• 前往官网下载flink-1.16.2-bin-scala_2.12.tgz 安装包：

        https://archive.apache.org/dist/flink/flink-1.16.2/

• 将安装包上传至服务器的/opt/software目录，解压文件：

tar -zxvf /opt/software/flink-1.16.2-bin-scala_2.12.tgz -C  /opt/software

• 修改masters

hadoop01:8081

• 修改workers

hadoop01
hadoop02
hadoop03

• 修改  flink-conf.yaml

jobmanager.rpc.address: hadoop01
jobmanager.bind-host: 0.0.0.0
taskmanager.bind-host: 0.0.0.0
taskmanager.numberOfTaskSlots: 4
parallelism.default: 4
blob.server.port: 0
blob.server.ssl.enabled: false
blob.service.ssl.enabled: false
blob.service.cleanup.interval: 3600
# ==================== 高可用配置 ====================
high-availability: NONE
rest.bind-address: 0.0.0.0

• 分发集群

scp -r /opt/software/flink-1.16.2/ hadoop02:/opt/software
scp -r /opt/software/flink-1.16.2/ hadoop03:/opt/software

• 配置环境变量

vim /etc/profile.d/my_env.sh 

#添加flink信息
export FLINK_HOME=/opt/software/flink-1.12.7
export PATH=$FLINK_HOME/bin:$PATH

• 分发环境变量

 scp /etc/profile.d/my_env.sh hadoop02:/etc/profile.d/
 scp /etc/profile.d/my_env.sh hadoop03:/etc/profile.d/

• source 环境变量

hadoop01节点：source /etc/profile.d/my_env.sh
hadoop02节点：source /etc/profile.d/my_env.sh
hadoop03节点：source /etc/profile.d/my_env.sh

• 启动flink集群

./opt/software/flink-1.16.2/bin/start-cluster.sh

• 检查服务状态

在各个集群节点输入jps命令查看对应服务是否存在

hadoop01:

hadoop02:

hadoop03:

• 进入web页面

输入地址：http://hadoop01:8081(ip替换为真实ip)

执行jar中指定作业命令（示例）：

flink run -c  com.sa.scala.OracleToPgCDC /opt/software/flink-1.16.2/ks-stream-collect-plugins.jar

完成部署！