Home > Articles > Security > Network Security

Dissecting Snort

This chapter is from the book

This chapter is from the book

Preprocessors

Snort's preprocessors fall into two categories. They can be used to either examine packets for suspicious activity or modify packets so that the detection engine can properly interpret them. A number of attacks cannot be detected by signature matching via the detection engine, so "examine" preprocessors step up to the plate and detect suspicious activity. These types of preprocessors are indispensable in discovering non-signature-based attacks. The other preprocessors are responsible for normalizing traffic so that the detection engine can accurately match signatures. These preprocessors defeat attacks that attempt to evade Snort's detection engine by manipulating traffic patterns.

Additionally, Snort cycles packets through every preprocessor to discover attacks that require more than one preprocessor to detect them. If Snort simply stopped checking for the suspicious attributes of a packet after it had set off an alert via a preprocessor, attackers could use this deficiency to hide traffic from Snort. Suppose a black hat intentionally encoded a malicious remote exploit attack in a manner that would set off a low priority alert from a preprocessor. If processing is assumed to be finished at this point and the packet is no longer cycled through the preprocessors, the remote exploit attack would register only an encoding alert. The remote exploit would go unnoticed by Snort, obscuring the true nature of the traffic.

Preprocessor parameters are configured and tuned via the snort.conf file. The same snort.conf file lets you add or remove preprocessors as you see fit.

frag2

The frag2 preprocessor is Snort's weapon against IP fragmentation attacks. Fragmentation is a normally occurring phenomenon in IP networks. It is necessary to successfully send traffic over different types of network media. Different network-based protocols have divergent rules for the maximum allowable size or maximum transmission unit (MTU) for datagrams on their networks.

Fragmentation occurs normally when a packet's payload exceeds the MTU. On an ethernet network, any IP datagram larger than 1,500 bytes has to be fragmented. Fragmentation does not have to happen exclusively at the origination point; it can occur at an intermediate router.

After packets are fragmented, they must be reassembled at the target host. It is the reassembly process that hackers exploit to perpetrate attacks. The receiving host expects the sender to follow some rules when fragmenting. They are:

  • The receiving host reassembles packets by associating a fragment with an identical fragment identification number, or fragment ID. The fragment ID is a copy of the IP identification number in the IP header.

  • Each fragment must carry its position or offset in the original unfragmented packet (that is, the first fragment will have an offset of 0).

  • Each fragment must display the amount of data carried in the corresponding fragment.

  • If the fragment is not the last fragment to be received, it must flag the more fragments (MF) bit.

As you can see, any host that expected these rules to be followed at all times could be exploited by a malicious hacker. One simple attack devised to bypass firewalls and filtering devices uses fragmentation to overwrite TCP header data. The attack works by sending a fragmented TCP packet with header information that is allowed through the firewall. Subsequent fragments contain malicious data that would not otherwise be allowable through the firewall. If the fragmentation offset is small enough, though, the malicious packet overwrites the header information, allowing the malicious traffic through the firewall. An example is overwriting the header to change the destination port number. The attacker might change traffic from normal port 80 HTTP traffic to attack RPC on TCP port 111.

A good proportion of fragmentation attacks are DoS attacks. Host operating systems have been discovered to react unexpectedly to crafted fragmented packets. Probably the most famous type of these attacks was the Ping of Death attack. This attack used many small fragmented ICMP packets, which, when reassembled, exceeded the maximum allowable size for an IP datagram of 65,535 bytes. This caused most operating systems vulnerable to the Ping of Death to crash unexpectedly.

A hacker tool named fragroute uses fragmentation techniques to hide malicious traffic from Snort. Fragroute combines many of the previously known fragmentation evasion tactics into one tool. The various techniques can be mixed and matched together, making reconstruction of the fragmented attacks difficult. Fortunately, frag2 can now detect these types of attacks with some additional configuration options.

The frag2 preprocessor can detect attack types that are related to fragmentation, whether they are IDS evasion techniques or malicious DoS attacks. The frag2 preprocessor is important and should never be disabled. There are five options for configuring frag2. They are described in the following sections.

timeout Seconds

This option sets the number of seconds that a fragment is to be saved. If the fragment is not completed in the defined time allowance, it is dropped. Sixty seconds is the default, and should work well for most networks. If you are monitoring a network prone to IP fragmentation, such as a network with an abundance of NFS traffic, it is recommended that you reduce this threshold to avoid false positives. Most external monitoring sensors should not see much fragmented traffic. Some crafty black hats know that 60 seconds is the default setting for frag2, and use this information to evade frag2. It is recommended that you increase this threshold slightly to 65 seconds to combat this.

memcap bytes

The memory cap option limits the amount of memory the frag2 preprocessor can utilize. The default is 4,194,304 bytes, which is suitable for most fragmentation situations. Once again, if you are monitoring a network with heavy fragmentation traffic, it would be prudent to increase this amount.

detect_state_problems

This simple configuration option enables detection of overlapping fragments. Overlapping fragments have an offset that overwrites a previous fragment. They are used in firewall bypassing attacks, described previously in this section.

min_ttl number

The min_ttl configuration option specifies the minimum time to live (TTL) that frag2 will accept. Anything lower than [number] will be dropped. The default setting is 0. TTL is a death counter for packets. Every time a router forwards a packet, it decrements the TTL value and forwards it on. If the TTL reaches 0, the packet is dropped and no longer routed. TTLs were put in place to avoid situations where packets would get caught in endless loops, which could potentially create a Denial of Service situation.

