Technology is radically changing the way companies conduct business, opening up new possibilities that enable
efficiencies and growth on a global scale. But for everything that technology facilitates, it also opens up new risks,
forcing companies to think about how to protect the assets they are working so hard to build. Security and IT
administrators are faced daily with the challenge of successfully implementing technology that supports the
company’s success, while maintaining the security of the organization’s critical resources.
The first step that organizations generally take is to control who and what gets in and out of the network by
deploying a firewall. Firewalls perform access control, user authentication, traffic management and policy
enforcement to ensure only appropriate users and services are able to traverse the network and that business
applications are given priority. Firewalls, however, are no longer relegated to just perimeter deployments. Rather
organizations are increasingly taking advantage of firewall capabilities throughout the network to segment it and
apply security policies between different segments. These segments, or zones, could represent geographically
distributed networks, such as regional offices, different types of traffic, such as wireless or VPN connections,
different departments or even different servers. This segmentation enables the organization to create additional
levels of trust to protect sensitive resources and perform attack containment.
Firewalls also provide some protection against attacks, traditionally focusing on preventing network-level
such as Denial of Service attacks. But, as many organizations have come to realize, attackers are increasingly
attacking vulnerabilities found not at the network-level, but at the application-layer, and are actually leveraging traffic
“allowed” by the firewall to get into the network. As a result, some firewalls have started to look deeper into the
traffic they are allowing in and out of the network to try to identify and prevent attacks found at the application-layer.
Firewalls are also often coupled with virtual private network (VPN) functionality, which is designed to enable
organizations to provision site-to-site connectivity that takes advantage of the cost-benefits of the public Internet
infrastructure in a secure manner. The most commonly deployed site-to-site VPN technology is an IPSec VPN, so
this guide will focus on these solutions. IPSec VPNs encrypt traffic to maintain its confidentiality and protect against
tampering with or altering of the data. As a result, they enable organizations to securely extend their network
perimeter across the public Internet to facilitate secure communications between geographically distributed
locations.
As with any solution, an administrator needs to be aware of the potential impact that a device can have on their
network’s performance and availability, as well as the time and management implications that each solution
introduces. While VPN functionality can also be deployed as a standalone solution, it is always a good idea to apply
access controls to the VPN traffic. As a result, the tight integration of firewall and VPN functionality can reduce
network complexity, simplify deployment and management and reduce the overall total cost of ownership of an
organization’s connectivity and security.
Administrators need these solutions to enable business productivity, as well as network security, so this guide is
designed to help organizations find the balance they need between functionality and security, without compromising
one for the other. This guide provides a framework for evaluating firewall and VPN security solutions. It is organized
into three sections. The first is an executive level summary that splits the evaluation criteria into five different
categories and explains the impact of each category on the overall solution’s ability to deliver value. The next
section takes those five categories and provides a quick checklist for each that will help the evaluator start to ask
the questions that will differentiate the capabilities of products. Finally, the last section provides a detailed list of
features that make up each category to enable evaluators to really make product comparisons to ensure they can
select the one that best meets the needs and requirements of their organization.
Firewall/IPSec VPNs serve as the foundation upon which a strong security stance can be built, so the purchase
decision should be framed in terms that support a long-term investment that can be leveraged as the organization’s
needs change and grow. The chosen firewall/VPN solution should not only provide robust security functionality, but
also the networking and availability features that will support the company’s ongoing connectivity and expansion
requirements. In addition, the security solution needs to be easily integrated into the network and simple to manage,
so that it does not put a strain on already tight IT, security and networking budgets. There are so many firewall and
VPN vendors in the market that it can become overwhelming for a company to try and sort through them all and
determine what the best solution is for their environment. This section is designed to help decision-makers and
evaluators think, in broad terms, about the criteria that will be most helpful as they make their solution choice.
