CHAPTER 14

                                                SECURING AN EXTRANET

 

The are of a private network that hosts Internet-accessible resources is called an extranet.

 

Scenario:

 

  •   Market Florist plans to manage its own Internet site and allow it more flexibility in the services it

offers on the Internet.

  •   They plan to add a new firewall.
  •   You’ve been hired to design a secure Internet presence for Market Florist that provides accessibility

to all Internet-accessible resources while maintaining the security of the private network.

  •   The security design for the extranet must allow private network users to connect to the Internet and

to Internet-accessible resources without exposing the internal IP addressing scheme to the public

network.  Must use NAT to translate addresses from private to public.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

======================================================================

 

winsec14.html                                                 PAGE 2                                                       2002/05/07

 

 

 

Lesson 1:  Identifying Common Firewall Strategies

 

This configuration is commonly referred to as a Demilitarized Zone (DMZ), perimeter network, or

screened subnet.  This internal firewall ensures protection of the private network if the external

firewall or resources in the extranet are compromised.

 

What is an Extranet?

 

The challenge is to configure firewalls in such a way as to allow Internet users into a portion of the

private network without granting them access to the entire network.  The area of a private network

that hosts Internet-accessible resources is called an extranet.

 

Web server content is the most common network resource that organizations make available on the

Internet.

 

 

Identifying Firewall Features to Protect the Extranet

 

In the simplest deployment, a firewall is placed between the private and public networks to secure

the private network from the public network.

 

A firewall acts as a barrier against attacks launched from the public network.  To protects the private

network, firewalls can offer a suite of services, including:

 

Network Address Translation (NAT)  Translation of the source address of all outbound packets

from a private network address to a public network address.

 

Packet filtering.  Configuration of rules at a firewall that define which protocols are allowed to pass

through the firewall.

 

Static address mapping. 

Configuration of how incoming packets are re-routed to servers using private network addressing.

 

Stateful inspection of network traffic.  Verification that protocols are following basic rules of

communications. 

 

Advanced features that detect common attacks against the private network.  These include

setting time-outs for incomplete session establishment and inspecting the content of incoming packets.

 

 

 

 

======================================================================

 

winsec14.html                                                 PAGE 3                                                       2002/05/07

 

 

 

 

Protecting Private Network Addressing with NAT

 

NAT prevents exposure of the IP addressing scheme used on your private network.  An attacker

with knowledge of the IP addressing scheme can attempt an IP spoofing attack by sending packets

to the network with the falsified IP source address of a trusted private network address.

 

There are three ranges of IP addresses for private network addressing:

 

10.0.0.0 – 10.255.255.255 (10.0.0.0/8)

172.16.0.0 – 172.31.255.255 (172.16.0.0/12)

192.168.0.0 – 192.168.255.255 (192.168.0.0/16)

 

Packet Filters

 

Once you’ve deployed a firewall, you must establish firewall rules to define what data is allowed to

enter and exit the private network. Firewall rules are made up of individual packet filters.  A firewall

can send a packet filter to profile a protocol so that data transmitted using the protocol is identified.

 

A packet filter is typically composed of fields that profile a protocol and identify what action to take

if the protocol attempts to pass through the firewall.  The fields used by packet filters include:

 

Source address.  The individual IP address or network IP address from which the data

originates.

 

Source Port. Every IP-based protocol originates from a port at the source host.

 

Destination address. The individual IP address or network IP address to which the

data is sent.

 

Destination port.  The port that the server uses to listen for connections.  IP-based

servers and applications listen for connections on a predefined port.

 

Protocol.  Application and Services use a specific protocol ID or a transport layer

protocol for transporting application data.  If the application or service uses a transport layer

protocol, it will be either TCP or UDP.

 

Action.  Options include Allow or Deny.

 

Many firewalls include an option to mirror all packet filters.  Mirroring is necessary to allow

response packets to return to the source client computers. 

 

Mirroring simply switches the source and destination information to allow the response packets

to cross the firewall.

