[
https://issues.apache.org/jira/browse/CASSANDRA-4897?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13536859#comment-13536859
]
Sylvain Lebresne commented on CASSANDRA-4897:
---------------------------------------------
Not sure I understand the logic here as this seems to defeat the purpose of
size tiered compaction to me. If you cap the sstable size, then after a short
while, any old enough sstable will have this capped size, and compacting those
will yield same capped sized files. Meaning that after some time, all sstable
will end up in the 'maxSSTableSize' bucket and at this point you'll either:
* have an infinite compaction loop: v1 of the patch
* or stop compacting sstables that reach this max sstable size forever. Which
is, unless I'm missing something, what v2 does due to:
{noformat}
if (avgSize(bucket) > maxSSTableSize * bucketLow)
continue;
{noformat}
which literally mean 'stop compacting files once they've reached
maxSSTableSize. I'm not sure I see how this is much better than the infinite
compaction loop .
> Allow tiered compaction define max sstable size
> -----------------------------------------------
>
> Key: CASSANDRA-4897
> URL: https://issues.apache.org/jira/browse/CASSANDRA-4897
> Project: Cassandra
> Issue Type: Improvement
> Reporter: Radim Kolar
> Assignee: Radim Kolar
> Fix For: 1.2.1
>
> Attachments: cass-maxsize1.txt, cass-maxsize2.txt
>
>
> Lucene is doing same thing. Correctly configured max segment size will
> recycle old data faster with less diskspace.
--
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