1. Provide strong security.
The solution needs to provide robust security functionality to maximize the protection it provides to the
network. Some of the functionality that should be included is strong access control, user authentication,
attack protection - both at the network and application-layer - IPSec and encryption choices for data
integrity, and network segmentation for attack containment. Ideally, the functionality should be integrated to
maximize the security derived from the solution. Integrating the VPN functionality into the firewall, for
instance, requires fewer open ports and enables firewall policies to be easily applied to VPN traffic. It is
especially important, however, to scrutinize the feature set of products that integrate multiple functions to
ensure they are not too simplistic in their approach and are not lacking all of the robust, proven features that
are required for strong security. While initially appealing because they seem to be easy to manage, an
integrated solution that does not marry best-of-breed functionality can actually end up creating more work
due to the security holes they allow. For example, how effective is it to have intrusion prevention integration
that can only stop network-layer attacks? In response, it is more important that the solution provides the
granularity and flexibility needed to identify differences in traffic and appropriately process that traffic than to
satisfy a checklist. In addition, it is important to identify potential vulnerabilities that could be introduced by
the device itself, such as those associated with general-purpose platforms and operating systems. It is also
important that the solution accommodate the different requirements of different network segments, from the
smallest remote office to the largest central site, to ensure security can be uniformly deployed and eliminate
any weak links. The solution should be designed for and deliver security to justify its deployment.
2. Offer predictable performance.
The solution needs to be an enabler to network connectivity rather than a barrier. If the solution cannot
keep up with the performance requirements of the network segment that it is designed to protect, its value
will be significantly diminished. Not surprisingly, it must be able to efficiently process traffic and deliver
predictable performance under load. The performance should be sustainable for both large and small
packets. It should also minimize latency and accommodate the necessary concurrent sessions and VPN
tunnels that are required for that particular network segment. In order to provide adequate Denial of Service
(DoS) protection the solution needs to support a high ramp rate to handle attempts at performance
overload. The solution must be able to handle the performance requirements of the network and function
without degradation.
3. Deliver a high level of fault tolerance to ensure the solution is always available.
Being able to survive a failure and maintain both connectivity and the security stance of the organization is
the sign of good solution. The solution needs to provide redundancy at all levels to give an organization the
flexibility to choose the level of availability they want for each of their network segments, based on their cost
and connectivity requirements. The device, itself, needs to offer solid-state performance and component
redundancy. It then needs to support a high availability configuration that is able to maintain session and
VPN state information and survive a failure both up and down stream of the device, offering an
active/active, full mesh architecture. It needs to include network redundancy, leveraging the resiliency of
dynamic routing and supporting path redundancy to multiple ISPs or a dial-back up line. At the VPN level, it
needs to support multiple tunnels and minimize failover time to ensure optimal connectivity. Only a solution
that is able to provide all of the redundancy pieces is truly fault tolerant.
4. Offer ease of use and management.
The real costs of a solution are tied not to the initial capital outlay, but to the ongoing management and
operational costs associated with keeping the solution up and running. If a solution requires a lot of time
and resources to maintain, it is going to take away from other activities and increase the management
burden on the organization. The solution needs to be easy to interact with to ensure changes can be quickly
made to keep the security policy in force. An administrator should be able to manage the device, network
and security aspects of the solution, from a single interface, as opposed to having to go to one interface to
make routing changes and another interface to set security policies. It should automate as much as
possible to minimize human intervention, using tools such as templates and auto-configurations to
maximize consistent security deployments throughout the network. It should also, however, provide
granular controls to ensure that specific sites have a configuration that is most appropriate to their
environment. It should enable different people in the organization to efficiently do their jobs, without
introducing any risk to the security at large. For example, a NOC administrator should be able to get access
to device status, but shouldn’t be able to make security policy changes, a CIO should be able to see
reports, but not make routing changes, etc. It should also be easy to troubleshoot to enable organizations to
quickly resolve problems. Organizations don’t want to waste a lot of time on managing, rather they want an
easy to use solution that enables them to spend time on activities core to their business success.
5. Enable quick and simple deployment and installation.
IT, network and security managers are expected to do more with less, so it is important to be able to get
solutions up and running quickly. It needs to seamlessly integrate into the network environment, without
introducing interoperability issues. It should be intuitive, so that it doesn’t require a lot of training or security
expertise to use. Updates need to be easy to accomplish, without having to worry about overriding custom
configurations or introducing new vulnerabilities. For instance, an organization doesn’t want to have to
worry about how a newly applied patch to the operating system will affect the underlying platform or the
applications that it is running. The solution should be designed with everything working together, to
minimize complexity and simplify deployment and installation.
This section builds upon the framework for evaluating firewall and VPN products that was described in the previous
section, providing a quick checklist of some of the top questions to pose in each criteria category. For more indepth questions that enable a side-by-side comparison of different solutions, go to the Detailed Buyer’s Checklist
that follows this section.