 

======================================================================

 

winsec14.html                                                 PAGE 4                                                       2002/05/07

 

 

 

 

Static Address Mapping

 

Use static address mapping at a firewall to redirect incoming traffic to Internet accessible resources

hidden behind the firewall.  The resources are advertised on the Internet with publicity accessible

P addresses.  When a firewall receives the packets the firewall translates the destination address

to the true IP address of the resource behind the firewall and redirects the data to the resource.

 

Combined with packet filtering rules, static address mapping allows you to define authorized

protocols and redirect the protocols to servers located in a DMZ.

 

 

Stateful Inspection

 

In higher-security networks, simple packet filters may not provide enough security.  Packet filters

define which ports are left open at the firewall to redirect network traffic to Internet-accessible

resources.  Many protocols utilize random ports above port 1024 at the client computer side. 

Opening up all ports above 1024 can leave both the firewall and the private network resources

susceptible to attack.

 

Stateful inspection allows firewall rules to be established so that UDP-based protocols such as

SNMP can pass through firewalls successfully.

 

 

Advanced Techniques

 

In addition to these services, firewalls provide advanced security through the following features:

 

Configuring Time-out tolerance.  Some attacks attempt to lock up the firewall by flooding it with

incomplete TCP sessions.  The time-out tolerance allows the firewall to disconnect these sessions

before the SYN queue overflows.

 

Content Scanning.  Although packet filtering can restrict protocols, it can’t inspect commands

within a protocol.

 

 

 

======================================================================

 

winsec14.html                                                 PAGE 5                                                       2002/05/07

 

 

 

 

Designing Firewall Features to Protect Internet-Accessible Resources

 

=====================================================================

Use This Firewall                   To do the following

Feature

=====================================================================

NAT                                   Prevent The private Network addressing scheme from

                                           Being revealed.

 

                                           Hide the true IP address of private network resources

                                           When accessing Internet-based resources.

 

Packet Filters                     Manage what protocols are allowed to cross between

                                            The private network and the public network.

                       

                                            Define what action to take if a protocol is identified

                                            when crossing the firewall.

 

                                            Define a default action to take if a protocol doesn’t

                                            meet any of the defined packet filters.

 

Static Address Mapping     Advertise Internet-accessible resources that have

                                             private network addresses using public network

                                             addressing.

 

                                            Hide the true IP address of Internet-accessible

                                            resources.

 

Stateful inspection              Protect UDP-based protocols that must enter the

                                             private network.

 

                                             Detect session hijacking attempts.

                                   

                                             Detect application-level attacks that attempt to bypass

                                             the packet filters established for a protocol.

 

Time-out tolerance             Prevent SYN Flood attacks by closing sessions that

                                            That have timed out.

 

                                             Free up connections for new connection attempts if

                                             The connection is left idle.

 

Content scanning                Prevent specific application commands from being.

                                             Issued within a protocol.  Detect viruses.

 

 

======================================================================

 

winsec14.html                                                 PAGE 6                                                       2002/05/07

 

 

 

Comparing DMZ Configurations

 

DMZs have a variety of common names, including screened subnet and perimeter network. 

Screened subnet refers to the function of a DMZ in network security.  All network traffic that

attempts to enter or exit the DMZ is screened by packet filters to determine whether they’re

allowed.

 

Networks use one of the following DMZ designs:  a three-pronged firewall DMZ, a mid-ground

DMZ, or a hybrid (or multizone) DMZ.

 

 

Three-pronged firewall DMZ:

 

A three-pronged firewall DMZ consists of a firewall with three network interfaces.  One interface

is connected to the private network, another is connected to the public network, and the final

interface is connected to the DMZ.

 

A DMZ is part of the private network and part of the public network.  In a sense the DMZ is part

of the public network, because resources that are accessible to the public network are placed in

he DMZ.  But the DMZ is also part of the private network because packet filters enforce which

protocols can be used to connect to each server located in the DMZ.

