[
https://issues.apache.org/jira/browse/CASSANDRA-10276?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14734051#comment-14734051
]
Jeff Jirsa commented on CASSANDRA-10276:
----------------------------------------
If your most recent window/bucket with compactable sstables has more than
max_threshold sstables, but they're sized in such a way that stcs won't combine
any of them, {{SizeTieredCompactionStrategy.mostInterestingBucket}} may return
empty, which then causes {{trimToThreshold}} and {{newestBucket}} to return
empty lists. If you may have older buckets that could otherwise be compacted
in some way, they'll starve. If you check to see if
{{SizeTieredCompactionStrategy.mostInterestingBucket(...).isEmpty()}} , you can
continue to check older time windows/buckets/targets.
This also impacts calculation of {{estimatedRemainingTasks}}
Other than that it looks good.
> With DTCS, do STCS in windows if more than max_threshold sstables
> -----------------------------------------------------------------
>
> Key: CASSANDRA-10276
> URL: https://issues.apache.org/jira/browse/CASSANDRA-10276
> Project: Cassandra
> Issue Type: Sub-task
> Components: Core
> Reporter: Marcus Eriksson
> Assignee: Marcus Eriksson
> Fix For: 3.x, 2.1.x, 2.2.x
>
>
> To avoid constant recompaction of files in big ( > max threshold) DTCS
> windows, we should do STCS of those files.
> Patch here: https://github.com/krummas/cassandra/commits/marcuse/dtcs_stcs
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
