CEH: Certified Ethical Hacker: 2012

Preventing Session Hijacking

To defend against session hijack attacks, a network should employ several defenses. The most effective protection is encryption, such as Internet Protocol Security (IPSec). This also defends against any other attack vectors that depend on sniffing. Attackers may be able to passively monitor your connection, but they won't be able to interpret the encrypted data. Other countermeasures include using encrypted applications such as Secure Shell (SSH, an encrypted telnet) and Secure Sockets Layer (SSL, for HTTPS traffic).

You can help prevent session hijacking by reducing the potential methods of gaining access to your network—for example, by eliminating remote access to internal systems. If the network has remote users who need to connect to carry out their duties, then use virtual private networks (VPNs) that have been secured with tunneling protocols and encryption (Layer 3 Tunneling Protocol [L3TP]/Point-to-Point Tunneling Protocol [PPTP] and IPSec).

The use of multiple safety nets is always the best countermeasure to any potential threat. Employing any one countermeasure may not be enough, but using them together to secure your enterprise will make the attack success rate minimal for anyone but the most professional and dedicated attacker. The following is a checklist of countermeasures that should be employed to prevent session hijacking:

Use encryption.
Use a secure protocol.
Limit incoming connections.
Minimize remote access.
Have strong authentication.
Educate your employees.
Maintain different username and passwords for different accounts.
Use Ethernet switches rather than hubs to prevent session hijacking attacks.

Dangers Posed by Session Hijacking

TCP session hijacking is a dangerous attack: most systems are vulnerable to it, because they use TCP/IP as their primary communication protocol. Newer operating systems have attempted to secure themselves from session hijacking by using pseudo-random number generators to calculate the ISN, making the sequence number harder to guess. However, this security measure is ineffective if the attacker is able to sniff packets, which gives all the information required to perform this attack.

The following are reasons why it's important for a CEH to be aware of session hijacking:

Most computers are vulnerable.
Few countermeasures are available to adequately protect against it.
Session hijacking attacks are simple to launch.
Hijacking is dangerous because of the information that can be gathered during the attack.

Sequence Prediction | Session Hijacking

TCP is a connection-oriented protocol, responsible for reassembling streams of packets into their original intended order. Every packet has to be assigned a unique session number that enables the receiving machine to reassemble the stream of packets into their original and intended order; this unique number is known as a sequence number. If the packets arrive out of order, as happens regularly over the Internet, then the SN is used to stream the packets correctly. As just illustrated, the system initiating a TCP session transmits a packet with the SYN bit set. This is called a synchronize packet and includes the client's ISN. The ISN is a pseudo-randomly generated number with over 4 billion possible combinations, yet it is statistically possible for it to repeat.

When the ACK packet is sent, each machine uses the SN from the packet being acknowledged, plus an increment. This not only properly confirms receipt of a specific packet, but also tells the sender the next expected TCP packet SN. Within the three-way handshake, the increment value is 1. In normal data communications, the increment value equals the size of the data in bytes (for example, if you transmit 45 bytes of data, the ACK responds using the incoming packet's SN plus 45).

Figure 1 illustrates the sequence numbers and acknowledgments used during the TCP three-way handshake.

Figure 1: Sequence numbers and acknowledgment during the TCP three-way handshake

Hacking tools used to perform session hijacking do sequence number prediction. To successfully perform a TCP sequence prediction attack, the hacker must sniff the traffic between two systems. Next, the hacker or the hacking tool must successfully guess the SN or locate an ISN to calculate the next sequence number. This process can be more difficult than it sounds, because packets travel very fast.

When the hacker is unable to sniff the connection, it becomes much more difficult to guess the next SN. For this reason, most session-hijacking tools include features to permit sniffing the packets to determine the SNs.

Hackers generate packets using a spoofed IP address of the system that had a session with the target system. The hacking tools issue packets with the SNs that the target system is expecting. But the hacker's packets must arrive before the packets from the trusted system whose connection is being hijacked. This is accomplished by flooding the trusted system with packets or sending an RST packet to the trusted system so that it is unavailable to send packets to the target system.

Juggernaut is a network sniffer that can be used to hijack TCP sessions. It runs on Linux operating systems and can be used to watch for all network traffic, or it can be given a keyword such as a password to look for. The program shows all active network connections, and the attacker can then choose a session to hijack.

Hunt is a program that can be used to sniff and hijack active sessions on a network. Hunt performs connection management, Address Resolution Protocol (ARP) spoofing, resetting of connections, monitoring of connections, Media Access Control (MAC) address discovery, and sniffing of TCP traffic.

TTYWatcher is a session-hijacking utility that allows the hijacker to return the stolen session to the valid user as though it was never hijacked. TTYWatcher is only for Sun Solaris systems.

IP Watcher is a session-hijacking tool that lets an attacker monitor connections and take over a session. This program can monitor all connections on a network, allowing the attacker to watch an exact copy of a session in real time.