1. Provide Strong Security
• Does the solution integrate best-of-breed technologies?
o How long have the technologies been in the market?
o Are there any third party verifications of viability available?
o Are the technologies based on open source solutions?
• Does the solution provide strong access control – stateful inspection?
• What kind of user authentication does the solution support?
• What network-level attacks does the solution protect against?
o DoS attacks
o DDoS attacks
• Does it have the ability to make determinations on whether to allow or deny traffic based on applicationlayer information?
o What kind of application-level attacks can it detect?
o What kind of application-level attacks can it prevent?
• What kind of encryption does the VPN support?
• Can the solution apply policies to internal traffic to establish additional layers of trust and contain
attacks?
• What type of security certifications does the product have?
• What kind of platform is the solution built on?
o Is it a general-purpose platform that could introduce security risks?
• Can the solution scale to meet the different security needs of small to large sites?
2. Offer Predictable Performance
• What are the performance (large and small packet size) capabilities of the solution to ensure that
performance remains predictable?
• What has the solution done to optimize its traffic processing?
• How does the solution minimize latency to ensure real-time applications are not degraded (e.g. VoIP)?
• How does the solution handle very fast session ramp rates to protect against DoS attacks?
• How does the architecture of the solution enable performance under load?
• How does the solution handle multiple concurrent sessions to ensure user connectivity is not lost or
slowed?
• How does the solution accommodate additional functionality, without degrading performance?
• How does the solution accelerate the VPN negotiation to set up the VPN tunnels to make the time
imperceptible to the user?
• How can the solution quickly create and then maintain VPN tunnels to ensure they are always available
for the user?
3.Deliver a high level of fault tolerance to ensure the solution is always available
• Does the solution support high availability (HA) configurations, including active/active, full mesh, to
reduce the chance of a single point of failure?
• Does the HA solution maintain both session and VPN state information to ensure that both the
connection and VPN security association are maintained in the event of a failure?
• Can the solution take advantage of dynamic routing as part of VPN resiliency?
• Can the solution support redundant paths? If so, what kind – multiple ISPs, dial back-up?
• What redundancy features have been built into the VPN configuration?
• What are the mechanisms used to minimize fail-over latency and ensure maximum uptime?
4. Offer ease of use and management.
• Are there multiple ways to interact and manage the system?
• How easy is it to perform management tasks?
o Can device, network and security configurations be managed using the same interface?
• Does the system grant different people in the organization different access privileges?
o How does the system ensure that people are only accessing what they need to access?
o How easy is it to set up or change a role to ensure access privileges map to current
employee activity?
• How quickly can changes be made in a large distributed network?
• Are there configuration templates to simplify deployments?
o How easy is it to customize the template information for specific site deployments?
• How easy is it to troubleshoot problems?
o Is there a way to roll back to a previous configuration if changes affect the connectivity of
the solution?
• How much manual intervention is needed when a VPN connection goes down?
• Can firewall policies be easily applied to VPN traffic, without a lot of additional configuration?
• How easy is it to add a network to the VPN?
• How easy is it to configure complex VPN configurations, such as a hybrid full-mesh and hub and
spoke?
5. Enable quick and simple deployment and installation
• Are there different options that accommodate administrator preferences for installing and configuring
the system?
• What kind of platform is the solution running on?
o Is the solution based on a general-purpose platform?
o Is the solution delivered as an appliance for easy deployment?
• How easy is it to deploy a device in the field?
o What level of technical expertise is required?
o Can it be managed centrally?
• Does the solution have a transparent deployment mode that does not require routing changes to the
network?
• What routing protocols does the solution support?
• What networking features does the solution support to facilitate a timely deployment?
This section provides a feature/functionality checklist for each of the criteria categories to help evaluators determine
the true capabilities of vendor solutions they are considering.
Evaluation Date:
Evaluated By:
Feature
Juniper Networks
Firewall / IPSec VPN /
Deep Inspection
Alternate
Solution:
Notes
Solutions*
1. Strong Security
Performs Stateful Inspection
Protects against networklevel attacks
Protects against DoS and
DDoS attacks
Protects against transport
layer attacks
Protects against applicationlayer attacks:
e-mail
Web
FTP
DNS
Uses proxies for attack
detection
Uses Stateful signatures for
attack detection
Uses protocol enforcement
for attack detection
Blocks malicious URLs
Protects against viruses
Options for strong user
authentication:
Ability to process traffic of
varying packet sizes to meet
the performance
requirements of the network
Accelerates intensive
processing with hardware
Ability to support applications
with a low tolerance for
latency/jitter, such as VoIP,
multimedia, etc.
