How to Sniff a Network : Raw Sockets and Sniffing
Network sniffers allow you to see packets entering and exiting a target machine. As a result, they have many practical uses before and after exploitation. In some cases, you’ll be able to use Wireshark (https://wireshark.org/) to monitor traffic or use a Pythonic solution like Scapy (which we’ll explore in the later posts). Nevertheless, there’s an advantage to knowing how to throw together a quick sniffer to view and decode network traffic. Writing a tool like this will also give you a deep appreciation for the mature tools that can painlessly take care of the finer points with little effort on your part. You will also likely pick up some new Python techniques and perhaps a better understanding of how the low-level networking bits work.
In the previous chapter, we covered how to send and receive data using TCP and UDP, and arguably this is how you will interact with most network services. But underneath these higher-level protocols are the fundamental building blocks of how network packets are sent and received. You will use raw sockets to access lower-level networking information such as the raw IP and ICMP headers. In our case, we are only interested in the IP layer and higher so we won’t decode any Ethernet information. Of course, if you intend to perform any low-level attacks such as ARP poisoning or you are developing wireless assessment tools, you need to become intimately familiar with Ethernet frames and their use. Let’s begin with a brief walkthrough of how to discover active hosts on a
Building a UDP Host Discovery Tool
The main goal of our sniffer is to perform UDP-based host discovery on a target network. Attackers want to be able to see all of the potential targets on a network so that they can focus their reconnaissance and exploitation attempts. We’ll use a known behavior of most operating systems when handling closed UDP ports to determine if there is an active host at a particular IP address. When you send a UDP datagram to a closed port on a host, that host typically sends back an ICMP message indicating that the port is unreachable. This ICMP message indicates that there is a host alive because we’d assume that there was no host if we didn’t receive a response to the UDP datagram. It is essential that we pick a UDP port that will not likely be used, and for maximum coverage, we can probe several ports to ensure we aren’t hitting an active UDP service. Why UDP? There’s no overhead in spraying the message across an entire subnet and waiting for the ICMP responses to arrive accordingly. This is quite a simple scanner to build with most of the work going into decoding and analyzing the various network protocol headers. We will implement this host scanner for both Windows and Linux to maximize the likelihood of being able to use it inside an enterprise environment. We could also build additional logic into our scanner to kick off full Nmap port scans on any hosts we discover to determine if they have a viable network attack surface. These are exercises left for the reader, and I look forward to hearing some of the creative ways you can expand this core concept.
Let’s get started.