flink yarn 部署指南

主要分紅兩部分，yarn的安裝與flink的安裝， 共3臺機器

10.10.10.125
10.10.10.126
10.10.10.127

----------------------------------------------------yarn 安裝

wget 'http://mirror.bit.edu.cn/apache/hadoop/common/hadoop-2.8.5/hadoop-2.8.5.tar.gz'
tar -zxvf hadoop-2.8.5.tar.gz -C /home/zr/hadoop/

vim /etc/profile
export HADOOP_HOME=/home/zr/hadoop/hadoop-2.8.5
export PATH=$HADOOP_HOME/bin:$PATH
source /etc/profile

vim /etc/sysconfig/network host名千萬不要搞成下劃線 不然就等着完蛋吧
HOSTNAME=flink125（不一樣的機器配置不一樣）

vim /etc/hosts
10.10.10.125 flink125
10.10.10.126 flink126
10.10.10.127 flink127

vim yarn-site.xml 增長如下內容

<property>
<name>yarn.nodemanager.aux-services</name>
<value>mapreduce_shuffle</value>
</property>
<property>
<name>yarn.nodemanager.aux-services.mapreduce.shuffle.class</name>
<value>org.apache.hadoop.mapred.ShuffleHandler</value>
</property>

<property>
<name>yarn.resourcemanager.hostname</name>
<value>flink125</value>
</property>

<property>
<name>yarn.resourcemanager.am.max-attempts</name>
<value>4</value>
<description>The maximum number of application master execution attempts</description>
</property>

<property>
<name>yarn.nodemanager.vmem-check-enabled</name>
<value>false</value>
</property>

vi etc/hadoop/core-site.xml  增長如下內容
<configuration>
<property>
<name>fs.default.name</name>
<value>hdfs://flink125:9000</value>
</property>
</configuration>

vi etc/hadoop/hdfs-site.xml  增長如下內容
<property>
<name>dfs.namenode.name.dir</name>
<value>file:///data/hadoop/storage/hdfs/name</value>
</property>

<property>
<name>dfs.datanode.data.dir</name>
<value>file:///data/hadoop/storage/hdfs/data</value>
</property>

$ vi etc/hadoop/mapred-site.xml 增長如下內容
<property>
<name>mapreduce.framework.name</name>
<value>yarn</value>
</property>

<property>
<name>dfs.replication</name>
<value>2</value>
</property>

vi  slaves  增長如下內容

flink126
flink127

vim hadoop/etc/hadoop/hdoop-env.sh

將語句 export JAVA_HOME=$JAVA_HOME

修改成 export JAVA_HOME=/usr/java/jdk1.8.0_101

以上操做都是在三臺機器上執行

下面是免登操做

1 在125機器上執行

rm -r ~/.ssh

    ssh-keygen

　　scp ~/.ssh/id_rsa.pub 126，127 機器上

2　在126 127上分別執行

cat id_rsa.pub>>.ssh/authorized_keys

最後在125上執行如下操做

啓動hadoop
start-dfs.sh
啓動yarn
start-yarn.sh

----------------------------------------------------yarn 配置完畢

------------------------------------------------------flink 安裝

wget 'http://mirrors.tuna.tsinghua.edu.cn/apache/flink/flink-1.7.2/flink-1.7.2-bin-hadoop28-scala_2.12.tgz'
tar zxvf flink-1.7.2-bin-hadoop28-scala_2.12.tgz -C /home/zr/module/

修改flink/conf/masters，slaves，flink-conf.yaml

vi masters
flink125:8081

flink126:8081

vi slaves
flink126
flink127

vi flink-conf.yaml
taskmanager.numberOfTaskSlots：2
jobmanager.rpc.address: flink125

sudo vi /etc/profile
export FLINK_HOME=/opt/module/flink-1.6.1
export PATH=$PATH:$FLINK_HOME/bin
source /etc/profile

conf/flink-conf.yaml

high-availability: zookeeper
high-availability.zookeeper.quorum: flink125:2181,flink126:2181,flink127:2181
high-availability.storageDir: hdfs:///home/zr/flink/recovery
high-availability.zookeeper.path.root: /home/zr/flink
yarn.application-attempts: 4

zoo.cfg
server.1=flink125:2888:3888
server.2=flink126:2888:3888
server.3=flink127:2888:3888

以上操做都在三臺機器上執行

--------------------------------------------flink 配置完畢

在125機器上執行：

啓動ZooKeeper仲裁

./start-zookeeper-quorum.sh

啓動flink
./flink run -m yarn-cluster -yn 2 -ytm 2048 /home/zr/module/flink-1.7.2/examples/zr/flink-data-sync-0.1.jar

參考文章

http://www.javashuo.com/article/p-pegbsoan-e.html