Sniffers Seminar Report

1. INTRODUCTION:-

A Sniffer is a program or a device that eavesdrops on the network traffic by grabbing information traveling over a network. Sniffers basically are "Data Interception" technology. They work because the Ethernet was built around a principle of sharing. Most networks use broadcast technology wherein messages for one computer can be read by another computer on that network. In practice, all the other computers except the one for which the message is meant, will ignore that message. However, computers can be made to accept messages even if they are not meant for them. This is done by means of a Sniffer!

Many people assume computers connected to a switch are safe from sniffing. Nothing could be further from the truth. Computers connected to switch are just as vulnerable to sniffing as those connected to a hub.

2. How a Sniffer works?

A computer connected to the LAN has 2 addresses. One is the MAC (Media Access Control) address that uniquely identifies each node in a network and is stored on the network card itself. It is the MAC address that gets used by the Ethernet protocol while building `frames' to transfer data to and from a machine.

The other is the IP address which is used by applications. The Data Link Layer uses an Ethernet header with the MAC address of the destination machine rather than the IP Address. The Network Layer is responsible for mapping IP network addresses to the MAC address as required by the Data Link Protocol.

It initially looks up the MAC address of the destination machine in a table, usually called the ARP (Address Resolution Protocol) cache. If no entry is found for the IP address, the Address Resolution Protocol broadcasts a request packet (ARP request) to all machines on the network. The machine with that address responds to the source machine with its MAC address. This MAC address then gets added to the source machines ARP Cache. This MAC address is then used by the source machine in all its communications with the destination machine.

There are two basic types of Ethernets environments and how sniffers work in both these cases is slightly different.

SHARED ETHERNET:

In a shared Ethernet environment, all hosts are connected to the same bus and compete with one another for bandwidth. In such an environment packets meant for one machine are received by all the other machines.

Thus when a machine Venus wants to talk to Cupid in such an environment, it sends a packet on the network with the destination MAC address of Cupid along with its own source MAC address. All the computers on the shared Ethernet compare frame's destination MAC address with their own. If the two don't match, the frame is quietly discarded.

A machine running a sniffer breaks this rule and accepts all frames. Such a machine is said to have been put into promiscous mode and can effectively listen to all the traffic on the network. Sniffing in a Shared Ethernet environment is totally passive and hence extremely difficult to detect.

SWITCHED ETHERNET:

An Ethernet environment in which the hosts are connected to switch instead of a hub is called a Switched Ethernet. The switch maintains tables keeping track of each computer’s MAC address and the physical port on the switch to which that MAC address is connected and delivers packets destined for a particular machine correspondingly. The switch is an intelligent device which sends packets to the destined computer only and does not broadcast it to all the machines on the network, as in the previous case.

 As a result of this the process followed earlier, of putting the machine into promiscous mode, to gather packets does not work. As a result of this, even many experienced Systems Administrators fall into the belief that switched networks are secure and cannot be sniffed.

Why has the Internet been vulnerable to sniffers for so long?

The situation has taken this long to fix for a number of reasons.

Part of the problem is that software companies are fighting their own battles for market share and trying to bring users fancy new features. They will say the prefer delivering "user friendly" systems over ones that come with tight security features, which often make the systems harder to set up or less convenient to use. (It is often said that security is inversely proportional to ease of use.) This is the same logic car manufacturers used in the early 1900s to justify not installing seat belts in all vehicles sold.

Part of the problem is added cost for Ethernet switches, hubs, interface cards that don't support promiscuous mode, and new software. If there is barely enough money to buy the cheapest hardware to get a minimal network up and running, and practically none for skilled network administrators, its going to be very hard to have a secure network. With security, you [don't] get what you [don't] pay for.

Part of the problem is incompatibilities -- vendor A software to vendor B software, and older software to newer software -- between software products.

Often the priorities for dealing with these problems are set by "market forces" -- in other words "money"; they don't have enough, or you won't pay enough -- and software vendors set these priorities in response to user demand, or the company's perception of user demand. More people want a new widget in their word processor, or support for some new sound card or 62X CD-ROM drive, than are asking for abolishing clear text passwords in network services and other security features.

One thing is certain. You can be sure that as more of our economic, academic, and personal lives are lived on ever more interconnected computer networks, the more responsibility we have to protect our online assets and our privacy.

Either that or we had better be prepared to lose them.

3. METHODS TO SNIFF ON A SWITCH

3.1 ARP Spoofing:

We have explained earlier how ARP is used to obtain the MAC address of the destination machine with which we wish to communicate. The ARP is stateless, you can send a ARP reply even if one has not been asked for and such a reply will be accepted. Ideally when you want to sniff the traffic originating from machine Venus, you can ARP Spoof the gateway of the network.

