使用python构建httpserver，python+tornado+pika+rabbitmq

一、搭建环境

1.1安装Erlang

     RabbitMQ是基于Erlang的，所以首先必须配置Erlang环境。从Erlang的官网 http://www.erlang.org/download.html 下载最新的erlang安装包，我下载的是otp_src_17.4.tar.gz

然后：

$ tar xvzf otp_src_17.4.tar.gz
$ cd otp_src_17.4
$ ./configure

编译后出现如下图

提示没有wxWidgets和fop，但是问题不大。继续：
$ make
$ sudo make install

1.2 安装python

需要安装一个比较新的Python版本。安装略。

1.3 安装simplejson

需要安装simplejson。从此处下载最新的版本： http://pypi.python.org/pypi/simplejson#downloads 。我下载的版本是 simplejson-3.6.5.tar.gz

$ tar xvzf simplejson-3.6.5.tar.gz
$ cd simplejson-3.6.5
$ sudo python setup.py install

1.4安装RabbitMQ-Server

开始安装RabbitMQ-Server，从此处下载源代码版本的RabbitMQ： http://www.rabbitmq.com/server.html。我下载的版本是 rabbitmq-server-3.4.2.tar.gz

然后：

$ tar xvzf rabbitmq-server-3.4.2.tar.gz
$ cd rabbitmq-server-3.4.2
$ sudomake TARGET_DIR=/usr/local SBIN_DIR=/usr/local/sbin MAN_DIR=/usr/local/manDOC_INSTALL_DIR=/usr/local/man install

在sbin/目录下出现了三个命令：
rabbitmqctl  rabbitmq-env  rabbitmq-server

安装成功。

运行

找到sbin/目录，运行程序：
/usr/local/sbin/rabbitmq-server –detached 

停止程序：
/usr/local/sbin/rabbitmqctl stop

1.5 安装pika

安装pika可以使用pip来进行安装，pip是python的软件管理包

sudo pip install pika

1.6 安装tornado

从此处下载源代码版本的tornado，https://pypi.python.org/pypi/tornado/，我下载的版本是tornado-4.0.2.tar.gz，然后

tar xvzf tornado-4.0.2.tar.gz
cd tornado-4.0.2
python setup.py build
sudo python setup.py install

二、编写简单样例程序

发送端：helloword.py

功能：对外提供http服务，用户通过调用，往rabbitmq队列插入记录，代码如下

import tornado.ioloop import tornado.web import pika class MainHandler(tornado.web.RequestHandler):     def get(self):         #self.write("Hello, world")         connection = pika.BlockingConnection(pika.ConnectionParameters('127.0.0.1'))         channel = connection.channel()         channel.queue_declare(queue='hello')         channel.basic_publish(exchange='', routing_key='hello', body='Hello World!')         print " [x] Sent 'Hello World!'"         connection.close() application = tornado.web.Application([     (r"/", MainHandler), ]) if __name__ == "__main__":     application.listen(9999)     tornado.ioloop.IOLoop.instance().start()

接收端：receive.py

功能：从消息队列取出消息，代码如下

import pika connection = pika.BlockingConnection(pika.ConnectionParameters(                '127.0.0.1')) channel = connection.channel() channel.queue_declare(queue='hello') def callback(ch, method, properties, body):     print " [x] Received %r" % (body,) channel.basic_consume(callback, queue='hello', no_ack=True) print ' [*] Waiting for messages. To exit press CTRL+C' channel.start_consuming()

三、代码调试

3.1开启rabbitmq，一定要开启

3.2启动tornado web服务

另开一个终端，启动web服务

3.3http访问

另开一个终端，发起http请求

3.4查看队列

发起http请求后，调用到helloword.py，往消息队列插入了一个消息，队列名字正如代码中所示为“hello”，到底是不是这样呢，我们可以看一下

3.5接收端接收消息

确实收到了发送的消息

四.远程链接rabbitmq

官网上面给的例子都是在本地使用系统默认的guest用户连接的。没有给出远程连接的例子，于是阅读文档发现：

When the serverfirst starts running, and detects that its database is uninitialised or hasbeen deleted, it initialises a fresh database with the following resources: avirtual host named / a user named guest with a default password of guest, granted full access to the/ virtual host.

也就是刚刚安装好rabbitmq-server,系统会自动创建一个名为“/”的virtual host，同时也会创建一个用户名和密码都是guest的用户，并且应用"/virtual host"的所有访问权限。因此在rabbitmq安装的机器上使用官网给出的例子是没有问题的，正如上面讲述的例子。如果要切换到远程机器访问的话，单纯的修改

connection= pika.BlockingConnection(pika.ConnectionParameters('127.0.0.1'))

是不行的。因为 guest 用户只是被容许从 localhost 访问。官网文档描述如下：

"guest"user can only connect via localhost By default, the guest user is prohibited from connecting to thebroker remotely; it can only connect over a > loopback interface (i.e.localhost). This applies both to AMQP and to any other protocols enabled viaplugins. Any > other users you create will not (by default) be restricted inthis way. This is configured via the loopback_users item in theconfiguration file. If you wish to allow the guest user to connect from a remotehost, you should set the loopback_users configuration item to []. A completerabbitmq.config which does this would look like: [{rabbit,[{loopback_users, []}]}].

默认情况下，使用下面的命令

sudo rabbitmqctl environment

会发现：

{default_permissions,[<<".*">>,<<".*">>,<<".*">>]},   {default_user,<<"guest">>},   {default_user_tags,[administrator]},   {default_vhost,<<"/">>},   {loopback_users,[<<"guest">>]},   {tcp_listeners,[5672]},

远程链接不想使用guest用户，于是我创建了一个moqian.ydd用户，密码为root，然后授予所有权限，使用下面的命令

Sudorabbitmqctl add_user rollen root Sudorabbitmqctl set_user_tags rollen administrator Sudorabbitmqctl set_permissions -p / rollen ".*"".*"".*"

然后使用下面的代码，远程访问

import pika creds=pika.PlainCredentials('moqian.ydd','root') parameters =pika.ConnectionParameters(host = '10.189.193.186', port = 5672 , credentials =creds) connection =pika.BlockingConnection(parameters) channel = connection.channel() channel.queue_declare(queue='hello') def callback(ch, method, properties,body):    print " [x] Received %r" % (body,) channel.basic_consume(callback,queue='hello', no_ack=True) print ' [*] Waiting for messages. Toexit press CTRL+C' channel.start_consuming()

结果如下，收到了消息

sudo python  receive.py  [x] Received 'Hello World!'  [*] Waiting for messages. To exit press CTRL+C

附录：

推荐几个不错的网站

http://www.tornadoweb.cn/documentation#overview

http://tengine.taobao.org/book/

http://demo.pythoner.com/itt2zh/index.html

http://www.rabbitmq.com/tutorials/tutorial-two-python.html

http://www.01happy.com/python-pika-rabbitmq-summary/

http://blog.csdn.net/historyasamirror/article/details/6827870

http://blog.sina.com.cn/s/blog_3f4791a60102vbpb.html