T-Sight is a session-monitoring and -hijacking tool for Windows that can assist when an attempt at a network break-in or compromise occurs. With T-Sight, a system administrator can monitor all network connections in real time and observe any suspicious activity that takes place. T-Sight can also hijack any TCP session on the network. For security reasons, En Garde Systems licenses this software only to predetermined IP addresses.

The Remote TCP Session Reset Utility displays current TCP session and connection information such as IP addresses and port numbers. The utility is primarily used to reset TCP sessions.

Session Hijacking

Session hijacking is when a hacker takes control of a user session after the user has successfully authenticated with a server. Session hijacking involves an attack identifying the current session IDs of a client/server communication and taking over the client's session. Session hijacking is made possible by tools that perform sequence-number prediction. The details of sequence-number prediction will be discussed later in this chapter in the sequence prediction section.

Spoofing attacks are different from hijacking attacks. In a spoofing attack, the hacker performs sniffing and listens to traffic as it's passed along the network from sender to receiver. The hacker then uses the information gathered to spoof or uses an address of a legitimate system. Hijacking involves actively taking another user offline to perform the attack. The attacker relies on the legitimate user to make a connection and authenticate. After that, the attacker takes over the session, and the valid user's session is disconnected.

Session hijacking involves the following three steps to perpetuate an attack:

Tracking the Session The hacker identifies an open session and predicts the sequence number of the next packet.
Desynchronizing the Connection The hacker sends the valid user's system a TCP reset (RST) or finish (FIN) packet to cause them to close their session.
Injecting the Attacker's Packet The hacker sends the server a TCP packet with the predicted sequence number, and the server accepts it as the valid user's next packet.

Hackers can use two types of session hijacking: active and passive. The primary difference between active and passive hijacking is the hacker's level of involvement in the session. In an active attack, an attacker finds an active session and takes over the session by using tools that predict the next sequence number used in the TCP session.

In a passive attack, an attacker hijacks a session and then watches and records all the traffic that is being sent by the legitimate user. Passive session hijacking is really no more than sniffing. It gathers information such as passwords and then uses that information to authenticate as a separate session.

Two of the key features of TCP are reliability and ordered delivery of packets. To accomplish these goals, TCP uses acknowledgment (ACK) packets and sequence numbers. Manipulating these numbers is the basis for TCP session hijacking. To understand session hijacking, let's review the TCP three-way handshake described in earlier chapters:

The valid user initiates a connection with the server. This is accomplished by the valid user sending a packet to the server with the SYN bit set and the user's initial sequence number (ISN).
The server receives this packet and sends back a packet with the SYN bit set and an ISN for the server, plus the ACK bit set identifying the user's ISN incremented by a value of 1.
The valid user acknowledges the server by returning a packet with the ACK bit set and incrementing the server's ISN by 1.

This connection can be closed from either side due to a timeout or upon receipt of a package with the FIN or RST flag set.

Upon receipt of a packet with the RST flag set, the receiving system closes the connection, and any incoming packets for the session are discarded. If the FIN flag is set in a packet, the receiving system goes through the process of closing the connection, and any packets received while closing the connection are still processed. Sending a packet with the FIN or RST flag set is the most common method hijackers use to close the client's session with the server and take over the session by acting as the client.

DoS/DDoS Countermeasures

There are several ways to detect, halt, or prevent DoS attacks. The following are common security features:

Network-Ingress Filtering All network access providers should implement network-ingress filtering to stop any downstream networks from injecting packets with faked or spoofed addresses into the Internet. Although this doesn't stop an attack from occurring, it does make it much easier to track down the source of the attack and terminate the attack quickly. Most IDS, firewalls, and routers provide network-ingress filtering capabilities.
Rate-Limiting Network Traffic A number of routers on the market today have features that let you limit the amount of bandwidth some types of traffic can consume. This is sometimes referred to as traffic shaping.
Intrusion Detection Systems Use an intrusion detection system (IDS) to detect attackers who are communicating with slave, master, or agent machines. Doing so lets you know whether a machine in your network is being used to launch a known attack but probably won't detect new variations of these attacks or the tools that implement them. Most IDS vendors have signatures to detect Trinoo, TFN, or Stacheldraht network traffic.
Automated Network-Tracing Tools Tracing streams of packets with spoofed addresses through the network is a time-consuming task that requires the cooperation of all networks carrying the traffic and that must be completed while the attack is in progress.
Host-Auditing and Network-Auditing Tools File-scanning tools are available that attempt to detect the existence of known DDoS tool client and server binaries in a system. Network-scanning tools attempt to detect the presence of DDoS agents running on hosts on your network.

Find_ddos is a tool that scans a local system that likely contains a DDoS program. It can detect several known DoS attack tools.

SARA gathers information about remote hosts and networks by examining network services. This includes information about the network information services as well as potential security flaws, such as incorrectly set up or configured network services, well-known bugs in the system or network utilities system software vulnerabilities listed in the Common Vulnerabilities and Exposures (CVE) database, and weak policy decisions.

RID is a free scanning tool that detects the presence of Trinoo, TFN, or Stacheldraht clients.

