cassandra-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Fabien Rousseau (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (CASSANDRA-12015) Rebuilding from another DC should use different sources
Date Sat, 18 Jun 2016 21:25:05 GMT

    [ https://issues.apache.org/jira/browse/CASSANDRA-12015?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15338193#comment-15338193
] 

Fabien Rousseau commented on CASSANDRA-12015:
---------------------------------------------

Hum, I first tried adding a method "sortByDcRack" but it still had the same behaviour...

I finally tried another approach (inspired by NetworkTopologyStrategy.calculateNaturalEndpoints
method) using the TokenMetadata.ringIterator() which returns the next endpoints after a "startToken".

Note: a basic unit test is attached in the patch 

> Rebuilding from another DC should use different sources
> -------------------------------------------------------
>
>                 Key: CASSANDRA-12015
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-12015
>             Project: Cassandra
>          Issue Type: Improvement
>            Reporter: Fabien Rousseau
>         Attachments: 12015-2.1.patch
>
>
> Currently, when adding a new DC (ex: DC2) and rebuilding it from an existing DC (ex:
DC1), only the closest replica is used as a "source of data".
> It works but is not optimal, because in case of an RF=3 and 3 nodes cluster, only one
node in DC1 is streaming the data to DC2. 
> To build the new DC in a reasonable time, it would be better, in that case, to stream
from multiple sources, thus distributing more evenly the load.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Mime
View raw message