PRODUCT contents

PRODUCT contents
PRODUC T
Reviews 2007
Guide
TESTING & ANALYSIS
TO HELP YOU
MAKE PURCHASING
DECISIONS
contents
DATA PROTECTION
Application Security
2
Watchfire (IBM)
Database Security
3
4
5
Guardium
Lumigent
Oracle
Data Loss Prevention
6
7
Code Green
Workshare
ENDPOINT SECURITY
8
9
10
11
12
13
14
15
16
17
18
Bit9
CA
Checkpoint
eEye Digital
ESET
F-Secure
Norman
Novell
ScriptLogic
Trustware
Yoggie
Comparative Review
19 Color Me Complex
NETWORK SECURITY
IPS
37 Juniper Networks
38 Stonesoft
Log Management
39 LogLogic
Unified Threat Management
40
41
42
43
Information Security
magazine’s 2007
Product Review Guide
is a compilation of the
single and comparative
reviews published in
2007, an indispensable
guide for information
security managers
tasked with evaluating
and purchasing security
hardware and software
in 2008.
Fortinet
Secure Computing
SonicWALL
WatchGuard
VPN
44 SonicWALL
Web Security Gateway
45 Facetime
Comparative Reviews
46 Gone in a Flash
53 Universal Control
SECURITY TESTING AND ANALYSIS
Forensics
61 Paraben
Security Testing
IDENTITY MANAGEMENT
29
30
31
32
33
34
35
36
Apere
BeyondTrust
Cyber-Ark
DigitalPersona
e-DMZ
Identity Engines
Sun Microsystems
Symark
62 Core Security
63 Metasploit
VULNERABILITY/RISK
MANAGEMENT
64
65
66
67
Elemental Security
nCircle
Patchlink (Lumension)
RedSeal
I N F O R M A T I O N
S ECURITY
I N F O R M AT I O N S E C U R I T Y
®
1
P RODUC T
Reviews
A P P L I C ATI O N S E C U R IT Y
AppScan 7.0
R E V I E W E D B Y P H O R A M M E H TA
Watchfire
www.watchfire.com
Price: Starts at $14,400; Reporting Console
(including AppScan 7.0) starts at $35,000
existing policies to recording two-factor login information. Unfortunately, the login information is not stored
in an encrypted format.
The failure to incorporate sound security practices into
software development has left business-critical Web
applications open to attack, but that’s changing as corporations adopt secure coding requirements. To that
end, Watchfire’s AppScan 7.0 provides sound application security testing for developers, quality assurance
teams and penetration testers.
Installation/Usage
B+
The wizard-driven installation took five minutes;
AppScan runs on Windows XP, Vista or 2003 Server.
To initiate a scan, a wizard walks you through the
information required, from assessment type (Web application or Web service), starting URL, login parameters,
test policy (default, app only, infrastructure, invasive)
and scan options (full scan or explore/crawl). There are
plenty of advanced settings and customization options,
like two-factor recorded login and privilege escalation.
There are more than 75,000 individual security checks
distributed across various policy files; advanced users can
create custom tests in a few steps.
Advanced Features
B
AppScan has tried to create a one-stop solution for Web
application and services assessment by incorporating
multiple advanced techniques. Tools like HTTP Request
Editor, Encode/Decode and Regex Tester come in handy
for vulnerability assessment and other QA tests. You can
add external tools by linking to the executable.
Above all, AppScan gives you a single interface to
open all the tools and techniques required to test your
Web apps. Users have lots of options, from customizing
Testing methodology: AppScan 7.0 was run multiple times using
default and custom settings against two production Web applications
based on .NET, PHP, Apache Tomcat, Oracle and others.
Performance
B
The AppScan dashboard gives users multiple real-time
views of the structure, results summary and details of
vulnerabilities discovered. The number and severity levels of vulnerabilities are displayed in the bottom taskbar.
We ran the tool against two production Web applications, both of which handle sensitive data and use different application and infrastructure technologies. AppScan
discovered common issues, and a few subtle flaws.
We weren’t blown away with the scanning speed, but
were impressed with the adaptive scanning technique:
Once the tool determines that a particular technology,
say IIS, is not used, it removes all the corresponding tests
from the queue.
If you elect to report a false positive to Watchfire,
AppScan generates an unencrypted email to the tech
support team, so be sure to scrub any sensitive data from
the files before sending the email.
Reporting
A
AppScan’s reporting capabilities are as good as we’ve
seen in any tool. Report categories include security,
industry standard, regulatory compliance and delta
analysis. Each of these categories has multiple templates
and options to customize reports. Reports can be exported in numerous formats.
AppScan Reporting Console (sold separately) enables
users to consolidate vulnerability data into one centralized location to better control who has access to sensitive
data. Because it is Web-based, you can create dashboards
and for multiple users, such as QA and development.
Verdict
Consultants and in-house app security testers will appreciate AppScan’s accuracy and efficiency. The reporting
options alone are enough to wow management.w
I N F O R M AT I O N S E C U R I T Y
2
P RODUC T
Reviews
ning a sniffer; installing agents will require admin
credentials and console-level access.
The classification feature helps you identify potentially
sensitive information on a live database. You can create
rules based on SQL Guard’s Perl Compatible Regular
Expression (PCRE) engine to search for data, specific
permissions, or even conduct a catalog search. The results
can be categorized and assigned additional rules for protection.
You can create any number of levels of classification
depending on the complexity of your environment or
business (low, medium, high, or severe, critical, sensitive,
compliance, etc.).
DATABAS E S E C U R ITY
Guardium SQL
Guard 6.0
R EVI EWE D BY JAM E S C. F O STE R
Guardium
www.guardium.com
Price: Starts at $50,000
A
Reporting
In an industry flush with
products for securing the network perimeter, Guardium’s SQL
Guard 6.0 serves as an important
addition for monitoring and managing connections to and from a wide
variety of enterprise database products.
SQL Guard continues to address one of the most typical database audit failure points. Most auditors will not
issue a “pass” if you leverage a database’s native logging
features because they are owned and controlled by the
groups you are trying to monitor (for example, DBAs
should not be responsible for configuring and monitoring DBAs). SQL Guard ensures a system of checks and
balances between the security and database engineering
teams.
The solution consists of local database agents, network-based appliances to passively gather traffic or to
actively work as a firewall, and aggregation servers that
collect and analyze data.
Installation/Configuration
B
The preconfigured Linux-based 1U Dell appliances can
be plugged directly into the span port on a switch that
controls traffic to the databases. The administrative
account is created during installation, along with a series
of default user roles—common users, administrators,
DBAs, security, application developers, auditors, network
engineering—that can be used to create other accounts.
Passively collecting network traffic is as easy as runTesting methodology: We tested a Guardium G2000 appliance
testing a lab that contained DB2 8 and Oracle 9i and 10g on Linux 2.6,
Informix 7 on AIX 5.3, SQL Server 2000 and 2005 on Windows Server
2003, and Sybase 15 on Sun Solaris 9.
Guardium has all of the bases covered here. Reports are
grouped and labeled under three tabs for templates, custom reports and alerts. Templates include high-level or
technical information on database activities, sensitive
object usage, data markup language exceptions, overall
performance and permanent schema changes.
The strong custom reporting is built atop a SQL querying engine.
The new incident management dashboard provides a
clear-cut summary on policy violations and incidents. It
permits you to quickly dig deep into the incident, via a
click, to identify the timestamp, source/destination IP,
user, full SQL string, technical incident specifics and
more. The breadth of information is impressive.
Management/Monitoring
B+
In addition to monitoring database connections, 6.0 has
added application layer monitoring, providing JD
Edwards, Oracle, PeopleSoft, SAP and Siebel filters.
