详解Python Celery和RabbitMQ实战教程

pip install celery

brew install rabbitmq
#如果是Ubuntu的话使用apt-get安装

PATH=$PATH:/usr/local/sbin

# 添加用户跟密码
$ rabbitmqctl add_user test test123
# 添加虚拟主机
$ rabbitmqctl add_vhost test_vhost
# 为用户添加标签
$ rabbitmqctl set_user_tags test test_tag
# 设置用户权限
$ rabbitmqctl set_permissions -p test_vhost test ".*" ".*" ".*"

from __future__ import absolute_import
from celery import Celery

app = Celery('test_celery',
broker='amqp://test:test123@localhost/test_vhost',
backend='rpc://',
include=['test_celery.tasks'])

from __future__ import absolute_import
from test_celery.celery import app
import time

@app.task
def add_longtime(a, b):
  print 'long time task begins'
  # sleep 5 seconds
  time.sleep(5)
  print 'long time task finished'
  return a + b

from .tasks import add_longtime
import time

if __name__ == '__main__':
  result = add_longtime.delay(1,2)
	#此时，任务还未完成，它将返回False
  print 'Task finished"htmlcode">

celery -A test_celery worker --loglevel=info


Celery成功连接到RabbitMQ，你会看到这样的东西:



运行任务
再项目文件中输入以下命令运行它：


python -m test_celery.run_tasks


查看Celery控制台，看到运行任务:

[2020-05-15 17:15:21,508: INFO/MainProcess] 

Received task: test_celery.tasks.add_longtime[25ba9c87-69a7-4383-b983-1cefdb32f8b3]

[2020-05-15 17:15:21,508: WARNING/Worker-3] long time task begins

[2020-05-15 17:15:31,510: WARNING/Worker-3] long time task finished

[2020-05-15 17:15:31,512: INFO/MainProcess]

Task test_celery.tasks.add_longtime[25ba9c87-69a7-4383-b983-1cefdb32f8b3] succeeded in 15.003732774s: 3

当Celery收到一个任务，它打印出任务名称与任务id(在括号中):

Received task: test_celery.tasks.add_longtime[7d942984-8ea6-4e4d-8097-225616f797d5]

在这一行下面是我们的任务add_longtime打印的两行，时间延迟为5秒:


long time task begins
long time task finished


最后一行显示我们的任务在5秒内完成，任务结果为3:


Task test_celery.tasks.add_longtime[7d942984-8ea6-4e4d-8097-225616f797d5] succeeded in 5.025242167s: 3


在当前控制台中，您将看到以下输出:



实时监控Celery
Flower是一款基于网络的Celery实时监控软件。使用Flower，可以轻松地监视任务进度和历史记录
使用pip来安装Flower:


pip install flower


要启动Flower web控制台，需要运行以下命令：


celery -A test_celery flower


Flower将运行具有默认端口5555的服务器，可以通过http://localhost:5555访问web控制台