基于docker的自制蜜罐

捕捉到的东西地址

https://github.com/yingshang/Legacy-of-intrusion.git

蜜罐地址

https://github.com/yingshang/honeypot.git

我发现使用docker去搞蜜罐是一种挺不错的选择，相对于说什么低交互，中交互的蜜罐系统来说，简直无敌。首先起码可以在安全性能够得到保障，假如被人逃逸了，瞬间就有CVE了。

我聊聊一下蜜罐的架构，一般来说，在整个互联网中存在太多扫描器，一部分是爆破服务，一部分是exp漏洞，那么首先我们要考虑收益的问题，也就是我要攻击者能够很快的攻击到我的蜜罐系统，进而留下攻击痕迹，让我去分析和溯源，进而建立攻击模型。

所以我选择ssh和mysql设置弱密码，让黑客进行暴力破解进来。在一个月前，我分别在不同区域设置了蜜罐系统，让我惊讶的是，在同一天我居然捕捉到相同的攻击样本。在这个过程中，外网攻击的流程一般是使用工具爆破成功后上传样本（挖矿，ddos病毒，自我复制病毒等）。

安装ssh的时候，我对源码进行一些修改，用来记录攻击者爆破的密码

Mar 29 10:57:09 85740b4003de sshd[120]: Username: root, Password: 123123 Mar 29 10:57:09 85740b4003de sshd[120]: Failed password for root from 174.138.56.93 port 45482 ssh2 Mar 29 10:57:09 85740b4003de sshd[120]: error: Received disconnect from 174.138.56.93: 11: Normal Shutdown, Thank you for playing Mar 29 10:57:23 85740b4003de sshd[121]: error: Received disconnect from 223.111.139.211: 11:  Mar 29 10:57:58 85740b4003de sshd[122]: error: Received disconnect from 222.186.30.71: 11:  Mar 29 10:58:49 85740b4003de sshd[123]: Invalid user ubuntu from 159.89.116.97 Mar 29 10:58:49 85740b4003de sshd[123]: input_userauth_request: invalid user ubuntu Mar 29 10:58:49 85740b4003de sshd[123]: Username: ubuntu, Password: ubuntu123 Mar 29 10:58:49 85740b4003de sshd[123]: error: Could not get shadow information for NOUSER Mar 29 10:58:49 85740b4003de sshd[123]: Failed password for invalid user ubuntu from 159.89.116.97 port 57664 ssh2 Mar 29 10:58:49 85740b4003de sshd[123]: error: Received disconnect from 159.89.116.97: 11: Normal Shutdown, Thank you for playing Mar 29 11:02:07 85740b4003de sshd[124]: Invalid user oracle from 46.105.30.20 Mar 29 11:02:07 85740b4003de sshd[124]: input_userauth_request: invalid user oracle Mar 29 11:02:07 85740b4003de sshd[124]: Username: oracle, Password: qwe@123 Mar 29 11:02:07 85740b4003de sshd[124]: error: Could not get shadow information for NOUSER Mar 29 11:02:07 85740b4003de sshd[124]: Failed password for invalid user oracle from 46.105.30.20 port 42954 ssh2 Mar 29 11:02:08 85740b4003de sshd[124]: error: Received disconnect from 46.105.30.20: 11: Normal Shutdown, Thank you for playing Mar 29 11:03:46 85740b4003de sshd[125]: Invalid user support from 103.120.226.12 Mar 29 11:03:46 85740b4003de sshd[125]: input_userauth_request: invalid user support Mar 29 11:03:46 85740b4003de sshd[125]: Username: support, Password: password Mar 29 11:03:46 85740b4003de sshd[125]: error: Could not get shadow information for NOUSER Mar 29 11:03:46 85740b4003de sshd[125]: Failed password for invalid user support from 103.120.226.12 port 48932 ssh2 Mar 29 11:03:46 85740b4003de sshd[125]: error: Received disconnect from 103.120.226.12: 11: Normal Shutdown, Thank you for playing Mar 29 11:05:17 85740b4003de sshd[126]: error: Received disconnect from 36.156.24.94: 11:  Mar 29 11:06:15 85740b4003de sshd[127]: Username: root, Password: uClinux Mar 29 11:06:15 85740b4003de sshd[127]: Failed password for root from 95.189.253.93 port 59721 ssh2 Mar 29 11:06:16 85740b4003de sshd[127]: Username: root, Password: root Mar 29 11:06:16 85740b4003de sshd[127]: Failed password for root from 95.189.253.93 port 59721 ssh2 Mar 29 11:06:16 85740b4003de sshd[127]: Username: root, Password: system Mar 29 11:06:16 85740b4003de sshd[127]: Failed password for root from 95.189.253.93 port 59721 ssh2 Mar 29 11:06:16 85740b4003de sshd[127]: Username: root, Password: 123456 Mar 29 11:06:16 85740b4003de sshd[127]: Accepted password for root from 95.1

