[Supplement] AirFlow installation to help industrial Internet of Things and industrial big data

[Supplement] AirFlow installation to help industrial Internet of Things and industrial big data

Install directly on node1

1. Install Python

• Install dependencies

  yum -y install zlib zlib-devel bzip2 bzip2-devel ncurses ncurses-devel readline readline-devel openssl openssl-devel openssl-static xz lzma xz-devel  sqlite sqlite-devel gdbm gdbm-devel  tk tk-devel gcc
  yum install mysql-devel -y
  yum install libevent-devel -y
  

• Add Linux users and groups

  # 添加py用户
  useradd py
  # 设置密码 '123456'
  passwd py
  # 创建anaconda安装路径
  mkdir /anaconda
  # 赋予权限
  chown -R py:py /anaconda
  

• Upload and execute the Anaconda installation script

  cd /anaconda
  rz
  chmod u+x Anaconda3-5.3.1-Linux-x86_64.sh
  sh Anaconda3-5.3.1-Linux-x86_64.sh
  

  • Custom installation path

    Anaconda3 will now be installed into this location:
    /root/anaconda3
    
      - Press ENTER to confirm the location
      - Press CTRL-C to abort the installation
      - Or specify a different location below
    
    [/root/anaconda3] >>> /anaconda/anaconda3 
    

• Add to system environment variables

  # 修改环境变量
  vi /root/.bash_profile
  # 添加下面这行
  export PATH=/anaconda/anaconda3/bin:$PATH
  # 刷新
  source /root/.bash_profile
  # 验证
  python -V
  

• Configure pip

  mkdir ~/.pip
  touch ~/.pip/pip.conf
  echo '[global]' >> ~/.pip/pip.conf
  echo 'trusted-host=mirrors.aliyun.com' >> ~/.pip/pip.conf
  echo 'index-url=http://mirrors.aliyun.com/pypi/simple/' >> ~/.pip/pip.conf
  # pip默认是10.x版本，更新pip版本
  pip install PyHamcrest==1.9.0
  pip install --upgrade pip
  # 查看pip版本
  pip -V
  

2. Install AirFlow

• Install

  pip install --ignore-installed PyYAML
  pip install apache-airflow[celery]
  pip install apache-airflow[redis]
  pip install apache-airflow[mysql]
  pip install flower
  pip install celery
  

• verify

  airflow -h
  ll /root/airflow
  

3. Install Redis

• Download and install

  wget https://download.redis.io/releases/redis-4.0.9.tar.gz
  tar zxvf redis-4.0.9.tar.gz -C /opt
  cd /opt/redis-4.0.9
  make
  

• start up

  cp redis.conf src/
  cd src
  nohup /opt/redis-4.0.9/src/redis-server redis.conf > output.log 2>&1 &
  

• verify

  ps -ef | grep redis
  

4. Configure and start AirFlow

• Modify the configuration file: airflow.cfg

  [core]
  #18行:时区
  default_timezone = Asia/Shanghai
  #24行:运行模式
  # SequentialExecutor是单进程顺序执行任务，默认执行器，通常只用于测试
  # LocalExecutor是多进程本地执行任务使用的
  # CeleryExecutor是分布式调度使用（可以单机），生产环境常用
  # DaskExecutor则用于动态任务调度，常用于数据分析
  executor = CeleryExecutor
  #30行:修改元数据使用mysql数据库，默认使用sqlite
  sql_alchemy_conn = mysql://airflow:airflow@localhost/airflow
  
  [webserver]
  #468行:web ui地址和端口
  base_url = http://localhost:8085
  #474行
  default_ui_timezone = Asia/Shanghai
  #480行
  web_server_port = 8085
  
  [celery]
  #735行
  broker_url = redis://localhost:6379/0
  #736
  celery_result_backend = redis://localhost:6379/0
  #743
  result_backend = db+mysql://airflow:airflow@localhost:3306/airflow
  

• Initialize metadata database

  • Enter mysql

    mysql -uroot -p
    set global explicit_defaults_for_timestamp =1;
    exit
    

  • initialization

    airflow db init
    

• Configure web access

  airflow users create --lastname user --firstname admin --username admin --email [email protected] --role Admin --password admin
  

• start up

  # 以后台进程方式，启动服务
  airflow webserver -D
  airflow scheduler -D
  airflow celery flower -D
  airflow celery worker -D
  

• Close【No need to execute】

  # 统一杀掉airflow的相关服务进程命令
  ps -ef|egrep 'scheduler|flower|worker|airflow-webserver'|grep -v grep|awk '{print $2}'|xargs kill -9
  # 下一次启动之前
  rm -f /root/airflow/airflow-*
  

5. Verify AirFlow

• Airflow Web UI：node1:8085

  

• Airflow Celery Web：node1:5555