一、spark-shell

1.把hive的hive-site.xml複製到spark的conf下面

2.開啟spark-shell

bin/spark-shell --master local[2] --jars /opt/datas/mysql-connector-java-5.1.27-bin.jar

3.執行測試語句

spark.sql("select * from imooc.emp").show
spark.sql("show tables from imooc").show
spark.sql("select * from imooc.emp e join imooc.dept d on e.deptno=d.deptno").show
 

二、spark-sql

1.啟動

bin/spark-sql --master local[2] --jars /opt/datas/mysql-connector-java-5.1.27-bin.jar

 2.直接執行sql語句

select * from imooc.emp;
show tables from imooc;
select * from imooc.emp e join imooc.dept d on e.deptno=d.deptno;、

三、thriftserver&beeline

1.啟動thriftserver

sbin/start-thriftserver.sh  \
--master local[2] \
--jars /opt/datas/mysql-connector-java-5.1.27-bin.jar  \
--hiveconf hive.server2.thrift.port=14000 
 

2.啟動beeline

bin/beeline -u jdbc:hive2://bigdata.ibeifeng.com:14000 -n bigdata.ibeifeng.com

其中

“bigdata.ibeifeng.com:14000”是thriftserver設定埠

“bigdata.ibeifeng.com:14000”是伺服器hostname

SparkSQL（二）spark-shell和spark-sql以及thriftserver&beeline訪問hive表