Tools such as fragroute use small TTL values to set the expirations for datagrams before they reach the target host, but Snort still picks them up. This scenario is possible when a sensor lies at an intermediate point between the attacker and the target. If the target is two hops away from the sensor, the sensor and not the target interprets any packets with a TTL of 1. Packets with a TTL of 1 expire and are not forwarded on to the target. This can have the effect of inserting garbage data into an attack signature, making the signature not discernable by Snort. Snort is unable to detect the attack. Imagine that three packets make up an attack on an FTP server:

Packet 1 - QUOTE SITE         - TTL = 55

Packet 2 – inserted_garbage data      - TTL = 1 

Packet 3 - EXEC EXEC echo toor::0:0::/:/bin/sh >> /etc/passwd - TTL = 122

Snort sees the attack as the following:

QUOTE SITEinsert_garbage dataEXEC EXEC echo toor::0:0::/:/bin/sh >> /etc/passwd 

Because of the insert garbage data, Snort cannot match a rule to this signature. However, the host still sees the real attack because the packet containing the garbage data will expire and not be forwarded:

QUOTE SITE EXEC EXEC echo toor::0:0::/:/bin/sh >> /etc/passwd 

Dropping packets that have a very low TTL can avert this type of evasion. The recommended setting is 3.

ttl_limit number

This option specifies the maximum difference in TTL values that fragmented packets with the same fragment ID can have. The default is 5. This option was added because tools like Fragroute utilize inconsistent values for fragmented packets to evade Snort. They are vastly different than what is found in normal Internet traffic. The average change in TTL values for normal Internet traffic is somewhere in the 5–7 range. Leaving this set at the default of 5 creates some false positive noise; it is recommended to set ttl_limit to 8.

stream4

The stream4 preprocessor is what Snort uses to maintain the state of TCP streams, which is used in detecting some types of information gathering attacks. Stateful inspection with stream4 helps Snort better match attack signatures across multiple packets. This is important because attacks can be spread over many individual packets or exhibit anomalies in terms of TCP connections. The stream4 preprocessor also lets you create rules that are state-aware via the flow option.

A simple example can be demonstrated with the WU-FTPDexploit. A vulnerability was discovered for all versions of WU-FTPD, which is included by default in most major Linux distributions. With this exploit, any user who can log into a vulnerable version of the WU-FTPD server can execute arbitrary code remotely with root access. The vulnerability is due to a combination of coding errors, one located within the function responsible for the remote command globbing feature, which fails to properly signal an error to its caller under certain conditions. The globbing function does not properly handle the string "~{" as an illegal parameter.

To detect the first phase of this attack, a rule must be created to trigger any time a signature matched "~{". This works fine if the entire content of the attack is in the same packet. If the attacker causes the signature to be split up, so that ~ is in one packet and the { is in the next, the signature does not trigger. With stream4, Snort stores the ~ portion of the signature. When the attacker sends the final piece, {, in a subsequent packet, Snort completes the signature and an alert is generated. Without stream reassembly this would not be possible.

The stream4 preprocessor can also detect attackers doing unusual things with TCP streams in an attempt to hide information-gathering traffic. One such method of reconnaissance is a TCP stealth or half-open scan. In this method, the TCP three-way handshake is never fully completed. The attacker first sends a SYN packet as normal. When a SYN-ACK is received from the host, the service is assumed to be listening. The attacker never completes the handshake by responding with the expected ACK packet. This type of scanning could trick an IDS that is not monitoring state.

The stream4 preprocessor is the next evolutionary step up from the stream preprocessor. It had to be expanded to deal with a type of stateless attack specifically created to create a Denial of Service condition with Snort. Snot attacks take a Snort's own ruleset and generate a flood of randomly chosen attack signatures. When this flood is aimed at Snort, it creates a massive amount of false positives. A Snot attack can be used to bury a legitimate attack in the noise. When a Snot attack is run, it completes only the first phase of the three-way handshake, similar to the stealth scan. You can configure stream4 in conjunction with a command-line option to defeat Snot attacks. With the –z command-line option specified with the "est" argument, Snort ignores TCP streams that do not demonstrate normal bidirectional activity.

