Table Of Contents

Flexible Packet Matching XML Configuration

Contents

Prerequisites for the Flexible Packet Matching XML Configuration

Restrictions for the Flexible Packet Matching XML Configuration

Information About the Flexible Packet Matching XML Configuration

Traffic Classification Definition Files for the Flexible Packet Matching XML Configuration

Protocol Header Definition Files for Traffic Classification Definitions

Traffic Classification Description File Format and Use

Traffic Class Definitions for a Traffic Classification Definition File

Class Element Attributes for a Traffic Classification Definition File

Match Element for a Traffic Classification Definition File

Operator Element Attributes for a Traffic Classification Definition File

Policy Definitions for a Traffic Classification Definition File

Policy Element Attributes for a Traffic Classification Definition File

Action Element for a Traffic Classification Definition File

How to Create and Load Traffic Classification Definition Files for the FPM XML Configuration

Creating a Traffic Classification Definition File for the FPM XML Configuration

Traffic Classification Definition File Syntax Guidelines

Loading a Traffic Classification Definition File for the FPM XML Configuration

Examples

What to Do Next

Associating a Traffic Classification Definition File with an Interface or Subinterface

Prerequisites

Displaying TCDF-Defined Traffic Classes and Policies

Configuration Examples for Creating and Loading Traffic Classification Definition Files for the FPM XML Configuration

Creating and Loading a Traffic Classification Definition File for Slammer Packets for the FPM XML Configuration: Example

Creating and Loading a Traffic Classification Definition File for MyDoom Packets for the FPM XML Configuration: Example

Additional References

Related Documents

Standards

MIBs

RFCs

Technical Assistance

Command Reference

Glossary

Feature Information for Flexible Packet Matching XML Configuration

Flexible Packet Matching XML Configuration

---

First Published: March 3, 2006

Last Updated: March 3, 2006

The Flexible Packet Matching XML Configuration feature allows the use of eXtensible Markup Language (XML) to define traffic classes and actions (policies) to assist in blocking network attacks. The XML file used by Flexible Packet Matching (FPM) is called the traffic classification definition file (TCDF).

The TCDF gives you an alternative to the command-line interface (CLI) as a method to define traffic classification behavior. Traffic classification behavior is identical regardless of the method you use.

Finding Feature Information in This Module

Your Cisco IOS software release may not support all of the features documented in this module. To reach links to specific feature documentation in this module and to see a list of the releases in which each feature is supported, use the "Feature Information for Flexible Packet Matching XML Configuration" section.

Finding Support Information for Platforms and Cisco IOS Software Images

Use Cisco Feature Navigator to find information about platform support and Cisco IOS software image support. Access Cisco Feature Navigator at http://www.cisco.com/go/fn. You must have an account on Cisco.com. If you do not have an account or have forgotten your username or password, click Cancel at the login dialog box and follow the instructions that appear.

Contents

•Prerequisites for the Flexible Packet Matching XML Configuration

•Restrictions for the Flexible Packet Matching XML Configuration

•Information About the Flexible Packet Matching XML Configuration

•How to Create and Load Traffic Classification Definition Files for the FPM XML Configuration

•Configuration Examples for Creating and Loading Traffic Classification Definition Files for the FPM XML Configuration

•Additional References

•Command Reference

•Glossary

•Feature Information for Flexible Packet Matching XML Configuration

Prerequisites for the Flexible Packet Matching XML Configuration

The Flexible Packet Matching XML Configuration feature has the following prerequisites:

•A protocol header definition file (PHDF) relevant to the TCDF must be loaded on the router.

•Although access to an XML editor is not required, using one might make the creation of the TCDF easier.

•You must be familiar with XML file syntax.

Restrictions for the Flexible Packet Matching XML Configuration

The Flexible Packet Matching XML Configuration has the following restrictions:

•The FPM TCDF cannot be used to mitigate an attack that requires stateful classification.

•Because FPM is stateless, it cannot keep track of port numbers being used by protocols that dynamically negotiate ports. Thus, when using the FPM TCDF, port numbers must be explicitly specified.

•FPM cannot perform IP fragmentation or TCP flow reassembly.

Information About the Flexible Packet Matching XML Configuration

Before you create and load the TCDF XML configuration files for use with FPM, you should understand the following concepts.

•Traffic Classification Definition Files for the Flexible Packet Matching XML Configuration

•Protocol Header Definition Files for Traffic Classification Definitions

•Traffic Classification Description File Format and Use

•Traffic Class Definitions for a Traffic Classification Definition File

•Policy Definitions for a Traffic Classification Definition File

Traffic Classification Definition Files for the Flexible Packet Matching XML Configuration

FPM uses a TCDF to define policies that can block attacks on the network. FPM is a packet classification feature that allows users to define one or more classes of network traffic by pairing a rich set of standard matching operators with user-defined protocol header fields. FPM users can create their own stateless packet classification criteria and define policies with multiple actions (such as drop, log, or send Internet Control Message Protocol [ICMP] unreachable) to immediately block new viruses, worms, and attacks on the network.

