ERROR Org.apache.hadoop.yarn.server.resourcemanager.ResourceManager:RECEIVED SIGNAL 15:sigterm

Rebuild the Hadoop cluster, when everything is ready, after starting the cluster, you can see through the JPS command on master and all the slave nodes that the process that the cluster should start has been started. You can also view all Datanode in live status via Port 50070 and upload the downloaded file to HDFs as normal.

However, on port 8088, no nodes information is visible, the task is submitted to the cluster, and the task is stuck in the accepted state.

The following error is prompted by viewing the log discovery:

2016-03-15 15:51:01,740 ERROR Org.apache.hadoop.yarn.server.resourcemanager.ResourceManager:RECEIVED SIGNAL 15: SIGTERM
2016-03-15 15:51:01,743 ERROR Org.apache.hadoop.security.token.delegation.AbstractDelegationTokenSecretManager:ExpiredTokenRemover received Java.lang.InterruptedException:sleep interrupted
2016-03-15 15:51:01,743 INFO org.mortbay.log:Stopped httpserver2$selectchannelconnectorwithsafestartup@0.0.0.0:8088
2016-03-15 15:51:01,744 ERROR Org.apache.hadoop.security.token.delegation.AbstractDelegationTokenSecretManager:ExpiredTokenRemover received Java.lang.InterruptedException:sleep interrupted
2016-03-15 15:51:01,744 ERROR Org.apache.hadoop.security.token.delegation.AbstractDelegationTokenSecretManager:ExpiredTokenRemover received Java.lang.InterruptedException:sleep interrupted

 
Workaround: Add the following configuration parameters in Yarn-site.xml:

 That is, specify the IP that Yarn.resourcemanager.hostname is the master node.

<property>
<name>yarn.resourcemanager.hostname</name>
<value>192.168.0.100</value>
</property>