Fast session ramp rates to
protect against DoS attacks
Provide additional
functionality without
degrading performance
Turning on all applications
does not affect the solutions
ability to meet the
performance needs of the
deployment
Traffic prioritization to ensure
business critical applications
are available
Deliver Quality of Service
(QoS):
o Control bandwidth
o Set priority field in
the Type of Service
(TOS) byte to reflect
traffic class priority
See Tolly Reports for
third party verification
Yes, including custom
security ASICs
Yes, hardware is
optimized for streamlined
Yes, Dedicated hardware,
allowing separate paths
for session set up and
established flows
• ASIC/FPGAs offload
intensive processing,
making CPU available
for new/additional
functions
• Programmability in
ASIC to accelerate
future functions
See spec sheets for
performance numbers
Yes,
processing
Yes,
Yes
Yes
Yes
VPN Specific
Accelerate IKE
negotiations for quick
tunnel set up
Minimal latency to ensure
real-time applications are
not degraded:
Maintain large numbers of
tunnels to ensure
availability
Yes, OS and Hardware
designed specifically to
negotiate security
associations
Yes,
o Provides fast path for
established flows
o Packets are quickly
processed without
unnecessary
traversals of PCI
busses
Yes
Purpose built solutions can
develop process efficiencies
over general purpose OS’
Unnecessary traversals of
PCI busses is a common
problem with PC-based
platforms using VPN
acceleration cards, adding
latency to application.
o Port Density
Supports dynamic routing
protocols:
o OSPF
o BGP
o RIP
High Availability (HA)
Configurations to reduce single
point of failure:
o Stateful (sharing
session information) to
maintain connections
o VPN sync (sharing
VPN information to
maintain security
association in the
event of a failure)
o Active-passive HA
(one device
processing traffic, with
the second device as a
back-up)
o Active-active HA (both
devices processing
traffic)
o Active-active, full-
mesh HA to survive a
failure up or
downstream from
device
Redundant physical
connections (e.g. connections
to different service providers)
Alternate transport options,
such as:
o DSL
o Dial back-up
A high Mean Time Before
Failure (MTBF) expectancy
VPN Specific
Ability to run dynamic routing
through its tunnels to
automatically learn the network
and route around failures
Product’s HA performs VPN
sync (sharing VPN state
information) to maintain the
VPN connection in the event of
a failure
Yes, using Bellcore
MTBF calculations
Yes, Dynamic Route-
based VPNs (Best
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Path VPNs)
Yes
Enables the survival of
Note: most routers cannot
failures at the transport level
–needed for other
components of resiliency
Supports different VPN
deployment modes:
Rule-based/Policy-based
Route-based
Dynamic Route-based (Best
Path)
Support multiple VPN
gateways to enable VPN to
persist in the event of a failure
Supports multiple tunnels,
running the same services,
between VPN gateways
Supports fail-over between
tunnels based on alternate
static routes defined in the
route table
Supports fail-over between
redundant tunnels using
dynamic routing
Supports fail-over between
redundant tunnels using
another mechanism
R-associate VPN with another
tunnel without having to
renegotiate the encryption
keys
Yes, custom VPN
configurable interval
to allow fail-over in
Association mirroring
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Path Monitor-
seconds
Yes, Security
mechanism
For rule-based or policy-
based VPNs
Note: rule-based or policy-
based VPNs cannot do this,
only route-based and
dynamic route-based VPNs
monitoring
Ability to manage the device,
network and security
functionality from a single
console
Ability to view all logs in central
location
Ability to assign different
people in the organization
different read/write privileges
Policy changes can be
distributed quickly to one or
many devices
Firewall policies can be easily
applied to VPN traffic, without
having to define the network
(IP-based) within that policy
Policies can be easily applied
to new networks/interfaces
Offers VLAN support to
integrate subnets easily
Different network segments
can have different policy sets,
effectively segmenting the
network
Administrators can apply
universal rules to multiple
security zones
Different network segments,
departments, offices, etc. can
manage their own security,
completely separate from each
other:
o Separate management
devices
o Separate “view”
Built in troubleshooting
features:
o Contextual information
Networks NetScreen-
