[Flink] Flink builds clusters in standalone mode

What I use here is Flink1.13.0 Version of ,JDK1.8

Here are two servers , One is used to manage tasks (JobManager), One is used to perform tasks (TaskManager ),

One server for task management , The server that executes the task can expand the node unlimited according to the actual demand ！！！

First go Flink Download the package on the official website  https://archive.apache.org/dist/flink/flink-1.13.0/flink-1.13.0-src.tgz

Then upload the file to liunx System and unzip

  Input instruction  cd flink-1.13.0/conf/  Get into conf Catalog

Input instruction  vim flink-conf.yaml edit conf file , This file is the core configuration file

Change the address here to Manage the task server by yourself IP Save the address and exit

Input vim masters Edit the file , This file is used to specify Flink The console address of , Default localhost

Write to the management task server IP address

Input  vim workers Edit the file , This file is used to configure flink Cluster sub nodes , Default localhost

Write to the execution task server IP Address , Multiple nodes wrap and write multiple

  And then through scp Distribute the unzipped package to the sub server ！

Enter the management task server bin Catalog

Input  ./start-cluster.sh Open clusters

  Enter the service password of the child node to start synchronously

Input  ./stop-cluster.sh Shut down the cluster

Enter the console to view the cluster status  

xxx.xxx.xxx.xxx:8080

attach :

Flink The card slot of each server in can be in conf Configuration in file , The default is 1

If the console cannot be opened, the port may be occupied :

modify conf Change the file to an unused port

  At the same time to modify  masters file

  Just visit the console again ！！！