Before the release of the Flexible Packet Matching XML Configuration feature, FPM defined traffic classes (class maps), policies (policy maps), and service policies (attach policy maps to a class maps) through the use of CLI commands. With TCDFs, FPM can use XML as an alternative to the CLI to define classes of traffic and specify actions to apply to the traffic classes. Traffic classification behavior is the same whether you create the behavior using a TCDF or configure it using CLI commands. Once a TCDF is created, it can be loaded on any FPM-enabled device in the network.

For more information on FPM, see the Flexible Packet Matching feature module.

Protocol Header Definition Files for Traffic Classification Definitions

TCDFs require that a relevant PHDF is already loaded on the device. A PHDF defines each field contained in the header of a particular protocol. Each field is described with a name, optional comment, an offset (the location of the protocol header field in relation to the start of the protocol header), and the length of the field. The total length is specified at the end of each PHDF.

The description of a traffic class in a TCDF file can contain header fields defined in a PHDF. If the PHDF is loaded on the router, the class specification to match begins with a list of the protocol headers in the packet. In the TCDF, the traffic class is associated with a policy that binds the match to an action, such as drop, log, or send ICMP unreachable.

FPM provides ready-made definitions for these standard protocols, which can be loaded onto the router with the load protocol command: ether.phdf, ip.phdf, tcp.phdf, and udp.phdf. You can also write your own custom PHDFs using XML if one is required for the TCDF.

---

Note Because PHDFs are defined via XML, they are not shown in a running configuration.

---

For more information about PHDFs, see the Flexible Packet Matching feature module.

Traffic Classification Description File Format and Use

In the TCDF, you can define one or more classes of traffic and policies that describe specified actions for each class of traffic. The TCDF is an XML file that you create in a text file or with an XML editor. The file that you create must have a filename that has the .tcdf extension.

The TCDF has the following basic format. XML tags are shown in bold text for example purposes only.

<tdcf>

       <class ...> ... </class>

         ...

       <policy> ... </policy>

         ...

</tdcf>

For a traffic class, you can identify a match for any field or fields against any part of the packet.

---

Note FPM is stateless and cannot be used to mitigate an attack that requires stateful classification, that is classify across IP fragments, across packets in a TCP stream, or peer-to-peer protocol elements.

---

Policies can be anything from access control, quality of service (QoS), or even routing decisions. For FPM, the associated actions (policies) might include permit, drop, log, or send ICMP unreachable.

Once loaded, the TCDF-defined classes and policies can be applied to any interface or subinterface and behave in an identical manner as the CLI-defined classes and policies. You can define policies in the TCDF and apply then to any entry point to the network to block new attacks.

Traffic Class Definitions for a Traffic Classification Definition File

A class can be any traffic stream of interest. You define a traffic stream of interest by matching a particular interface or port, a source address or destination IP address, a protocol or an application. The following sections contain information you should understand before you define the traffic class in the TCDF for FPM configuration:

•Class Element Attributes for a Traffic Classification Definition File

•Match Element for a Traffic Classification Definition File

•Operator Element Attributes for a Traffic Classification Definition File

Class Element Attributes for a Traffic Classification Definition File

Table 1 lists and describes the attributes that you can associate with the class element in a TCDF for the FPM XML configuration. The class element contains attributes you can use to specify the traffic class name, its description and type, where to look in the packet, what kind of match, and when the actions should apply to the traffic.

Table 1 Attributes for Use with the Class Element in a TCDF for the FPM XML
Configuration 

Attribute Name	Use	Type
name (required)	Specifies the name of the class. Note When you use the class element inside policy elements, you need specify the name attribute only.	String
type (required)	Specifies the type of class.	Keywords: stack or access-control
stack start	Specifies where to look in the packet. By default, the match starts at Layer 3.	Keyword: l2-start
match	Specifies the type of match to be performed on the class.	Keywords: all or any •all—All class matches must be met to perform the policy actions. •any—One or more matches within the class must be met to perform the policy actions.
undo	Directs the device to remove the class-map when set to true.	Keywords: true or false

For example, XML syntax for a stack class describing an IP, User Datagram Protocol (UDP), Simple Management Protocol (SNMP) stack might look like this:

<class name="snmp-stack" type="stack">

    <match>

       <eq field="ip.protocol" value="x"></eq>

       <eq field="udp.dport" value="161"></eq>

    </match>

</class>

Match Element for a Traffic Classification Definition File

The match element in the TCDF for FPM XML configuration contains operator elements. Operator elements are the following: eq (equal to), neq (not equal to), lt (less than), gt (greater than), range (a value in a specific range, for example, range 1 - 25), and regex (regular expression string with a maximum length of 32 characters).