 

While deploying a three-pronged firewall, ensure that your firewall solution supports three

network interfaces.  Each interface will be assigned to a zone and you must establish packet

filters that define the interaction allowed between each zone.

 

 

Mid-Ground DMZ:

 

You establish a mid-ground DMZ by using two firewalls.  You place the first firewall between

the public network and the DMZ and the second firewall between the DMZ and the private

network.

 

As with a three-pronged firewall DMZ, the IP addressing used in the mid-ground DMZ can be

either private or public network addressing.  Generally you will use private network addressing

in the DMZ unless there’s a requirement to establish IPSec connections through the external

firewall.

 

 

Hybrid DMZ:

 

 

Sometimes a single DMZ may not meet your business requirements.  In these cases you can use

Hybrid DMZ.  See page 554.  In this scenario you create two DMZs to support the need for

IPSec connections and for protection of the private network addressing configuration for all other

 

 

 

======================================================================

 

winsec14.html                                                 PAGE 7                                                       2002/05/07

 

 

 

 

Internet-accessible resources. One DNZ, containing the remote access server that will accept

IPSec connections, uses public network addressing.  The second DMZ uses private network

addressing and contains all other Internet-accessible resources.  The firewall prevents static

address mapping for all incoming traffic to the network segment. 

 

You can also have multiple firewalls set up to establish two or more DMZs between the private

network and the public network.

 

 

Choosing Between DMZ configurations

 

===================================================================

Use                              To Meet These Business Objectives

===================================================================

Three-pronged             To reduce the costs associated with deploying firewalls. 

Firewall DMZ              Only a single firewall is required for this design.

 

                                    To maintain a single packet filter list.  The packet filter

                                    rules determine which interface a packet filter is applied

                                    to.

 

Mid-ground DMZ        Top provide physical separation of the private network

                                    from the public network with the DMZ being placed

                                    between the private and public networks.  An attacker

                                    must breach two firewalls to access the private network.

 

                                    To reduce the change of breach of the external

                                    firewall will lead to access of the private network.  If

                                    you use two different manufacturers for the internal and

                                    external firewall, you gain security because different

                                    methods must be used to breach the second firewall.

 

Hybrid                          To provide both private network and public network

                                    addressing to DMZ segments.

 

                                    To categorize Internet-accessible resources into different

                                    levels of access that can be protected by firewall strategies.

 

 

====================================================================

 

 

Lesson Summary:

 

  •   Knowing the features of firewalls and how they are commonly deployed will assist you in

ensuring access to Internet-accessible resources is secure. 

  •   Your design should take advantage of your firewall’s configuration options to ensure that

only authorized data transactions can take place.

 

 

 

======================================================================

 

winsec14.html                                                 PAGE 8                                                       2002/05/07

 

 

 

 

Lesson 2:  Securing Internet-Accessible Resources in a DMZ

 

With the DMZ you must secure resources to ensure that only authorized access takes place.

 

Securing IIS

 

The most common network resource exposed to the Internet is the content on a Web server. 

Internet Information Service (IIS) 5.0 is included with Windows 2000 Server and allows an

organization to host Web sites.

 

You should consider additional security configurations for securing and IIS server.

 

  •   Change all default account names.  Make sure that the Administrator and Guest accounts are

renamed, but don’t choose names that are easy to guess when you rename them.

  •   Ensure that the Web server isn’t a member of the same forest as the private network.  If the
  •   Web server is compromised, your Active Directory could become compromised, too.
  •   Separate available content into different folders by type.  By separating the Web site by content

type, you can apply security specific to each content type.

  •   Secure available content by type.  Once you’ve separated the Web content by type, you can

then apply discretionary access control lists specific to the Web content type.

  •   Remove all sample applications from the Web server.  The sample application files that are

included with IIS 5.0 are installed by default and could provide tools to an attacker.

  •   Disable unnecessary services.  If the Web server is dedicated to hosting Web applications,

disable all unnecessary service, such as the Telnet or FTP service.

