You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@cassandra.apache.org by "Ariel Weisberg (JIRA)" <ji...@apache.org> on 2018/09/12 15:36:00 UTC
[jira] [Commented] (CASSANDRA-14729) Transient Replication: rapid
transient write protection writes to all uncontacted replicas
[ https://issues.apache.org/jira/browse/CASSANDRA-14729?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16612346#comment-16612346 ]
Ariel Weisberg commented on CASSANDRA-14729:
--------------------------------------------
This refers to {{[AbstractWriteResponseHandler.maybeTryAdditionalReplicas|https://github.com/apache/cassandra/blob/2046c30adec194fb07bc5dd1c31fc19a64e7895c/src/java/org/apache/cassandra/service/AbstractWriteResponseHandler.java#L316]}} which tries all available uncontacted replicas instead of just 1 more necessary replica where we might be missing one.
This would happen 1% of the time even in a healthy scenario.
> Transient Replication: rapid transient write protection writes to all uncontacted replicas
> ------------------------------------------------------------------------------------------
>
> Key: CASSANDRA-14729
> URL: https://issues.apache.org/jira/browse/CASSANDRA-14729
> Project: Cassandra
> Issue Type: Improvement
> Components: Core
> Reporter: Benedict
> Priority: Major
> Labels: performance, transient-replication
> Fix For: 4.x
>
>
> This is wasteful both on a per-node level, but especially at the DC level, as we will now double our inter-DC traffic when there is a struggling node anywhere for the range.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@cassandra.apache.org
For additional commands, e-mail: commits-help@cassandra.apache.org