Flink 1.17 standalone 模式安装教程

1. 官网下载

wget https://dlcdn.apache.org/flink/flink-1.17.1/flink-1.17.1-src.tgz

2. 解压

tar -zxvf flink-1.17.1-bin-scala_2.12.tgz -C /data1/

3. 修改配置文件

#打开 flink 安装目录下 conf/flink-conf.yaml

vim conf/flink-conf.yaml

#修改参数

#jobmanage地址

jobmanager.rpc.address: flink-master-01

jobmanager.bind-host: 0.0.0.0

#taskmanage地址
taskmanager.bind-host: 0.0.0.0
taskmanager.host: flink-master-01

#metric监控通信地址

rest.address: flink-master-01

rest.bind-address: 0.0.0.0
# ssh 默认端口不是22的需添加

env.ssh.opts: -p 10241

#rocksdb相关

state.backend.type: rocksdb
state.backend.incremental: true
state.backend.rocksdb.localdir: /data1/flink-1.17.1/dataDir/rocksdb

#修改master文件

vim conf/masters

   flink-master-01:8081

#修改 workers 文件

vim conf/workers

  flink-master-01

  flink-worker-01

  flink-worker-02

4. 将 flink 文件分发到其他worker节点

scp -r -P 10241 /data1/flink-1.17.1 root/flink-worker-01:/data1/

scp -r -P 10241 /data1/flink-1.17.1 root/flink-worker-02:/data1/

# -P指定ssh端口号 默认22

5. 修改 worker 节点 conf/flink-conf.yaml配置

#将 taskmanager.host 和 rest.address 修改成本机ip

taskmanager.host: flink-worker-01

rest.address: flink-worker-01

taskmanager.host: flink-worker-02

rest.address: flink-worker-02

6. 启动 flink 集群

bin/start-cluster.sh

7. 官方测试用例

./bin/flink run examples/streaming/WordCount.jar

8. web ui 查看集群使用状态

flink-worker-01:8081