Python Socket单线程+阻塞模式
Python之旅】第五篇(二):Python Socket单线程+阻塞模式
摘要: 前面第五篇(一)中的一个Socket例子其实就是单线程的,即Server端一次只能接受来自一个Client端的连接,为了更好的说明socket单线程和阻塞模式,下面对前面的例子做修改。 1.单线程+阻塞+交互式 前面的例子是单线程阻塞和非交互式的,现在改写为交互式的,即不会执行一次就结...
前面第五篇(一)中的一个Socket例子其实就是单线程的,即Server端一次只能接受来自一个Client端的连接,为了更好的说明socket单线程和阻塞模式,下面对前面的例子做修改。
1.单线程+阻塞+交互式
前面的例子是单线程阻塞和非交互式的,现在改写为交互式的,即不会执行一次就结束,希望达到的效果是,发送的数据由User输入,然后Server端进行接收。
Server端:与上个例子一样,并没有什么变化
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
|
import socket #导入socket类 HOST = '' #定义侦听本地地址口(多个IP地址情况下),这里表示侦听所有,也可以写成 0.0 . 0.0 PORT = 50007 #Server端开放的服务端口 s = socket.socket(socket.AF_INET, socket.SOCK_STREAM) #选择Socket类型和Socket数据包类型 s.bind((HOST, PORT)) #绑定IP地址和端口 s.listen( 1 ) #定义侦听数开始侦听(实际上并没有效果) conn, addr = s.accept() #定义实例,accept()函数的返回值可以看上面的socket函数说明 print 'Connected by' , addr while 1 : data = conn.recv( 1024 ) #接受套接字的数据 if not data: break #如果没有数据接收,则断开连接 print 'revc:' ,data #发送接收到的数据 conn.sendall(data) #发送接收到的数据 conn.close() #关闭套接字 |
Client端:
1
2
3
4
5
6
7
8
9
10
11
12
13
14
|
import socket HOST = '192.168.1.13' PORT = 50007 s = socket.socket(socket.AF_INET, socket.SOCK_STREAM) s.connect((HOST, PORT)) while True: user_input = raw_input( 'msg to send:' ).strip() #由User输入要发送的数据 s.sendall(user_input) data = s.recv( 1024 ) print 'Received' , repr(data) s.close() |
演示:
步骤1:Server端运行服务端程序
1
2
|
xpleaf@xpleaf-machine:/mnt/hgfs/Python/day5$ python server4.py ===>光标在此处处于等待状态 |
步骤2:Client A端运行客户端程序
1
2
3
4
5
6
7
8
9
10
|
xpleaf@xpleaf-machine:/mnt/hgfs/Python/day5$ python client4.py msg to send:The first msg. ===>User输入数据 Received 'The first msg.' ===>Server端返回的数据 msg to send:The second msg. Received 'The second msg.' msg to send:The third msg. Received 'The third msg.' msg to send:I'm A. Received "I'm A." msg to send: ===>继续等待User输入数据 |
步骤3:在Server端中观察现象
1
2
3
4
5
6
7
8
|
xpleaf@xpleaf-machine:/mnt/hgfs/Python/day5/[ 2 ]sec_4_ver2(单线程,交互式,阻塞模 一般演示)$ python server4.py Connected by ( '192.168.1.13' , 52645 ) revc: The first msg. ===>接收到用户发送的数据 revc: The second msg. revc: The third msg. revc: I'm A. ===>光标在此处处于等待状态 |
如果此时有另一个Client B端再连接进来,会有下面的情况:
1
2
3
|
xpleaf@xpleaf-machine:/mnt/hgfs/Python/day5$ python client4.py msg to send:I'm B ===>光标在此处处于等待状态 |
这时如果在Client A端断开连接,则服务端也会关闭套接字,Client B端发送的数据仍然无法被Server端接收。
此时服务端即出现阻塞情况,因为服务端还和Client A处于连接状态,无法接收Client B发送的数据,这也说明了此时的Server端是单线程的。
2.单线程+阻塞+交互式的进阶演示
把上面的例子中的代码再做进一步的修改,以使得阻塞模式的现象更加明显。
Server端:
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
|
