[
https://issues.apache.org/jira/browse/CASSANALYTICS-50?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Yifan Cai moved CASSANDRA-20075 to CASSANALYTICS-50:
----------------------------------------------------
Component/s: Reader
Writer
(was: Analytics Library)
Impacts: (was: None)
Key: CASSANALYTICS-50 (was: CASSANDRA-20075)
Platform: (was: All)
Project: Apache Cassandra Analytics (was: Apache Cassandra)
> [Analytics] Add support for vnodes
> ----------------------------------
>
> Key: CASSANALYTICS-50
> URL: https://issues.apache.org/jira/browse/CASSANALYTICS-50
> Project: Apache Cassandra Analytics
> Issue Type: Improvement
> Components: Reader, Writer
> Reporter: James Berragan
> Assignee: Andrew Johnson
> Priority: Normal
> Time Spent: 4h 50m
> Remaining Estimate: 0h
>
> Analytics currently assumes 1 token per node, but most people in the
> community use vnodes by default. We should add support for vnodes, at the
> most basic level we can improve to issue a Spark task per vnode range. After
> that optimizations can be added to more intelligently merge contiguous
> ranges, or sort tasks to avoid the Spark JVM connecting to the entire cluster.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