The ARP cache of Venus will now have a wrong entry for the gateway and is said to be Poisoned. This way all the traffic destined for the gateway will pass through your machine. Another trick that can be used is to poison a hosts ARP cache by setting the gateway's MAC address to FF:FF:FF:FF:FF:FF(also known as the broadcast MAC). An excellent tool for this is the arpspoof utility which comes with the dsniff suite. Using arpspoof to poison the ARP cache of a machine is accomplised by giving the command

[root@tachyon dhar]# arpspoof -t 203.199.66.243 203.199.66.193

0:80:ad:7c:7:3a 52:54:5:f3:95:1 0806 42: arp reply \

              203.199.66.193 is-at 0:80:ad:7c:7:3a

0:80:ad:7c:7:3a 52:54:5:f3:95:1 0806 42: arp reply \

              203.199.66.193 is-at 0:80:ad:7c:7:3a

The -t flag specifies the target whose ARP cache we wish to poison and the other argument is the IP address of the gateway which we wish to spoof. So now all the data destined for the gateway from the target machine will have to pass through our machine. Before you do this, it is essential you turn on IP Forwarding on your machine. You can do this by giving the command:

[root@tachyon dhar]# echo 1 > /proc/sys/net/ipv4/ip_forward 

[root@tachyon dhar]# cat /proc/sys/net/ipv4/ip_forward

1

[root@tachyon dhar]#

If the cat command returns a value of 1, then IP Forwarding has been enabled but if it returns 0, it means IP Forwarding has not been enabled. It is important to enable IP Forwarding or else the network will die.

3.2 MAC Flooding:

 Switches keep a translation table which maps various MAC addresses to the physical ports on the switch. As a result of this it can intelligently route packets from one host to another. The switch has a limited memory for this work. MAC flooding makes use of this limitation to bombard the switch with fake MAC addresses till the switch can't keep up.

 The switch then enters into what is known as a `failopen mode' wherein it starts acting as a hub by broadcasting packets to all the machines on the network. Once that happens sniffing can be performed easily. MAC flooding can be performed by using macof, a utility which comes with dsniff suite.

[root@tachyon dhar]# macof

77:6b:e1:6e:5e:8c 93:2d:ed:45:f9:e3 0.0.0.0.45702 > \

      0.0.0.0.11000: S 1847390231:1847390231(0) win 512

84:a4:d3:57:ef:8 12:56:52:42:dc:95 0.0.0.0.16630 >  \

      0.0.0.0.3031: S 1484147693:1484147693(0) win 512

88:f0:9:3f:18:89 d:86:53:53:d7:f8 0.0.0.0.15535 > \ 

      0.0.0.0.7466: S 293820390:293820390(0) win 512

The idea is to cause the switch to start forwarding to the intruder the frames destined tothe victim. One can do this simply by sending out frames with the source address of thevictim. The "auto-learning" feature will make the switch now believe that the one thatshould receive the packets is the intruder, and will send the frames the wrong way.

The obvious problem is that if the victim doesn't receive the frame, then its communications breaks, and there won't be anything more to sniff. Another problem is that the victim itself will still send out frames with its MAC address, causing the switch to revert.

There are a few solutions to these problems. One is by subverting an authenticated

connection, in which case the eavesdropper has to DoS the victim (taking it offline), redirect the switch, and continue on with the connection as if nothing happened. This can

be achieved easily.

 For example, in the case that the victim has a Telnet connection to the server, the

intruder can DoS the victim’s machine, taking him/her off line. Then, the eavesdropper sends out packets with the victim’s MAC address, causing the switch to send to the  eavesdropper all the packets destined for the victim. In order to pick up the victim’s connection, the intruder causes the server to send him a TCP packet (i.e. use the talk service to prompt the server for a connection). At this point, the intruder simply starts

reversing the sequence and acknowledgement numbers to continue the Telnet connection.

A similar solution is that when the intruder receives an incoming packet, he turns around   and broadcasts it back. This way the victim still receives the packet. A steady stream of outgoing traffic and broadcasts of the incoming traffic will allow the intruder to recover a good percentage of the original traffic.

Concluding, it is obvious that a switch cannot stop a malicious person from sniffing the network.

3.3 IP-based sniffing

This is the original way of packet sniffing. It works by putting the network card into promiscuous mode and sniffing all packets matching the IP address filter. Normally, the IP address filter isn’t set so it can capture all the packets. This method only works in non-switched networks. 

4. What Does Sniffed Data Look Like?

It is easy to grasp the concepts discussed above by watching a sniffer in action. The information in the following example was derived using tcpdump, a program that has been around for quite sometime and is available for many platforms. This particular snippet is an abbreviated exchange between a machine and the SecurityFocus Web server.