Zombie Zapper instructs zombie routines to go to sleep, thus stopping their attack. You can use the same commands an attacker would use to stop the attack.

Smurf and SYN Flood Attacks

A smurf attack sends a large amount of ICMP Echo (ping) traffic to a broadcast IP address with the spoofed source address of a victim. Each secondary victim's host on that IP network replies to the ICMP Echo request with an Echo reply, multiplying the traffic by the number of hosts responding. On a multiaccess broadcast network, hundreds of machines might reply to each packet. This creates a magnified DoS attack of ping replies, flooding the primary victim. IRC servers are the primary victim of smurf attacks on the Internet.

A SYN flood attack sends TCP connection requests faster than a machine can process them. The attacker creates a random source address for each packet and sets the SYN flag to request a new connection to the server from the spoofed IP address. The victim responds to the spoofed IP address and then waits for the TCP confirmation that never arrives. Consequently, the victim's connection table fills up waiting for replies; after the table is full, all new connections are ignored. Legitimate users are ignored as well and can't access the server.

A SYN flood attack can be detected through the use of the netstat command. An example of the netstat output from a system under a SYN flood is shown in Figure 1.

Figure 1: netstat output under a SYN flood attack

Here are some of the methods used to prevent SYN flood attacks:

SYN Cookies SYN cookies ensure the server does not allocate system resources until a successful three-way handshake has been completed.
RST Cookies Essentially the server responds to the client SYN frame with an incorrect SYN ACK. The client should then generate an RST packet telling the server that something is wrong. At this point, the server knows the client is valid and will now accept incoming connections from that client normally.
Micro Blocks Micro blocks prevent SYN floods by allocating only a small space in memory for the connection record. In some cases, this memory allocation is as small as 16 bytes.
Stack Tweaking This method involves changing the TCP/IP stack to prevent SYN floods. Techniques of stack tweaking include selectively dropping incoming connections or reducing the timeout when the stack will free up the memory allocated for a connection.

In Exercise 1, you will learn how to prevent SYN flood attacks on Windows 2000 servers.

Exercise 1: Preventing SYN Flood Attacks on Windows 2000 Servers

Run the Windows Registry editor by clicking Start ð Run and typing Regedit.
Navigate to the HKLM\SYSTEM\CurrentControlSet\Services\Tcpip\Parameters Registry key.
Add the SynAttackProtect=2 DWORD value to the Registry key.
Close the regedit program.

This change will allow the operating system to handle more SYN requests. When the value of SynAttackProtect is 2, Windows delays the creation of a socket until the three-way handshake is completed. This change will effectively prevent SYN flood attacks from tying up resources on a Windows server.

How BOTs/BOTNETs Work

A BOT is short for web robot and is an automated software program that behaves intelligently. Spammers often use BOTs to automate the posting of spam messages on newsgroups or the sending of emails. BOTs can also be used as remote attack tools. Most often, BOTs are web software agents that interface with web pages. For example, web crawlers (spiders) are web robots that gather web page information.

The most dangerous BOTs are those that covertly install themselves on users' computers for malicious purposes.

Some BOTs communicate with other users of Internet-based services via instant messaging, Internet Relay Chat (IRC), or another web interface. These BOTs allow IRQ users to ask questions in plain English and then formulate a proper response. Such BOTs can often handle many tasks, including reporting weather; providing zip code information; listing sports scores; converting units of measure, such as currency; and so on.

A BOTNET is a group of BOT systems. BOTNETs serve various purposes, including DDoS attacks; creation or misuse of Simple Mail Transfer Protocol (SMTP) mail relays for spam; Internet marketing fraud; and the theft of application serial numbers, login IDs, and financial information such as credit card numbers. Generally a BOTNET refers to a group of compromised systems running a BOT for the purpose of launching a coordinated DDoS attack. See Figure 1.

Figure 1: Anatomy of a Distributed DoS Attack

How DDoS Attacks Work

DDoS is an advanced version of the DoS attack. Like DoS, DDoS tries to deny access to services running on a system by sending packets to the destination system in a way that the destination system can't handle. The key of a DDoS attack is that it relays attacks from many different hosts (which must first be compromised), rather than from a single host like DoS. DDoS is a large-scale, coordinated attack on a victim system.

Trinoo is a tool that sends User Datagram Protocol (UDP) traffic to create a DDoS attack. The Trinoo master is a system used to launch a DoS attack against one or more target systems. The master instructs agent processes (called daemons) on previously compromised systems (secondary victims) to attack one or more IP addresses. This attack occurs for a specified period of time. The Trinoo agent or daemon is installed on a system that suffers from a buffer overflow vulnerability. WinTrinoo is a Windows version of Trinoo and has the same functionality as Trinoo.

Shaft is a derivative of the Trinoo tool that uses UDP communication between masters and agents. Shaft provides statistics on the flood attack that attackers can use to know when the victim system is shut down; Shaft provides UDP, ICMP, and TCP flooding attack options.

