Flink on Yarn no starting application as amIfStarted exceeds amLimit

Deployment took more than 60 seconds. Pleaseck if the requested resources are available in the YARN cluster

查看 resource manager yarn 日志

no starting application as amIfStarted exceeds amLimit

yarn Scheduler webui

yarn 集群运行着3个任务，其 jobmanager 内存都是指定的 1 G

yarn.scheduler.capacity.maximum-am-resource-percent 控制 yarn 集群 AM 使用资源比例

root@yarn-resourcemanager-0:/etc/hadoop# cat capacity-scheduler.xml

<configuration>

    ...

  <property>
    <name>yarn.scheduler.capacity.maximum-am-resource-percent</name>
    <value>0.1</value>
    <description>
      Maximum percent of resources in the cluster which can be used to run
      application masters i.e. controls number of concurrent running
      applications.
    </description>
  </property>

  ...

</configuration>

3 > 24 * 0.1

因此被限制提交

### 更新 resouremanager | nodemanager capacity-scheduler.xml 
root@yarn-resourcemanager-0:/etc/hadoop# cat capacity-scheduler.xml | grep 0.1
    <value>0.1</value>
...
root@yarn-nodemanager-0:/# sed -i 's/0.1/0.5/g' /opt/hadoop-2.7.4/etc/hadoop/capacity-scheduler.xml
...

### 更新配置
root@yarn-resourcemanager-0:/# yarn rmadmin -refreshQueues
root@yarn-resourcemanager-0:/#

转自：https://blog.csdn.net/weixin_29208327/article/details/113537459
参考：https://stackoverflow.com/questions/33465300/why-does-yarn-job-not-transition-to-running-state
https://hadoop.apache.org/docs/current/hadoop-yarn/hadoop-yarn-site/CapacityScheduler.html
https://blog.csdn.net/weixin_43878293/article/details/98526311