21:06:30.786814 0:1:3:e5:46:6b 0:4:5a:d1:46:ad 0800 650: 192.168.1.3.32946 >

66.38.151.10.80: P [tcp sum ok] 1:585(584) ack 336 win 64080 <nop,nop,timestamp 608776

899338> (DF) (ttl 64, id 7468, len 636)

0x0000   4500 027c 1d2c 4000 4006 8074 c0a8 0103      E..|.,@.@..t....

0x0010   4226 970a 80b2 0050 54ac b070 78ef d6c3      B&.....PT..px...

0x0020   8018 fa50 c663 0000 0101 080a 0009 4a08      ...P.c........J.

0x0030   000d b90a 4745 5420 2f63 6f72 706f 7261      ....GET./corpora

0x0040   7465 2f69 6d61 6765 732f 6275 696c 642f      te/images/build/

0x0050   626c 6c74 5f72 645f 312e 6769 6620 4854      bllt_rd_1.gif.HT

0x0060   5450 2f31 2e31 0d0a 486f 7374 3a20 7777      TP/1.1..Host:.ww

0x0070   772e 7365 6375 7269 7479 666f 6375 732e      w.securityfocus.

0x0080   636f 6d0d 0a55 7365 722d 4167 656e 743a      com..User-Agent:

0x0090   204d 6f7a 696c 6c61 2f35 2e30 2028 5831      .Mozilla/5.0.(X1

0x00a0   313b 2055 3b20 4c69 6e75 7820 6936 3836      1;.U;.Linux.i686

21:06:30.886814 0:4:5a:d1:46:ad 0:1:3:e5:46:6b 0800 402: 66.38.151.10.80 >  

192.168.1.3.32949: P [tcp sum ok] 2363393025:2363393361(336) ack 1437810754 win 8616

<nop,nop, timestamp 899338 608766> (ttl 61, id 10825, len 388)

0x0000   4500 0184 2a49 0000 3d06 b74f 4226 970a      E...*I..=..OB&..

0x0010   c0a8 0103 0050 80b5 8cde 8401 55b3 4042      .....P......U.@B

0x0020   8018 21a8 0543 0000 0101 080a 000d b90a      ..!..C..........

0x0030   0009 49fe 4854 5450 2f31 2e31 2032 3030      ..I.HTTP/1.1.200

0x0040   204f 4b0d 0a41 6765 3a20 320d 0a41 6363      .OK..Age:.2..Acc

0x0050   6570 742d 5261 6e67 6573 3a20 6279 7465      ept-Ranges:.byte

0x0060   730d 0a44 6174 653a 2054 7565 2c20 3132      s..Date:.Tue,.12

0x0070   2046 6562 2032 3030 3220 3033 3a30 343a      .Feb.2002.03:04:

0x0080   3538 2047 4d54 0d0a 436f 6e74 656e 742d      58.GMT..Content-

0x0090   4c65 6e67 7468 3a20 3433 0d0a 436f 6e74      Length:.43..Cont

0x00a0   656e 742d 5479 7065 3a20 696d 6167 652f      ent-Type:.image/

0x00b0   6769 660d 0a53 6572 7665 723a 2041 7061      gif..Server:.Apa

0x00c0   6368 652f 312e 332e 3232 2028 556e 6978      che/1.3.22.(Unix

0x00d0   2920 6d6f 645f 7065 726c 2f31 2e32 360d      ).mod_perl/1.26.

This excerpt shows two packets: an HTTP request by the client and the server’s response. Note that the first few lines of each sniffed packet provide a summary of the transaction: timestamps, source and destination MAC addresses, source and destination IP addresses and several other bits of information. The numbered lines (0x00##) show the data transmitted by each packet in hexadecimal format. Additionally, an ASCII decode of the payload is located off to the right - a convenient feature for crackers and nosy neighbors watching you on the network.

5. Why Should Users Be Concerned? 

On a normal LAN there are thousands of packets exchanged by multiple machines every minute, ample supply for any attacker. Anything transmitted in plaintext over the network will be vulnerable - passwords, web pages, database queries and messaging to name a few. A sniffer can easily be customized to capture specific traffic like telnet sessions or e-mail. Once traffic has been captured, crackers can quickly extract the information they need - logins, passwords and the text of messages. And the users will likely never know they were compromised - sniffers cause no damage or disturbance to a network environment.

6. PACKET SNIFFERS

On non-switched Ethernet networks all frames are broadcasted by network equipment to all machines, but only the computer that the packets are destined for will recognize them. All of the other machines on that network still see the packet, but if they are not the intended receiver, they will disregard it.

