yarn上运行flink环境搭建

主要完成hadoop集群搭建和yarn上运行flink

1.搭建hadoop伪集群

主要是搭建hadoop MapReduce(yarn)和HDFS

1.1 下载&配置环境变量

这里下载的hadoop二进制包为 2.7.7,下载后解压到本地，假设是/usr/hadoop/hadoop-2.7.7

#HADOOP VARIABLES START
export HADOOP_INSTALL=/usr/hadoop/hadoop-2.7.7
export HADOOP_HOME=$HADOOP_INSTALL
export PATH=$PATH:$HADOOP_INSTALL/bin
export PATH=$PATH:$HADOOP_INSTALL/sbin
export HADOOP_MAPRED_HOME=$HADOOP_INSTALL
export HADOOP_COMMON_HOME=$HADOOP_INSTALL
export HADOOP_HDFS_HOME=$HADOOP_INSTALL
export YARN_HOME=$HADOOP_INSTALL
export HADOOP_COMMON_LIB_NATIVE_DIR=$HADOOP_INSTALL/lib/native
export HADOOP_OPTS="-Djava.library.path=$HADOOP_INSTALL/lib"
#HADOOP VARIABLES END

1.2 设置ssh

运行命令

ssh localhost

若是出现 “Warning: Permanently added 'localhost' (ECDSA) to the list of known hosts.” 相似的错误则须要作以下配置

$ ssh-keygen -t rsa -P '' -f ~/.ssh/id_rsa # 若是已经生成了公私钥对则跳过改步骤

$ cat ~/.ssh/id_rsa.pub >> ~/.ssh/authorized_keys

$ chmod 0600 ~/.ssh/authorized_keys

1.3 配置hdfs

cd hadoop-2.7.7

vim etc/hadoop/core-site.xml

修改core-site.xml文件内容为

<configuration>
    <property>
        <name>hadoop.tmp.dir</name>
        <value>file:/usr/hadoop/hadoop-2.7.2/tmp</value>
        <description>Abase for other temporary directories.</description>
    </property>
    <property>
        <name>fs.defaultFS</name>
        <value>hdfs://localhost:9000</value>
    </property>
</configuration>

vim etc/hadoop/hdfs-site.xml

修改hdfs-site.xml内容为

<configuration>
    <property>
        <name>dfs.replication</name>
        <value>1</value>
    </property>
    <property>
        <name>dfs.namenode.name.dir</name>
        <value>file:/usr/hadoop/hadoop-2.7.2/tmp/dfs/name</value>
    </property>
    <property>
        <name>dfs.datanode.data.dir</name>
        <value>file:/usr/hadoop/hadoop-2.7.2/tmp/dfs/data</value>
    </property>
</configuration>

注意: 官网的配置只配置了fs.defaultFS和dfs.replication，这样即可以启动起来，可是若没有配置hadoop.tmp.dir参数，则默认使用的临时目录为 /tmp/hadoo-hadoop，而这个目录在重启时有可能被系统清理掉，致使必须从新执行format才行。

vim etc/hadoop/hadoop-env.sh

须要显示的声明JAVA_HOME, 即便环境变量里已经有了JAVA_HOME。不然会报错:JAVA_HOME is not set and could not be found

## 修改此处为jdk的home目录
    export JAVA_HOME=/opt/jdk/jdk1.8

1.4 格式化和启动hdfs

$ bin/hdfs namenode -format

$ sbin/start-dfs.sh

成功启动后能够经过 http://localhost:50070/ 访问hdfs web页面。使用jps查看进程能够看到DataNode、NameNode、SecondaryNameNode 三个进程，若是没有看到NameNode，能够排除下是否是端口有冲突，而后修复core-site.xml中fs.defaultFS配置的端口号重试下。

1.4 配置yarn

vim etc/hadoop/mapred-site.xml

<configuration>
    <property>
        <name>mapreduce.framework.name</name>
        <value>yarn</value>
    </property>
</configuration>

vim etc/hadoop/yarn-site.xml

<configuration>
    <property>
        <name>yarn.nodemanager.aux-services</name>
        <value>mapreduce_shuffle</value>
    </property>
</configuration>

启动yarn

$ sbin/start-yarn.sh

启动后能够经过 http://localhost:8088/ 访问ResourceManager

到此hadoop伪集群已经搭建完毕

1 在yarn上运行flink

flink要下载和hadoop版本对应的flink版本，不然会出现错误， 这里咱们下载 Apache Flink 1.7.2 with Hadoop® 2.7 for Scala 2.11。下载后解压为flink-1.7.2。直接运行以下命令便可:

$ flink-1.7.2/bin/flink run -m yarn-cluster -yn 2 ../my-flink-project-0.1.jar

其中yarn-cluster表示在yarn上运行flink集群, my-flink-project-0.1.jar是本身写的flink程序。

提交后能够经过ResourceManager http://localhost:8088/ 查看yarn任务运行.

3 参考

1. http://hadoop.apache.org/docs/stable/hadoop-project-dist/hadoop-common/SingleCluster.html
2. https://blog.csdn.net/Evankaka/article/details/51612437