此脚本可以用来统计某个端口上连接的IP的数量，统计连接到这一端口的所有IP、最多的IP和次数以及TCP连接状态。
涉及到Python读取网络连接统计信息以及统计计算的一些基本操作。在编写脚本的过程中预先定义了统计信息的数据结构，在向最终结果中添加统计信息时需要用到list去重功能，因此临时创建了一个列表使用set()函数去重。set()函数不能直接add字典类型，因此先将字典转成可哈希的字符串，再将去重后的字符串转成字典。其中字典、列表和集合都属于不可哈希的类型。
此脚本可以用于Windows、Linux以及OSX，其中OSX上运行需要使用root权限（由于psutil的原因），使用时直接使用python运行此脚本文件即可。如果提示‘ImportError’，则使用pip安装所缺的模块，非特权用户使用pip安装模块时需要使用sudo。
脚本内已经设定port为22,可以自己修改代码，使它允许成接收命令行位置参数或者手动输入。
运行效果图如下：
1.使用root用户运行
2.使用非特权用户运行
脚本文件可以通过GitHub获取：
脚本内容如下：
#!/usr/bin/python# encoding: utf-8# -*- coding: utf8 -*-"""Created by PyCharm.File: LinuxBashShellScriptForOps:portStatistics.pyUser: GuodongCreate Date: 2016/10/27Create Time: 10:51Note: Usage: Using user as you want in Linux/Windows system. The python module 'psutil' and 'prettytable' is required, using pip install them. ``` pip install psutil prettytable ``` On OSX this function requires root privileges. # python portStatistics.py Total connections of port 22 is 10. +--------------+-------------------+-------------------+-----------------+--------------+ | Total Counts | Remote IP Address | Established Conns | Time_wait Conns | Others Conns | +--------------+-------------------+-------------------+-----------------+--------------+ | 5 | 10.6.28.46 | 5 | 0 | 0 | | 1 | 10.6.28.35 | 1 | 0 | 0 | | 1 | 10.6.28.27 | 1 | 0 | 0 | | 2 | 10.6.28.135 | 2 | 0 | 0 | | 1 | 10.6.28.125 | 1 | 0 | 0 | +--------------+-------------------+-------------------+-----------------+--------------+ Elapsed time: 0.0104579925537 seconds. # """import psutilimport prettytableimport timestartTime = time.time()port = 22 # ssh -i /etc/ssh/ssh_host_rsa_key root@10.6.28.28# define data structure for each connection, each ip is unique unitipaddress = { 'ipaddress': None, 'counts': 0, 'stat': { 'established': 0, 'time_wait': 0, 'others': 0 }}# define data structure for statisticsstatistics = { 'portIsUsed': False, 'portUsedCounts': 0, 'portPeerList': [ { 'ipaddress': None, 'counts': 0, 'stat': { 'established': 0, 'time_wait': 0, 'others': 0 }, }, ]}tmp_portPeerList = list()portPeerSet = set()netstat = psutil.net_connections()# get all ip address only for statistics datafor i, sconn in enumerate(netstat): if port in sconn.laddr: statistics['portIsUsed'] = True if len(sconn.raddr) != 0: statistics['portUsedCounts'] += 1 ipaddress['ipaddress'] = sconn.raddr[0] tmp_portPeerList.append(str(ipaddress)) # dict() list() set() is unhashable type, collections.Counterfor ip in tmp_portPeerList: portPeerSet.add(str(ip)) # remove duplicated ip address using set()for member in portPeerSet: statistics['portPeerList'].append(eval(member))# add statistics data for each ip addressfor sconn in netstat: if port in sconn.laddr: if len(sconn.raddr) != 0: for i, item in enumerate(statistics['portPeerList']): if item['ipaddress'] == sconn.raddr[0]: statistics['portPeerList'][i]['counts'] += 1 if sconn.status == 'ESTABLISHED': statistics['portPeerList'][i]['stat']['established'] += 1 elif sconn.status == 'TIME_WAIT': statistics['portPeerList'][i]['stat']['time_wait'] += 1 else: statistics['portPeerList'][i]['stat']['others'] += 1# print statistics result using prettytableif statistics['portIsUsed']: print "Total connections of port %s is %d." % (port, statistics['portUsedCounts']) table = prettytable.PrettyTable() table.field_names = ["Total Counts", "Remote IP Address", "Established Conns", "Time_wait Conns", "Others Conns"] for i, ip in enumerate(statistics['portPeerList']): if ip['ipaddress'] is not None: table.add_row([ip['counts'], ip['ipaddress'], ip['stat']['established'], ip['stat']['time_wait'], ip['stat']['others']]) print table.get_string(sortby=table.field_names[1], reversesort=True)else: print 'port %s has no connections, please make sure port is listen or in use.' % portendTime = time.time()print "Elapsed time: %s seconds." % (endTime - startTime) tag:端口统计,python TCP连接数统计,Python统计连接数
--end--