Nameko是Python下的一个微服务框架，小巧简洁，通过RabbitMq消息组件来实现RPC服务

Github:NameKo

一、准备工作

1.RabbitMq 使用docker安装

docker pull rabbitmq：3.7-management
docker run --hostname my-rabbit --name rabbitmq-borg -p 15672:15672 -p 25672:25672 -p 5672:5672 -d rabbitmq:3.7-management

docker 管理页面为 localhost:15672，默认用户名密码 guest、 guest。

2.Nameko安装，直接使用pip3 install nameko安装即可

二、测试代码

1、服务端 service

from nameko.rpc import rpc

class GreetingService:
    name="greeting_service"

    @rpc
    def hello(self,name):
        return name

使用@rpc 装饰器定义RPC服务

2、配置文件 config.yml

AMQP_URI: 'pyamqp://guest:guest@localhost'
WEB_SERVER_ADDRESS: '0.0.0.0:8000'
rpc_exchange: 'nameko-rpc'
max_workers: 10
parent_calls_tracked: 10

LOGGING:
    version: 1
    handlers:
        console:
            class: logging.StreamHandler
    root:
        level: DEBUG
        handlers: [console]

运行服务 nameko run –config config.yml service1

3、客户端 使用

import cProfile
from nameko.standalone.rpc import ClusterRpcProxy
config = {
    'AMQP_URI': "pyamqp://guest:guest@localhost"
}

def test():
    with ClusterRpcProxy(config) as cluster_rpc:
        rs=cluster_rpc.greeting_service.hello("hellø")

if __name__ == '__main__':
    cProfile.run('test()')

4、上面使用cProfile测试性能

      5914 function calls (5783 primitive calls) in 0.076 seconds

   Ordered by: standard name

   ncalls  tottime  percall  cumtime  percall filename:lineno(function)
    27/12    0.000    0.000    0.002    0.000 <frozen importlib._bootstrap>:1009(_handle_fromlist)

1次调用RPC开销在76毫秒

5、封装成http

import json

from nameko.web.handlers import http

class HttpService:
    name = "http_service"

    @http('GET', '/get/<int:value>')
    def get_method(self, request, value):
        return json.dumps({'value': value})

    @http('POST', '/post')
    def do_post(self, request):
        return u"received: {}".format(request.get_data(as_text=True))

6、使用wrt测试，mac使用brew install wrt 安装即可

wrk -t1 -d20s -c100 http://127.0.0.1:8000/get/42

1个线程，持续20秒100个连接

Running 20s test @ http://127.0.0.1:8000/get/42
  1 threads and 100 connections
  Thread Stats   Avg      Stdev     Max   +/- Stdev
    Latency    74.17ms   11.06ms 141.23ms   63.00%
    Req/Sec     1.35k   204.27     1.72k    62.50%
  26934 requests in 20.03s, 3.34MB read
Requests/sec:   1344.62
Transfer/sec:    170.70KB

响应时间和直接RPC请求差不多

QPS并发量1344/sec，只能说一般般吧