[
https://issues.apache.org/jira/browse/CASSANDRA-9470?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
C. Scott Andreas updated CASSANDRA-9470:
----------------------------------------
Component/s: Coordination
> Employ MS Negative Acks to report dropped operations to coordinators
> --------------------------------------------------------------------
>
> Key: CASSANDRA-9470
> URL: https://issues.apache.org/jira/browse/CASSANDRA-9470
> Project: Cassandra
> Issue Type: Improvement
> Components: Coordination
> Reporter: Benedict
> Priority: Major
>
> Along with CASSANDRA-9469, we can improve Cassandra's stability under
> overload by immediately reporting to coordinators that mutations or responses
> they are waiting on will not be delivered due to system overload. Initially
> this just permits query failure, reouting or hinting to occur promptly,
> eliminating temporary state on the coordinator. However if failure reasons
> are reported along with the NegAck, we can proactively reroute messages to
> nodes that have had the fewest NegAck's due to overload recently (perhaps as
> a new, or modified, DynamicSnitch)
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org
For additional commands, e-mail: commits-h...@cassandra.apache.org