Tribal Flood Network (TFN) allows an attacker to use both bandwidth-depletion and resource-depletion attacks. TFN does UDP and ICMP flooding as well as TCP SYN and smurf attacks. TFN2K is based on TFN, with features designed specifically to make TFN2K traffic difficult to recognize and filter. It remotely executes commands, hides the source of the attack using IP address spoofing, and uses multiple transport protocols (including UDP, TCP, and ICMP).

Stacheldraht is similar to TFN and includes ICMP flood, UDP flood, and TCP SYN attack options. It also provides a secure telnet connection (using symmetric key encryption) between the attacker and the agent systems (secondary victims). This prevents system administrators from intercepting and identifying this traffic.

Mstream uses spoofed TCP packets with the ACK flag set to attack a target. It consists of a handler and an agent portion, but access to the handler is password protected.

The services under attack are those of the primary victim; the compromised systems used to launch the attack are secondary victims. These compromised systems, which send the DDoS to the primary victim, are sometimes called zombies or BOTs. They're usually compromised through another attack and then used to launch an attack on the primary victim at a certain time or under certain conditions. It can be difficult to track the source of the attacks because they originate from several IP addresses.

Normally, DDoS consists of three parts:

Master/handler
Slave/secondary victim/zombie/agent/BOT/BOTNET
Victim/primary victim

The master is the attack launcher. A slave is a host that is compromised by and controlled by the master. The victim is the target system. The master directs the slaves to launch the attack on the victim system. See Figure 1.

Figure 1: Master and Slaves in a DDoS Attack

DDoS is done in two phases. In the intrusion phase, the hacker compromises weak systems in different networks around the world and installs DDoS tools on those compromised slave systems. In the DDoS attack phase, the slave systems are triggered to cause them to attack the primary victim. See Figure 2.

Figure 2: Bots or Zombie systems

Denial of Service

A DoS attack is an attempt by a hacker to flood a user's or an organization's system. As a CEH, you need to be familiar with the types of DoS attacks and should understand how DoS and DDoS attacks work. You should also be familiar with robots (BOTs) and robot networks (BOTNETs), as well as smurf attacks and SYN flooding. Finally, as a CEH, you need to be familiar with various DoS and DDoS countermeasures.

There are two main categories of DoS attacks:

Attacks sent by a single system to a single target (simple DoS)
Attacks sent by many systems to a single target (distributed denial of service, or DDoS)

The goal of DoS isn't to gain unauthorized access to machines or data, but to prevent legitimate users of a service from using it. A DoS attack may do the following:

Flood a network with traffic, thereby preventing legitimate network traffic.
Disrupt connections between two machines, thereby preventing access to a service.
Prevent a particular individual from accessing a service.
Disrupt service to a specific system or person.

Different tools use different types of traffic to flood a victim, but the result is the same: a service on the system or the entire system is unavailable to a user because it's kept busy trying to respond to an exorbitant number of requests.

On the evening of May 28, 2008, the company I was working for (alfasystems.com) suddenly dropped off the Internet. Their web servers were no longer accessible from the Internet.

Within a minute of the start of the attack, it was clear to the Alpha Systems engineers that they were experiencing a "packet flooding" attack of some sort. After looking at the log files of their Cisco router, it showed that both of their two T1 trunk interfaces to the Internet were receiving some sort of traffic at their maximum 1.54 megabit rate, while their outbound traffic had fallen to nearly zero. They were drowning in a flood of malicious traffic and valid traffic was unable to get out. Alpha Systems was the victim of a denial-of-service attack, more commonly referred to as a DoS. The engineers knew they had to do something quickly to stop the attack and get the web servers back up and accessible for their customers. But no one really knew what to do as this had never happened to the systems before. Then someone thought of the packet filtering capabilities of the router.

Luckily, because this DoS attack was prone to filtering, Alpha Systems was able to weed out the bad packets and return their service to almost normal operation. In two minutes Alpha Systems engineers applied "brute force" filters to their routers, shutting down all UDP and ICMP traffic, and alfasystems.com instantly popped back onto the Internet.

It was finally determined that their server had been attacked by 474 security-compromised Windows PCs containing remote-control attack "zombies," in a classic DoS attack generated by the coordinated efforts of these hundreds of individual PCs.

A DoS attack is usually an attack of last resort. It's considered an unsophisticated attack because it doesn't gain the hacker access to any information but rather annoys the target and interrupts their service. DoS attacks can be destructive and have a substantial impact when sent from multiple systems at the same time (DDoS attacks).

Ping of Death is an attack that can cause a system to lock up by sending multiple IP packets, which will be too large for the receiving system when reassembled. Ping of Death can cause a DoS to clients trying to access the server that has been a victim of the attack.

SSPing is a program that sends several large fragmented, Internet Control Message Protocol (ICMP) data packets to a target system. This will cause the computer receiving the data packets to freeze when it tries to reassemble the fragments.

A LAND attack sends a packet to a system where the source IP is set to match the target system's IP address. As a result, the system attempts to reply to itself, causing the system to create a loop—which will tie up system resources and eventually may crash the OS.

