spark3.1.1安装（spark on yarn）

目录

1 安装scala2.12.12

1.1 下载scala2.12.12

1.2 上传解压

1.3 配置环境变量SCALA_HOME

1.4 验证

1.5 分发解压包配置node2,node3机器

2 其他安装

3 spark 安装配置

3.1 下载上传解压

3.2 配置spark-env.sh

3.3 配置spark-defaults.conf

3.4 配置SPARK_HOME

3.5 验证spark 自带的例子PI

4 spark  JobHistoryServer 配置

4.1 现在hdfs 上创建日志聚合目录

4.2 配置spark-defaults.conf

4.3 配置spark-env.sh

4.4 启动

4.5 验证结果

---

• 环境scala2.12.12
• hadoop3.2.2
• centos7

1 安装scala2.12.12

1.1 下载scala2.12.12

https://www.scala-lang.org/download/2.12.12.html

1.2 上传解压

node1

[liucf@node1 softfile]$ tar -xvf scala-2.12.12.tgz -C ../soft

1.3 配置环境变量SCALA_HOME

[liucf@node1 scala-2.12.12]$ sudo vim /etc/profile

[liucf@node1 scala-2.12.12]$ source /etc/profile

1.4 验证

1.5 分发解压包配置node2,node3机器

scp -r scala-2.12.12 liucf@node2:/home/liucf/soft
scp -r scala-2.12.12 liucf@node3:/home/liucf/soft

sudo scp /etc/profile root@node2:/etc
sudo scp /etc/profile root@node3:/etc

source /etc/profile

scala  安装完成

2 其他安装

hadoop3.2.2 安装参考

https://blog.csdn.net/m0_37813354/article/details/116561918

3 spark 安装配置

3.1 下载上传解压

下载

https://mirrors.bfsu.edu.cn/apache/spark/spark-3.1.1/spark-3.1.1-bin-hadoop3.2.tgz

解压

[liucf@node1 softfile]$ tar -xvf spark-3.1.1-bin-hadoop3.2.tgz -C ../soft

修改解压目录名称

[liucf@node1 soft]$ mv spark-3.1.1-bin-hadoop3.2 spark-3.1.1

3.2 配置spark-env.sh

[liucf@node1 conf]$ cp spark-env.sh.template spark-env.sh
 

export JAVA_HOME=/home/liucf/soft/jdk1.8.0_121
export SCALA_HOME=/home/liucf/soft/scala-2.12.12

# Options read in YARN client/cluster mode
# - SPARK_CONF_DIR, Alternate conf dir. (Default: ${SPARK_HOME}/conf)
export SPARK_CONF_DIR=/home/liucf/soft/spark-3.1.1/conf
# - HADOOP_CONF_DIR, to point Spark towards Hadoop configuration files
export HADOOP_CONF_DIR=/home/liucf/soft/hadoop-3.2.2
# - YARN_CONF_DIR, to point Spark towards YARN configuration files when you use YARN
export YARN_CONF_DIR=/home/liucf/soft/hadoop-3.2.2/etc/hadoop
# - SPARK_EXECUTOR_CORES, Number of cores for the executors (Default: 1).
# - SPARK_EXECUTOR_MEMORY, Memory per Executor (e.g. 1000M, 2G) (Default: 1G)
# - SPARK_DRIVER_MEMORY, Memory for Driver (e.g. 1000M, 2G) (Default: 1G)

3.3 配置spark-defaults.conf

[liucf@node1 conf]$ cp spark-defaults.conf.template spark-defaults.conf

这个配置暂时只修改名称内容暂时不做修改

3.4 配置SPARK_HOME

export SPARK_HOME=/home/liucf/soft/spark-3.1.1
PATH=$SPARK_HOME/bin:$SCALA_HOME/bin:$HBASE_HOME/bin:$HIVE_HOME/bin:$ZOOKEEPER_HOME/bin:$HADOOP_HOME/bin:$JAVA_HOME/bin:$PATH

分发到其他两台机器

 scp -r /etc/profile root@node2:/etc/profile
 scp -r /etc/profile root@node3:/etc/profile

使生效，三台机器都执行

source /etc/profile

3.5 验证spark 自带的例子PI

spark-submit --master yarn --deploy-mode cluster --class org.apache.spark.examples.SparkPi --name SparkPi --num-executors 1 --executor-memory 1g  --executor-cores 1 /home/liucf/soft/spark-3.1.1/examples/jars/spark-examples_2.12-3.1.1.jar 100

yarn 上能看到任务正常执行完毕

日志可以聚合到hdfs 能够使用yarn 上执行的日志

但是spark 自己的历史执行日志是没有被记录下来的。

4 spark  JobHistoryServer 配置

4.1 现在hdfs 上创建日志聚合目录

hdfs dfs -mkdir -p /spark/sparkhistory

后面会使用到不然会报错目录找不到

4.2 配置spark-defaults.conf

# Example:
# spark.master                     spark://master:7077
# spark.eventLog.enabled           true
# spark.eventLog.dir               hdfs://namenode:8021/directory
# spark.serializer                 org.apache.spark.serializer.KryoSerializer
# spark.driver.memory              5g
# spark.executor.extraJavaOptions  -XX:+PrintGCDetails -Dkey=value -Dnumbers="one two three"
spark.yarn.historyServer.address node1:18080
spark.history.ui.port 18080
spark.eventLog.enabled true
spark.eventLog.dir hdfs://node1:8020/spark/sparkhistory
spark.eventLog.compress true

4.3 配置spark-env.sh

增加如下配置

# - SPARK_HISTORY_OPTS, to set config properties only for the history server (e.g. "-Dx=y")
export SPARK_HISTORY_OPTS="-Dspark.history.ui.port=18080 -Dspark.history.retainedApplications=3  -Dspark.history.fs.logDirectory=hdfs://node1:8020/spark/sparkhistory"

注意其中：Dspark.history.fs.logDirectory对应的目录必须是和spark-defaults.conf里配置的 spark.eventLog.dir hdfs://node1:8020/spark/sparkhistory 保持一致。

4.4 启动

/home/liucf/soft/spark-3.1.1/sbin/start-history-server.sh

4.5 验证结果

jps验证

18080 端口页面验证

完成