[
https://issues.apache.org/jira/browse/CASSANDRA-1222?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Edward Capriolo updated CASSANDRA-1222:
---------------------------------------
Attachment: CompactionManager.java
This is just a rough draft of the back off idea.
> Implement adaptive backoff for compaction/repairs
> -------------------------------------------------
>
> Key: CASSANDRA-1222
> URL: https://issues.apache.org/jira/browse/CASSANDRA-1222
> Project: Cassandra
> Issue Type: New Feature
> Reporter: Edward Capriolo
> Attachments: CompactionManager.java
>
>
> When nodes are in repair/streaming/compact they tend use all available IO and
> hurt the performance of read requests. In our use case we do not compact very
> long and we always want reads to have a priority. Our suggestion throttle
> compaction based on the row-read-stage.pending count.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
