[
https://issues.apache.org/jira/browse/FLINK-25066?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
jocean.shi updated FLINK-25066:
-------------------------------
Description:
if the hdfs-site.xml like this
<property>
<name>dfs.nameservices</name>
<value>cluster1,cluster2</value>
</property>
and the core-site.xml like this
<property>
<name>fs.defaultFS</name>
<value>hdfs://cluster1</value>
</property>
flink only can use cluster1 to resolve dependences when we submit job to
yarn（include yarn session, yarn per job, yarn application-mode）
was:
if the hdfs-site.xml like this
<property>
<name>dfs.nameservices</name>
<value>namenode1,namenode2</value>
</property>
and the core-site.xml like this
<property>
<name>fs.defaultFS</name>
<value>hdfs://namenode1</value>
</property>
flink only can use namenode1 to resolve dependences when we submit job to
yarn（include yarn session, yarn per job, yarn application-mode）
> Support using multi hdfs namenode to resolved dependences when submitting job
> to yarn
> -------------------------------------------------------------------------------------
>
> Key: FLINK-25066
> URL: https://issues.apache.org/jira/browse/FLINK-25066
> Project: Flink
> Issue Type: Improvement
> Components: Client / Job Submission
> Reporter: jocean.shi
> Priority: Not a Priority
>
> if the hdfs-site.xml like this
> <property>
> <name>dfs.nameservices</name>
> <value>cluster1,cluster2</value>
> </property>
>
> and the core-site.xml like this
> <property>
> <name>fs.defaultFS</name>
> <value>hdfs://cluster1</value>
> </property>
>
> flink only can use cluster1 to resolve dependences when we submit job to
> yarn（include yarn session, yarn per job, yarn application-mode）
--
This message was sent by Atlassian Jira
(v8.20.1#820001)