A packet sniffer is either a wire-tap device or a software that plugs into the network and eavesdrops on the network traffic. When a computer is running a packet sniffer, its network interface is in promiscuous mode, which means that it listens for all traffic and it has the ability to receive all of the packets reaching the interface.

7. Usage of a packet sniffer

7.1 To help administrators

The initial usage of a packet sniffer was to help LAN/WAN administrators to maintain networks by analyzing network traffic and determining where a problem is on the network. A security administrator uses multiple sniffers, as a detection system by placing them strategically throughout the network. A packet sniffer converts the data to human readable format so that people can read the traffic. Moreover, a packet sniffer performs fault analysis in order to discover problems in the network, such as why computer A cannot talk to computer B and performance analysis to discover network bottlenecks.

Finally a packet sniffer can help administrators perform network intrusion detection in order to discover intruders and network traffic logging in order to create logs that eavesdroppers cannot break into and erase.

7.2 To help eavesdroppers

Since packet sniffing refers to the technique of monitoring each packet as it flows across the network, it implies that a intruder can use this technique to automatic sift of clear-text passwords and usernames from the network, in order to break into systems, or any other useful information such as credit card numbers. From this moment, packet sniffing becomes a threat that has to be taken into account.

8. Components of a packet sniffer

A packet sniffer consists of the following:

Ø Special hardware:-

The special hardware is needed for some products when analyzing hardware faults like CRC errors, voltage problems, cable problems, jitter, negotiation errors, and so forth is needed. But most products work from standard network adapters.

Ø Capture driver:-

This is the most important part of a packet sniffer because it captures the network traffic from the wire, filters it for the particular traffic one wants, and then stores the data in a buffer.

Ø Buffer:-

The frames are stored in a buffer, after being captured from the network. There are a couple captures modes: capture until the buffer fills up, or use the buffer as a circular buffer where the newest data replaces the oldest data.

Ø Real-time analysis:-

This feature does some minor bit of analysis of the frames as they come off the wire. This is able to find network performance issues and faults while capturing.

Ø Decoder:-

The decoder displays the contents of network traffic with descriptive text so that an analyst can figure out what is happening.

Ø Packet editing/transmission:-

This is a feature that is contained in some products that allow you to edit your own network packets and transmit them onto the network.

9. Steps in creation of a packet Linux sniffer

The main steps in the development of a packet sniffer are:

Ø  Creating a socket stream.

Ø  Setting the NIC into Promiscuous mode.

Ø  Reading data from the open socket stream.

The rest of the steps only deal with interpreting the headers and formatting the

data (and redirecting the data to the output stream).

Creating a socket:-

On UNIX and its clones, communication points called sockets can be created.

These help in the communication of end systems present in the network.

When a socket is created, a socket stream, similar to the file stream, is created, through which data is read.

Setting the NIC to promiscuous mode:-

First, a reference to a structure called ifreq is needed. This is done by the statement,

struct ifreq ifr;

This is an interface request structure used for socket ioctl. The ifreq structure is a

rather large structure. The main members that will be used in this structure are

the members that hold the interface name and the interface flags.

Next, the NIC name, viz. eth0 (usually), is copied to the structure member

ifr_name.

The next step is to get the flags of the specified interface using the ioctl system

call. The ioctl system call takes 3 arguments, viz.

Ø  The socket stream descriptor

Ø  The function that the ioctl function is supposed to perform.

Here, the macro used is - SIOCGIFFLAGS.

Ø  Reference to the ifreq member.

In the next step, the promisc flag must be set. This is done by using the statement,

Ifr.ifr_flags |= IFF_PROMISC;

The last step in this process is to set these flags to the NIC using the ioctl call. The

same first and third parameters are used, but the second parameter to ioctl is changed to SIOCSIFFLAGS.

Now comes the cumbersome part - protocol interpretation. To do this, the user is

required to have some basic knowledge about the protocol that he/she intends to sniff. The protocol contents, its fields, field lengths must be known and the user must know what kind of data to expect in those fields.

To interpret the headers, the fields of the protocol must be accessible. On a Linux

machine, by including the headers <linux/ip.h> and <linux/tcp.h>, the IP and TCP protocols can be interpreted. These headers contain structures that represent the IP and TCP headers. The user can refer to these headers and interpret the information that is passing across the network. A more complete sniffer would not just be able to capture a single protocol, but all types of packets.