Alerts are triggered in one of two ways: statistical
or real time. Both save the same type and amount of
data; however, one is merely logged into the back-end
Guardium database and the other is logged and then
passed to one of four notification mechanisms.
Organizations looking to monitor databases in real
time will be best served leveraging SQL Guard’s integration capabilities as opposed to its console. SQL Guard can
easily integrate with SIEM or aggregation platforms via
SMTP, SNMP, syslog, or a custom Web-based Java class.
Verdict
SQL Guard has evolved from an impressive technology
to an enterprise-class data security product that
should be on every organization’s radar.w
I N F O R M AT I O N S E C U R I T Y
3
P RODUC T
Reviews
databases, it’s probably best to monitor them centrally
as opposed to leveraging individual agents. Each component’s policy can be tailored to the rules that matter
to the organization or geared toward a regulation.
Lumigent supports all major database platforms
including Microsoft SQL Server, IBM DB2, Sybase and
Oracle, all with good documentation.
DATABAS E S E C U R ITY
Lumigent
Audit DB 6.0
Lumigent
www.lumigent.com
Price: Starts at $10,500
Lumigent Audit DB 6.0 helps organizations, particularly those with significant compliance-related issues and
failed audits, protect and audit production databases.
The overarching problem remains
that database administrators and engineers do not have the security background required to lock down their
databases. Audit DB 6.0 is designed to
audit, monitor and protect databases.
Audit DB has two main auditing components. The
first captures network traffic to and from the database.
Matching patterns or out-of-policy actions are identified, triggering alerts. The second mode uses agents to
pull database log files, activity information and general
database configuration information.
Installation/Configuration
B+
Reporting
R EVI EWE D BY JAM E S C. F O STE R
B
Installing and configuring Audit DB is not an afternoon
project. Plan for key stakeholders within the database,
network and security teams to provide input into the
Lumigent product. You will need system and database
administrative credentials for the target databases as well
as admin-level access for the system that is going to
house the Audit DB reporting engine. Network administrators will help identify placement of the Audit DB
NetWatch sniffer agents, which can reside on the target
databases or on nearby systems. Your design will depend
on the number of databases you must audit; for multiple
Audit DB 6.0 is built atop a role-based reporting engine
that allows you to create and schedule reports based
upon organizational components, specific servers or
technologies, and audit requirements. Its strength is its
compliance/audit reporting capability, which supports
SOX, GLBA, SAS 70, HIPAA, PCI, SB 1386 and Basel II.
It also includes frameworks such as COBIT, COSO and
ITIL. These preconfigured reports are easily customized.
Executive and/or managerial dashboard-level views
allow you to drill down to the audited systems.
Of particular interest is Audit DB’s user reporting
capability. Most regulations place heavy emphasis on
user provisioning and decommissioning. Audit DB has
strong features meeting compliance standards for user
management and can be a valuable tool for identifying
obsolete or dormant users, validating password policies,
and identifying privileged users and entitlements.
Management/Monitoring
A
The most beneficial feature of Audit DB 6.0 is its API,
which comes with a complete 98-page reference, flush
with SOAP interface details and code samples. This integration allows database administrators and developers to
leverage Audit DB’s functionality natively within its
environments. Through the API, database and application developers have the ability to write events, logs and
other data to the Lumigent repositories, while administrators can automate common maintenance.
The SOAP interface is efficient and clean. If you’ve
ever implemented an RSS or XML feed you won’t have
an issue integrating this feature.
The tabs in the Web-based management interface
allow you to access all data collected by Audit DB. Drilldown reports take you into the details of an event with
timestamps, user information, data sources and activity.
Verdict
Testing methodology: We tested Lumigent Audit DB 6.0 running on
an Intel-Red Hat Enterprise Linux machine auditing an Oracle 10g R2
database on Red Hat Linux and SQL 2000 on Windows 2003 Server.
Audit DB is a strong tool for organizations that are
mandated to achieve and report compliance on their
database servers.w
I N F O R M AT I O N S E C U R I T Y
4
P RODUC T
Reviews
properly. A Realm is similar to a database software firewall. They can be put around an entire application or a
particular table within an application.
Vault uses two technology concepts to control application access, Command Rules and Factors. Factors are
properties or elements—users, IP addresses, network
ranges and specific databases—that can be included
within Command Rules.
Vault implements Command Rules to control the
execution of SQL commands and can control Data
Definition Language and Data Manipulation Language
SQL commands. This level of protection can be useful in
locking down permissions and accessibility for application service accounts and internal users alike. For example, a rule could be created to disallow any application
user from executing a CREATE DATABASE LINK command on a particular database, a command that is typically reserved when creating applications. Or, you might
prohibit any application user or service account from
leveraging SQL INJECT commands to thwart injection
attacks.
D ATA B A S E S E C U R IT Y
Oracle
Database Vault
R EVI EWE D BY JAM E S C. F O STE R
Oracle
www.oracle.com
Price: Starts at $20,000 per CPU or $400 per user
Auditing/Reporting
Oracle Database Vault enables advanced separation of
duty to help organizations meet compliance and data
security business challenges.
While database administrators and engineers may be
responsible for securing, managing, backing up and performance tuning, they shouldn’t need access to data.
Vault allows admins and application owners to manage
databases and applications without accessing credit card
numbers, customer information, company secrets, etc.
Installation/Configuration
B+
Set aside one morning to complete the installation; you’ll
be installing it on your current Oracle server (Oracle 9i
R2, 10g and 11 are supported), and will need both system
and database admin accounts—strong passwords are
required. The configuration agent helps automatically
configure the key components—adapter configurations,
DNS name, host name and host file updates.
Security Features
A
User Realms ensure data protections are implemented
Testing methodology: We tested Oracle Database Vault on Oracle
10g with Red Hat Linux. All components of the application were tested
to include user administration and application development.
B
Each created Realm can include auditing, or in this case,
event logging. If enabled, auditing comes in two flavors,
audit on failure and audit on success or failure. The audit
on failure option enables you to see who is attempting
to break the rules/Realms, while the more robust audit
on success or failure option will give you a picture of
everyone who successfully or unsuccessfully attempts to
conduct an operation that is protected by Vault.
While the reporting options are straightforward
and somewhat effective, there is room for significant
improvement if you intend to use these for daily operations. For instance, it would be beneficial to run operational reports within specific windows of time, or to
correlate events across all databases throughout the
enterprise.
Global reporting allows you to analyze results from
the entire database. The auditor and executive reports
include high-level statistics such as number of successes
and failures as well as user permission reports. User
reports and statistics can help you identify users, their
corresponding roles and access levels.
Verdict
Surprisingly mature for a first release, Oracle
Database Vault may prove valuable for large environments that have made heavy investments in Oracle.w
I N F O R M AT I O N S E C U R I T Y
5
P RODUC T
Reviews
I N F O R M ATI O N L E A K A G E
Content Inspection
Appliance 1500
R EVI EWE D BY M I K E C HAP P LE
Code Green Networks
www.codegreennetworks.com
Price: Starts at $25,000 for networks with up to 250 users
As organizations increasingly turn security focus
from outside attackers to
the threat from within,
they are beginning to consider information leakage tools. Code Green’s Content
Inspection Appliance 1500 (CI-1500) is among the stillmaturing handful of products designed to detect sensitive information leaving the enterprise.
Policy Control
A-
Code Green’s primary detection engine uses proprietary
technology to develop many short “fingerprints” of each
piece of submitted content.
Code Green also supports the use of regular expression matching rules to protect against users extracting
content from databases and other structured data sources.
The challenge here is determining and managing
rules for what constitutes sensitive data. This may be a
straightforward task for companies with clearly defined
document-classification policies, but it could pose a considerable challenge for less mature organizations.
You create policy rules based on content match, traffic source and destination, protocol, and desired action.