  •   Block commonly attacked ports with IPSec.
  •   Enable IIS Logging.  IIS logging will help you determine whether your IIS server has been

attacked.  By using the World Wide Web Consortium (W3) Extended logging you choose

exactly which properties will be logged for each connection attempt to the IIS server.  Enable

  ODBC log file.

  •   Implement Secure Socket Layer (SSL) protocol to protect secure areas of the Web Server.
  •   Any areas of the Web server that require visitors to enter confidential information such as

addresses, credit cards, or password information should be protected by using application-

layer SSL protection.

  •   Deploy an intrusion detection system.  The intrusion detection system should inform you when

the attack occurred, what kind of attack was attempted, and whether the attack was successful.

  •   Disable the use of parent paths.  IIS off the Administrative Menu, Properties of the default Web

  Site.

  •   Apply the IIS5.0 security checklist.  The IIS 5.0 security checklist includes a high-security Web

server security template (Hisecweb.inf) that you can apply to secure the Web server.

  •   Mitigate against successful attacks.  Maintain against successful attacks.  By configuring the Web

server to participate in an NLBS cluster, you can host the Web site on multiple servers

simultaneously.  When a public network user connects to the shared address of the NLBS

cluster, the connection is directed to one of the nodes in the cluster.  If a server in the NLBS

cluster is rendered inaccessible by an attack, the NLBS cluster recognizes that the node is

down and directs future traffic to the remaining servers in the cluster.

 

 

 

 

======================================================================

 

winsec14.html                                                 PAGE 9                                                       2002/05/07

 

 

 

 

NOTE:  NLBS allows weighting of nodes.  This means that you can direct a higher percentage

of incoming traffic to the server with the most resources.  For example, if your NLBS cluster is

comprised of a Pentium 166 MHz server and a Pentium 933 MHz, you’d want a higher

percentage of the connections to be established with a higher performance Web server.

 

Maintain the latest service packs and hot fixes for the Web server.   Ensure that regular

updates to the operating system are applied so that security fixes included in the updates

are applied to the Web server.

 

 

Securing Other Services Within the DMZ

 

Within a DMZ you can expect to find several common resources.  These include external

FTP servers, Telnet servers and DNS servers.

 

FTP services.  The FTP service isn’t loaded by default in a Windows 2000 installation. 

If you install the FTP service, be aware that the C:\Internet\Ftproot folder allows the

Everyone group Full Control by default.  Consider changing the permissions and enabling

disk quotas to limit the amount of data that can be written to the FTP root. 

 

Telnet servers.  Telnet provides text-based access to a server’s disk subsystem.  If it’s enabled,

all users could access the Telnet service by default.  You can restrict access to telnet servers

by creating a group named TelnetClients in either the local Security Account Management

(SAM) database of the Telnet server or, if the Telnet server is a member of a domain, within

Active Directory.  Create a TelnetClient group.

 

 

======================================================================

 

winsec14.html                                                 PAGE 10                                                     2002/05/07

 

 

 

DNS Services.  Only externally accessible DNS resource records should be stored on a

DNS server located in the DMZ.  Try to protect DNS.

 

Terminal Services.  Restrict Terminal Services to administrators of the network by configuring

Terminal Services to operate in Remote Administration mode.

 

All Services.  Use IPSec to limit access to these well-known ports by creating IPSec block

policies that will block access to other ports found on the server.  The firewall can do the

filtering to protect the system.   

 

IMPORTANT:  If you want private network computers to connect to other ports on the server,

change the IPSec action to negotiate and define the encryption protocols that must be used for

the connection.  Only members on the private network should be able to negotiate a security

association (SA) with the Web server.

 

In addition to securing the additional servers within the DMZ, you can also use IPSec to

configure the security of transmitted data between the servers located in the DMZ.  Data

between the Web Server and the SQL Server is transferred and possibly encrypted.  All of

this is between the DMZ Internal & External Walls.