10. CHALLENGES FOR DETECTING PACKET SNIFFERS

a) A sniffer does not send any packets apart from those that would originate from the machine under normal network connectivity circumstances. It is a common misconception that a packet sniffer reads all packets from a broadcast network, and then re-sends them. This is, in fact not the case. A packet sniffer copies packets from the network, allowing the original packets to proceed in a normal manner.

b) A sniffer is a machine, which is not necessarily under our administrative control. Another common misconception, and one also made by CERT, in advisory CA-94.01 is that all machines on which illegitimate sniffers are running are under local administrative control. This is clearly not the case. Over 80% of network attacks come from inside the firewall. With the increasing power and reduction in cost of a laptop computer, we can now no longer make the assumption that every machine on a Local Area Network is controllable by the local administrative team.

c) The intruder is under complete control (root level or equivalent) of the machine on which the sniffer is running. This is known to be true, as in order for a network card to be put into promiscuous mode, the user must have administrative privileges on the local machine. This is true of both Unix, and Microsoft Operating Systems.

d) The sniffer is running on a standard machine, with no special services running to give it a unique “fingerprint”.

In theory, it is impossible to detect sniffing programs because they are passive; they only collect packets, they do not transmit anything. However, in practice it is sometimes possible to detect sniffing programs, based on the fact that the networking code of the computer running the packet sniffer generates traffic.

With these considerations in mind, there now follows a set of possible solutions to the problem of detecting an illegitimate packet sniffer on a network, and these will be described at a high level. There are many more methods than these available, but many are left out for the sake of brevity.

11. METHODS TO DETECT PACKET SNIFFERS

11.1) ICMP echo response

If we send an ICMP echo (or “ping”) request packet containing a correct IP address and an incorrect MAC address to a normal networked machine, then we would expect to receive no response. This is because the physical layer of the recipient checks for a packet with the MAC address of either the local machine, or the broadcast Ethernet address.

If the machine we are “pinging” is running in promiscuous mode the packet should be replied to. This is because a promiscuous mode device will accept all packets, and consequently the fake packet we have created will also be accepted. When this packet is received, it is passed up through the layers, where the IP address would be recognised as its own. Therefore, the intruder’s machine would respond to the “ping” request.

 

11.2) Ping method, part2

The ping method can be enhanced in a number of ways. Any protocol that generates a response can be used, such as a TCP connection request or a UDP protocol or any protocol that might generate an error on the target machine might be used. For example, bad IP header values might be used to generate an ICMP error. Moreover, sometimes a broadcast address needs to be used in order to bypass software IP address filtering.

11.3) DNS Lookup Test

If we send a packet to a remote machine, it is possible that a sniffer is using DNS in order to establish a machine name from the IP address of the packet. This would be desirable in order to decide whether a machine was worth gaining access to. For example, an intruder is more likely to attempt to gain access for “payroll.uwe.ac.uk” than to “joebloggs.uwe.ac.uk.”

If this is the case, then the IP address of any packet that is received by the intruder must be resolved via DNS to a machine name. We can potentially use this to our advantage. If we send a packet to a fictitious host, we can then see if there are any DNS requests for the fake IP address. If there are, we can see who has sent the DNS request and find the IP and MAC address of the host, which has looked up the fake address (i.e. the intruder.)

11.4) ARP METHOD

The ARP method is similar to the ping method, but an ARP packet is used instead. The simplest ARP method transmits an ARP to a non-broadcast address. If a machine responds to such an ARP of its IP address, then it must be in promiscuous mode.

11.5) The decoy method

The decoy method consists of setting up a connection between a client and a virtual server. The client runs a script to logon to the server using a plain-text protocol, allowing  any eavesdropper to sift the usernames and the passwords from the wire. When the eavesdropper will attempt to log on using this information, standard intrusion detection systems will alert the fact that a sniffing intruder has found the traffic and attempted to use the information.

The decoy method works everywhere, unlike the ping and ARP methods, which only work on the local network.

11.6) Source-route method

This method functions by configuring the source-route information inside the IP header. A ping packet is created, while a loose-source route is used to force it by another machine on the same segment. This machine should have routing disabled, so that it will not forward it to the target. If you get a response (from the networking code of the computer that runs the sniffer), then it is likely the target sniffed the packet off the wire.

11.7) Time Domain Reflectometers (TDR)

A Time Domain Reflectometer is basically radar for the wire. It sends a pulse down the wire and graphs the reflections that come back. By looking at the graph of the response an expert can figure out if any devices that should not be attached to the wire, are attached. They also roughly tell where, in terms of distance along the wire, the tap is

located.

11.8) SNMP monitoring

Another way to figure out whether a hardware packet sniffer is attached to the wires is by using smart hubs with SNMP management. Such hubs can provide automated monitoring of Ethernet hubs. In other words a packet sniffer can be tracked down by configuring the system with the information where all the cables terminate.

