You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@flink.apache.org by "ASF GitHub Bot (Jira)" <ji...@apache.org> on 2022/08/09 06:43:00 UTC
[jira] [Updated] (FLINK-28027) Initialise Async Sink maximum number of in flight messages to low number for rate limiting strategy
[ https://issues.apache.org/jira/browse/FLINK-28027?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
ASF GitHub Bot updated FLINK-28027:
-----------------------------------
Labels: pull-request-available (was: )
> Initialise Async Sink maximum number of in flight messages to low number for rate limiting strategy
> ---------------------------------------------------------------------------------------------------
>
> Key: FLINK-28027
> URL: https://issues.apache.org/jira/browse/FLINK-28027
> Project: Flink
> Issue Type: Bug
> Components: Connectors / Common, Connectors / Kinesis
> Affects Versions: 1.15.0
> Reporter: Zichen Liu
> Priority: Minor
> Labels: pull-request-available
> Fix For: 1.16.0
>
>
> *Background*
> In the AsyncSinkWriter, we implement a rate limiting strategy.
> The initial value for the maximum number of in flight messages is set extremely high ({{{}maxBatchSize * maxInFlightRequests{}}}).
> However, in accordance with the AIMD strategy, the TCP implementation for congestion control has found a small value to start with [is better]([https://en.wikipedia.org/wiki/TCP_congestion_control#Slow_start]).
> *Suggestion*
> A better default might be:
> * maxBatchSize
> * maxBatchSize / parallelism
--
This message was sent by Atlassian Jira
(v8.20.10#820010)