe for windows by making it close the segments it's clearing - adds debug logging > truncate can still result in data being replayed after a restart > ---------------------------------------------------------------- > > Key: CASSANDRA-3297 > URL: https://issues.apache.org/jira/browse/CASSANDRA-3297 > Project: Cassandra > Issue Type: Bug > Components: Core > Affects Versions: 0.8.0 > Reporter: Jonathan Ellis > Assignee: Jonathan Ellis > Labels: commitlog > Fix For: 0.8.7, 1.0.0 > > Attachments: 3297.txt > > > Our first stab at fixing this was CASSANDRA-2950. -- 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