11.9) Antisniffer

In 1999, L0pht Heavy Industries released a product called Antisniff. AntiSniff is a

network security software program that attempts to scan one’s network and determine if a

computer is running in promiscuous mode, using some of the detection methods described above. AntiSniff, which operates on both Windows NT and UNIX, is just a detection tool; it does not protect against packet sniffers.

11.10) Fake user name & password login attempt test

We can create and known (fake) login sequence to a host, containing a username and password with no credentials, but named something tempting like “Admin” We can then check to see if a login attempt is made in future with that username and password. If there is a login attempt using this username and password, then we know there is a sniffer on the network “somewhere” (although the login attempt is highly unlikely to come from the same machine that is sniffing.) This can be achieved with a combination of auditing and logging techniques.

11.11) Unrecognized MAC address test

Many networks use standard machines in order to make administration easier. This means that the Network Interface Cards (NIC’s) in the machines are likely to be of the same manufacturer and type. As each manufacturer of NIC is allocated a specified range of MAC addresses to use, each and every NIC is uniquely identifiable. So, the network cards in a set of standard machines will be similar, and therefore the first two octets of the MAC address should also be similar. Therefore, we can use pattern matching to scan the network for any unfamiliar MAC addresses in order to ascertain if a machine has been brought in from outside.

11.12) ICMP Echo Latency Test

If we send a series of sensitive ICMP echo requests to every machine in a given subnet, then we can find the average response time for each machine. This will be referred to as the machines “ICMP echo response profile.” If we then flood the network with packets destined for a fake IP address, we would expect that this would make little difference to the ICMP echo response profile of each machine.

This is because even though the network is running at a higher usage level, normal networked machines will not be affected considerably as the network interface and CPU load of the machine are not considerably affected by even a massive increase of network traffic. This is in spite of the greater amount of collisions on the network. However, an interface running in promiscuous mode will see a noticeable increase in the amount of packets that it will be handling, as the fake packets are queued up on the network card awaiting processing. If, then we perform the ICMP echo response profile again while the network is flooded, we would expect a “normal” machine to show no significant increase in response time when comparing the two timings taken.

 A machine running in promiscuous mode however may show a significant time delay in responding to an ICMP echo request due to the quantity of packets queued on its Ethernet device. So, if we gain a profile of each machine, and then compare it with a repeat of the profile taken when the network is flooded, there should be no significant time difference for a legitimate machine. For a sniffer however, we may see up to three to four times echo response time difference.

12. Protection against packet sniffers

Even though configuring the local area network can make sniffing hard, there is pretty much nothing one can do in order to stop people from out on the Internet from sniffing the traffic.

A lot of people believe that a good way to protect your network against sniffers is to segment it as much as possible using Ethernet switches instead of regular hubs. They claim that the only drawback to this solution is cost. As mentioned earlier this is not enough. It can make sniffing harder, but it cannot make sniffing impossible.

But there are some other ways one can be protected if not from letting hackers sniff the traffic, from letting hackers read the traffic. The following paragraphs describe such methods.

12.1. Encryption

Encryption is the best protection against any form of traffic interception. It is reasonable to assume that at some point along a path, data can always be compromised. Therefore, your best defense is to ensure that traffic is essentially unreadable to everyone but the intended receiver. This isn’t difficult to do, since many organizations have deployed services that make use of Secure Socket Layers (SSL), Transport Layer Security (TLS) and other methods that provide secure messaging, web browsing and more. Only the payloads are scrambled, ensuring that packets reach the correct destinations. So an attacker can see where traffic was headed and where it came from, but not what it carries.

         21:09:04.599289 192.168.1.3.32933 > opensource-01.ee.ethz.ch.https: . [tcp sum ok] 

         793:793(0) ack 7011 win 20104         (DF) (ttl 64, id 12206, len 40)

         0x0000  4500 0028 2fae 4000 4006 c059 c0a8 0103      E.(/.@.@..Y....

         0x0010  8184 0799 80a5 01bb 19a2 0520 be10 d77f      ................

         0x0020  5010 4e88 dfd0 0000                          P.N.....

         21:09:04.599289 opensource-01.ee.ethz.ch.https > 192.168.1.3.32933: P [tcp sum ok]

         7011:7135(124) ack 793 win 10052 (DF) (ttl 237, id 65192, len 164)

         0x0000  4500 00a4 fea8 4000 ed06 43e2 8184 0799      E.....@...C.....

         0x0010  c0a8 0103 01bb 80a5 be10 d77f 19a2 0520      ................

         0x0020  5018 2744 8303 0000 4d3a a587 805e e2bc      P.'D....M:...^..

         0x0030  9a2a 8ff3 fe95 46d4 930e b2bc 74f0 a484      .*....F.....t...

         0x0040  fcae 33ad 6d1f 0198 6020 aee5 0c26 908e      ..3.m...`....&..