The stream4 preprocessor was also rewritten to incorporate higher bandwidth conditions. With modern hardware, stream4 can handle stateful inspection for up to 64,000 simultaneous connections. stream4 supports 10 different configurable options, described in the following sections.

detect_scans

This option simply enables stream4 for portscan detection. This option detects normal TCP connect scans and stealthy scans, such as the Xmas Tree, Half Open, and SYN-FIN scans. It can detect a variety of other attempts at stealthy scanning.

detect_state_problems

With this option enabled, stream4 alerts on various state inconsistencies that black hats have in their arsenals. One of the many possible attacks this configuration option detects is a TCP reset packet evasion attempt. This works by sending forged TCP reset packets each time a TCP connection is made. If this option is not enabled, Snort might perform stream reassembly incorrectly and assume that the connection has been terminated. Snort would then think that the session had been torn down and not log future data relating to this stream. Malicious traffic could be hidden from Snort in this manner.

disable_evasion_alerts

This configuration option disables some of the alerting that Snort does for aged, less widely used IDS evasion attempts. TCP overlapping is one of these methods. It works by creating connections with overlapping but incongruous data. This may cause older intrusion detection systems to misinterpret the intent of the connection. TCP overlapping was used to create both false positives and negatives in older IDSs. Recent TCP implementations of Windows 2000 and XP send status information in this manner, which results in a noticeable number of false positives. For this reason, it may be advisable to enable this option if the number of false positives becomes too much to bear. With any of the options beginning with "disable" in Snort, it is always best to leave them enabled unless they present an unmanageable number of false positives.

min_ttl number

Nearly the same as the option in frag2. This option drops packets with a TTL less than number. A short TTL attack has a short enough TTL set for some packets in a stream. A router may then cause some of the packets with TTLs of zero to expire before they reach the target. If Snort is only one hop away from the target and does not check the TTLs, it will reassemble the TCP stream badly. The recommended setting for number is 3.

ttl_limit number

This option specifies the maximum difference in TTL values that fragmented packets with the same fragment ID can have. The default is 5. Fragroute utilized inconstant TTL values for fragmented packets that were way out of the range of normal Internet traffic. The average change in TTL values for normal Internet traffic is somewhere in the 5–7 range. Leaving this set at the default of 5 creates some false positive noise; it is recommended to set number equal to 8.

noinspect

This option disables stateful inspection of TCP streams. The noinspect command should be utilized only in special situations, such as testing in very high bandwidth or limited hardware.

keepstats machine or binary

This option logs session summary information to session.log in Snort's log directory. The machine switch logs in a flat file; the binary switch logs in unified binary output format. This log is useful for additional event correlation and forensics.

timeout seconds

This option sets the number of seconds for a which a session is to be saved. If the session is not completed in the defined time allowance, it is flushed. Thirty seconds is the default. If you are monitoring a network with heavy session traffic, it is recommended to increase this threshold to avoid false negatives. Some crafty black hats know that 30 seconds is the default setting for stream4, and use this to evade stream4. It is recommended that you increase this threshold slightly to 35 seconds to combat this.

memcap bytes

The memory cap option limits the amount of memory the stream4 preprocessor can utilize. The default is 8,388,608 bytes, which is suitable for normal networks. Once again, if you are monitoring a network with heavy session traffic, it would be prudent to increase this amount.

log_flushed_streams

This option is used to log complete streams to disk that contain a packet that matches a signature. Without this option, Snort logs only the specific packet(s) that match the signature. This option works only when Snort is run in pcap mode.

stream4_reassemble

The stream4_reassemble preprocessor is closely related to stream4. It performs TCP stream reassembly as described earlier. Malformed packet injection is one of the many reassembly attacks. Malicious hackers attack the TCP stream reassembly process by creating packets with some valid header data such as sequence numbers, ACK numbers, source ports, and destination ports. The forged packets have a bogus payload and a checksum that fails. These crafted packets are inserted between totally valid packets. The target drops the injected crafted packets, and IDSs that do not do TCP checksum verification when reassembling packets miss the attack. Fortunately, Snort has stream4_reassemble to catch these attacks.

The stream4_reassemble preprocessor has five different configuration options, described in the next few sections.

clientonly

The default setting for stream4_reassemble is clientonly. This setting reassembles sessions only if they originate from the client side. The client network is defined elsewhere in the snort.conf file. Chapter 6 examines the definition of clients, servers, and other important hosts.

serveronly

The inverse of the clientonly option is serveronly. The stream4_reassemble preprocessor reassembles only sessions originating from the server side.

both

This reassembles from both client and server directions.

ports list

The ports option is used to specify which TCP destination port traffic is reassembled. The default is set for ports 21/FTP, 23/Telnet, 25/SMTP, 53/DNS, 80/HTTP, 143/IMAP, 110/POP, 111/RPC, and 513/rlogin. You can enable this port list by specifying default for the [list] parameter. You can enable reassembly for every port by specifying all. If you have a custom list of TCP ports on which you would like reassembly, you can list them, separated by spaces.

noalerts

This option disables alerting for reassembly attacks and evasions. If you are generating an unhealthy number of false positives related to TCP session reassembly, this should be your last resort. It is recommended that you attempt to narrow down the traffic with the ports, serveronly, and clientonly configuration options before disabling alerts altogether.

HTTP_decode

The HTTP_decode preprocessor is responsible for detecting abnormal HTTP traffic and normalizing it so that the detection engine can properly interpret it. Normalizing traffic is the process of translating an obscure character set, such as Unicode or hex, to a character set that Snort can recognize. This is necessary for Snort to be able to match signatures to malicious content. HTTP_decode works specifically with the URI string of an HTTP request. It generates an alert if it encounters traffic that requires decoding.