 

 

Lesson Summary:

 

Don’t depend exclusively on firewalls to protect resources exposed to the Internet.

Ensure that each server in the DMZ is properly configured so that security will be maintained

even if a firewall is compromised.

 

 

Lesson 3:  Securing Data Flow Through a DMZ

 

Firewalls can regulate which protocols are allowed to enter and exit the DMZ.

By configuring the firewall to allow only the authorized protocols to connect to each server in

the DMZ, you can prevent hackers from attaching to the servers using alternate protocols.

 

Determining a Firewall Strategy

 

When configuring a firewall, decide between the following basic firewall strategies based on

your desired security level:

 

 

======================================================================

 

winsec14.html                                                 PAGE 11                                                     2002/05/07

 

 

 

 

  •   Specify allowed protocols and prohibit everything else.  In this strategy the packet filters identify

all protocols that are allowed to pass through the firewall.

 

  •   Specify prohibited protocols that allow everything else.  In this strategy the packet filters identify

all protocols that must be dropped at the firewall.

 

 

What about the Order of the Packet Filters?

 

The order in which packet filters are processed depends on the specific firewall product. 

Firewalls typically process the packet filters using one of two common methods.

 

The first method is to process the packet filters in the order in which they’re entered.

 

The second method is to process the most specific packet filters before the more general packet

filters.  In other words, a packet filter that restricts access to a specific server will be processed

before a packet filter that defines traffic between network segments.

 

DNS must be in internal & external, so you can allow “forwarding”.

 

In High Security, do this:  Allow certain protocols and packs and Deny everythingelse.

 

 

Securing Web Traffic

 

A Web server is one of the most common network resources made available for Internet Access. 

A Web server listens for connections from external client computers on TCP port 80 for HTTP

connections and on TCP port 443 for HTTPS connections.  In High Security do this:  Allow

certain protocols and packs and Deny everythingelse.  This gives you control of the traffic.

 

 

Securing FTP Traffic

 

FTP allows data to be transferred to and from a central location.

 

FTP uses two separate channels for FTP sessions.  FTP uses a control stream (a connection to

TCP port 21) to send FTP commands from the FTP client software to the FTP server and a

data stream (a connection to TCP port 20)  for transferring data).

To restrict FTP connections from the Internet, you typically place the FTP server in a DMZ.

 

 

 

======================================================================

 

winsec13.html                                                 PAGE 12                                                     2002/05/04

 

 

 

 

NOTE:  The packet filters listed on page 576 assume that active FTP clients exist on the

network.  The last two packet filters allow data transfers initiated by the FTP server to pass

through the external firewall.  If the FTP clients were passive, you could remove the last two

filters from the packet filter listing.

 

 

Securing Mail Traffic

 

Many organizations use e-mail to communicate within the organization and with other organizations. 

Ensure that mail servers are secured  against attacks by allowing only authorized protocols to

connect to the mail server.

 

Carefully plan mail access for an organization so that e-mail doesn’t cause security problems. 

Typically, a mail server must support several protocols for e-mail access.  Common protocols

supported by mail servers include:

 

  • Post Office Protocol v3 (POP3).  Authentication is performed using plaintext, a security risk. 

Allows you to receive mail.

 

  • Internet Mail Access Protocol v4(IMAPv4).  Used by newer e-mail client software to retrieve

messages form the mail server.  Like POP3 also uses plaintext authentication.  Allows you to

receive mail.

 

  • Simple Message Transfer Protocol (SMTP).  Used by e-mail client software to send messages

to other e-mail recipients.  SMTP is subject to spamming attacks where unsolicited messages

are relayed using an unprotected e-mail server.  Sends the data clear text, not the user and

password.

 

  • Lightweight Directory Access Protocol (LDAP).  Used by e-mail client software as a directory

for the e-mail server.  LDAP provides the ability to find e-mail addresses on a mail server. 

LDAP uses plaintext authentication by default.

