Spark 从零到开发（二）Spark安装和集群搭建

准备工作

首先得安装scala：CentOS7.x 安装scala

伪集群搭建没做过的参考：CentOS7.x Hadoop集群搭建

下载解压

配置

1. 配置环境变量

/etc/profile

export SPARK_HOME=/home/fantj/spark
export PATH=$PATH:$SPARK_HOME/bin
export CLASSPAHT=.:$CLASSPATH:$JAVA_HOME/lib:$JAVA_HOME/jre/lib

2. 配置`/conf/spark-env.sh`

cp spark-env.sh.template spark-env.sh

给尾部添加环境变量：

export JAVA_HOME=/home/fantj/jdk
export SCALA_HOME=/home/fantj/scala
export SPARK_MASTER_IP=s166
export SPARK_WORKER_MEMORY=1g
export HADOOP_CONF_DIR=/home/fantj/hadoop/etc/hadoop

3. 配置`/conf/slaves.conf`

cp slaves.template slaves.conf

新添数据：

spark2
spark3
spark4

同步配置到slave节点

将spark和scala 和配置文件拷贝到每个slave节点。

 1099  scp -r scala-2.11.7 spark-1.5.1-bin-hadoop2.4/ s168:/home/fantj/download/
 1100  scp -r scala-2.11.7 spark-1.5.1-bin-hadoop2.4/ s169:/home/fantj/download/

 1135  scp /etc/profile s167:/etc/profile
 1136  scp /etc/profile s168:/etc/profile
 1137  scp /etc/profile s169:/etc/profile

启动spark

首先得启动hadoop或者只启动hdfs。start-dfs.sh命令。
jps查看并确保主从机的hadoop的dfs都启动后。（主：NameNode，从：DataNode）
在spark的根目录下执行./sbin/start-all.sh，如果想要slave节点也跟着启动，需要做免密码登录。没有做的话可以用相同的命令一个一个节点去启动。

[root@s166 spark]# ./sbin/start-all.sh 
starting org.apache.spark.deploy.master.Master, logging to /home/fantj/download/spark-1.5.1-bin-hadoop2.4/sbin/../logs/spark-root-org.apache.spark.deploy.master.Master-1-s166.out
localhost: starting org.apache.spark.deploy.worker.Worker, logging to /home/fantj/download/spark-1.5.1-bin-hadoop2.4/sbin/../logs/spark-root-org.apache.spark.deploy.worker.Worker-1-s166.out
localhost: starting org.apache.spark.deploy.worker.Worker, logging to /home/fantj/download/spark-1.5.1-bin-hadoop2.4/sbin/../logs/spark-root-org.apache.spark.deploy.worker.Worker-1-s167.out
localhost: starting org.apache.spark.deploy.worker.Worker, logging to /home/fantj/download/spark-1.5.1-bin-hadoop2.4/sbin/../logs/spark-root-org.apache.spark.deploy.worker.Worker-1-s168.out
localhost: starting org.apache.spark.deploy.worker.Worker, logging to /home/fantj/download/spark-1.5.1-bin-hadoop2.4/sbin/../logs/spark-root-org.apache.spark.deploy.worker.Worker-1-s169.out

再查看jps

-------s166 jps -------
1397 NameNode
52854 Worker
1559 SecondaryNameNode
53671 Jps
52719 Master
-------s167 jps -------
1764 DataNode
29092 Jps
28414 Worker
-------s168 jps -------
33921 Worker
1756 DataNode
34063 Jps
-------s169 jps -------
27384 Jps
1754 DataNode
27242 Worker

可以看到，一个Master三个Worker。
然后再访问主节点ip的8080端口。

打开Spark-shell

[root@s166 bin]# spark-shell 
18/07/30 12:34:16 WARN util.NativeCodeLoader: Unable to load native-hadoop library for your platform... using builtin-java classes where applicable
18/07/30 12:34:20 INFO spark.SecurityManager: Changing view acls to: root
18/07/30 12:34:20 INFO spark.SecurityManager: Changing modify acls to: root
18/07/30 12:34:20 INFO spark.SecurityManager: SecurityManager: authentication disabled; ui acls disabled; users with view permissions: Set(root); users with modify permissions: Set(root)
18/07/30 12:34:22 INFO spark.HttpServer: Starting HTTP Server
18/07/30 12:34:23 INFO server.Server: jetty-8.y.z-SNAPSHOT
18/07/30 12:34:23 INFO server.AbstractConnector: Started [email protected]:35005
18/07/30 12:34:23 INFO util.Utils: Successfully started service 'HTTP class server' on port 35005.
...
...
18/07/30 12:38:39 INFO session.SessionState: Created local directory: /tmp/2c350bb0-1297-40d8-a9bd-47446b116bf3_resources
18/07/30 12:38:39 INFO session.SessionState: Created HDFS directory: /tmp/hive/root/2c350bb0-1297-40d8-a9bd-47446b116bf3
18/07/30 12:38:39 INFO session.SessionState: Created local directory: /tmp/root/2c350bb0-1297-40d8-a9bd-47446b116bf3
18/07/30 12:38:40 INFO session.SessionState: Created HDFS directory: /tmp/hive/root/2c350bb0-1297-40d8-a9bd-47446b116bf3/_tmp_space.db
18/07/30 12:38:40 INFO repl.SparkILoop: Created sql context (with Hive support)..
SQL context available as sqlContext.

scala>

这就证明开启成功了，同理访问4040端口。