CPUHog is a DoS attack tool that uses up the CPU resources on a target system, making it unavailable to the user.

WinNuke is a program that looks for a target system with port 139 open, and sends junk IP traffic to the system on that port. This attack is also known as an out-of-bounds (OOB) attack and causes the IP stack to become overloaded—eventually the system crashes.

Jolt2 is a DoS tool that sends a large number of fragmented IP packets to a Windows target. This ties up system resources and eventually locks up the system. Jolt2 isn't Windows specific; many Cisco routers and other gateways may be vulnerable to the Jolt2 attack.

Bubonic is a DoS tool that works by sending TCP packets with random settings, in order to increase the load of the target machine so that it eventually crashes.

Targa is a program that can be used to run eight different DoS attacks. The attacker has the option to either launch individual attacks or try all of the attacks until one is successful.

RPC Locator is a service that, if unpatched, has a vulnerability to overflows. Details on patching a system to prevent RPC vulnerabilities will be covered later in the chapter. The RPC Locator service in Windows allows distributed applications to run on the network. It is susceptible to DoS attacks, and many of the tools that perform DoS attacks exploit this vulnerability.

Note

Because DoS attacks are so powerful and can cripple a production system or network, this chapter does not include any DoS tool exercises. If you want to test the tools listed here, ensure that you are not using them on a production network or system. The DoS tools could render the target systems unusable.

DDoS attacks can be perpetrated by BOTs and BOTNETs, which are compromised systems that an attacker uses to launch the attack against the end victim. The system or network that has been compromised is a secondary victim, whereas the DoS and DDoS attacks flood the primary victim or target.

Hacking Tools

EtherFlood is used to flood an Ethernet switch with traffic to make it revert to a hub. By doing this, a hacker is able to capture all traffic on the network rather than just traffic going to and from their system, as would be the case with a switch.

Dsniff is a collection of Unix-executable tools designed to perform network auditing as well as network penetration. The following tools are contained in dsniff: filesnarf, mailsnarf, msgsnarf, urlsnarf, and webspy. These tools passively monitor a vulnerable shared network (such as a LAN where the sniffer sits behind any exterior firewall) for interesting data (passwords, email, files, and so on).

Sshmitm and webmitm implement active man-in-the-middle attacks against redirected Secure Shell (SSH) and HTTPS sessions.

Arpspoof, dnsspoof, and macof work on the interception of switched network traffic that is usually unavailable to a sniffer program because of switching. To get around the Layer 2 packet-switching issue, dsniff spoofs the network into thinking that it's a gateway that data must pass through to get outside the network.

IP Restrictions Scanner (IRS) is used to find the IP restrictions that have been set for a particular service on a host. It combines ARP poisoning with a TCP stealth or half-scan technique and exhaustively tests all possible spoofed TCP connections to the selected port of the target. IRS can find servers and network devices like routers and switches and identify access-control features like access control lists (ACLs), IP filters, and firewall rules.

sTerm is a telnet client with a unique feature: it can establish a bidirectional telnet session to a target host, without ever sending the real IP and MAC addresses in any packet. Using ARP poisoning, MAC spoofing, and IP spoofing techniques, sTerm can effectively bypass ACLs, firewall rules, and IP restrictions on servers and network devices.

Cain & Abel is a multipurpose hacking tool for Windows. It allows easy recovery of various kinds of passwords by sniffing the network; cracking encrypted passwords using dictionary or brute-force attacks; recording Voice over IP, or VoIP, conversations; decoding scrambled passwords; revealing password boxes; uncovering cached passwords; and analyzing routing protocols. The latest version contains a lot of new features like ARP Poison Routing (APR), which enables sniffing on switched LANs and man-in-the-middle attacks. The sniffer in this version can also analyze encrypted protocols such as SSH-1 and HTTPS, and it contains filters to capture credentials from a wide range of authentication mechanisms.

Packet Crafter is a tool used to create custom TCP/IP/UDP packets. The tool can change the source address of a packet to do IP spoofing and can control IP flags (such as checksums) and TCP flags (such as the state flags, sequence numbers, and ack numbers).

SMAC is a tool used to change the MAC address of a system. It lets a hacker spoof a MAC address when performing an attack.

MAC Changer is a tool used to spoof a MAC address on Unix. It can be used to set the network interface to a specific MAC address, set the MAC randomly, set a MAC of another vendor, set another MAC of the same vendor, set a MAC of the same kind, or display a vendor MAC list to choose from.

WinDNSSpoof is a simple DNS ID spoofing tool for Windows. To use it on a switched network, you must be able to sniff traffic of the computer being attacked. Therefore, it may need to be used in conjunction with an ARP spoofing or flooding tool.

Distributed DNS Flooder sends a large number of queries to create a DoS attack, disabling DNS. If DNS daemon software logs incorrect queries, the impact of this attack is amplified.