 

 

IMPORTANT:  You can protect POP3, IMAPv4, and LDAP by implementing SSL.  The

protocols will use SSL encryption to protect user account and password verification during

the authentication process.

 

 

Making the Decision

 

When configuring a firewall to allow access to a mail server, include the following items in your

security design:

 

 

 

 

======================================================================

 

winsec14.html                                                 PAGE 13                                                     2002/05/07

 

 

 

Determine which protocols will be allowed to access the mail server from the public network.

 

Establish packet filters at the external firewall to allow only the necessary ports to connect to

he mail server.  Ports to be opened can include POP3 (TCP port 110), Pop3-ssl (TCP port

995….. BLA, BLA BLA see page 580.

Establish restrictions on SMTP relaying to prevent the mail server from becoming a source for

unsolicited bulk e-mail.

Restrict which protocols can be used to connect to the mail server from the private network.

 

Restrict access to specific protocols, such as allowing only POP3 and SMTP protocol access.

Restrict access to specific users.  Use this strategy if the organization has multiple mail servers

and the mail server in the DMZ only acts as a gateway to the other e-mail system.

Allow private network client computers to connect using any protocol, provided they’re located

on the private network.

 

 

Securing Application Traffic

 

Servers in the DMZ are often required to store or access data from an application server, such

as an SQL server, in the private network. 

 

It’s too risky to place the application server in the DMZ because data stored on the application

server might be compromised.

 

You can manage the connection between the two servers by using one of three methods:

 

Attach the server in the DMZ directly to the computer on the private network using a crossover

cable and protocol other than TCP/IP.

Open the firewall to allow the native protocol to transfer between whatever SQL needs to

communicate , TCP/IP etc. and  the server in the DMZ and the application server in the private

network.

 

Use IPSec to encrypt that data transmitted between the server in the DMZ and the application

server in the private network.  As long as the internal firewall isn’t performing NAT on the

transmitted data, you can use IPSec to encrypt all data transmitted between the two servers.

 

 

 

======================================================================

 

winsec14.html                                                 PAGE 14                                                     2002/05/07

 

 

 

The risk with this configuration is that the data transmitted between the two servers may be

passed in plaintext to allow confidential data to be inspected by a network sniffer.  Network

sniffers are able to view the contents of any unencrypted data packets transmitted on the network.

 

NOTE:  IPSec encryption prevents the firewall from inspecting the actual protocol that’s being

transmitted through the firewall.  Be sure to set up your rules so that only server-to-server

connections are allowed to pass using IPSec.  This prevents other servers form establishing

IPSec agreements and attempting to pass though the firewall.

 

 

It’s the Web server that actually performs the query on behalf of the external customer.  For the

transaction to occur, the external customer only needs to connect to the Web server using HTTP

or HTTPS.

 

 

**** Review the charts on page 583 ****

 

 

Making the Decision

 

When configuring a firewall to allow access to an application server on the private network,

configure the firewall rules to match how data transmissions take place.  Your design should

include the following:

 

To go through the Firewall you need port 500 (p583) to IPSec to go through the firewall.

 

  •   Determine which protocols are required to access the server-based component in the DMZ.
  •   Configure the external firewall to allow only public network client computers to connect to the

server in the DMZ using the protocols required for the server-based component.

  •   Determine which protocols the server-based component uses to connect to the application

server in the private network.

  •   Determine the securest method to connect the server in the DMZ to the application server in

the private network.

  •   Define the necessary packet filters at the internal firewall to allow only the required protocols

to exchange data between the DMZ an the private network.

 

NOTE:  IKE uses port 500 and protocol UDP.

 

 

Securing Terminal Server Traffic

 

Terminal Services allows an administrator to connect to servers on the network by using Remote

Desktop Protocol (RDP).  You can restrict Terminal Services to be used for administrative

purposes only by configuring Terminal Services to run in Remote Administration mode rather

than Application Services mode.