A document triggers an alert if it contains at least
one fingerprint from the RedList of restricted content
(the GreenList holds permitted content). Actions
include notifying the administrator, retaining a copy of
the offending document, sending a syslog report and
blocking/rerouting email messages.
We were impressed with the range of file types
covered—390 according to Code Green—including all
standard productivity software, such as Microsoft
Office, Acrobat and WordPerfect, as well as other formats such as AutoCAD, Flash and .exe/.dll binary files.
A
Configuration/Management
The appliance ships with default policies and settings. We
simply booted it up, attached it to a network tap and were
monitoring traffic.
We then explored the content registration and policy
creation process. The CI-1500 allows you to register individual content files through a Web upload, scan Windows
and NFS file shares, and integrates with Stellent and EMC
Documentum content management systems.
We evaluated both the Web upload and Windows file
share registration options and had no difficulty adding
new content. We also created policies based on customized regular expressions.
B
Effectiveness
We ran the CI-1500 through a variety of tests designed to
thoroughly evaluate its content-inspection capabilities.
It successfully detected Microsoft Office, Adobe Acrobat
and text documents that we attempted to send via Web
upload and email. We tried cutting and pasting document sections into other formats, and the appliance
detected our circumvention attempt. It also scanned and
detected protected content in .zip files.
The appliance’s Achilles’ heel, common to contentinspection products, is its inability to inspect encrypted
traffic. An insider aware the organization performs content filtering can simply encrypt outbound traffic (for
example, using Gmail) to bypass the appliance’s scrutiny.
The ideal solution would be to integrate a Web proxy
server to allow for SSL decryption and inspection.
Reporting
B+
The CI-1500 provides a fairly robust reporting mechanism. Administrators may use one of nine predefined
reports (such as top-matched policies, policy violations
by protocol and top-source emails of policy matches) or
create their own based on specified criteria (such as time,
policy violated, source/destination IP, content).
Verdict
Testing methodology: The CI-1500 was tested in a network of
Windows and Macintosh clients using a variety of common file formats,
including Microsoft Office files and Adobe Acrobat documents.
Code Green’s CI-1500 could prove valuable to organizations with well-defined classification schemes.w
I N F O R M AT I O N S E C U R I T Y
6
P RODUC T
Reviews
a large set of policy violation responses. For example, we
created a policy to intercept all outbound email including Microsoft Office attachments containing reviewer
comments, and scrub those comments from the document before transmitting it to the end user. In addition
to cleaning documents, administrators may create policies that block transmission, alert administrators to
the violation, send the document for review prior to
transmission, convert the document to a secure PDF
or compress the file in a secure ZIP. Administrators
may create policies based upon document contents
(keywords, phrases or regular expression matches) or
attributes (classification, hidden metadata, file name,
file type, file size).
D O C U M E NT C O NTR O L
Workshare
Professional 5
R EVI EWE D BY M I K E C HAP P LE
Workshare
www.workshare.com
Price: Starts at $349 per workstation
Workshare Professional allows
organizations to maintain control over information throughout the document lifecycle. It
provides the ability to send
documents to internal and
external reviewers, manage
changes with a strong audit
trail, and protect documents
from data leakage.
Workshare has strong integration with Microsoft Office
2007 but is significantly limited in effectiveness by its
ties to specific email platforms.
Configuration/Management
A-
Installation of Workshare Professional on the client uses
an intuitive wizard-driven process. Workshare automatically installs Office integration features, allowing the
user to quickly begin operating in a Workshare environment. Office 2007 users will find a Workshare pane in
the Office Ribbon that offers all Workshare functionality
in a familiar format. Administrators create and manage
policy through the Workshare Policy Designer in the
desktop version of the product. Enterprise installations
may distribute policies through a centralized server.
Policy Control
A
Workshare allows for an impressive degree of granular
policy control over user actions, allowing administrators to specify broad or narrow criteria and choose from
Testing methodology: We tested Workshare Professional 5 in
a Windows XP environment using Microsoft Office 2007.
C
Effectiveness
Workshare met all of its stated document control objectives and correctly enforced our defined policies when
we attempted to send protected content via Outlook
email. The biggest drawback to this product is its sole
reliance upon three supported email environments:
Microsoft Outlook, Lotus Notes and Novell GroupWise.
This limitation makes the content distribution control
features of Workshare useful only to prevent accidental
data leakage or deter a novice attacker. In fact, we were
able to defeat the content controls by simply uploading
the file via a Web-based email connection. Workshare
does offer a complementary network appliance that filters
all SMTP and HTTP connections, but this device does
not support encrypted connections, making it straightforward for a determined attacker to undermine the
system’s controls.
Reporting
B+
Workshare offers three types of reports: an audit report
that itemizes all changes made to a document by any
reviewer, a history report that shows participation in the
document review cycle, and a risk report that shows
hidden metadata that has the potential for inadvertent
information disclosure. Workshare will generate reports
in either HTML or XML format.
Verdict
Workshare Professional is a good document control
solution for organizations seeking to prevent inadvertent
disclosure of sensitive information. Organizations should
supplement it with strict controls on outbound network
traffic to filter all content leaving an enterprise.w
I N F O R M AT I O N S E C U R I T Y
7
P RODUC T
Reviews
EN DPOINT SECURITY
Bit9 Parity 3.5
R EVI EWE D BY G R E G BALA Z E
Bit9
www.bit9.com
Price: $35 per desktop
Bit9’s Parity 3.5 is designed to give you control over what
users can do on company computers, and prevent executables from unauthorized or malicious apps from running on your desktops.
Configuration/Management
B
Bit9 Parity Server was easy to install and didn’t have
much to configure. A step-by-step screen walks you
through setting everything from IP addresses and ports
to selecting the creation of a self-signed or previously
generated certificate. It automatically installs SQL Server
2005 and Apache Web Server, which is used for remote
administration.
Small client agents for Windows XP/2000 (Vista is
coming) are generated or updated automatically when a
policy is created or modified for a group. The agents can
also be downloaded off the Web, or distributed by application deployment software such as SMS. The agent and
server communicate via a SSL tunnel.
Policy Control
B
Policies are applied based on groups set up within Parity
Server that specify the file types it will block. Security
condition levels, set by group, determine what happens
when a file violates policy—various combinations of
allowing or prohibiting file execution with or without
notification. For policy enforcement, you can identify
executables by name, or hash the file. Although malware
can use an altered name to pose as a legitimate app,
Parity will report on renamed programs. We recommend
using hashes, though this means additional administrative overhead before deploying software.
Programs can be authorized to run from trusted indiTesting methodology: We installed Parity Server on a Windows
2003 SP1 machine to manage several fully patched XP and Windows
2000 VMware clients. We used a variety of applications, such as
Skype, Kazaa and µTorrent, to test executable blocking.
viduals, trusted directories or trusted deployment
applications, eliminating the need to manually add to
the policy for each software deployment.
Recognizing the problems posed by mobile workforces, Bit9 allows for different security conditions
when attached to the local network, and when disconnected.
Effectiveness
A
Parity is effective at stopping programs from executing,
as the agent goes through a lengthy process of inventorying the host workstation and reporting executable files
to the Parity Server. This can take a while, especially in
large enterprises with many clients.
Parity Server uses a combination of blacklisted applications and Bit9’s signature database of known malware.
The latter prevents the rapid spread of viruses and spyware from host to host by identifying the offending program and preventing its subsequent execution on other
protected systems.
The Parity agent allowed executables to run according to policy, and quickly caught changes we made to a
file. For example, we renamed Kazaa, a prohibited app,
but still couldn’t run it.
Reporting
C
Bit9 has some work to do to beef up reporting capabilities. While several canned reports give quick access to
important information, the sparse main reports page
gives limited statistics on important file activities. We
were disappointed that there’s no way to graph the
statistics, which would be especially useful for trending
reports. There’s no syslog support, nor can reports be
exported to another format.
