[
https://issues.apache.org/jira/browse/CASSANDRA-1909?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12975833#action_12975833
]
Gary Dusbabek commented on CASSANDRA-1909:
------------------------------------------
It took me a while to realize that replicate-on-write and counters were
orthogonal concepts. Knowing that puts this ticket in a different light since
it seems that you can have counters that don't use replicate-on-write, in which
case you'd want to rely on normal replication. But then how is a counter that
doesn't use replicate-on-write useful? The example conf doesn't use it.
Either way in the default case (ROW is not being used), we send a ROWTask to
the ROW-Stage for every RowMutation that ends up being mostly busy-work (it
exits as soon as it realizes it's dealing with a non-ROW CF). There ought to
be a way to improve that so as not to penalize normal writes.
> normal replication shouldn't happen on counter CFs.
> ---------------------------------------------------
>
> Key: CASSANDRA-1909
> URL: https://issues.apache.org/jira/browse/CASSANDRA-1909
> Project: Cassandra
> Issue Type: Bug
> Components: Core
> Reporter: Gary Dusbabek
> Assignee: Kelvin Kakugawa
> Fix For: 0.8
>
>
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
