[ https://issues.apache.org/jira/browse/CASSANDRA-5077?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Jonathan Ellis updated CASSANDRA-5077: -------------------------------------- Attachment: 5077-close-writers.txt Hmm, this gets messy because now we can have # next-to-last row causes hits the segment threshold so we create a new writer # last row is empty, calls writer.closeAndOpenReader, which errors out since no rows have been written Attached patch splits the logic up into new-segment and end-of-rows sections to handle this. > Simplify CompactionIterable.getReduced > -------------------------------------- > > Key: CASSANDRA-5077 > URL: https://issues.apache.org/jira/browse/CASSANDRA-5077 > Project: Cassandra > Issue Type: Improvement > Reporter: Jonathan Ellis > Assignee: Jonathan Ellis > Fix For: 1.2.1 > > Attachments: 5077-close-writers.txt, 5077.txt, 5077.txt > > > CompactionIterable turns empty rows into null, and > purge-compacted-rows-from-cache logic is duplicated (and different!) across > implementations. Let's simplify to just return empty rows as-is and let > CompactionTask handle the purge logic. -- 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