Verdict
Bit9 Parity 3.5 does a good job of preventing unwanted
programs from running, although we didn’t see any
new methodologies or technologies that make it
stand out from established competitors.w
I N F O R M AT I O N S E C U R I T Y
8
P RODUC T
Reviews
HIPS
Policy Control
CA Host-Based
Intrusion Prevention
System
R EVI EWE D BY B RAD CAU S EY
CA
www.ca.com
Price: $40 per client with enterprise-level maintenance
CA Host-Based Intrusion
Prevention System (CA
HIPS) combines standalone firewall and intrusion
detection and prevention
technologies to provide
security, access control,
policy enforcement, intrusion prevention management and deployment from a central console.
Configuration/Management
A
Management server setup was simple and fast. The server
will automatically enumerate all LDAP users and groups,
ultimately providing the ability to control policies and
rule sets by Active Directory group membership. The
management server is configured by default to check for
LDAP changes every 180 seconds.
The real beauty of CA HIPS is its Learning Mode,
which allows you to monitor a group of systems to determine what constitutes acceptable behavior. This will help
avoid the problem of a policy deployment that is either
too restrictive, causing application/network outages, or
too liberal, limiting the effectiveness of the product.
Client deployment is a cinch for desktops and servers.
Log in to the intuitive, Web-based console and configure
the client with the options you desire. Click “Build” and
you are presented with an installation package.
Testing methodology: We assigned policies to users and computer
groups in a single Active Directory domain. Attacks were launched on
the clients in the form of viruses, worms, remote exploits (some with
no known patches) and spyware.
A
The intuitive policy creation and deployment workflow
simplifies what could be a complicated process.
You begin by creating “Common Objects”—basically,
the targets of security policies, such as USB drives, registry entries or network protocols. There are thousands
of default objects, which can be easily customized. Next,
you define the list of rules associated to the objects. Rules
are broken down into categories such as application, firewall, operating system and IDS/IPS.
Customizing or creating rules is simple, and they can
be grouped into manageable collections such as highsecurity laptops and DMZ Web servers.
Policies allow you to apply groups of rules to subnets,
hosts, users, groups and a number of other criteria. A
simple deployment wizard pushes the policy to clients.
Effectiveness
B
CA HIPS provides effective defense against threats,
known and unknown. Clients are heavily protected, but
in a way that will not adversely affect the functionality of
the system. By sandboxing an application or questionable device, it protects against unknown threats.
We executed a number of viruses, Trojans and other
malware, many of which didn’t have signatures. Software
that wasn’t approved was sandboxed based on the policy,
and escalated for approval/restriction. Known malware
was restricted from accessing the OS and network.
You’ll still need an antivirus product to remove viruses
and Trojans. CA’s antivirus product works with CA HIPS
to remove malicious code, but the products aren’t completely integrated yet.
Reporting
D
Reporting seems immature. There are around 100 prebuilt reports that offer a plethora of information, but
there’s no method to modify them or create custom
reports from inside the Web interface. The only criteria
that can be supplied to filter the results are based on time
frames, such as the last week or month.
CA does provide an API to create custom reports, but
that shouldn’t be necessary to create a report from data
that is stored on the management server.
Verdict
CA HIPS offers comprehensive threat defense in a
flexible and easy-to-use product. Its reporting weaknesses may rule it out for some organizations.w
I N F O R M AT I O N S E C U R I T Y
9
P RODUC T
Reviews
HotPick
Pointsec Protector
D E V I C E MANAG E M E NT
TION
R MAIT Y ®
O
F
IN
UR
SEC
R E V I E W E D B Y S A N D R A K AY M I L L E R
Check Point Software Technologies
www.checkpoint.com
Price: Starts at $45 per seat
Pointsec Protector (formerly Device Protector prior to Check Point’s acquisition of
Pointsec) addresses the growing problem of
unsecured ports and endpoint devices while
transparently delivering encryption, filtering
content, enforcing policies and maintaining an
audit trail, even when mobile devices are disconnected from the network.
Configuration/Management
A
Getting the product running was effortless, thanks to
well-designed wizards and a straightforward installation
process. Protector ties in with Microsoft Active Directory
and Novell eDirectory for user and group synchronization when assigning device access rights, encryption and
policies.
The administration console is intuitive, and multiple
tiers of administrative access can be assigned for distributed management. We were able to easily manage users,
groups and devices, policies, alerts and encryption, and
create and view audits, logs and reports.
Policy Control
A
We began by editing Protector’s default profile through a
series of tabs to choose what types of devices and removable media to permit/deny access, define encryption, create email alerts, and assign stringent policies for groups
that fell under compliance regulations (e.g., finance) and
less stringent for others.
Policies are layered, so the default policy is applied to
every group to which it is assigned. When another policy profile is created, it can inherit from the default poliTesting methodology: The testing environment included Windows
clients, AD and SQL Server. We tested the use of portable storage
devices, including USB flash drives, FireWire external hard drives,
CD-RW drives and floppy disk drives.
cy or become a new profile. For example, in the default
profile we globally banned iPods and enabled encryption
on all USB storage devices. The next policy, while
it inherits the default profile, may define access to
approved devices, such as portable hard drives, on which
encryption from the default policy will be enforced.
Policies can be assigned on a user, group or device
basis. Administrators can restrict the types of files that
may be transferred or the launching of unauthorized
applications from removable media.
Protector uses combinations of whitelists and blacklists to block access to devices and files without any legitimate business purpose, while still allowing users access
to critical tools, applications and data defined by brand,
model and file type.
A
Reporting
Protector excels in logging and reporting. With detailed
auditing, administrators can determine what devices are
being used and in what way. Alerts are easily set up to be
sent via email; we assigned each AD group a different
notification recipient simulating department managers
being alerted to their employees doing such things as
downloading music at work or copying sensitive files to
portable media.
Logs can be customized, filtered by column heading
and exported to CSV. Reports are equally flexible and
can be exported in HTML.
B
Effectiveness
Protector enforces all policies and offers a high level of
control and auditing over offline devices. Even with local
admin rights, we were prevented from disabling or uninstalling the client software from our test laptop thanks to
anti-tampering controls.
The encryption feature works transparently when the
user is logged on to the network. For offline machines
and mobile devices, users simply drag and drop files on
or off of the encrypted device through a password or
challenge/response.
Protector lacks centralized control for Linux and
Mac, and doesn’t have data shadowing, meaning administrators could record all information sent to a particular
device or port.
Verdict
Pointsec Protector is an affordable and scalable
solution that will work well in both SMB and
enterprise environments.w
I N F O R M AT I O N S E C U R I T Y
10
P RODUC T
Reviews
EN DPOINT SECURITY
Blink Professional 3.0
R EVI EWE D BY STEVE N WE I L
eEye Digital Security
www.eeye.com
Price: $59 per computer per year
Antivirus software is no longer
enough to protect your company’s
computers. Prolific laptops, desktops
and critical servers are facing threats
from many fronts: malware; insecure
protocols and applications; lost,
stolen or misused portable storage
devices; and network traffic. Host
IPS, antivirus and storage device
control programs can mitigate certain threats, but force security managers to install and manage multiple applications.
eEye’s Blink Professional 3.0 is among the increasing
number of host-based endpoint security products that
use a layered, consolidated approach to defend Windows
computers against different attacks.
Configuration/Management
B
Following eEye’s well-written documentation, we were
able to quickly and easily install Blink. Blink’s interface
is intuitive and easy to use; we were able to effectively
navigate among the many local settings.
We liked the well-designed wizard programs that are
used to create rules and signatures. We also liked being
able to add references, such as CVE and Bugtraq IDs, to
IPS rules. Blink can be configured to automatically check
for software and signature updates.