Understanding MAC Flooding and DNS Spoofing

A packet sniffer on a switched network can't capture all traffic as it can on a hub network; instead, it captures traffic either coming from or going to the system. It's necessary to use an additional tool to capture all traffic on a switched network. There are essentially two ways to perform active sniffing and make the switch send traffic to the system running the sniffer:

ARP Spoofing This method involves using the MAC address of the network gateway and consequently receiving all traffic intended for the gateway on the sniffer system. A hacker can also flood a switch with so much traffic that it stops operating as a switch and instead reverts to acting as a hub, sending all traffic to all ports. This active sniffing attack allows the system with the sniffer to capture all traffic on the network.

Note

Many switches have been patched or redesigned to not be susceptible to the flooding vulnerability.

DNS Spoofing (or DNS Poisoning) This is a technique that tricks a DNS server into believing it has received authentic information when in reality it hasn't. Once the DNS server has been poisoned, the information is generally cached for a while, spreading the effect of the attack to the users of the server. When a user requests a certain website URL, the address is looked up on a DNS server to find the corresponding IP address. If the DNS server has been compromised, the user is redirected to a website other than the one that was requested, such as a fake website.

To perform a DNS attack, the attacker exploits a flaw in the DNS server software that can make it accept incorrect information. If the server doesn't correctly validate DNS responses to ensure that they come from an authoritative source, the server ends up caching the incorrect entries locally and serving them to users that make subsequent requests.

This technique can be used to replace arbitrary content for a set of victims with content of an attacker's choosing. For example, an attacker poisons the IP address's DNS entries for a target website on a given DNS server, replacing them with the IP address of a server the hacker controls. The hacker then creates fake entries for files on this server with names matching those on the target server. These files may contain malicious content, such as a worm or a virus. A user whose computer has referenced the poisoned DNS server is tricked into thinking the content comes from the target server and unknowingly downloads malicious content.

The types of DNS spoofing techniques are as follows:

Intranet Spoofing Acting as a device on the same internal network
Internet Spoofing Acting as a device on the Internet
Proxy Server DNS Poisoning Modifying the DNS entries on a proxy server so the user is redirected to a different host system
DNS Cache Poisoning Modifying the DNS entries on any system so the user is redirected to a different host

Wireshark Filters

Wireshark is a freeware sniffer that can capture packets from a wired or wireless LAN connection. It is a very powerful tool which can provide network and upper layer protocol data captured on a network. Like a lot of other network programs, Wireshark uses the pcap network library to capture packets.

Wireshark was called Ethereal until 2006 when the main developer decided to change its name because of copyright reasons with the Ethereal name, which was registered by the company he decided to leave in 2006.

In Exercise 1 you installed and began capturing packets using Wireshark. To narrow down the amount of information gathered by Wireshark, you can use filters. These filters limit the amount of information captured or displayed.

Here are some examples of Wireshark filters:

ip.dst eq www.eccouncil.org This sets the filter to capture only packets destined for the web server www.eccouncil.org.
ip.src == 192.168.1.1 This sets the filter to capture only packets coming from the host 192.168.1.1.
eth.dst eq ff:ff:ff:ff:ff:ff This sets the filter to capture only Layer 2 broadcast packets.
host 172.18.5.4 This sets the filter to capture only traffic to or from IP address 172.18.5.4.
net 192.168.0.0/24 This sets the filter to capture traffic to or from a range of IP addresses.
port 80 This sets the filter to capture traffic to destination port 80 (HTTP).
port 80 and tcp[((tcp[12:1] & 0xf0) >> 2):4] = 0x47455420 This sets the filter to capture HTTP GET requests. The filter looks for the bytes "G", "E", "T", and " " (hex values 47, 45, 54, and 20) just after the TCP header. "tcp[12:1] & 0xf0) >> 2" figures out the TCP header length.

Exercise 1 shows you how to write filters in Wireshark.

Exercise 1: Create a Wireshark filter to capture only traffic to or from an IP address

Open Wireshark.
Click the active Network Interface to capture traffic.
Click Capture, then select filters.
Click the new button to create a new filter.
Name the new filter in the filter name field.
Type host IPaddress in the filter string field.
Click OK.
Select the capture menu and click start to begin the capture.

Repeat the above steps to create filters using the following strings:

net 192.168.0.0/24 To capture traffic to or from a range of IP addresses.
src net 192.168.0.0/24 To capture traffic from a range of IP addresses.
dst net 192.168.0.0/24 To capture traffic to a range of IP addresses.
port 53 To capture only DNS (port 53) traffic.
host www.example.com and not (port 80 or port 25) To capture non-HTTP and non-SMTP traffic on your server.
port not 53 and not arp To capture all except ARP and DNS traffic.
tcp portrange 1501-1549 To capture traffic within a range of ports.
not broadcast and not multicast Capture only unicast traffic. Useful to get rid of noise on the network if you only want to see traffic to and from your machine.

Practice writing filters in Wireshark that capture only one type of protocol traffic or traffic from a specific source IP or MAC address. Use your PC's IP or MAC address to test that the filter is working.

It's important to understand how to create these filters before you attempt the CEH exam.