Encoding or obfuscating HTTP traffic is a method that hackers can use to disguise an attack from an IDS or even the human eye. Without HTTP_decode, an attack that Snort would normally catch can be obfuscated in a manner that does not match a signature, but that the target Web server still accepts as a valid URL string. Using Microsoft's IIS %u encoding, an .ida attack that would normally match a signature can be easily hidden. Suppose you have a Snort rule designed to trigger on any URI content that matches .ida. An attacker could use %u encoding to hide the a and evade Snort with the following URL request:

GET /vulnerable.id%u0061 HTTP/1.0

After the request evaded Snort, IIS would translate this to

GET /vulnerable.ida HTTP/1.0

Encoding URLs in this manner can be used to obfuscate any type of malicious HTTP URI request. Encoded URLs are also used to coax unsophisticated Internet users into clicking on malicious hyperlinks. Cross-site scripting attacks sometime require a user to click on a link for an authentication token to be delivered to a host controlled by the attacker. These links are often encoded in hex to increase the chances of success. The HTTP_decode preprocessor can detect users clicking on hex-encoded links.

Moving beyond obfuscation, some malicious attacks utilize encoded strings. Programmers implementing application-level access controls and bug fixes often forget about encoded URI strings. The Unicode directory traversal sample attack in Chapter 1 is a good example of a coding error that resulted in a security exposure. Even though directory traversal attacks had been patched, they were still vulnerable when the / symbol was encoded in Unicode.

HTTP_decode has five different configuration options that relate to different encoding strategies attackers employ.

port list

This option is the list of source ports where HTTP_decode is to look for encoded HTTP URI requests. The default is 80. Any Web server ports should be listed here, delimited by a space.

unicode

Specifying the unicode option normalizes all Unicode strings to ASCII. Leave this option off if you do not want alerts for Unicode URI traffic.

iis_alt_unicode

Enabling this option allows HTTP_decode to detect %u encoding. Microsoft's IIS supports a little-known, proprietary method of HTTP URI encoding known as %u encoding. The legitimate purpose of this %u encoding was to represent true Unicode character strings. Because %u encoding is proprietary and not an IETF standard, most IDS systems are unaware that %u encoding existed. Attackers can use %u encoding to hide IIS attacks that would have otherwise triggered an alert. Not including this configuration option causes Snort to not generate alerts for %u encoding.

double_encode

This option detects possible double encoding attempts and normalizes double decoding attempts. In a double encoding attack, the cracker forwards an encoded HTTP URI request that has been through two rounds of encoding. The IDS detects the first round of encoding and normalizes it. The attack still evades the IDS, however, because the result of the first normalization outputs an encoded string that does not match a signature. Without this option enabled, Snort decodes a string only once and assumes there is no more nested encoding. For example, if an attacker wanted to pass in the infamous \ character with double encoding, she would encode the character in hex as %5c. To double encode, the attacker hex encodes the %5c string again. The attacker hex encodes % to %25, 5 to %35, and c to %63, resulting in %25%35%63. Thus, the cracker has double encoded the \ character as %25%35%63.

iis_flip_slash

This option detects and decodes directory separators that are obscured through the use of a Microsoft proprietary directory separator. Microsoft OSs separate directories using a \ instead of the standard /. A HTTP request requires that the direction separator be a /. This means that IIS, as well as all other Win32-based Web servers, must convert the directory separator to a /. IIS still allows the use of \ in HTTP requests, because it is considered a valid directory separator. With IIS, directory separators can be encoded as /scripts\admin.asp, which does not match a typical /scripts/admin.asp signature.

full_whitespace

This option generates alerts and normalizes strings that are obscured by the use of the /t method for formatting HTTP requests to an Apache Web server. A standard HTTP request uses spaces to separate different portions of the request. It looks like this:

method <whitespace> URI <whitespace> HTTP/ Version

The <whitespace>s are used as delimiters for extracting specific portions of the HTTP request, such as the URI string. Snort utilizes this method for extracting URI strings to match to signatures.

Apache 1.3.6 and newer permits a different nonstandard request:

method /t URI /t HTTP/ Version 

This request evades any IDS that expects an HTTP request to be separated by <whitespace> rather than /t or a <tab>.

RPC_decode

This preprocessor has a similar function as HTTP_decode, but with another protocol: RPC. RPC can be used by black hats for both reconnaissance and remote exploit attacks. Attackers can use port mapping applications, such as rpcbind and portmapper, that make dynamic binding of remote services possible. The attacker can use information gathered from rpcbind to find additional targets for buffer overflows, or the attacker can attack the RPC service itself.

Malicious hackers desiring to hide RPC traffic can break up the RPC signature. The RPC signature, 0186A0, can be split up to obscure an RPC request. If the signature is spread out over many packets, Snort cannot match a valid signature. For this reason, RPC requests are normalized with the RPC_decode preprocessor. RPC_decode has one configuration option: [ports list] is the list of RPC ports to be normalized by RPC_decode. The default is 111 and 32771.

BO

The BO preprocessor detects Back Orifice. Back Orifice is a remote control Trojan for Windows systems. Back Orifice was a very popular Trojan when it was released, but has recently been replaced by other more advanced Trojans. Back Orifice still remains a considerable threat; it can be used by script kiddies to take total control of a remote system.