import socket HOST = '' PORT = 50007 s = socket.socket(socket.AF_INET, socket.SOCK_STREAM) s.bind((HOST, PORT)) s.listen( 1 ) while 1 : conn, addr = s.accept() #在循环中接受Client端连接的请求 print 'Connected by' , addr while True: #再做一个内部的循环 data = conn.recv( 1024 ) print 'Received' ,data if not data: break conn.sendall(data) conn.close() |
Client端:与前面例子的代码一样
1
2
3
4
5
6
7
8
9
10
11
12
13
14
|
import socket HOST = '192.168.1.13' PORT = 50007 s = socket.socket(socket.AF_INET, socket.SOCK_STREAM) s.connect((HOST, PORT)) while True: user_input = raw_input( 'msg to send:' ).strip() s.sendall(user_input) data = s.recv( 1024 ) print 'Received' , repr(data) s.close() |
演示:
步骤1:Server端运行服务端程序
1
2
|
xpleaf@xpleaf-machine:/mnt/hgfs/Python/day5$ python server4.py ===>光标在此处处于等待状态 |
步骤2:Client A端运行客户端程序
1
2
3
4
5
6
|
xpleaf@xpleaf-machine:/mnt/hgfs/Python/day5$ python client4.py msg to send:Hello! Received 'Hello!' msg to send:I'm Client A. Received "I'm Client A." msg to send: ===>继续等待User输入数据 |
步骤3:在Server端中观察现象
1
2
3
4
5
|
xpleaf@xpleaf-machine:/mnt/hgfs/Python/day5$ python server4.py Connected by ( '192.168.1.13' , 52647 ) Received Hello! Received I'm Client A. ===>光标在此处处于等待状态 |
如果此时有另一个Client B端再连接进来,会有下面的情况:
1
2
3
|
xpleaf@xpleaf-machine:/mnt/hgfs/Python/day5$ python client4.py msg to send:I'm Client B. ===>光标在此处处于等待状态 |
Server端的状态依然为:
1
2
3
4
5
|
xpleaf@xpleaf-machine:/mnt/hgfs/Python/day5$ python server4.py Connected by ( '192.168.1.13' , 52647 ) Received Hello! Received I'm Client A. ===>光标在此处处于等待状态 |
这时试图把Client A端断开:
1
2
3
4
5
6
7
8
9
|
xpleaf@xpleaf-machine:/mnt/hgfs/Python/day5$ python client4.py msg to send:Hello! Received 'Hello!' msg to send:I'm Client A. Received "I'm Client A." msg to send:^CTraceback (most recent call last): File "client4.py" , line 10 , in <module> user_input = raw_input( 'msg to send:' ).strip() KeyboardInterrupt |
再看看Server端的情况:
1
2
3
4
5
6
7
8
|
xpleaf@xpleaf-machine:/mnt/hgfs/Python/day5$ python server4.py Connected by ( '192.168.1.13' , 52647 ) Received Hello! Received I'm Client A. Received Connected by ( '192.168.1.13' , 52648 ) Received I'm Client B. ===>成功接收到来自Client B端发送的数据 ===>光标在此处处于等待状态 |
再看看Client B端的情况:
1
2
3
4
|
xpleaf@xpleaf-machine:/mnt/hgfs/Python/day5$ python client4.py msg to send:I'm Client B. Received "I'm Client B." msg to send: ===>光标在此处处于等待状态 |
以上的现象,再根据Server端的程序代码,就可以非常好理解单线程模式和阻塞的细节情况了,在这里是这样的:Server端接受Client A端的连接后,即把接受连接的线程释放,但此时仍然占用接收和发送数据的线程,所以Client B端虽然可以连接上Server端,但数据是无法成功被Server端接收的;当Client A端断开与Server端的连接后,Server端的接收和发送数据的线程立即被释放,之后就可以正常接收来自Client B端发送的数据了。
单线程,即数据的串行发送,会导致阻塞,上面的两个例子就非常好地演示了这个阻塞的过程,如果要解决这个问题,当然在Server端就需要支持多线程,即数据折并发。