[
https://issues.apache.org/jira/browse/CASSANDRA-5213?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14078326#comment-14078326
]
Michael Shuler commented on CASSANDRA-5213:
-------------------------------------------
[~yukim] is this still an issue and do you have some version details, perhaps?
> Failed to flush SSTable can cause commit log not to reply
> ---------------------------------------------------------
>
> Key: CASSANDRA-5213
> URL: https://issues.apache.org/jira/browse/CASSANDRA-5213
> Project: Cassandra
> Issue Type: Bug
> Reporter: Yuki Morishita
> Priority: Minor
>
> When flushing failed(e.g. out of disk space), that memtable was kept in
> memtablePendingFlush and might trigger next flush inside MeteredFlusher since
> it would be always calculated in countFlushingBytes.
> If that next flush succeeded(e.g. this time it fit on disk), ReplayPosition
> was recorded at that point, and would cause commit log not to replay for the
> data that was not flushed.
--
This message was sent by Atlassian JIRA
(v6.2#6252)
