[
https://issues.apache.org/jira/browse/FLINK-32775?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Archit Goyal updated FLINK-32775:
---------------------------------
Description:
Currently with {*}yarn.provided.lib.dirs{*}, Flink libs can be copied to HDFS
location in each cluster and when set Flink tries to reuse the same jars
avoiding uploading it every time and YARN also caches it in the nodes.
This works fine with jars but if we try to add the xml file parent directory to
path, Flink job submission fails. If I add the parent directory of the xml to
the
{noformat}
yarn.ship-files{noformat}
Flink job is submitted successfully.
was:Yarn provide application acls mechanism to be able to provide specific
rights to other users than the one running the job (view logs through the
resourcemanager/job history, kill the application)
> Support yarn.provided.lib.dirs to add parent directory to classpath
> -------------------------------------------------------------------
>
> Key: FLINK-32775
> URL: https://issues.apache.org/jira/browse/FLINK-32775
> Project: Flink
> Issue Type: New Feature
> Components: Deployment / YARN
> Reporter: Archit Goyal
> Assignee: Archit Goyal
> Priority: Minor
> Labels: pull-request-available
>
> Currently with {*}yarn.provided.lib.dirs{*}, Flink libs can be copied to HDFS
> location in each cluster and when set Flink tries to reuse the same jars
> avoiding uploading it every time and YARN also caches it in the nodes.
>
> This works fine with jars but if we try to add the xml file parent directory
> to path, Flink job submission fails. If I add the parent directory of the xml
> to the
> {noformat}
> yarn.ship-files{noformat}
> Flink job is submitted successfully.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
