[
https://issues.apache.org/jira/browse/KAFKA-8543?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16864951#comment-16864951
]
ASF GitHub Bot commented on KAFKA-8543:
---------------------------------------
ZhaoBo02 commented on pull request #6946: [KAFKA-8543]Optimize fetch thread
allocation
URL: https://github.com/apache/kafka/pull/6946
*More detailed description of your change,
if necessary. The PR title and PR message become
the squashed commit message, so use a separate
comment to ping reviewers.*
*Summary of testing strategy (including rationale)
for the feature or bug fix. Unit and/or integration
tests are expected for any behaviour change and
system tests should be considered for larger changes.*
### Committer Checklist (excluded from commit message)
- [ ] Verify design and implementation
- [ ] Verify test coverage and CI build status
- [ ] Verify documentation (including upgrade notes)
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
> Optimize Fetch thread allocation strategy to ensure full utilization of Fetch
> thread resources
> ----------------------------------------------------------------------------------------------
>
> Key: KAFKA-8543
> URL: https://issues.apache.org/jira/browse/KAFKA-8543
> Project: Kafka
> Issue Type: Improvement
> Components: core
> Affects Versions: 2.2.0
> Reporter: zhaobo
> Priority: Major
>
> The Follow TP on the Broker needs to fetch data from the Leader broker. The
> original fetch thread allocation strategy has two areas that can be optimized:
> （1）The configuration file has a parameter "num.consumer.fetchers" which
> limits the number of threads fetching data from a source broker. In fact,
> Fetch resources are not fully used, and only a part of the threads will be
> used with a high probability. When the traffic of the Leader TP increases,
> the fetch capability may be bottlenecked.
> （2）The number of TPs responsible for each Fetch thread may be uneven.
> Improve: We propose a polling allocation method to ensure that Fetch thread
> resources can be fully utilized. And the number of TPs responsible for each
> fetch thread is balanced.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
