[
https://issues.apache.org/jira/browse/CASSANDRA-6912?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13957076#comment-13957076
]
Tyler Hobbs commented on CASSANDRA-6912:
----------------------------------------
Overall this looks good. The only part I'm not sure about is running the
SSTableDeletingTask and deleting the data/index files directly in tidy().
Instead of calling deletingTask.run() directly, deletingTask.schedule() should
be used to properly handle failures. I believe there's a similar problem with
calling deleteWithConfirm() directly on the files; see the comment near the top
of SSTableDeletingTask.
Adding a comment to tidy() with your summary would also be good.
> SSTableReader.isReplaced does not allow for safe resource cleanup
> -----------------------------------------------------------------
>
> Key: CASSANDRA-6912
> URL: https://issues.apache.org/jira/browse/CASSANDRA-6912
> Project: Cassandra
> Issue Type: Bug
> Reporter: Benedict
> Assignee: Benedict
> Fix For: 2.1 beta2
>
>
> There are a number of possible race conditions on resource cleanup from the
> use of cloneWithNewSummarySamplingLevel, because the replacement sstable can
> be itself replaced/obsoleted while the prior sstable is still referenced
> (this is actually quite easy with compaction, but can happen in other
> circumstances less commonly).
--
This message was sent by Atlassian JIRA
(v6.2#6252)
