[
https://issues.apache.org/jira/browse/KAFKA-17217?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Shivsundar R updated KAFKA-17217:
---------------------------------
Description:
In ShareConsumeRequestManager, currently every time we perform a commitSync or
commitAsync, we create one ShareAcknowledge RPC for the same. Here, we can
optimise the number of RPC calls by batching the acknowledgements before the
next poll is invoked per node.
This will ensure that between 2 calls, the acknowledgements are accumulated in
one request per node and then sent during poll, resulting in lesser RPC calls.
was:
In ShareConsumeRequestManager, currently every time we perform a commitSync or
commitAsync, we create one ShareAcknowledge RPC for the same. Here, we can
optimise the number of RPC calls by batching the acknowledgements before the
next poll is invoked per node.
This will ensure that between 2 calls, the acknowledgements are accumulated in
one request per node and then sent during poll, resulting in lesser RPC calls,
> Clients : Optimise batching of requests per node in ShareConsumeRequestManager
> ------------------------------------------------------------------------------
>
> Key: KAFKA-17217
> URL: https://issues.apache.org/jira/browse/KAFKA-17217
> Project: Kafka
> Issue Type: Sub-task
> Reporter: Shivsundar R
> Assignee: Shivsundar R
> Priority: Major
>
> In ShareConsumeRequestManager, currently every time we perform a commitSync
> or commitAsync, we create one ShareAcknowledge RPC for the same. Here, we can
> optimise the number of RPC calls by batching the acknowledgements before the
> next poll is invoked per node.
> This will ensure that between 2 calls, the acknowledgements are accumulated
> in one request per node and then sent during poll, resulting in lesser RPC
> calls.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