Bypassing the Limitations of Switches

Because of the way Ethernet switches operate, it is more difficult to gather useful information when sniffing on a switched network. Since most modern networks have been upgraded from hub to switches, it takes a little more effort to sniff on a switched network. One of the ways to do that is to trick the switch into sending the data to the hackers' computer using ARP poisoning.

How ARP Works

ARP allows the network to translate IP addresses into MAC addresses. When one host using TCP/IP on a LAN tries to contact another, it needs the MAC address or hardware address of the host it's trying to reach. It first looks in its ARP cache to see if it already has the MAC address; if it doesn't, it broadcasts an ARP request asking, "Who has the IP address I'm looking for?" If the host that has that IP address hears the ARP query, it responds with its own MAC address, and a conversation can begin using TCP/IP.

ARP poisoning is a technique that's used to attack an Ethernet network and that may let an attacker sniff data frames on a switched LAN or stop the traffic altogether. ARP poisoning utilizes ARP spoofing, where the purpose is to send fake, or spoofed, ARP messages to an Ethernet LAN. These frames contain false MAC addresses that confuse network devices such as network switches. As a result, frames intended for one machine can be mistakenly sent to another (allowing the packets to be sniffed) or to an unreachable host (a denial-of-service, or DoS, attack). ARP spoofing can also be used in a man-in-the-middle attack, in which all traffic is forwarded through a host by means of ARP spoofing and analyzed for passwords and other information.

ARP Spoofing and Poisoning Countermeasures

To prevent ARP spoofing, permanently add the MAC address of the gateway to the ARP cache on a system. You can do this on a Windows system by using the ARP -scommand at the command line and appending the gateway's IP and MAC addresses. Doing so prevents a hacker from overwriting the ARP cache to perform ARP spoofing on the system but can be difficult to manage in a large environment because of the number of systems. In an enterprise environment, port-based security can be enabled on a switch to allow only one MAC address per switch port.

In Exercise 1 you will use Wireshark to sniff traffic.

Exercise 1: Use Wireshark to Sniff Traffic

Download and install the latest stable version of Wireshark from www.wireshark.org.
Click on the Capture menu and then select interfaces.
Click the Start button next to the interface that shows packets being sent and received. If you have multiple interfaces with packet activity, choose one of them—preferably the interface with the most activity.
Click on a packet to analyze that single packet. The detailed headers will be displayed beneath the packet capture screen.
Expand each header (IP, TCP) of a packet and identify the address information.

This exercise will provide much more network traffic if performed on a hub rather than a switch. A wireless network can be used, as a wireless LAN is a shared network segment similar to how a hub operates.

Wireshark is a freeware sniffer that can capture packets from a wired or wireless LAN connection. The software was previously called Ethereal. Wireshark is a common and popular program because it is free, but it has some drawbacks. An untrained user may find it difficult to write filters in Wireshark to capture only certain types of traffic.

Snort is an intrusion detection system (IDS) that also has sniffer capabilities. It can be used to detect a variety of attacks and probes, such as buffer overflows, stealth port scans, Common Gateway Interface (CGI) attacks, Server Message Block (SMB) probes, and OS fingerprinting attempts.

WinDump is the Windows version of tcpdump, the command-line network analyzer for Unix. WinDump is fully compatible with tcpdump and can be used to watch, diagnose, and save to disk network traffic according to various rules.

EtherPeek is a great sniffer for wired networks with extensive filtering and TCP/IP conversation tracking capabilities. The latest version of EtherPeek has been renamed OmniPeek.

WinSniffer is an efficient password sniffer. It monitors incoming and outgoing network traffic and decodes FTP, POP3, HTTP, ICQ, Simple Mail Transfer Protocol (SMTP), telnet, Internet Message Access Protocol (IMAP), and Network News Transfer Protocol (NNTP) usernames and passwords.

Iris is an advanced data- and network-traffic analyzer that collects, stores, organizes, and reports all data traffic on a network. Unlike other network sniffers, Iris is able to reconstruct network traffic, such as graphics, documents, and emails including attachments.

How a Sniffer Works

Sniffer software works by capturing packets not destined for the sniffer system's MAC address but rather for a target's destination MAC address. This is known as promiscuous mode. Normally, a system on the network reads and responds only to traffic sent directly to its MAC address. However, many hacking tools change the system's NIC to promiscuous mode. In promiscuous mode, a NIC reads all traffic and sends it to the sniffer for processing. Promiscuous mode is enabled on a network card with the installation of special driver software. Many of the hacking tools for sniffing include a promiscuous-mode driver to facilitate this process. Not all Windows drivers support promiscuous mode, so when using hacking tools ensure that the driver will support the necessary mode.

Any protocols that don't encrypt data are susceptible to sniffing. Protocols such as HTTP, POP3, Simple Network Management Protocol (SNMP), and FTP are most commonly captured using a sniffer and viewed by a hacker to gather valuable information such as usernames and passwords.

