When running sql in hive, mapreduce fails to launch the task, check the log that APPmaster cannot find the hadoop path.
Solution:
Add the following codes in mapred-site.xml, the error will be solved.
<configuration> <property> <name>yarn.app.mapreduce.am.env</name> <value>HADOOP_MAPRED_HOME=/opt/apps/hadoop-3.1.1</value> </property> <property> <name>mapreduce.map.env</name> <value>HADOOP_MAPRED_HOME=/opt/apps/hadoop-3.1.1</value> </property> <property> <name>mapreduce.reduce.env</name> <value>HADOOP_MAPRED_HOME=/opt/apps/hadoop-3.1.1</value> </property> </configuration>