Policy Control
B
Blink deploys a single agent and common management
of its multiple security capabilities: a host firewall, monitoring inbound and outbound traffic; an application
firewall that controls the network activity of installed
applications; signature- and protocol analysis-based host
Testing methodology: Our test network included a Windows 2003
laptop, an unmanaged switch and three Windows 2003 servers.
intrusion prevention; antimalware protection against
worms, viruses and spyware; antiphishing capabilities;
system protection against buffer overflows; and controls
over which applications may access the registry and/or
be launched.
Blink can also block the use of storage devices, such as
USB flash drives, and conduct local vulnerability assessment scans.
Security managers can configure Blink locally, or configure it to regularly check and download a centralized
policy. Blink can also be integrated with eEye’s REM
Security Management Console for creation and management of dynamic policies. It also centralizes logging.
We were able to create numerous granular firewall
rules, IPS signatures and system protection rules, which
defined the actions to be taken (allow, log, block, alert).
A
Effectiveness
Blink did an excellent job of protecting host computers.
We ran numerous manual and automated attacks
against our test computers; our attacks included sending
malicious data and executing unexpected protocol
actions. Blink always took the correct action, such as
blocking or logging attacks. Permitted traffic was correctly allowed. Blink also correctly blocked the use of
prohibited storage devices and detected malware we
installed on the test computers.
B
Reporting
Integration with REM enhances Blink’s reporting abilities. In standalone mode, Blink locally logs system, firewall and IPS events, and can send SNMP traps. Individual
log events are easy to understand, but the logs can only
be exported as a .csv file.
Via Blink’s local event log interface, an administrator
can select an individual log event and, as appropriate,
block an IP address, go to the rule that logged the event
or create a new rule in response to an event, such as
allowing traffic that was blocked. Administrators can
configure Blink to pop up a user alert when a specific
event occurs, such as an RDP connection to a server.
Blink also generates useful reports after an antimalware and/or vulnerability assessment scan is run, but
they cannot be exported.
Verdict
Blink is well-designed, and its multilayered approach
makes it a good choice for protecting Windows
computers throughout an organization.w
I N F O R M AT I O N S E C U R I T Y
11
P RODUC T
Reviews
ANTIV IRUS
NOD32 Antivirus 2.7
R EVI EWE D BY M I K E C HAP P LE
ESET
www.eset.com
Price: Starts at $39 per workstation for a one-year license
NOD32 Antivirus provides effective virus and spyware protection,
albeit without many of the bells
and whistles found in competing
products.
Our review supports ESET’s
claim that NOD32’s lightweight
design allows it to scan systems
more efficiently than the competition. While it might appeal to
administrators seeking to scratch
out every last cycle of CPU performance, it’s not a product for those
seeking a highly integrated enterprise solution.
Configuration/Management
B
Client installation uses standard deployment techniques—push the package onto client systems, create a
login script or distribute installation packages through
traditional means. The central administration capabilities require installing the remote administration server,
and the remote administration console on administrator
systems. We used the wizard-guided installation processes to install all these packages without difficulty.
Policy Control
C
NOD32 allows administrators to apply policy to individual systems or groups of systems through the remote
administrator feature. Policy control is quite granular: you
can specify scan types and frequency, excluded files, etc.
Testing methodology: We tested NOD32 in a Windows XP virtual
machine environment with separate machines serving as the client
and administrator. We performed the competitor comparison on an
identically configured virtual machine.
You can apply different policies on a group basis.
However, the policy control GUI is very awkward. To
create or modify a policy, you must open a separate policy editor and then save the resulting policy as an XML
file. You can then return to a separate window to apply
the XML file to systems/groups.
NOD32’s Remote Administrator package provides
limited integration with Active Directory. You are able
to synchronize NOD32’s internal groups with an Active
Directory server, but you can’t manage NOD32 within
the native AD environment. Additionally, you can only
import machine information, not user information. This
highlights one of the package’s significant shortcomings:
NOD32 does not allow you to create individual administrator accounts, so admins must use a shared server
password—hardly a best security practice.
A-
Effectiveness
NOD32 delivers on its promise. It successfully detected
the test viruses we placed on the evaluation system and
alerted the administrator to their presence.
NOD32 met its efficiency claims. We ran it against a
leading product in a head-to-head test on identical environments. Both packages detected our test viruses, but
NOD32 completed its scans 16 percent faster.
An active monitor watches for file system changes,
while the on-demand scanner performs full system
scans. The Internet monitor scans HTTP and POP3
traffic. While NOD32 does not support scanning of
IMAP connections, an add-on supports Microsoft
Exchange. Another module scans Microsoft Office
documents.
NOD32 doesn’t include a centrally managed host
firewall service or a host-based IPS. These shortcomings
may limit the product’s viability for some organizations.
A
Reporting
NOD32’s smorgasbord of reports will please even
the pickiest data hound. Predefined templates include
reports on top viruses, top clients with most alerts and
alerts by module, among others. Administrators can filter
reports based upon specific clients, servers or viruses
and the HTML-based reports can be exported to a .csv
file for use in other analysis packages.
Verdict
NOD32 is an efficient, effective antivirus solution with
decent central administration tools. However, it lacks
strong enterprise-level features.w
I N F O R M AT I O N S E C U R I T Y
12
P RODUC T
Reviews
EN DPOI NT SECU RITY
Management
F-Secure Client
Security 7.0
The Java-based Policy Manager console interface is
clean, with tabbed windows allowing easy navigation.
The console allows you to install F-Secure remotely, as
well as manage configuration, updating and monitoring. The policy management tool auto-detects Windows
machines, facilitating installation.
The console allows admins to centrally control all
installed systems. Pressing one button allows you to
scan every system for rootkits with the new BlackLight
feature (rootkit and spyware detection), and remotely
update all clients with the latest virus definitions.
R E V I E W E D BY B R E N T H U STO N
F-Secure
www.f-secure.com
Price: $29.75 per user for 100 licenses
In the face of criminal zero-day
exploits, targeted attacks and mobile
workforces, companies are increasingly seeking comprehensive endpoint
security tools. F-Secure Client Security, a business-grade, centrally managed suite, addresses emerging threats,
as Version 7 adds rootkit scanning
and host-based intrusion prevention
(HIPS).
Installation & Features
B
B+
Both the Policy Manager and client software install in
a snap. By default, the antivirus scanner does not look
inside archives, or force a complete system scan upon
installation, saving system resources and time. It removes
other antivirus or firewall products, although it left
Comodo Firewall on one of our test systems. Client
Security suite consumed about 80 MB of memory on
our test systems—a little high, but not excessive.
In addition to new rootkit scanning and HIPS capabilities, Client Security suite features an antivirus and
antispyware scanner, mail and Web proxies, and client
firewall. The proxies act as a middleman, inspecting
email and data passed via Web browsers. This allows
Client Security to remove malicious software before it
has completely reached the system. F-Secure’s excellent
antivirus engine is known for its speed and efficiency; it
displays the same competence in this suite.
Testing methodology: The management server was run on Windows
Server 2003, the Client Security Suite on Windows XP SP2 clients. We
tested against common viruses and malware, as well as newer malicious
software.
Effectiveness
A
Client Security is unobtrusive during normal desktop
operations. It caught everything we threw at it: common
viruses and rootkits, as well as hacking tools planted on
test systems. We browsed Web sites that install malicious
software to test the HTTP streaming scanner, which
stopped all threats before they could be downloaded.
Client Security also performed well in cleaning up
already infected systems.
The new HIPS component, DeepGuard, which is
designed to protect against new threats, blocked malware from installing when the antivirus scanner was
disabled. It also stopped software downloads most
organizations block by policy, such as some browser
toolbars. DeepGuard can also be configured to prompt
anytime something tries to change the registry, instead
of relying on its AI to detect if the change is malicious
or not.
