PySpark实战（二）——Spark环境配置

# 下载spark安装包
cd /export/software
wget https://dlcdn.apache.org/spark/spark-3.3.2/spark-3.3.2-bin-hadoop3.tgz

# 解压
tar -zxvf spark-3.3.2-bin-hadoop3.tgz -C /export/servers/

安装Python3环境

下载安装包

wget https://www.python.org/ftp/python/3.8.15/Python-3.8.15.tgz

解压

tar -zxvf Python-3.8.15.tgz -C /export/servers/

配置

# 安装依赖
 yum -y  install zlib-devel bzip2-devel openssl-devel ncurses-devel sqlite-devel readline-devel tk-devel gdbm-devel db4-devel libpcap-devel xz-devel libffi-devel gcc make


# 更新
cd /export/servers/Python-3.8.15
./configure --prefix=/export/servers/Python-3.8.15
make && make install

# 添加软连接，保证每个目录下都能使用
ln -s  /export/servers/Python-3.8.15/bin/pip3.8 /usr/bin/pip3
ln -s /export/servers/Python-3.8.15/bin/python3.8 /usr/bin/python3

# 配置镜像源
 pip3 config --global set global.index-url https://pypi.tuna.tsinghua.edu.cn/simple/

# 检验环境是否成功
cd /export/servers/spark-3.3.2-bin-hadoop3/bin
./spark-submit ../examples/src/main/python/pi.py

Spark环境

配置日志文件

 cd /export/servers/spark-3.3.2-bin-hadoop3/conf

重命名log4j2.properties.template文件为log4j2.properties

修改内容如下

 保存后再执行，就没有多余的输出信息了

进入Spark交互界面

安装配置Spark集群

master机，切换目录，修改配置文件

重名名为spark-env.sh 

# 切换目录
 cd /export/servers/spark-3.3.2-bin-hadoop3/conf

# 修改配置文件，内容如下图
vi spark-env.sh

 # 添加以下内容

PYSPARK_PYTHON=/export/servers/Python-3.8.15/bin/python3
export JAVA_HOME=/export/servers/jdk1.8.0_361
export HADOOP_HOME=/export/servers/hadoop-3.3.3
export HADOOP_CONF_DIR=$HADOOP_HOME/etc/hadoop
SPARK_MASTER_IP=master
SPARK_LOCAL_DIRS=/export/servers/spark-3.3.2-bin-hadoop3

修改配置文件

vi /etc/profile

# 添加以下内容
export SPARK_HOME=/export/servers/spark-3.3.2-bin-hadoop3
export PATH=$PATH:$SPARK_HOME/bin:$SPARK_HOME/sbin

 更新配置文件

修改下面这个文件的名字为workers

 更改里面的内容为

server01

server02

【去三台机器上进行相同的配置】

# 启动Hadoop
cd /export/servers/hadoop-3.3.3/sbin
start-all.sh

# 启动Spark
cd /export/servers/spark-3.3.2-bin-hadoop3/sbin
start-master.sh
start-slaves.sh

# 提交Spark命令
 cd /export/servers/spark-3.3.2-bin-hadoop3/bin
spark-submit --master=spark://master:7077 $SPARK_HOME/examples/src/main/python/pi.py

进入页面就能看到一条已经执行了的任务记录

【上述是在Standalone集群下提交任务，在YARN提交如下所示】

# 执行命令
spark-submit --master yarn --deploy-mode client $SPARK_HOME/examples/src/main/python/pi.py

Hive 和 Spark整合

# master机
# 将Hive的配置文件hive-site.xml复制到spark的conf目录
cd $HIVE_HOME/conf
cp hive-site.xml /export/servers/spark-3.3.2-bin-hadoop3/conf/

# 分发到其他电脑
scp -rq hive-site.xml server01:/export/servers/spark-3.3.2-bin-hadoop3/conf/
scp -rq hive-site.xml server02:/export/servers/spark-3.3.2-bin-hadoop3/conf/

scp /export/servers/apache-hive-2.3.9-bin/lib/HikariCP-2.5.1.jar master:/export/servers/spark-3.3.2-bin-hadoop3/jars
scp /export/servers/apache-hive-2.3.9-bin/lib/HikariCP-2.5.1.jar server01:/export/servers/spark-3.3.2-bin-hadoop3/jars
scp /export/servers/apache-hive-2.3.9-bin/lib/HikariCP-2.5.1.jar server02:/export/servers/spark-3.3.2-bin-hadoop3/jars

# 上传Mysql的JDBC驱动
cd /export/servers/mysql-connector-java-5.1.46
scp mysql-connector-java-5.1.46.jar master:/export/servers/spark-3.3.2-bin-hadoop3/jars
scp mysql-connector-java-5.1.46.jar server01:/export/servers/spark-3.3.2-bin-hadoop3/jars
scp mysql-connector-java-5.1.46.jar server02:/export/servers/spark-3.3.2-bin-hadoop3/jars

【Hadoop引入了一个安全伪装机制，使得Hadoop不允许上层系统直接将实际用户传递到Hadoop层，而是将实际用户传递给一个超级代理，由代理在Hadoop上执行操作】

# 修改配置文件
vi /export/servers/hadoop-3.3.3/etc/hadoop/core-site.xml

添加以下内容，解决User:root is not allowed to impersonate root的错误

    hadoop.proxyuser.root.hosts
    *


    hadoop.proxyuser.roo.groups
    *

修改master的hive-site，添加以下内容

    
        datanucleus.connectionPoolingType
        dbcp
        
            Expects one of [bonecp, dbcp, hikaricp, none].
            Specify connection pool library for datanucleus
        
    

重启集群服务，使用Spark-sql

成功