Man In the Middle (MITM) ??


Can you please answer a couple of questions about Man In the Middle (MITM) ?

Background:  I'm a Terminal Server (TS) admin for my company, and one of our directors is questioning the security of our TS users coming in thru RDP.  We have "High" 128 encryption enabled, and good Password policies in place, Win2003 server all SPs and WUs.   Our TS remote users are scattered and usually located in their client company offices.  We have little control on their side, except knowing that they're using RDC RDP supporting 128-bit.

1.  Is MITM just a theoretical vulnerabilty or do real threats exist?   Has there ever been real attacks reported?
2.  Does the Attack(er) have to be somewhere nearby, physically near the source IP or target IP ?
3.  What communications are at risk, RDP, ICA, VNC, HTTPS, HTTP port 80, others?   All others?
4.  How do the big shops, like and Online banking protect themselves and their web browser customers?
5.  On the surface, MITM reads like a serious vulnerabilty impacting all internet communications, I just can't believe it.
6.  Is there a documented way to protect TS without going overly nuts?

Any links you can share will be helpful in my report, I tried to find web sites with documented information without much success.

Who is Participating?
Rich RumbleConnect With a Mentor Security SamuraiCommented:
MITM is real, and for RD/TS, please read this paper by the autor of such a tool (Cain&Abel)

Man-in-the-middle affects just about everything, and people are comming up with many new variations. Most man in the middle involve arp poisoning or ip spoofing.

It's hard to poison arp on the internet, as ip ranges must be advertised and passed on by each isp, so if your in another country and trying to spoof your ip to be a man in the middle for someone across the globe, it's unlikely to succeed. On the local lan however it's much easier to do, so that is your main area of man in the middle exposure. If you worked (or hacked into) at an ISP you could also easily intercept traffic with no need for MITM, you could sniff the connection from start to finish without the nedd to spoof, or you could spoof and get the info.

For mitm to work, 9 times out of 10, you need to capture or be part of the data connection from the very begining, for HTTPS or RDP for example, you need to capture the key exchange so you can decrypt the data without bruteforce.

You can secure communications with a VPN (which can be MITM'd as well if the key excahnge is captured) and send the packets over an encrypted tunnel. VNC can be wrapped in SSH etc...
JReamAuthor Commented:

Thanks richrumble,  I've carefully read your links, some of which I've already read.   Thanks.

It sounds like its mostly about ARP spoofing during ARP requests and replies phase, and almost always near the target, ie  on the local LAN.   VPN is no sure solution.  Got it.

I think you did good answering and leading me to answers on most of my initial question.

What about #1, has a real attact ever been reported?  I did read an interesting artical how how Hotels (or their ISPs) routinely do legitimate ARP spoofing to get guests to pay for web access.  With a clear real life example like that, I would think that a real attack could have occurred.   Maybe the victims stayed quiet for good reason.

Solution:  If it's all mostly about ARP, couldn't we find a Gateway/Router that somehow had our LAN MAC addresses pre-configued and thus eliminating ARP broadcasts entirely?  

Rich RumbleSecurity SamuraiCommented:
Well fortuneatly (or unfortuneatly from another perspective) there are better/easier attacks that can yield more info, spyware, vir, phishingi, sniffing. In the early day's ARP spoofing was a real big concern. Cisco has "ARP inspection" which is a whitelist of mac's and ip's
ARP attacks do happen, but not on the internet, again due to the nature of route advertising it's just not possible unless orcastrated on some very grand and expensive scale. It would be picked up quickly and corrected, people advertise the wrong subnets and Ip ranges quite often, and it's a big headache when it happens, I've seen 3 major ones over the past 6 years. They were never found to be arp attacks but config mistakes.
There are plenty of tools out there for arp spoofing, some better than others, but again, an attacker will want to keep a low-profile and take the path of least resistance. Email, IM, 70% of wi-fi, ftp, http, smb, etc... most protocols are plain-text and all you have to do is listen, rather sniff... Lot's of users (from my experience 99.9%) run as administrators which is the worst thing you can do in my opinion, you hardly need AV, or even up2date patches if your not admin (15+ years too late)

@stake did some good work on arp and vlan vulerabilites a few years back to:
JReamAuthor Commented:
Thanks richrumble,
You've been extremely helpful.

Are you ready for another?
   Win2K3 TS Certificate newbie questions

Question has a verified solution.

Are you are experiencing a similar issue? Get a personalized answer when you ask a related question.

Have a better answer? Share it in a comment.

All Courses

From novice to tech pro — start learning today.