• Distributed task processing system,similar to gearman,based on swoole
• High performance / dynamic multi woker process consumption queue to accelerate backend time consuming service
• There is no need to configure a crontab like gearman worker, swoole-jobs is responsible for managing all worker states

• Slower logic in web, such as statistical /email/ SMS / picture processing, etc.
• Support redis/rabbitmq/zeromq or any other queue message store.
• It is more stable and faster than the Yii / laravel framework itself.
• With yii2/phalcon/yaf/ThinkPHP5 integration example, other frameworks can refer to src/Action code.
• yii2 demo
• ThinkPHP5 demo

• job scheduling component based on swoole; distributed task processing system similar to gearman;

• redis/rabbitmq/zeromq and any other queue message store (currently only redis/rabbitmq).

• use swoole process to realize multi process management, the number of processes can be configured, and the worker process will automatically pull up after exiting.

• the number of cycles of child processes can be configured to prevent memory leakage from business code; the default stop command will wait for the child process to exit smoothly.

• support topic features, different job binding different topic;

• each topic starts the corresponding number of sub processes to eliminate the interaction between different topic.

• according to the queue backlog, the sub process starts the process dynamically, and the number of the largest sub processes can be configured.

• support composer, which can be integrated with any framework;

• log file automatic cutting, default maximum 100M, up to 5 log files, prevent log brush full disk;

• backlog, support for nail robot and other news alerts.

git clone https://github.com/kcloze/swoole-jobs.git
cd swoole-jobs

composer install

• git clone https://github.com/kcloze/swoole-jobs.git
• cd swoole-jobs and composer install
• Building a mirror based on the root directory Dockerfile
• docker build -t swoole-jobs .
• docker run -it -v ~/data/code/php:/data swoole-jobs /bin/bash
• After entering the docker container, enter the project directory:php swoole-jobs.php start

1.edit config.php

2.start service
php ./swoole-jobs.php start >> log/system.log 2>&1

3.push jobs
php ./tests/testJobs.php

NAME
      php swoole-jobs - manage swoole-jobs

SYNOPSIS
      php swoole-jobs command [options]
          Manage swoole-jobs daemons.


WORKFLOWS


      help [command]
      Show this help, or workflow help for command.

      status [command]
      Show master and workers status.

      restart
      Stop, then start swoole-jobs master and workers.

      start
      Start swoole-jobs master and workers.

      stop
      Wait all running workers smooth exit, please check swoole-jobs status for a while.

      exit
      Kill all running workers and master PIDs.

* * * * * /usr/local/bin/php /***/swoole-jobs.php start >> /***/log/system.log 2>&1

more

1. According to your own project path, modify： docs/systemd/swoole-jobs.service
2. sudo cp -f systemd/swoole-jobs.service /etc/systemd/system/
3. sudo systemctl --system daemon-reload
4. Service management
#start service
sudo systemctl start swoole-jobs.service
#reload service
sudo systemctl reload swoole-jobs.service
#stop service
sudo systemctl stop swoole-jobs.service

• change log

• If you embed your own framework, you can refer to src/Action code to inherit the abstract class Kcloze\Jobs\Action\BaseAction.
• Various framework services will start slightly different, for specific reference: Code for example/bin projects.

• Bottleneck: redis/rabbitmq queue storage itself and job execution speed

• swoole

qq group：141059677

• If this project really helps you, please click on the top right corner for a star.