[
https://issues.apache.org/jira/browse/CASSANDRA-8692?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14328036#comment-14328036
]
Benedict commented on CASSANDRA-8692:
-------------------------------------
[~aweisberg] I think it would be preferrable to move all of the
DEBUG_COALESCING logic into CoalescingStrategy; it has the ability to manage
it, and it can all be hidden in our abstract class then (we can always provide
the endpoint and other identifying details on construction. The wasted memory
won't be significant). It also reduces the pollution of existing 2.1 code paths.
One more tiny nit: I'd prefer to use "- 1" instead of "- out.size()" since we
can only have added one item, so it seems more consistent with honouring the
caller's provision of capacity.
> Coalesce intra-cluster network messages
> ---------------------------------------
>
> Key: CASSANDRA-8692
> URL: https://issues.apache.org/jira/browse/CASSANDRA-8692
> Project: Cassandra
> Issue Type: Improvement
> Components: Core
> Reporter: Ariel Weisberg
> Assignee: Ariel Weisberg
> Fix For: 2.1.4
>
> Attachments: batching-benchmark.png
>
>
> While researching CASSANDRA-8457 we found that it is effective and can be
> done without introducing additional latency at low concurrency/throughput.
> The patch from that was used and found to be useful in a real life scenario
> so I propose we implement this in 2.1 in addition to 3.0.
> The change set is a single file and is small enough to be reviewable.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