Security Manager
Yes, role-based
administration
Networks NetScreen-
Security Manager
Yes, using Security
Yes, using Security
Yes, using Security
Yes, unique to
Yes, Virtual Systems
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes, Juniper
Yes
Yes, Juniper
Zones
Zones
Yes
Zones
NetScreen
Yes
Yes
Yes
See Juniper Networks VPN
White Paper “Dynamic VPNs
Achieving Scalable, Secure
Site-to-Site Connectivity”
shooting
Offers roll-back option to lastknown “good” configuration, if
changes do not “work”
Ability to integrate with other
management and enterprise
platforms/systems:
o SNMP traps
o MIP
o MIB
o CLI via SSH for
configuration
o Syslog
o NTP
On-line help Yes
Broad array of support options Yes
Support is delivered by a
single vendor with a single
support contract
VPN Specific
New networks can be easily
added to the VPN
Reroute around problems with
minimal human intervention
Flexibility to do complex VPN
configurations (e.g. hybrid full
mesh, hub and spoke) using:
o Rule-based VPNs
o Route-based
o Dynamic Route-
Based
Yes, utilizing dynamic
routing and Security
o Dynamic routing
automatically finds
available routes
o Route-based
VPNs can switch
to alternate routes
in route table
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Zones
Yes,
Yes
Yes
Yes
Note: NTP integration allows
internal clocks to be
synchronized to ensure log
files have accurate time
stamps
Delivered as an appliance for
simple deployment
Delivered as software that has to
be loaded onto hardware
Multiple deployment options:
o Transparent mode
o Route mode
o BGP
o OSPF
o NAT
Offers multiple ways to interact
with the system:
o Command Line Interface
(CLI)
o Web interface
o Graphical User Interface
(GUI)/central
management platform
Wizards to guide an administrator
through tasks, such as initial
configuration, policy install, VPN
set up
Templates available for consistent
configuration of multiple devices
Integrated key networking
functionality for easy integration
into a network environment, such
as:
o Dynamic routing protocols
o Virtual Routers
o Support multiple
routing domains
o Multiple methods of
address translation
o Dynamic IPs (DIPs)
o Support Mapped
IPsVLANs (MIPs)
o Support Virtual IPs (VIPs)
o Supports NAT
o Policy-based
o PAT/NAT capabilities
Single patches that apply to the
platform, OS and applications
Ability to maintain the VPN
abstraction and continue to
leverage dynamic routing when
applying the firewall policy
Tools and services to facilitate
migration from other Firewall/VPN
products
Yes, can be done on
per policy basis
Networks NetScreen-
Security Manager
Yes, through Security
Yes
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes, Juniper
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Zones
Yes
Can introduce interoperability
issues
o Support of DIPs allows
policy-based address
translations using pools of
IP addresses to handle
overlapping IP addresses.
o MIPs provide one-to-one
IP mapping for internal
servers
o VIPs provides mapping of
protocols from one public
external IP to multiple
internal private IPs based
on the port. Allows one IP
address to support Web,
FTP, e-mail and other
servers.
Not possible if applications,
OS and hardware are not fully
integrated or from the same
vendor
If the firewall policy requires
the use of IP addresses then
the management advantages
of dynamic routing are lost.
Remote User solution including
VPN, firewall, virus and
application-level protection
Provides strong remote site
security:
o Integrated functionality
to apply access control
to remote traffic
o Ability to protect against
viruses and applicationlevel attacks
o Split tunneling support
o Separation of corporate
and personal traffic to
ensure personal/Internet
traffic cannot enter the
corporate network
through the VPN
Supports a dial-back-up option to
ensure connectivity at a remote
office
Remote office appliance for easy
installation
Ability to configure a device at
the corporate office, so that
technical resources are not
needed at a remote site
Easy to manage to ensure
security experts don’t need to be
on site:
o Managed using the
same console as large
central site solutions to
ensure consistent policy
enforcement is
consistent
o Can be managed
centrally
Yes, purpose-built
Yes
Yes
Yes
Yes
Yes
Yes
device
Yes, Rapid
Deployment
Yes
Yes
Eliminates “weak” links with
affordable solutions
Features for Wireless
Can separate wireless traffic and
apply a security policy to it to
control access
*Please see specific product data sheets for individual product features, available on the Web site at