You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@cassandra.apache.org by "Aleksey Yeschenko (JIRA)" <ji...@apache.org> on 2014/07/08 14:30:04 UTC
[jira] [Commented] (CASSANDRA-7343) CAS contention back off time
should be configurable
[ https://issues.apache.org/jira/browse/CASSANDRA-7343?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14054884#comment-14054884 ]
Aleksey Yeschenko commented on CASSANDRA-7343:
----------------------------------------------
Makes sense. Although I think I would be right to assume that if you use CAS with a table, you likely do exclusively SERIAL *or* LOCAL_SERIAL.
Would it make sense to make the setting per-table, and not global? If so, we could (probably) get by a single per-table config (not that I mind splitting it into two). But on the global-vs-table-level side the general trend in C* is to pick table-level granularity.
> CAS contention back off time should be configurable
> ----------------------------------------------------
>
> Key: CASSANDRA-7343
> URL: https://issues.apache.org/jira/browse/CASSANDRA-7343
> Project: Cassandra
> Issue Type: Improvement
> Reporter: sankalp kohli
> Assignee: sankalp kohli
> Priority: Minor
> Fix For: 2.0.10
>
> Attachments: cas20-7343.diff, trunk-7343.diff
>
>
> We are currently making the contention call sleep for upto 100 millis. This is not ideal for all situations specially if you are doing LOCAL_SERIAL.
> This value should be configurable based on CL.
--
This message was sent by Atlassian JIRA
(v6.2#6252)