还有我在history那里埋了个点

PS1="`whoami`@`hostname`:"'[$PWD]' history USER_IP=`who -u am i 2>/dev/null| awk '{print $NF}'|sed -e 's/[()]//g'` if [ "$USER_IP" = "" ] then USER_IP=`hostname` fi if [ ! -d /usr/operation ] then mkdir /usr/operation chmod 777 /usr/operation fi if [ ! -d /usr/operation/${LOGNAME} ] then mkdir /usr/operation/${LOGNAME} chmod 300 /usr/operation/${LOGNAME} fi export HISTSIZE=4096 DT=`date "+%Y-%m-%d_%H:%M:%S"`  if [ ! -d /usr/operation/${LOGNAME}/${USER_IP} ] then mkdir /usr/operation/${LOGNAME}/${USER_IP} chmod 300 /usr/operation/${LOGNAME}/${USER_IP} fi   export HISTFILE="/usr/operation/${LOGNAME}/${USER_IP}/$DT" chmod 600 /usr/operation/${LOGNAME}/${USER_IP}/** 2>/dev/null

操作日志

root@85740b4003de:/# cd /usr/operation/root/185.234.217.217/ root@85740b4003de:/usr/operation/root/185.234.217.217# ls 2019-03-29_11:06:19 root@85740b4003de:/usr/operation/root/185.234.217.217# cat 2019-03-29_11\:06\:19  /gisdfoewrsfdf sudo /bin/sh  root@85740b4003de:/usr/operation/root/185.234.217.217# 

接着我在dockerfile设置ssh和mysql的密码是123456，不过我没有开启mysql日志记录，主要是我懒~~~~~~~有空在搞

其实在一开始上面这样已经完结的，因为攻击者的日志会通过history来记录，但是后来发现还是太年轻了，像这种使用脚本ssh进来的是不会留有history日志的。所以我进行了改进，干脆监控全局的变化，想到用ossec，但是发现太过笨重了，于是就用python写了个监控脚本用来监控攻击者的行为和保存攻击样本。

import os import datetime import pyinotify import logging import shutil import random import hashlib import string  import sys path = "/usr/share/record/file/"   def mylog():     # 创建一个日志记录器     log = logging.getLogger("test_logger")     log.setLevel(logging.INFO)     # 创建一个日志处理器     ## 这里需要正确填写路径和文件名，拼成一个字符串，最终生成一个log文件     logHandler = logging.FileHandler(filename ="/usr/share/record/error.log")     ## 设置日志级别     logHandler.setLevel(logging.INFO)     # 创建一个日志格式器     formats = logging.Formatter('%(asctime)s %(levelname)s: %(message)s',                 datefmt='[%Y/%m/%d %I:%M:%S]')      # 将日志格式器添加到日志处理器中     logHandler.setFormatter(formats)     # 将日志处理器添加到日志记录器中     log.addHandler(logHandler)     return log   def GetFileMd5(filename):     if not os.path.isfile(filename):         return     myhash = hashlib.md5()     f = open(filename,'rb')     while True:         b = f.read(8096)         if not b :             break         myhash.update(b)     f.close()     return myhash.hexdigest()   def filecp(source,name,type):     day_name = path + datetime.datetime.now().strftime('%Y%m%d')     hour_name = day_name+'/' + datetime.datetime.now().strftime('%H')     if not os.path.exists(day_name):         os.mkdir(day_name)         if not os.path.exists(hour_name):             os.mkdir(hour_name)      try:         source_md5 = GetFileMd5(source)         status = 0          for root, dirs, files in os.walk(path, topdown=False):             for i in files:                 file_md5 = GetFileMd5(os.path.join(root, i))                 if  file_md5 == source_md5:                     status = 1          if status ==0:             fsize = int(os.path.getsize(source))              if fsize != 0:                 now = datetime.datetime.now().strftime("%M-%S")                 token = ''.join(random.sample(string.ascii_letters + string.digits, 8))                 shutil.copy(source,hour_name+'/'+name + "_" +token+"_"+now+"_"+type)     except FileNotFoundError:         pass     except OSError:         pass   class MyEventHandler(pyinotify.ProcessEvent):     logging.basicConfig(level=logging.INFO, filename='/usr/share/record/monitor.log')      logging.info("Starting monitor...")      def process_IN_ACCESS(self, event):         print("ACCESS event:", event.pathname)         logging.info("ACCESS event : %s  %s" % (os.path.join(event.path, event.name), datetime.datetime.now()))      def process_IN_ATTRIB(self, event):         print("ATTRIB event:", event.pathname)         logging.info("IN_ATTRIB event : %s  %s" % (os.path.join(event.path, event.name), datetime.datetime.now())) #        filecp(source=os.path.join(event.path, event.name),name=event.name,type="ATTRIB")   #    def process_IN_CLOSE_NOWRITE(self, event): #        print("CLOSE_NOWRITE event:", event.pathname) #        logging.info("CLOSE_NOWRITE event : %s  %s" % (os.path.join(event.path, event.name), datetime.datetime.now()))   #    def process_IN_CLOSE_WRITE(self, event): #        print("CLOSE_WRITE event:", event.pathname) #        logging.info("CLOSE_WRITE event : %s  %s" % (os.path.join(event.path, event.name), datetime.datetime.now())) #        filecp(source=os.path.join(event.path, event.name),name=event.name,type="CLOSE_WRITE")       def process_IN_CREATE(self, event):         print("CREATE event:", event.pathname)         logging.info("CREATE event : %s  %s" % (os.path.join(event.path, event.name), datetime.datetime.now()))         filecp(source=os.path.join(event.path, event.name),name=event.name,type="CREATE")       def process_IN_DELETE(self, event):         print("DELETE event:", event.pathname)         logging.info("DELETE event : %s  %s" % (os.path.join(event.path, event.name), datetime.datetime.now()))         #filecp(source=os.path.join(event.path, event.name),name=event.name,type="DELETE")       def process_IN_MODIFY(self, event):         print("MODIFY event:", event.pathname)         logging.info("MODIFY event : %s  %s" % (os.path.join(event.path, event.name), datetime.datetime.now()))         if event.name != "null":             filecp(source=os.path.join(event.path, event.name), name=event.name, type="MODIFY")      def process_IN_OPEN(self, event):         print("OPEN event:", event.pathname)         logging.info("OPEN event : %s  %s" % (os.path.join(event.path, event.name), datetime.datetime.now()))   def main():     # watch manager     excl_list = [         '/usr/share/record',         '/var/log',     ]     excl = pyinotify.ExcludeFilter(excl_list)     wm = pyinotify.WatchManager()     wm.add_watch('/tmp', pyinotify.ALL_EVENTS, rec=True,exclude_filter=excl)      eh = MyEventHandler()      # notifier     logger = mylog()      try:         notifier = pyinotify.Notifier(wm, eh)         notifier.loop()     except :         logger.exception(sys.exc_info())         logger.info("Error in log")  if __name__ == '__main__':     main()