There are two different types of sniffing: passive and active. Passive sniffing involves listening and capturing traffic, and is useful in a network connected by hubs; active sniffinginvolves launching an Address Resolution Protocol (ARP) spoofing or traffic-flooding attack against a switch in order to capture traffic. As the names indicate, active sniffing is detectable but passive sniffing is not detectable.

In networks that use hubs or wireless media to connect systems, all hosts on the network can see all traffic; therefore, a passive packet sniffer can capture traffic going to and from all hosts connected via the hub. A switched network operates differently. The switch looks at the data sent to it and tries to forward packets to their intended recipients based on MAC address. The switch maintains a MAC table of all the systems and the port numbers to which they're connected. This enables the switch to segment the network traffic and send traffic only to the correct destination MAC addresses. A switch network has greatly improved throughput and is more secure than a shared network connected via hubs.

Another way to sniff data through a switch is to use a span port or port mirroring to enable all data sent to a physical switch port to be duplicated to another port. In many cases, span ports are used by network administrators to monitor traffic for legitimate purposes.

Understanding Host-to-Host Communication

All Host-to-Host network communications is based upon the TCP/IP Data Communications Model. The TCP/IP Model is a 4 layer model. The TCP/IP Model maps to the older OSI model with 7 layers of data communication. Most applications use the TCP/IP suite for host-to-host data communications. See Figure 1.

Figure 1: TCP/IP Model

In normal network operations, the application layer data is encapsulated and a header containing address information is added to the beginning of the data. An IP header containing source and destination IP address are added to the data as well as a MAC header containing source and destination MAC addresses. IP addresses are used to route traffic to the appropriate IP network, and the MAC addresses ensure the data is sent to the correct host on the destination IP network. In this manner, traffic is sent from source host to destination host across the Internet and delivery to the correct host is ensured. The postal system works much the same way. Mail is routed to the appropriate area using the zip code, and then the mail is delivered within the zip code to the street and house number. The IP address is similar to the zip code to deliver mail to the regional area, and the street and house numbers are like the MAC address of that exact station on the network.

The address system ensures accurate delivery to the receiver. In normal network operations, a host should not receive data intended for another host as the data packet should only be received by the intended receiver. Simply said, the data should only be received by the station with the correct IP and MAC address. However, we know that sniffers do receive data not intended for them.

In the real world, sometimes mail is not delivered to the intended receiver. I'm sure you have all opened your mailbox to discover an envelope addressed to your neighbor or someone who used to live at your address. This happens on a fairly regular basis at my house. Most people will just leave the mail in the box for the postal carrier to redeliver or physically take the envelope to a neighbor. This same type of situation can occur in computer networking, where application layer data does not reach its intended recipient because of a delivery error or other network fault.

Another cause of mail not being received by the intended recipient is someone is performing reconnaissance and watching your mailbox. Let's assume you are not home and the postal carrier delivers your mail to the mailbox. Someone watching the mailbox from down the street or a nearby building could wait for the mail to be delivered to the mailbox, and they go take the mail or just a particular envelope out of the box. This would be especially effective if the hacker performed some reconnaissance and knew what time each day the mail was delivered. The hacker could then examine and read the information in the envelope, and if they were trying to cover their tracks simply reseal the envelope and put it back in the mailbox.

Sniffing data on a network occurs in much the same way. Data is intercepted, read, and either sent on to the intended recipient or just discarded.

In addition to understanding network addresses, it is also important to understand the format of the TCP Header. Figure 2 shows the TCP Header format.

Figure 2: TCP Header Format

The TCP Header is comprised of the following fields:

Source Port: 16 bits The source port number.
Destination Port: 16 bits The destination port number.
Sequence Number: 32 bits The sequence number of the first data octet in this segment (except when SYN is present). If SYN is present the sequence number is the initial sequence number (ISN) and the first data octet is ISN+1.
Acknowledgment Number: 32 bits If the ACK control bit is set this field contains the value of the next sequence number the sender of the segment is expecting to receive.
Data Offset: 4 bits The number of 32 bit words in the TCP Header. This indicates where the data begins.
Reserved: 6 bits Reserved for future use. Must be zero.
Control Bits: 6 bits
- URG: Urgent Pointer field significant
- ACK: Acknowledgment field significant
- PSH: Push Function
- RST: Reset the connection
- SYN: Synchronize sequence numbers
- FIN: No more data from sender
Window: 16 bits The number of data octets beginning with the one indicated in the acknowledgment field which the sender of this segment is willing to accept.
Checksum: 16 bits The checksum field is a computation of all fields to ensure all data was received and the data was not modified in transit.
Urgent Pointer: 16 bits This field communicates the current value of the urgent pointer as a positive offset from the sequence number in this segment. The urgent pointer points to the sequence number of the octet following the urgent data. This field is only be interpreted in segments with the URG control bit set.
Options: variable Options may occupy space at the end of the TCP header and are a multiple of 8 bits in length.

When referring to the length of the fields in the TCP Header, 8 bits comprises a single byte. A Nibble is less than a byte and a Word is more than a byte.

In the next section we will explore how a hacking tool manipulates normal network operations in order to capture traffic on a host that is not the intended receiver.