Reporting
B-
Client Security reports are simple HTML pages. The
Policy Manager console reports are a bit more extensive,
allowing current status, trend and detailed list reports.
Policy Manager stores data in a lightweight SQL database, but it too can only generate reports in HTML.
However, the reports are well designed, and included
an easily printable version. The bookmarking feature
generates a new report with the latest data for the
report type you specified every time you select a bookmark.
Verdict
Client Security is fast, efficient and reasonably priced.
The Policy Manager is free, a definite value-add.w
I N F O R M AT I O N S E C U R I T Y
13
P RODUC T
Reviews
M A L W A R E A N A LY S I S
Norman SandBox
Analyzer Pro
R E V I E W E D BY TO M L I STO N
Norman
www.norman.com
Price: Starts at $5,000 for 100 users
Relying solely on
antivirus to protect
you from malware is
no longer an option.
Antivirus software is
reactive; vendors only
release signatures for
malware they’ve seen. With the growing prevalence of
more targeted viruses, the bigger your company, the
more likely you are to be hit by something that no one,
not even an antivirus vendor, has seen before. In
response, many companies are developing in-house malware analysis capabilities.
Norman SandBox Analyzer Pro is a unique malware
analysis tool that allows potentially malicious code to
execute within a simulated environment that effectively
mimics a generic Windows installation. All actions taken
by the code under analysis are monitored. Any permanent changes that the test code attempts to make are
trapped by the sandbox (files don’t get written to the file
system, keys don’t get changed in the registry) but everything appears normal from the point of view of the code
under test.
Analysis Tools
A-
Analyzer Pro provides analysts with an almost overwhelming amount of information about the inner workings of the code under test. From the files it attempts to
Testing methodology: Analyzer Pro was tested on a Windows XP
Professional machine with a 1.8 GHz processor and 1MB of RAM.
Testing was done by analyzing a variety of sample code (from the
reviewer’s malware “zoo”) using the tools provided. Tests were
performed using known benign code and previously analyzed
malware samples.
create, to the registry entries it adds or changes, to the
network connections it attempts to make, Analyzer Pro
sees and logs all.
One incredibly useful feature is the ability to allow
mediated access to the Internet using powerful filtering
tools. Access to the Internet can be controlled in many
ways—remote connections can be “faked” by Analyzer
Pro, access to the real Internet can be allowed, or the analyst can alter packets being sent or received from the
Internet on-the-fly.
Recent malware often has a networking component
that can only be fully investigated using this feature. For
example, the behavioral aspects of a bot program can be
fully understood if it is allowed to contact its command
and control server.
B-
Usability
Analyzer Pro is a powerful tool for combining code-level
analysis with extensive behavioral monitoring and logging, but it has a steep learning curve. The main analysis
tool is a specialized debugger that allows the analyst full
control over the execution of the program at a granular
level.
This is not a tool for neophytes. Even with years of
experience using debuggers and code analysis tools, we
found Analyzer Pro to be very confusing at times. We
had to analyze several dozen pieces of code before we felt
reasonably comfortable with the tool’s quirks.
If your organization is looking to start analyzing
malicious code, we would suggest staying away from
Analyzer Pro until you hire experienced malware analysts or develop internal expertise.
D
Documentation
Perhaps the greatest problem is documentation. Analyzer
Pro was obviously originally developed by Norman as
an internal analysis tool, and that heritage is evident
in its documentation. It is poorly written, confusing
and assumes a level of expertise that makes Analyzer
Pro unsuitable for anyone but a seasoned malware
analyst.
Verdict
Although it lacks polish in its user interface and its
documentation, SandBox Analyzer Pro’s powerful and
flexible feature set makes it a desirable tool for seasoned malware analysts. Beginners will find it frustrating
and confusing, but mature code analysts will find it a
welcome addition to their toolkit.w
I N F O R M AT I O N S E C U R I T Y
14
P RODUC T
Reviews
like reporting and alerts accessible through a click.
E N D P O I NT S E C U R ITY
ZENworks Endpoint
Security Management 3.5
R E V I W E D B Y S A N D R A K AY M I L L E R
Novell
www.novell.com
Price: $69 per device
ZENworks Endpoint
Security Management
3.5 (formerly Senforce
Endpoint Security
Suite) is a comprehensive endpoint security management
solution that lets organizations control applications,
protocols and removable storage devices. It delivers
encryption to files and folders, and network access
control to ensure protection levels are current.
Configuration/Management
B-
We installed ESM’s three components—Policy Distribution Service, Management Server and Management
Console—on a server. Larger deployments require that
the Management Server and Policy Distribution Service
be installed on separate servers.
Installations were straightforward; the client required
us to choose between obtaining policy updates through
ESM or files. The Policy Distribution Service checks what
is sent out against the Management Server, which interfaces with directory services. Password protection for the
client prevents removal and tampering.
Setting up the server required extensive networking,
security and SQL knowledge.
Multiple installs on secured machines connected to
the server are possible, but a Web-based interface would
make configuration and management easier.
The console allows navigation through the taskbar
and expandable submenus, but we’d prefer to see items
Policy Control
A
ESM earned top marks for the granularity and scope of
security and control policies.
There are extensive policy options for wireless and
wired networks, communications hardware, firewall settings for multiple locations, antivirus/spyware and Microsoft patches. You can use advanced scripting rules for
customized rule sets and set features and alerts specific
to regulatory actions.
Policies are distributed via SSL through a Web services
application, pulling users and groups from directory services. Policies are easily edited and instantly updated.
Reporting
B
Alerting and reporting are tough to locate, hidden in the
Management Console’s Tasks menu.
Alert thresholds are adjustable. For example, we
enabled an alert if data in excess of 5 GB is copied to
removable storage media or device.
ESM offers 10 reporting categories—adherence, alert
drill-down, endpoint activity, encryption solution, client
self-defense, integrity enforcement, outbound content
compliance, administrative overrides, endpoint updates
and wireless enforcement. If you want to create custom
reports, however, you’ll have to use an ODBC-compliant
app such as Crystal Reports.
Effectiveness
B
While ESM provides a multitude of security and control
features in a single suite, there are a number of features
available in similar products we would like to have seen.
For instance, we could not assign storage device control
policies when encryption for the particular device was
required. Also, an additional USB Drive Scanner Tool had
to be installed separately from the initial installation to be
able to scan and identify devices attached to USB ports.
Despite those shortcomings, policies were automatically distributed to clients. ESM blocked noncompliant
clients that were assigned specific requirements (such as
up-to-date antivirus signatures).
Verdict
Testing methodology: The single-server installation was deployed on
a Windows-based network behind a firewall. Clients were installed on
a variety of endpoint systems located within and outside of the firewall.
Policies were enabled for a variety of scenarios, including remote and
mobile endpoints.
ZENworks Endpoint Security Management is a comprehensive solution for managing and enforcing security
policies on networked devices, regardless of their
location and connectivity.w
Review how we grade at searchsecurity.com/grading_criteria.
I N F O R M AT I O N S E C U R I T Y
15
P RODUC T
Reviews
(antispyware deployment relies on the installation of
the Shavlik module). ScriptLogic also offers its USB/
port control tool, which secures data against theft or
accidental loss.
EN DPOINT SECURITY
Desktop
Authority 7.5
Policy Control
R EVI EWE D BY HAR R I S WE I S MAN
ScriptLogic
www.scriptlogic.com
Price: $30-$40 per seat
As the name suggests, ScriptLogic’s Desktop Authority 7.5 is a centralized desktop
management solution. What makes this
product unique is the optional add-ons
that address patch management, spyware
removal and USB/port security. The total
package makes this a useful tool for dayto-day desktop and security management.
Configuration/Management
B+
Server installation was straightforward: The software
recognized that the base Windows 2000 software was
missing several core components and added them.
