5.启动spark-shell

1.安装配置JDK
2.安装配置Spark，修改Spark配置文件(两个配置文件spark-env.sh和slaves)
vim spark-env.sh
#指定JAVA_HOME位置
export JAVA_HOME=/usr/local/jdk1.7.0_80
#指定spark老大Master的IP
export SPARK_MASTER_IP=hadoop-01.itcast.cn
#指定spark老大Master的端口
export SPARK_MASTER_PORT=7077




3.在slaves文件中加入所有Work的地址
node3
node4
node5


4.(可选)配置两个Spark Master实现高可靠(首先要配置zookeeper集群，在spark-env.sh添加SPARK_DAEMON_JAVA_OPTS)
export JAVA_HOME=/usr/java/jdk1.7.0_45
export SPARK_DAEMON_JAVA_OPTS="-Dspark.deploy.recoveryMode=ZOOKEEPER -Dspark.deploy.zookeeper.url=192.168.80.10:2181 -Dspark.deploy.zookeeper.dir=/spark"




5.启动spark-shell
bin/spark-shell --master spark://hadoop-01.itcast.cn:7077,hadoop-02.itcast.cn:7077 --executor-memory 512m --total-executor-cores 7


//spark-shell执行WC程序
sc.textFile("hdfs://hadoop-02.itcast.cn:9000/wc").flatMap(_.split(" ")).map((_,1)).reduceByKey(_+_).sortBy(_._2, false).saveAsTextFile("hdfs://hadoop-02.itcast.cn:9000/out05")


bin/spark-submit  --master spark://hadoop-01.itcast.cn:7077,hadoop-02.itcast.cn:7077 --executor-memory 512m --total-executor-cores 7 --class cn.itcast.spark.WordCount /root/spark-1.0-SNAPSHOT.jar hdfs://hadoop-02.itcast.cn:9000/wc hdfs://hadoop-02.itcast.cn:9000/out0001