In following sections, these various operators are collectively called the operator element.

Operator Element Attributes for a Traffic Classification Definition File

Table 2 lists and describes direct matching attributes that you can associate with the operator element in a TCDF for the FPM XML configuration.

Table 2 Direct Matching Attributes to Use with a Match Element in a TCDF for the FPM XML Configuration 

Attribute Name	Use	Type
start	Begin the match on a predefined keyword or Protocol.Field, if given.	Keyword: l2-start or l3-start Otherwise, a field of a protocol as defined in the PHDF, for example, the source field in the IP protocol.
offset	Used with start attribute. Offset from the start point.	Hexadecimal or decimal number, or string constants, Protocol.Field, or combination of a constant and Protocol.Field with +, -, *, /, &, or |.
size	Used together with start and offset attributes. How much to match.	Specifies the size of the match in bytes.
mask	Number specifying bits to be matched in protocol or field attributes. Used exclusively with field type of bitset to specify the bits of interest in a bit map.	Decimal or hexadecimal number
value	Value on which to match.	String, number, or regular expression
field	Specifies the name of the field to be compared.	Name of field as defined in the PHDF
next	Identifies the next layer of the protocol. This attribute can be used only in stack type classes.	Keyword that is the name of a protocol defined in the PHDF.
undo	Directs the device to remove the particular match operator when set to true.	Keywords: true or false

Policy Definitions for a Traffic Classification Definition File

A policy is any action that you apply to a class. You should understand the following information before defining the policy in a TCDF for the FPM XML configuration:

•Policy Element Attributes for a Traffic Classification Definition File

•Action Element for a Traffic Classification Definition File

Policy Element Attributes for a Traffic Classification Definition File

Policies can be anything from access control, QoS, or even routing decisions. For FPM, the associated actions or policies might include drop, log, or send ICMP unreachable. Policies describe the action to take to mitigate attacks on the network.

Table 3 lists and describes the attributes that you can use with the policy element in the TDCF for FPM XML configuration.

Table 3 Attributes for Use with the Policy Element in a TCDF for the FPM XML Configuration 

Attribute Name	Use	Type
name	Name of the policy.	String
type	Specifies the type of policy map.	Keyword: access-control
undo	Directs the device to remove the policy map when set to true.	Keywords: true or false

The policy name in this example is sql-slammer, and the action defined for the policy is to drop the packet. This action is to be applied to the class that has the same name as the policy (class name= "sql-slammer").

<policy name="sql-slammer">

    <class name="sql-slammer"></class>

    <action>drop</action>

</policy>

Action Element for a Traffic Classification Definition File

The action element is used to specify actions to associate with a policy. The policy with the action element is applied to a defined class. The action element can contain any of the following: permit, drop, Log, SendBackIcmp, set, RateLimit, alarm, ResetTcpConnection, and DropFlow. For example:

<action>

   log

</action>

How to Create and Load Traffic Classification Definition Files for the FPM XML Configuration

Perform the following tasks to create and load TCDFs for the FPM XML configuration. You can define traffic classes and policies with multiple actions (such as drop, log, or send Internet Control Message Protocol [ICMP] unreachable) in a TCDF to assist in the blocking of new viruses, worms, and attacks on the network.

•Creating a Traffic Classification Definition File for the FPM XML Configuration (required)

•Loading a Traffic Classification Definition File for the FPM XML Configuration (required)

•Associating a Traffic Classification Definition File with an Interface or Subinterface (required)

•Displaying TCDF-Defined Traffic Classes and Policies (optional)

Creating a Traffic Classification Definition File for the FPM XML Configuration

Perform the following task to create a TCDF for FPM XML configuration. The TCDF is used to define traffic classes and the associated policies with specified actions for the purpose of blocking new viruses, worms, and attacks on the network.

The TCDF is configured in a text or XML editor. The syntax of the TCDF must comply with the XML Version 1.0 syntax and the TCDF schema. For information about Version 1.0 XML syntax, see the document at the following url:

http://www.w3.org/TR/REC-xml/

Traffic Classification Definition File Syntax Guidelines

The following list describes required and optional syntax for the TCDF:

•The TCDF filename must end in the .tcdf extension, for example, sql_slammer.tcdf.

•The TCDF contains descriptions for one or more traffic classes and one or more policy actions.

•The file is encoded in the XML notation.