When a user logs on to the network, Desktop Authority
executes a script, SLogic, that installs the client software
and configures the registry. This script is assigned to users
through User Manager for Domains or Microsoft Active
Directory, or from within Desktop Authority itself.
Through a dedicated central console, admins can
manage the configuration of the user’s environment,
maintain a complete inventory of the enterprise’s hardware and software, and support users remotely.
Granular profiles allow for group and individual user
management for various environments. As the user logs
on to the network—either on the network or by connecting remotely—the script activates and applies the
group’s rules to the workstation. This allows admins to
maintain consistent and updated policy controls.
Desktop Authority bundles several optional security
components, including patch management via Shavlik
Technologies and antispyware from Aluria Software
Testing methodology: Our test platforms included Windows 2000
Server, Windows XP Home and Windows XP Professional.
B+
Desktop Authority makes it easy to administer several
key security functions. Patches can be tailored to each
individual group’s needs and deployed accordingly. This
also works for the antispyware component. One nice feature is the ability to exempt specific “spyware” applications, which is how many security tools are classified.
The USB/port control component can manage use of
Bluetooth, CD/DVD drives, FireWire devices, floppy
disks, hard disks, infrared ports, Zip disks, MP3 players,
modems, PCMIA controllers, PDAs, USB devices, parallel/serial ports and Wi-Fi devices.
Administrators have granular control of what can be
accessed; for example, you can grant read-only access to
CD-ROM burners, allowing users to access portable disk
content but not copy sensitive data.
Effectiveness
B+
In our testing, we focused on the security modules, developing a series of groups that denied and granted access
to specific devices and media. The system prevented
us from violating the assigned access control lists. Using
these same groups, we also deployed patches to Windows
XP desktops and laptops. We deployed the antispyware
component using the patch management tool.
Reporting
A
The reporting tool is very effective, allowing the administrator to use either packaged or custom reports. All
three security components use Desktop Authority’s
reporting tool, which allowed us to generate canned
reports on antispyware, USB/port and patch activity,
as well as administrative desktop management. Each of
these areas features several detailed options; for example,
under USB/port management, you can generate reports
by a specific type of device or workstation. Each canned
report can be modified to suit the enterprise’s needs.
Verdict
Desktop Authority is an effective desktop management/
security tool. While it does not work with non-Windows
systems, it would be a cost-effective solution in a
Windows shop.w
I N F O R M AT I O N S E C U R I T Y
16
P RODUC T
Reviews
GPO and software installation packages by copying the
.MSI file to a shared drive from which it’s deployed to
workstations as users log in.
This should be fine for most organizations, though
some may prefer products with more robust proprietary
consoles.
V I RTU A L I Z ATI O N
BufferZone
Enterprise
Policy Control
R EVI EWE D BY G R E G BALA Z E
TrustWare
www.trustware.com
Price: Starts at $2,899 per 100 licenses
With the information
security field seemingly
saturated with every possible appliance and software, it would seem
there’s little room for
an innovative approach.
TrustWare’s BufferZone
belies that notion.
TrustWare’s BufferZone
works by quarantining
suspect or restricted applications, creating a protected environment for each
Web- or network-based
application, such as Web
browsers, IM, email and
P2P applications, preventing viruses or malware from entering and affecting the
rest of the workstation
Configuration/Management
B
Instead of designing a management console, BufferZone
relies on Microsoft’s native Group Policy Objects to
manage and deploy BufferZone and its installation file.
This allows easy integration with Active Directory and
reduces the learning curve, through a group policy
template that uses the familiar management console
(MMC). Simply copy the administration file to the
c:\windows\inf\ directory and add it to the GPO administration templates. Deployment was just as easy, using
Testing methodology: BufferZone was run on two workstations,
Windows XP SP2 and Windows 2000 SP4, that were in a standalone
AD domain.
B
Depending on the policy set, you can easily prevent P2P
applications, such as Kazaa, from being launched by a
user, or allow ActiveX applications to be run only in the
BufferZone-protected area. You can add applications or
files by simply typing the name of the executable or DLL
in the applicable dialog box. Any application or attachment that is launched by a Web navigator, P2P application, IM or mailer is quarantined by default.
You can choose from four policy settings under
which files can be run: BufferZone, so files run only in
the protected area (this prevents a file from affecting
other areas of the workstation hard drive or memory
space, or removable media such as a CD, flash drive or
MP3 player); Forbidden, in which users have no access
to the files; Confidential, which means any file or path
matching the policy is invisible to applications run in
BufferZone; or Trusted.
In testing, we found that using both file path and wild
cards was best for policy enforcement (for example,
\*\MY DOCUMENTS\*.doc and *torrent.exe). BufferZone includes a switch to allow digitally signed executables to run outside of the controls set for a certain media
or file.
Effectiveness
B
Although BufferZone was excellent at stopping potentially malicious executables and preventing CDs or USB
devices from being accessed, it was obvious that it was
only for files already known. A new policy would have
to be created each time an unknown application or executable was installed. Fortunately, in an enterprise environment where typically standardized applications are
installed, this shouldn’t be too much of an issue.
The lack of any reporting capabilities may give pause
to some enterprises.
Verdict
BufferZone does a remarkable job controlling which
files can be run or downloaded on a workstation. The
typical entry points, such as Web browsers, can be
locked down, preventing unwanted access.w
I N F O R M AT I O N S E C U R I T Y
17
P RODUC T
Reviews
integrated antivirus and automatic updating of the
device firmware and virus definitions.
These features are excellent capabilities bundled into
the basic product, but we would have liked more finegrained control.
E N DP OI NT SECU RITY
Yoggie
Gatekeeper Pro
A
Effectiveness
We ran several Nessus and Nmap scans as well as some
custom penetration testing tools against the Yoggie-protected test system to see if we could either take out the
device, or tunnel a way through it. The Gatekeeper Pro
passed with flying colors and shrugged off the attacks,
while allowing the test system continued access to the
Internet.
We enabled Web filtering and attempted to hit some
dubious Web sites. We were denied access, and all the
attempts were logged. Gatekeeper Pro’s antivirus likewise caught the EICAR test suite—as expected—as well
as the sample viruses we had in the lab. Gatekeeper Pro
worked as advertised.
R E V I E W E D B Y P E T E R G I A N N A C O P O U LO S
Yoggie
www.yoggie.com
Price: $220
Configuration/Management
Yoggie Gatekeeper Pro is an interesting new take on
the perennial problem of using personal firewalls to
secure individual PCs. It is a Linux-based USB device
that serves as a full-blown firewall, VPN gateway and
integrated antivirus/Web filter gateway. All of this capability is packed into a form factor that is approximately
the size of a pager.
We tested the standalone version to look at the core
capabilities of this unique product, but an enterprise version is available that provides centralized management,
policy control and reporting.
Policy Control
B
The policy control is rather basic, but effective. The
administrator can open any necessary firewall ports,
configure IP addresses and perform other tasks. Some
rather pleasant surprises included a fairly effective Web
filter control that can block access to malicious Web
sites or those that violate corporate policies, such as
adult entertainment and gambling sites. It also offers
Testing methodology: We installed Yoggie on a Windows XP SP2
test machine. We used the EICAR test suite, along with samples of
known viruses, to validate the antivirus capabilities, and used Nmap
to verify open and closed ports on the host.
B+
Installation is gloriously simple: Plug in the device
to the USB port of your PC or laptop, connect your
Ethernet port to the clearly labeled PC port on the
device using the bundled cable, connect your broadband connection to the Net port, and you’re ready
to surf the Web. If you plan to use the Gatekeeper Pro
with a WLAN connection, you’ll need to install a
driver.
Configuring the device is accomplished through a
built-in Web GUI that will immediately look familiar to
anyone who has set up any type of home networking
equipment from D-Link or Linksys.
Though effective, the Web GUI is somewhat Spartan.