This preprocessor detects Back Orifice UDP traffic when commands are issued to the Trojan. It works by detecting the "magic cookie" that Back Orifice servers and clients require to communicate with each other. A fairly weak encryption algorithm is used to encrypt BO traffic. The entire keyspace of the Trojan's encrypted communication protocol can be brute forced with little effort. The BO preprocessor supports the brute forcing of Back Orifice's entire keyspace. It can also be configured to search for packets that use the default installation encryption key, 31337, to search for packets. Searching solely for the default key improves the performance of BO and Snort in general, but allows Back Orifice communication to slip by unnoticed if a key other than the default is used for encryption. BO supports two configuration options.

number

This option is used to set the default key used to decode Back Orifice–encrypted communication to something other than the default of 31337.

-nobrute

As explained previously, this option disables the brute forcing of the Back Orifice keyspace for a valid encryption key. This is the default setting. This setting should not be enabled unless BO is causing a significant performance impact on Snort.

Telnet_decode

This is another one of the family of decoding preprocessors. This preprocessor specifically relates to Telnet and FTP protocols. Telnet_decode decodes or removes arbitrarily inserted binary Telnet control codes in a Telnet or FTP stream. Malicious hackers insert control codes into communication in an attempt to evade Snort's watchful eye. Control code insertion is often performed with the SITE EXEC FTP command and its associated vulnerabilities. The FTP command SITE EXEC is used by attackers to execute system commands via an FTP connection. An attacker could enter an FTP command like so to retrieve a passwd file:

QUOTE SITE EXEC EXEC echo toor::0:0::/:/bin/sh >> /etc/passwd 

More sophisticated hackers intersperse Telnet control codes into the command hoping to evade Snort. Telnet_decode prevents this by removing or decoding the malicious control codes. Telnet_decode has no configuration options.

ARPspoof

ARPspoof is a preprocessor designed to detect malicious Address Resolution Protocol (ARP) traffic. ARP is used on ethernet networks to map an IP address to a hardware MAC address. To reduce the number of ARP broadcasts on modern networks, operating systems of connected devices store a cache of ARP mappings. When a device receives an ARP reply, it updates its ARP cache with the new IP-to-MAC address mapping whether or not the device sent the ARP request.

Various attacks involve ARP; the basis of them is ARP spoofing. ARP spoofing is accomplished by crafting ARP request and reply packets. Forged ARP reply packets are stored in the ARP cache of the receiving device even if the device did not send the request.

ARP spoofing can be used to misdirect traffic, making sniffing possible on switched networks. Forged ARP replies can trick a target device into misdirecting ethernet frames intended for a legitimate device to the attackers choice of device. The legitimate device is not aware that misdirection had taken place. ARP spoofing can also be used to perform a simple yet effective DoS. The DoS happens when spoofed ARP replies that contain invalid IP-to-MAC mappings flood a device's ARP cache. This causes the target device to send traffic to non-existent devices.

The ARPspoof preprocessor can detect some of these types of attacks and ARP spoofing attempts. ARPspoof detects spoofing attempts by determining that an ethernet source address is different from the one included in the ARP message, indicating a spoofed ARP reply. Conversely, ARPspoof detects a destination address that differs from the one in the ARP message, indicating a spoofed ARP request.

Another type of attack ARPspoof detects is the ARP cache overwrite attack. The attack works by sending ARP packets received by the device for the device's own interface address but a different MAC address. This overwrites the device's own MAC address in the ARP cache with the malicious ARP request. This causes the device to be unable to send and receive any ARP packet. In turn, this causes the device and any other devices that depend on it for communication to be unable to send packets to each other.

Because ARP is a Layer 2 protocol, ARPspoof detects only those attacks occurring on the same physical segment as the Snort sensor. ARPspoof has two configuration options.

host IP address host MAC address

Each device that you wish to monitor with ARPspoof must be specified with its own IP-to-MAC address mapping. Each one of the devices is listed on a new line in the snort.conf file. Any time the mapping changes, you must reconfigure the file. Devices that obtain their IP address via DHCP should be converted to static IPs before ARPspoof is enabled.

-unicast

This option will enable detection of ARP unicast attacks. Most valid ARP requests are sent to broadcast addresses. ARP requests that are sent to a Unicast address are often the sign of an attack designed to modify ARP caches. This option is disabled by default, but should be enabled for serious ARP misuse monitoring.

ASN1_decode

ASN1_decode is a preprocessor designed to detect various inconsistencies in ASN.1 that may indicate malicious behavior. ASN.1 or Abstract Syntax Notation One is an international standard for coding and transmitting complex data structures. ASN.1 protocol is used by a diverse set of higher-level protocols including LDAP, SNMP, SSL, and X.509. ASN.1 is also used for communication in proprietary government and industrial applications. Portions of the U.S. power grid are controlled remotely by the ASN.1 protocol. The SS7 network that controls telephone call routing, parcel delivery, and credit card verification systems all use ASN.1. Vulnerabilities associated with ASN.1-based networks are a matter of national security and are a major concern.

