Flink 入门（二）：Mac 上搭建 Flink & Flink 配置项

Flink V1.9 官方中文文档地址：https://ci.apache.org/projects/flink/flink-docs-release-1.9/zh/。

一、 Mac 上安装 Flink

1.1 确保安装了 JDK 1.8+ 版本

Flink 的核心模块使用 Java 开发，因此其运行环境依赖 JDK，且 1.8+ 版本。

查看 java 版本命令：java -version

$ java -version
java version "1.8.0_171"
Java(TM) SE Runtime Environment (build 1.8.0_171-b11)
Java HotSpot(TM) 64-Bit Server VM (build 25.171-b11, mixed mode)

1.2 安装方法

直接使用 Mac 自带的 Homebrew 工具安装 Flink 即可：
命令：brew install apache-flink

# 默认安装最新版的 flink，安装路径默认/usr/local/Cellar/apache-flink/版本号
brew install apache-flink

......

######################################################################    98.5%
######################################################################## 100.0%
  /usr/local/Cellar/apache-flink/1.9.1: 166 files, 277MB, built in 176 minutes 47 seconds
==> `brew cleanup` has not been run in 30 days, running now...
Removing: /Users/ycaha/Library/Caches/Homebrew/telnet--60.high_sierra.bottle.tar.gz... (52.7KB)
Removing: /Users/ycaha/Library/Caches/Homebrew/Cask/jd-gui--1.6.0.tar... (1.4MB)

1.3 测试安装是否成功 ？

然后测试是否安装成功：
命令：flink --version 或者 brew info apache-flink

$ flink --version
Version: 1.9.1, Commit ID: 4d56de8

或者

$ brew info apache-flink
apache-flink: stable 1.9.1, HEAD
Scalable batch and stream data processing
https://flink.apache.org/
/usr/local/Cellar/apache-flink/1.9.1 (167 files, 277MB) *
  Built from source on 2019-11-26 at 14:16:21
From: https://github.com/Homebrew/homebrew-core/blob/master/Formula/apache-flink.rb
==> Requirements
Required: java = 1.8 ✔
==> Options
--HEAD
    Install HEAD version
==> Analytics
install: 1,585 (30 days), 4,592 (90 days), 18,565 (365 days)
install_on_request: 1,574 (30 days), 4,572 (90 days), 18,411 (365 days)
build_error: 0 (30 days)

---

二、启动 Flink

使用 Mac 的 Homebrew 工具安装的 flink 默认在路径/usr/local/Cellar/apache-flink/1.9.1，其中1.9.1 是版本号。通过脚本来启动 flink，在/usr/local/Cellar/apache-flink/1.9.1/libexec/bin/目录下存在很多脚本文件：

ll
total 176
drwxr-xr-x  24 ycaha  admin  768 Nov 26 14:16 ./
drwxr-xr-x  11 ycaha  admin  352 Nov 26 14:16 ../
-rwxr-xr-x   1 ycaha  admin  139 Nov 26 14:16 config.sh*
-rwxr-xr-x   1 ycaha  admin  148 Nov 26 14:16 find-flink-home.sh*
-rwxr-xr-x   1 ycaha  admin  135 Nov 26 14:16 flink*
-rwxr-xr-x   1 ycaha  admin  146 Nov 26 14:16 flink-console.sh*
-rwxr-xr-x   1 ycaha  admin  145 Nov 26 14:16 flink-daemon.sh*
-rwxr-xr-x   1 ycaha  admin  146 Nov 26 14:16 historyserver.sh*
-rwxr-xr-x   1 ycaha  admin  143 Nov 26 14:16 jobmanager.sh*
-rwxr-xr-x   1 ycaha  admin  152 Nov 26 14:16 mesos-appmaster-job.sh*
-rwxr-xr-x   1 ycaha  admin  148 Nov 26 14:16 mesos-appmaster.sh*
-rwxr-xr-x   1 ycaha  admin  150 Nov 26 14:16 mesos-taskmanager.sh*
-rwxr-xr-x   1 ycaha  admin  155 Nov 26 14:16 pyflink-gateway-server.sh*
-rwxr-xr-x   1 ycaha  admin  146 Nov 26 14:16 pyflink-shell.sh*
-rwxr-xr-x   1 ycaha  admin  143 Nov 26 14:16 sql-client.sh*
-rwxr-xr-x   1 ycaha  admin  147 Nov 26 14:16 standalone-job.sh*
-rwxr-xr-x   1 ycaha  admin  146 Nov 26 14:16 start-cluster.sh*
-rwxr-xr-x   1 ycaha  admin  150 Nov 26 14:16 start-scala-shell.sh*
-rwxr-xr-x   1 ycaha  admin  155 Nov 26 14:16 start-zookeeper-quorum.sh*
-rwxr-xr-x   1 ycaha  admin  145 Nov 26 14:16 stop-cluster.sh*
-rwxr-xr-x   1 ycaha  admin  154 Nov 26 14:16 stop-zookeeper-quorum.sh*
-rwxr-xr-x   1 ycaha  admin  144 Nov 26 14:16 taskmanager.sh*
-rwxr-xr-x   1 ycaha  admin  145 Nov 26 14:16 yarn-session.sh*
-rwxr-xr-x   1 ycaha  admin  142 Nov 26 14:16 zookeeper.sh*

可以通过运行脚本文件 start-cluster.sh 来启动 flink 集群：

$ sh start-cluster.sh 
Starting cluster.
Starting standalonesession daemon on host LMA23004071M.local.
Starting taskexecutor daemon on host LMA23004071M.local.

通过 ps -ef | grep flink命令可以查到运行的 flink 进程。
启动 Flink 服务后，client 还可以通过 REST接口连接到 flink server，直接在浏览器中输入：http://localhost:8081/#/overview 即可，其中localhost 是 rest.address，8081 是rest.port。

Flink web界面

rest.address和rest.port的默认值分别为localhost 和 8081。这两个参数可以在配置文件${flink_home}/libexec/conf/flink-conf.yaml 中修改，然后重启 flink server（通过先后执行脚本 start-cluster.sh 和 stop-cluster.sh）使得配置生效。

#==============================================================================
# Rest & web frontend
#==============================================================================

# The port to which the REST client connects to. If rest.bind-port has
# not been specified, then the server will bind to this port as well.
#
# rest.port: 8081

# The address to which the REST client will connect to
#
#rest.address: 0.0.0.0

---