         0x0050  a1b5 17b4 84b7 44bc 1b0b 434e bbae a483      ......D...CN....

         0x0060  1e23 38d3 520f 687e c5e3 b62e 5225 aa2f      .#8.R.h~....R%./

         0x0070  f747 1a71 669c 8fd1 55bd 511c 4988 b78a      .G.qf...U.Q.I...

         0x0080  a08d 554e a3fe bb7d 36ca e66b fb8b 0392      ..UN...}6..k....

         0x0090  a3f3 4cef 7b04 af5a 7a94 cb4c a1e6 e7fa      ..L.{..Zz..L....

         0x00a0  9610 a5ee                                             ....

Compare this sniffed sample of a web session with the OpenSSL Web server to the example earlier in the article. Notice how the header information remains readable, but the ASCII decode of the payload contains seemingly random characters - thanks to the encryption. The two participants in this exchange, however, can both decrypt and process the data once it is received. This type of safeguard can be applied to virtually any network process and should be employed whenever possible.

Using encryption does not disable the sniffer monitoring the traffic, but makes the traffic useless since it is displayed as garbled data. The drawbacks of using encryption are the speed and the chance of breaking a weak encryption standard. Almost all encryption will introduce delay into the network. Typically, the stronger the encryption, the slower the machines using it will communicate. On the other hand, there is always the chance that wrong choice of key size, of encryption algorithm or encryption mode will allow hackers to figure out the encryption key and decrypt the encrypted data. That is why system administrators and users have to compromise somewhere in the middle.

Yet still this method does not prevent traffic analysis.

12.2. Authentication Protection

12.2.1 Kerberos

Kerberos is an authentication service that performs two-way authentication between any pair of entities. The advantage of using Kerberos is that clear text passwords are not sent on the network. Besides that, client passwords are not stored on servers. But, still it remains a fact that this is just an authentication service; it can keep the login information secret, but everything else is sent on clear. This is why an intruder can see the contents of the rest of the packets, which might include confidential information (like credit card numbers). Moreover, this method is not effective against traffic analysis. Finally, it has to be mentioned that Kerberos is vulnerable to password guessing attacks.

12.2.2 One time password technology

One-time passwords provide secure authentication, over an insecure channel. The user takes a list of usable pass codes and scratches out each one as it is used. The system

keeps track of the usage, removing each pass code from the list when it is used. A lot of

different schemes are used to generate one-time passwords.

Some of them are Lamport’s scheme, challenge-response scheme, and time-synchronized scheme. Once again, since one time password technology provides authentication, it makes invisible to an intruder the login information but not the packets exchanged afterwards. Moreover, this method does not protect against traffic analysis.

12.2.3 Virtual Private Networks (VPNs)

The most secure solution is establishing an encrypted VPN. VPNs provide encrypted traffic across the Internet. Besides that, this way one can be protected against traffic analysis, since all packets destined to different recipients can be sent through the

computer or computers that one has established a VPN with.

There is though the threat that if an attacker compromises the end-nodes of a VPN connection, he can sniff the traffic. A typical scenario is an end-user who surfs the Internet normally and gets compromised with a Remote Access Trojan (RAT) that contains a sniffing plug-in. When the user establishes the VPN connection, the sniffing program is able to see not only the encrypted traffic that can be seen on the Internet, but also the unencrypted traffic before it gets sent to the VPN.

13. Sniffing Tools

List some of the commonly available sniffers for linux.

1. tcpdump: The granddaddy of packet sniffers. Ships by default on many linux distros!
2. sniffit:Robust packet sniffer with good filtering.
   
   
3. ethereal: A free network protocol analyzer for Unix and Windows. It allows you to examine data from a live network or from a capture file on disk.
   
   
4.  hunt: According to Pavel Krauz, the main goal of the HUNT project is to develop tool for exploiting well known weaknesses in the TCP/IP protocol suite. Well I think he comes pretty close to it.
   
   
5. Ettercap: Ettercap is a sniffer specifically designed for switched LANs. It allows you to peform mitm attacks against SSH and SSL. It has password collector for telnet, ftp, POP, rlogin, ssh1, icq, smb, mysql, http, NNTP, X11, napster, IRC, rip, bgp, socks 5, IMAP 4, VNC, LDAP, NFS etc.
   
   
6. dsniff:dsniff is a collection of tools for network auditing and penetration testing. dsniff, filesnarf, mailsnarf, msgsnarf, urlsnarf, and webspy passively monitor a network for interesting data (passwords, e-mail, files, etc.). arpspoof, dnsspoof, and macof facilitate the interception of network traffic normally unavailable to an attacker (e.g, due to layer-2 switching). sshmitm and webmitm implement active monkey-in-the-middle attacks against redirected SSH and HTTPS sessions by exploiting weak bindings in ad-hoc PKI.
7. LCRZOEX: It is a network library for administrators that supports spoofing,  sniffing, client and server creation. Over 200 possible examples are included in the package.