A Finnish group of researchers at Oulu University discovered a wide range of ASN.1-related vulnerabilities. They were able to exploit poor coding of ASN.1 on almost every SNMP-enabled device. The exploits operate by deliberately violating the ASN.1 standard in thousands of different ways. Most of these violations take the format of extreme oversizing or undersizing of ASN.1 messages. These cause the higher-level protocol dependent on ASN.1 to crash or create a situation where a buffer overflow occurs. This may have the effect of creating a Denial of Service condition or remotely controlling the attacked device.

The ASN1_decode preprocessor detects various attempts to attack ASN.1-enabled hosts by breaking the rules of the ASN.1 protocol. The OpenSSL ASN.1 DoS is an example of the type of attack ASN1_decode is designed to detect. OpenSSL had a vulnerability in its implementation of the ASN.1 library. This vulnerability, due to parsing errors, affects SSL, TLS, S/MIME, PKCS#7, and certificate creation routines. The certificate encodings can cause a Denial of Service to server and client implementations that depend on OpenSSL-to-secure communication over pubic networks.

ASN1_decode is an experimental beta preprocessor that may generate more false positives than other preprocessors. It has no configuration options at this time.

fnord

The fnord preprocessor is designed to detect and defeat polymorphic shell code evasion attempts. Polymorphic shell code is a new twist on an old idea. Virus creators have been designing polymorphic viruses for years. Polymorphic viruses change their signature randomly to avoid detection by static signature-based Antivirus applications. A similar method of randomizing remote exploit shell code was developed recently to evade IDSs, and released with the ADMutate tool.

ADMutate evades Snort or any other IDS by randomly polymorphing the buffer overflow signature by choosing random no-op instructions and encrypting the shellcode portion of the exploit. A buffer overflow is divided into two sections: a no-op pad or no-op sled and the shellcode payload. The no-op pad is a large section of no operation codes (that is 0x90 for Intel) specific to the target architecture. The shellcode payload is a portion of the buffer overflow where the shell gets executed and bound to a TCP port. Snort and other IDSs use the no-ops and the shellcode signature to detect buffer overflow exploits.

ADMutate performs two actions on the buffer overflow exploit to disguise it. It randomly replaces the no-op instructions with other no-effect instructions that are functionally similar to a no-op. It then encrypts the shellcode with a fairly simple encryption algorithm (usually xor or a double xor). The purpose of encrypting is not to protect the shellcode, but rather to obscure it in a random fashion. ADMutate has to include the decryption routine with the newly formulated exploit. The decryption routine is randomly created through the use of methods similar to those employed by polymorphic viruses. This process makes the no-ops, the shellcode, and the newly added decryption engine impossible to detect for normal signature-based IDSs. Any attacker using ADMutate on previously existing exploits can happily hack away without Snort noticing.

To defeat polymorphic shellcode attacks, the fnord preprocessor was created. It detects a large amount of no-effect instructions grouped together for Intel, Sparc, or HP hardware architectures. It is effective at detecting polymorphed exploits, but generates some false positives.

The fnord preprocessor is relatively new and is potentially computationally expensive. It has no configuration options.

conversation

The conversation preprocessor is similar to the stream4 and stream4_reassemble preprocessors. It is designed to extend session tracking beyond TCP. It tracks pseudo- sessions or conversations for any other IP protocol (such as ICMP). This is useful for detecting where communication has originated. It can also detect when new communication commences. The conversation preprocessor must be enabled for the portscan2 preprocessor to function.

As of now, conversation is used only in conjunction with portscan2 to detect information gathering attempts. It has three configuration options.

allowed_protocols port list or any

This is the list of IP protocols for which Conversation will track sessions. It can be configured with either a list of IP protocols or the any option. The number list corresponds to the IP protocol number defined in RFC 791. Some of the most common are 1/ICMP, 6/TCP, and 17/UDP. The any option tracks all protocols.

timeout

This option sets the number of seconds that a conversation will be saved. If the conversation is not completed in the defined time allowance, it is dropped. The default is 60 seconds. Some crafty black hats know that 60 seconds is the default setting for conversation and use this as an evasion technique. It is recommended to increase this threshold slightly to 65 seconds to combat this attack.

max_conversations numbers

This is the maximum conversations that conversation will support simultaneously. The default is set to 65,335, which is the recommended value.

portscan2

The portscan2 preprocessor is the successor to the original anomaly detection portscan preprocessor. It functions by anomaly detection as well. The original portscan detected reconnaissance attempts by alerting whenever packets were seen going to four different ports in less than three seconds. This set off a lot of false positives. Even though you could specify which IP addresses to ignore, a fair amount of Snort regulars would disable the preprocessor altogether. It was not a bad preprocessor by any means; it just needed some evolutionary improvements. Fortunately, portscan2 is here to avenge the death of his father.