•The TCDF file should begin with the following version encoding:

<?xml version="1.0" encoding="UTF-8"?>

SUMMARY STEPS

1. Open a text file or an XML editor and begin the file with the XML version and encoding declaration.

2. Identify the file as a TCDF.

3. Define the traffic class of interest.

4. Identify matching criteria for the defined class of traffic.

5. Define the action to apply to the defined class.

6. End the traffic classification definition.

7. Save the TCDF file with a filename that has a .tcdf extension.

DETAILED STEPS

---

Step 1 Open a text file or an XML editor and begin the file with the XML version and encoding declaration.

<?xml version="1.0" encoding="UTF-8"?> 

Step 2 Identify the file as a TCDF. For example:

<tcdf>

Step 3 Define the traffic class of interest.

For example, a stack class describing an IP and UDP stack might be described as follows. In this example, the name of the traffic class is "ip-udp," and the class type is "stack."

<class name="ip-udp" type="stack"></class>

In the following example, the name of the traffic class is slammer, the class type is access control, and the match criteria is all:

<class name="slammer" type="access-control" match="all"></class>

Step 4 Identify matching criteria for the defined classes of traffic. For example:

    <class name="ip-udp" type="stack">

       <match> 

          <eq field="ip.protocol" value="0x11" next="udp"></eq>

       </match>

    </class> 

    <class name="slammer" type="access-control" match="all">

       <match> 

          <eq field="udp.dest-port" value="0x59A"></eq>

          <eq field="ip.length" value="0x194"></eq>

          <eq start="l3-start" offset="224" size="4" value="0x00401010"></eq>

       </match>

    </class>

The traffic of interest in this TCDF matches fields defined in the PHDF files, ip.phdf and udp.phdf. The matching criteria for slammer packets is a UDP destination port number 1434 (0x59A), an IP length not to exceed 404 (0x194) bytes, and a Layer 3 position with a pattern 0x00401010 at 224 bytes from start (offset) of the IP header.

Step 5 Define the action to apply to the defined class. For example:

<policy name="fpm-udp-policy">

    <class name="slammer"></class>

    <action>Drop</action>

</policy>

The policy name in this example is fpm-udp-policy, and the action defined for the policy is to drop the packet. This action is to be applied to the class that has the name slammer.

Step 6 End the traffic classification definition. For example:

</tcdf>

Step 7 Save the TCDF file with a filename that has a .tcdf extension, for example: slammer.tcdf.

---

Loading a Traffic Classification Definition File for the FPM XML Configuration

Perform this task to load a TCDF for the FPM XML configuration. After the TCDF is successfully loaded, you can use service-policy CLI to attach TCDF policies to a specific interface or interfaces (see the "Associating a Traffic Classification Definition File with an Interface or Subinterface" section).

SUMMARY STEPS

1. enable

2. show protocol phdf protocol-name

3. configure terminal

4. load protocol location:filename

5. load classification location:filename

6. end

7. show class-map [type {stack | access-control}] [class-map-name]

DETAILED STEPS

	Command or Action	Purpose
Step 1	enable Example: Router> enable	Enables privileged EXEC mode. •Enter your password if prompted.
Step 2	show protocol phdf protocol-name Example: Router# show protocol phdf ip	Displays protocol information from a specific PHDF. •Use this command to verify that a PHDF file relevant to the TCDF is loaded on the device.
Step 3	configure terminal Example: Router# configure terminal	Enters global configuration mode.
Step 4	load protocol location:filename Example: Router(config)# load protocol localdisk1:ip.phdf	(Optional) Loads a PHDF onto a router. •The specified location must be local to the router. Note If the required PHDF is already loaded on the router (see Step 2), skip this step and proceed to Step 5).
Step 5	load classification location:filename Example: Router(config)# load classification localdisk1:slammer.tcdf	Loads a TCDF onto a router. •The specified location must be local to the router.
Step 6	end Example: Router(config)# end	Exits to privileged EXEC mode.
Step 7	show class-map [type {stack | access-control}] [class-map-name] Example: Router# show class-map sql-slammer	(Optional) Displays a class map and its matching criteria. •Use this command to verify that a class defined in the TCDF file is available on the device. •The class-map-name argument is the name of a class in the TCDF.

Examples

The following is sample output from a show class-map command that displays the traffic classes defined in the TCDF after it is loaded on the router:

Router# show class-map

.

.

.

class-map type stack match-all ip-udp

   match field IP protocol eq 0x11 next UDP

class-map type access-control match-all slammer 

   match field UDP dest-port eq 0x59A 

   match field IP length eq 0x194 

   match start l3-start offset 224 size 4 eq 0x4011010 

.

.

.

What to Do Next

After you have defined the TCDF, you must apply that policy to an interface as shown in the following task "Associating a Traffic Classification Definition File with an Interface or Subinterface."

Associating a Traffic Classification Definition File with an Interface or Subinterface

Perform the following task to associate a TCDF with an interface or subinterface.