下面是监控的日志

INFO:root:OPEN event : /tmp/hsperfdata_root  2019-03-25 06:27:05.780024 INFO:root:ACCESS event : /tmp/hsperfdata_root  2019-03-25 06:27:05.780818 INFO:root:CREATE event : /tmp/2tk8gg28p1pkzsc4i088gdu4zl  2019-03-25 12:07:45.775160 INFO:root:OPEN event : /tmp/2tk8gg28p1pkzsc4i088gdu4zl  2019-03-25 12:07:45.935427 INFO:root:MODIFY event : /tmp/2tk8gg28p1pkzsc4i088gdu4zl  2019-03-25 12:07:46.179364 INFO:root:OPEN event : /tmp/2tk8gg28p1pkzsc4i088gdu4zl  2019-03-25 12:07:46.357906 INFO:root:ACCESS event : /tmp/2tk8gg28p1pkzsc4i088gdu4zl  2019-03-25 12:07:46.358347 INFO:root:OPEN event : /tmp/2tk8gg28p1pkzsc4i088gdu4zl  2019-03-25 12:07:46.358630 INFO:root:ACCESS event : /tmp/2tk8gg28p1pkzsc4i088gdu4zl  2019-03-25 12:07:46.358835 INFO:root:OPEN event : /tmp/2tk8gg28p1pkzsc4i088gdu4zl  2019-03-25 12:07:47.577594 INFO:root:ACCESS event : /tmp/2tk8gg28p1pkzsc4i088gdu4zl  2019-03-25 12:07:47.578082 INFO:root:ACCESS event : /tmp/2tk8gg28p1pkzsc4i088gdu4zl  2019-03-25 12:07:47.578376 INFO:root:DELETE event : /tmp/2tk8gg28p1pkzsc4i088gdu4zl  2019-03-25 12:07:48.699488 INFO:root:CREATE event : /tmp/knrm  2019-03-25 12:07:50.316929 INFO:root:OPEN event : /tmp/knrm  2019-03-25 12:07:50.317885 INFO:root:OPEN event : /tmp/knrm  2019-03-25 12:07:50.318216 INFO:root:MODIFY event : /tmp/knrm  2019-03-25 12:07:51.220061 INFO:root:OPEN event : /tmp/knrm  2019-03-25 12:07:51.221799 INFO:root:ACCESS event : /tmp/knrm  2019-03-25 12:07:51.221942 INFO:root:OPEN event : /tmp/knrm  2019-03-25 12:07:51.222171 INFO:root:ACCESS event : /tmp/knrm  2019-03-25 12:07:51.222359 INFO:root:MODIFY event : /tmp/knrm  2019-03-25 12:07:51.446485 INFO:root:OPEN event : /tmp/knrm  2019-03-25 12:07:51.448138

接下来是安装和运行

docker build -t hon .

运行蜜罐

docker run  -d  -v /record:/usr/share/record  -p 22:22 hon

接下来就等攻击者来，一般512M的VPS都可以带起来的。

有几个构想还没有写，等有空再写把，一个是将蜜罐的日志实时传输到实体机上面，假如被删除或者有什么异动，邮件报警。另外一个就是监控流量，我有一台VPS没有注意看，病毒直接把我的流量都打光了。

转载请注明来自网盾网络安全培训，本文标题：《基于docker的自制蜜罐》

标签：蜜罐系统Docker