The portscan2 preprocessor works in concurrence with conversation to track sessions for IP protocols. It has configurable options that enable it to detect different methods of scanning that are anomalous for most IP networks. It can detect vertical portscans, or scans that sweep the entire range of ports for one host. It can also detect horizontal portscans, which scan one port across many hosts. The portscan2 preprocessor is not as prone to false positives because it has more configuration options. It utilizes conversation, and is therefore aware of state. It has five different configuration options.

scanners_max number

This option sets the maximum number of possible conversations to track for reconnaissance attempts. The default setting is 3,200.

targets_max number

This is the maximum number of target hosts to be monitored. The default setting is 5,000. Even if you do not have this many possible targets, it is a good idea to keep this setting high. The attacker may be blindly scanning for IP addresses that have no host bound to them, and you would not want to miss this type of information gathering attempt.

target_limit number

This configuration option is the horizontal portscan threshold. It is the maximum number of unique hosts the conversation can attempt to access before generating an alert. The default is 5, and is the recommended setting.

port_limit number

This is the vertical portscan threshold. It is the maximum number of unique ports the conversation can attempt to access before generating an alert. The ports do not have to be on the same host to generate the alert. The default setting is 20, and is recommended.

timeout seconds

This configuration option is the length of time to hold a conversation in memory before flushing it. This is the amount of time you are giving portscan2 to generate an alert. The default setting is 60, and is recommended.

SPADE

SPADE (Statistical Packet Anomaly Detection Engine) is a preprocessor that detects suspicious traffic via anomaly detection methods. SPADE determines whether traffic is anomalous by building a table to describe normal traffic patterns. When traffic does not match data stored in the table, SPADE outputs an alert.

SPADE functions by assigning all incoming packets an anomaly score. High anomaly scores are assigned to specific packets that are rare in occurrence. SPADE defines a unique packet by combining packet IP and port information. For instance, a packet with a destination IP of 192.168.1.1 and destination port of 443 is defined as one packet.

A table is constructed that reflects the occurrences of different kinds of packets over time. Recent packets are given precedent over aged packets. Over time, an alerting threshold is defined for SPADE. For each packet that exceeds this threshold, an alert is generated.

SPADE has five configuration options.

anom-report-thresh

This is the threshold at which packets are reported when the Snort process starts. If a packet has an anomaly score that exceeds this threshold, an alert is generated. If configured with -1, SPADE starts with no default alerting threshold. The default is –1.

filename

This is the filename to which alerts will be written. If SPADE is used in conjunction with an output plugin, alerts are sent as expected through the plugins.

statefile

This file is used to store the state of SPADE's anomaly table. The statefile is used to keep the anomaly table intact when the Snort process is refreshed or restarted. You can set this option to 0 if you do not wish to store a statefile.

checkpoint-freq

The option to configure how often the statefile is refreshed. Checkpoint-freq is the number of packets to accept before refreshing the table.

probability-mode

This is a configurable list of probability modes to use in detecting anomalous behavior. The four different modes define a SPADE "packet" differently. You configure this by choosing one of the related numbers.

  • 0: A Bayes network approximation of the source IP, source port, destination IP, and destination port.

  • 1: Source IP, source port, destination IP, and destination port.

  • 2: Source IP, destination IP, and destination port.

  • 3: Destination IP and destination port.

The default setting is 3.

InformIT Promotional Mailings & Special Offers

I would like to receive exclusive offers and hear about products from InformIT and its family of brands. I can unsubscribe at any time.

Overview


Pearson Education, Inc., 221 River Street, Hoboken, New Jersey 07030, (Pearson) presents this site to provide information about products and services that can be purchased through this site.

This privacy notice provides an overview of our commitment to privacy and describes how we collect, protect, use and share personal information collected through this site. Please note that other Pearson websites and online products and services have their own separate privacy policies.

Collection and Use of Information


To conduct business and deliver products and services, Pearson collects and uses personal information in several ways in connection with this site, including:

Questions and Inquiries

For inquiries and questions, we collect the inquiry or question, together with name, contact details (email address, phone number and mailing address) and any other additional information voluntarily submitted to us through a Contact Us form or an email. We use this information to address the inquiry and respond to the question.

Online Store

For orders and purchases placed through our online store on this site, we collect order details, name, institution name and address (if applicable), email address, phone number, shipping and billing addresses, credit/debit card information, shipping options and any instructions. We use this information to complete transactions, fulfill orders, communicate with individuals placing orders or visiting the online store, and for related purposes.

Surveys

Pearson may offer opportunities to provide feedback or participate in surveys, including surveys evaluating Pearson products, services or sites. Participation is voluntary. Pearson collects information requested in the survey questions and uses the information to evaluate, support, maintain and improve products, services or sites, develop new products and services, conduct educational research and for other purposes specified in the survey.

Contests and Drawings

Occasionally, we may sponsor a contest or drawing. Participation is optional. Pearson collects name, contact information and other information specified on the entry form for the contest or drawing to conduct the contest or drawing. Pearson may collect additional personal information from the winners of a contest or drawing in order to award the prize and for tax reporting purposes, as required by law.

Newsletters

If you have elected to receive email newsletters or promotional mailings and special offers but want to unsubscribe, simply email information@informit.com.

Service Announcements

On rare occasions it is necessary to send out a strictly service related announcement. For instance, if our service is temporarily suspended for maintenance we might send users an email. Generally, users may not opt-out of these communications, though they can deactivate their account information. However, these communications are not promotional in nature.

