[
https://issues.apache.org/jira/browse/FLINK-37900?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Maximilian Michels reassigned FLINK-37900:
------------------------------------------
Assignee: Zhenqiu Huang
> Support configurable jitter in rocksdb checkpoint file uploader
> ---------------------------------------------------------------
>
> Key: FLINK-37900
> URL: https://issues.apache.org/jira/browse/FLINK-37900
> Project: Flink
> Issue Type: Improvement
> Components: Runtime / State Backends
> Affects Versions: 2.1.0
> Reporter: Zhenqiu Huang
> Assignee: Zhenqiu Huang
> Priority: Minor
> Labels: pull-request-available
>
> For large Flink jobs (with 512 + containers), it will create lots concurrent
> file uploader workload to remote storage. We see some cases that write
> request are rate limited by remote object store. We may add a jitter
> configuration to smooth the requests into a small periods for example 10
> seconds.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