14. Programs to Detect Sniffers

1. Anti Sniff: From the L0pht Heavy Industries, comes the new program Anti Sniff. It has the ability to monitor a network and detect if a computer is in promiscuous mode.
    
2. Neped: It detects network cards on the network who are in promiscuous mode by exploiting a flaw in the ARP protocol as implemented on Linux machines.
3. ARP Watch: ARPWatch keeps track of ethernet/IP address pairings. This is useful when you suspect you are being arp-spoofed.
4. Snort: Snort is an excellent Intrusion Detection System and its arp-spoof preprocessor can be used to detect instances of  ARP Spoofing, which might be an indication that someone on the network is Sniffing.

None of these programs are foolproof. This has been the personal experience of many System Administrators.

15. Popular packet sniffers

Below are listed some popular packet sniffers of the public domain. They are divided into

categories based on the Operating System they are written for.

I. Windows

1. Ethereal: Ethereal is a UNIX-based program that also runs on Windows. It is probably

the best freeware solution available for sniffing on Windows.

2. WinDump: A version of “tcpdump” for Windows.

3. Network Associates Sniffer

4. WinNT Server: Microsoft's WinNT Server comes with a built-in program called

"Network Monitor". "Network Monitor Tools and Agent" is in the "Services" of the

Networking control panel.

5. BlackICE Pro: BlackICE is an intrusion detection system that can also log evidence

files to disk in a format that can be read by other protocol analyzers. However, it is non-promiscuous, and only sniffs the packets going into/out-of the machine.

6. EtherPeek

7. Intellimax LanExplorer

8. Triticom LANdecoder32

9. dsniff: dsniff is a collection of tools for network auditing and penetration testing. dsniff, filesnarf, mailsnarf, msgsnarf, urlsnarf, and webspy passively monitor a network for interesting data, arpspoof, dnsspoof, and macof facilitate the interception of network traffic normally unavailable to an attacker, sshmitm and webmitm implement active monkey-in-the-middle attacks against redirected SSH and HTTPS sessions by exploiting weak bindings in ad-hoc PKI. Powerful tool. Probably the best available.

II. Macintosh

1. EtherPeek

III. Unix

1. tcpdump: The oldest and most common wiretap program. It is the standard for UNIX

2. Ethereal: It currently looks like this is the best GUI-based sniffing program for UNIX.

3. sniffit: Useful when trying to analyze application-layer data.

4. snort: A libpcap based packet-sniffer/logger with extensive filtering.

5. trinux: Contains tcpdump and sniffit, among numerous other security utilities on a

floppy bootable disk.

6. karpski: A GUI Linux packet sniffer including a GTK interface.

7. SuperSniffer v1.3

IV. Linux, Solaris and SunOS

1. RealSecure: Real time monitoring, attack recognition and response.

2. SniffIt

3. dsniff: For Linux and Solaris. Described above, since available for Windows, as well.

V. SunOS, Dec-Mips, SGI, Alpha, and Solaris

1. Packetman, Interman, Etherman, Loadman: Packetman was designed to capture packets, while Interman, Etherman, and Loadman monitor traffic of various kinds

VI. Solaris

Since DOS is not an Operating System, it is in some ways more flexible as platform for

running sniffers.

1. Sniffer(r) Network Analyzer

2. Gobbler

3. Klos PacketView

4. ethdump v1.03

5. ethload v1.04

It has to be noted that there are also commercial Sniffers.

16. CONCLUSION:-

When computers communicate over networks, they normally listen only to the traffic destined to them. However, they also have the ability to enter promiscuous mode, which allows them to listen to the traffic that is destined to other computers. Sniffers put a computer’s network interface into promiscuous mode, allowing their users to view unauthorized information going through the Internet. They can see usernames, passwords, credit card numbers; any sensitive information. From this point, sniffing becomes a major threat for one’s privacy and has to be taken seriously into account. Computer users that are connected into the Internet have to protect themselves against intruders. Unfortunately, the way Internet is constructed favors the eavesdroppers. If they are capable enough and persistent enough, they can overcome any kind of protection. No solution is perfect. Still, though, it does not mean that no perfect protection equals to no protection. On the contrary, good protection will make an attacker’s life difficult, discouraging him, most of the times. Easier victims will be seeked.

In fewer words, Sniffers are an invisible major threat that has to be taken into account. Even though there is no perfect solution, anyone connected into the Internet has

to take the appropriate measures against it.