Customer Service

We communicate with users on a regular basis to provide requested services and in regard to issues relating to their account we reply via email or phone in accordance with the users' wishes when a user submits their information through our Contact Us form.

Other Collection and Use of Information


Application and System Logs

Pearson automatically collects log data to help ensure the delivery, availability and security of this site. Log data may include technical information about how a user or visitor connected to this site, such as browser type, type of computer/device, operating system, internet service provider and IP address. We use this information for support purposes and to monitor the health of the site, identify problems, improve service, detect unauthorized access and fraudulent activity, prevent and respond to security incidents and appropriately scale computing resources.

Web Analytics

Pearson may use third party web trend analytical services, including Google Analytics, to collect visitor information, such as IP addresses, browser types, referring pages, pages visited and time spent on a particular site. While these analytical services collect and report information on an anonymous basis, they may use cookies to gather web trend information. The information gathered may enable Pearson (but not the third party web trend services) to link information with application and system log data. Pearson uses this information for system administration and to identify problems, improve service, detect unauthorized access and fraudulent activity, prevent and respond to security incidents, appropriately scale computing resources and otherwise support and deliver this site and its services.

Cookies and Related Technologies

This site uses cookies and similar technologies to personalize content, measure traffic patterns, control security, track use and access of information on this site, and provide interest-based messages and advertising. Users can manage and block the use of cookies through their browser. Disabling or blocking certain cookies may limit the functionality of this site.

Do Not Track

This site currently does not respond to Do Not Track signals.

Security


Pearson uses appropriate physical, administrative and technical security measures to protect personal information from unauthorized access, use and disclosure.

Children


This site is not directed to children under the age of 13.

Marketing


Pearson may send or direct marketing communications to users, provided that

  • Pearson will not use personal information collected or processed as a K-12 school service provider for the purpose of directed or targeted advertising.
  • Such marketing is consistent with applicable law and Pearson's legal obligations.
  • Pearson will not knowingly direct or send marketing communications to an individual who has expressed a preference not to receive marketing.
  • Where required by applicable law, express or implied consent to marketing exists and has not been withdrawn.

Pearson may provide personal information to a third party service provider on a restricted basis to provide marketing solely on behalf of Pearson or an affiliate or customer for whom Pearson is a service provider. Marketing preferences may be changed at any time.

Correcting/Updating Personal Information


If a user's personally identifiable information changes (such as your postal address or email address), we provide a way to correct or update that user's personal data provided to us. This can be done on the Account page. If a user no longer desires our service and desires to delete his or her account, please contact us at customer-service@informit.com and we will process the deletion of a user's account.

Choice/Opt-out


Users can always make an informed choice as to whether they should proceed with certain services offered by InformIT. If you choose to remove yourself from our mailing list(s) simply visit the following page and uncheck any communication you no longer want to receive: www.informit.com/u.aspx.

Sale of Personal Information


Pearson does not rent or sell personal information in exchange for any payment of money.

While Pearson does not sell personal information, as defined in Nevada law, Nevada residents may email a request for no sale of their personal information to NevadaDesignatedRequest@pearson.com.

Supplemental Privacy Statement for California Residents


California residents should read our Supplemental privacy statement for California residents in conjunction with this Privacy Notice. The Supplemental privacy statement for California residents explains Pearson's commitment to comply with California law and applies to personal information of California residents collected in connection with this site and the Services.

Sharing and Disclosure


Pearson may disclose personal information, as follows:

  • As required by law.
  • With the consent of the individual (or their parent, if the individual is a minor)
  • In response to a subpoena, court order or legal process, to the extent permitted or required by law
  • To protect the security and safety of individuals, data, assets and systems, consistent with applicable law
  • In connection the sale, joint venture or other transfer of some or all of its company or assets, subject to the provisions of this Privacy Notice
  • To investigate or address actual or suspected fraud or other illegal activities
  • To exercise its legal rights, including enforcement of the Terms of Use for this site or another contract
  • To affiliated Pearson companies and other companies and organizations who perform work for Pearson and are obligated to protect the privacy of personal information consistent with this Privacy Notice
  • To a school, organization, company or government agency, where Pearson collects or processes the personal information in a school setting or on behalf of such organization, company or government agency.

Links


This web site contains links to other sites. Please be aware that we are not responsible for the privacy practices of such other sites. We encourage our users to be aware when they leave our site and to read the privacy statements of each and every web site that collects Personal Information. This privacy statement applies solely to information collected by this web site.

Requests and Contact


Please contact us about this Privacy Notice or if you have any requests or questions relating to the privacy of your personal information.

Changes to this Privacy Notice


We may revise this Privacy Notice through an updated posting. We will identify the effective date of the revision in the posting. Often, updates are made to provide greater clarity or to comply with changes in regulatory requirements. If the updates involve material changes to the collection, protection, use or disclosure of Personal Information, Pearson will provide notice of the change through a conspicuous notice on this site or other appropriate way. Continued use of the site after the effective date of a posted revision evidences acceptance. Please contact us if you have questions or concerns about the Privacy Notice or any objection to any revisions.

Last Update: November 17, 2020