[
https://issues.apache.org/jira/browse/HUDI-4096?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17855845#comment-17855845
]
Qijun Fu commented on HUDI-4096:
--------------------------------
I think this [pr |https://github.com/apache/hudi/pull/9651] have already fixed
this?
> Sync timeline from embedded timeline server in flink pipline
> ------------------------------------------------------------
>
> Key: HUDI-4096
> URL: https://issues.apache.org/jira/browse/HUDI-4096
> Project: Apache Hudi
> Issue Type: New Feature
> Components: flink
> Reporter: sivabalan narayanan
> Assignee: Danny Chen
> Priority: Major
>
> At present, in the Flink-Hudi pipeline, each task will scan the meta
> directory to obtain the latest timeline, which will cause frequent get
> listing operations on HDFS and cause a lot of pressure.
> A proposal is we can modify the way to get the timeline in the Flink-Hudi
> pipeline and pull the active timeline through the embedded timeline server.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