 

 

 

======================================================================

 

winsec14.html                                                 PAGE 15                                                     2002/05/07

 

 

 

Configure the terminal server to require strong encryption of the connection.  If the Windows

2000 High Encryption Pack is installed at both the Terminal Services client computer and the

server running Terminal Services, 128-bit RC4 encryption is used to protect transmitted data.

 

 

Making the Decision

 

When configuring a firewall to allow access to a terminal server located in the DMZ, consider

 the following design points:

 

 

  •   Configure the firewall to allow only connections to TCP port 3389, and RDP protocol, to pass

through the firewall.

  •   Configure the terminal server to use the highest level of encryption supported by the client

computers, subject to local import and export laws.

  •   If you only require administrative access to the terminal server, configure the terminal server

use Remote administration mode.

 

 

Securing VPN Traffic

 

When an organization wants its network to permit VPN traffic, they must consider how the

protocol they use affects their firewall and network infrastructure design.  IPSec is unable to

pass through a firewall that performs NAT on incoming and outgoing packets.

 

 

Securing PPTP Tunnel Traffic

 

You don’t have to take any special considerations into account when placing a PPTP tunnel

server in your network.  It doesn’t matter if the external firewall performs NAT on incoming

and outgoing packets.

 

To support the placement of PPTP tunnel server in the DMZ, configure the external firewall

to allow connections to the tunnel server using PPTP (TCP port 1723) and the Generic Routing

Encapsulation (GRE) protocol (protocol ID 47).

 

The first packet filter allows PPTP connections from any computer on the public network to

the PPTP Server (10.30.30.45).  The second packet filter allows GRE packets, which are

used by PPTP to encapsulate the original data packets, to pass through the external firewall

to the PPTP server.

 

 

 

 

======================================================================

 

winsec14.html                                                 PAGE 16                                                     2002/05/07

 

 

 

To protect the Active Directory database, you can deploy the PPTP server as a member of

a workgroup rather than as a member of the domain.  To support domain authentication,

configure the tunnel server as a RADIUS client to a RADIUS server on the private network. 

To support this, configure the internal firewall to allow RADIUS authentication (UDP port

1812) packets and RADIUS accounting (UDP port 1813) packets to pass from the tunnel

server to the RADIUS server on the private network.

 

 

Securing L2TP/IPSec Tunnel Traffic

 

L2TP tunnel connections require that the tunnel server not be placed behind a firewall that

performs NAT.  You can modify the configuration of the DMZ to met this requirement. 

Because the firewall doesn’t perform NAT, you can place the L2TP tunnel server in the

DMZ like a PPTP tunnel server.

 

In this configuration the outer DMZ (p591) uses public network addressing so that NAT isn’t

performed on any of the packets destined for the L2TP tunnel server.

 

In this configuration the L2TP tunnel server is a dual-homed server with the external network

interface card (NIC) on the public network, and the internal NIC is attached to the DMZ.

 

IMPORTANT:  To protect the external NIC, you must establish packet filters to allow only

L2TP/IPSec connections.  All other connection attempts must be denied.

 

 

Making the Decision

 

When configuring a firewall to allow access to an L2TP tunnel server located in the DMZ,

consider the following:

 

  •   Use public network addressing if the L2TP tunnel server is placed in the DMZ.
  •   If the DMZ uses private network addressing as defined in RFC 1918, consider establishing

an outer DMZ that used public network addressing or configuring the L2TP server as a

perimeter server with an interface on the public network to accept tunnel connections.

  •   Configure the external firewall to pass the tunneling protocol used by the tunnel server in

the DMZ.

  •   Configure the internal firewall to allow RADIUS authentication to a RADIUS server on

the private network.  RADIUS authentication provides domain logon capabilities for remote

access when the tunnel server is located in a DMZ.

 

 

Lesson Summary:

 

  •   Creating packet filters at external and internal firewalls takes careful planning and design. 

after deciding on a firewall strategy, you must define the packet filters to fit your strategy.

  •   If additional protocols can pass through the firewall, you may be leaving your network

vulnerable to an attacker.