For example, the user is allowed to choose from low,
medium or high security levels, but there is no clear
explanation of the specific differences between the
levels, either in the online help or in the product documentation. We set the security level to high and were
able to perform common tasks such as instant messaging, Web browsing and email without any difficulties.
Verdict
We see a definite sweet spot in the market for
Yoggie’s offerings, especially the enterprise version
in large environments that want to protect disparate
endpoints without getting bogged down with
configuring individual hosts.w
I N F O R M AT I O N S E C U R I T Y
18
COLOR ME COMPLEX
REVIEWED BY ED SKOUDIS & MATT CARPENTER
Image by JOHN KUCZALA
P R ODUCT
R EVIE W
TODAY’S DESKTOP AND LAPTOP COMPUTERS ARE
a complex melange of software: office applications, specialized business programs, homegrown client apps, and
ever more complex browsers and operating systems. As
an enterprise security pro, you’re lucky if you can get at
least two agent-based security tools included in this zoo
of your standard builds to protect all of that software.
Some of us get only one.
Yet desktop security technology is rapidly advancing, with host-based intrusion
prevention systems (HIPS), personal firewalls and other
defenses augmenting traditional antivirus and antispyware tools. Because of the severe constraints on the number of host security products our enterprises can deploy
and manage, major security vendors have responded with
integrated endpoint security suites, rolling a bunch of
desktop defenses into a single package (see “End-to-End
These endpoint security
Endpoint Security,” p. 24).
products have introduced a new dynamism into our
industry, as antivirus vendors augment their wares with
fresh features to compete against each other and hungry challengers. To help sort out all of this, Information
Security evaluated seven enterprise endpoint security
solutions. We graded each on its management capabilities,
reporting, ability to detect and block malware, detecting
and thwarting exploit attempts, and integration of the
various desktop security capabilities in one package.
Funneling integrated endpoint security
features into one product makes for a
murky mix of complexity and immaturity.
I N F O R M AT I O N S E C U R I T Y
20
TE ST B E D
Specifically, we tested CA Threat Manager 8.1 and Host-Based Intrusion Prevention System 8; eEye Digital Security
Blink Enterprise Edition; IBM ISS Proventia Desktop Endpoint Security 9.0;
McAfee Total Protection for Enterprise;
Sophos Endpoint Security and Control
7.0; Symantec Endpoint Protection 11.0;
and Trend Micro OfficeScan 8.0.
Bearing witness to the rapidly evolving
nature of the endpoint security space, the
three giants of the information security
industry—McAfee, Symantec and Trend
Micro—responded with beta versions of
their suites that were nearly finalized for
shipping. (We requested every product we
analyzed be available for general purchase
by our publication date.)
Many of the problems we encountered
with testing and, in some cases, retesting
updated versions of these products reflected the difficulties in dealing with beta
builds of highly complex packages. But,
further, our testing suggests this class of
integrated endpoint security products is,
for the most part, far from mature.
ABOUT THIS
CA Threat Manager 8.1,
Host-Based Intrusion
Prevention System 8
eEye Digital Security Blink
Enterprise Edition
IBM ISS Proventia Desktop
Endpoint Security 9.0
McAfee Total Protection
for Enterprise
Sophos Endpoint Security
and Control 7.0
Symantec Endpoint
Protection 11.0
MANAGEMENT
REVIEW
We tested solutions that provided a minimum of
signature-based antivirus and antispyware, personal firewall, host-based IPS, and central management and reporting capabilities. We selected
a mix of leading traditional antivirus vendors and
relative newcomers in the market, testing comprehensive endpoint security products from CA,
eEye Digital Security, IBM ISS, McAfee, Sophos,
Symantec and Trend Micro. (Because of space
and resource limits, we opted not to include products from four other companies that responded
positively to our invitation to apply for consideration for testing: Fortinet, F-Secure, Grisoft and
Kaspersky Lab. In addition, ESET, Norman Data
Defense Systems and Third Brigade declined our
invitation to be considered.)
Our analysis test bed consisted of a Windows 2000 Server with Service Pack 4 running
the enterprise management software, managing
and protecting four Windows XP client systems.
Each XP client had no service pack or patches.
As a control, for each test, we utilized one target
Windows XP machine with exactly the same configuration but lacking the endpoint security protection.w
he immense complexity of these
tools can be overwhelming, with
—E D S KO U D I S & MAT T CAR P E NTE R
Trend Micro OfficeScan 8.0
more features than almost any distributed system in today’s enterprise.
If a given product provides really good
could be used to configure systems, quickly determine their
security, but cannot be managed across an enterprise in a
security state, and update settings based on attack activity.
coherent fashion, it just isn’t useful.
We were particularly impressed with Symantec’s manThe typical endpoint security suite must parse hundreds
agement capabilities, but McAfee’s completely new ePolicy
of types of files for antivirus and antispyware functionality,
Orchestrator (ePO) is a major disappointment.
analyze numerous packet types for firewall and some IPS
Symantec is top-notch from a large enterprise perspecactivities, police hundreds
tive, with intuitive GUIs for policy configuration and status
of operating system settings
checking. Its overall dashboard clearly identifies potential
and running processes to
THE IMMENSE
problems associated with infection, out-of-date signatures
detect malicious action, and
COMPLEXITY
or disabled functionality on managed hosts, offering advice
much more. All of these feaOF THESE TOOLS
to an administrator on how to fix each issue. The managetures are controlled with
CAN BE OVERment GUI comes in two flavors: a full-blown Java-based
thousands of settings applied
interface for all aspects of the administration console, and
via default policy templates WHELMING, WITH
a scaled-down Web-based interface that can be used for
or custom policies for specifMORE FEATURES
status checks and reporting, but not policy management.
ic enterprise needs. Admins
THAN ALMOST
Sophos also provided very solid management capabilroll out these policies to a
ities, organized, the vendor told us, around the KISS
multitude of managed work- ANY DISTRIBUTED
principle, which we assume stands for “Keep It Simple,
stations and servers across
SYSTEM IN
Sysadmin.” Sophos’ GUI is designed to reduce the time and
the network, all to be manTODAY’S
effort needed to configure and deploy the product. Sure,
aged in real time.
ENTERPRISE.
you don’t have access to a lot of the fine-grained policy
We looked at the ease
settings, but the overall options available for configuration
with which each product
[
]
T
I N F O R M AT I O N S E C U R I T Y
21
are excellent. Checking the status of managed workstations
is snappy, and alerts about systems that deviate from policy
are easy to understand.
The Trend Micro management interface worked quite
well in configuring and analyzing managed systems, especially for antivirus and antispyware. The new product architecture enables Trend Micro to package new endpoint
product building blocks into plug-ins for rapid deployment,
a design decision that will benefit Trend and its customers.
However, we were concerned that we couldn’t discern
client signature updates for its Intrusion Defense Firewall
(the component that implements the firewall and HIPS
functionality). Such information is vital in signature-based
IPS products such as this one, which applies network-based
IPS signatures to traffic going into the protected host. Trend
Micro licensed this functionality from Third Brigade to
create the first plug-in for its new architecture.
eEye Digital Security features a well-organized, intuitive
management interface. However, the client GUI is clearly
more mature than the enterprise management console
itself, offering finer-grained insight into the configuration
and alerts generated by the tool.
CA’s management console has improved significantly
since we last looked at it in our antispyware analysis in May
2006. Its latest version is much faster and more interactive
than previous versions. Still, checking the status of different
workstations required moving between different screens,
and policy configuration of this purely Web-based GUI was
more difficult than with other products.
We found the IBM ISS product quite difficult to manage.
Determining the current status of clients from the management console was cumbersome, and managing all of the
separate features was complicated and confusing. Also, at
several points we encountered cryptic error messages that
didn’t explain the problems we encountered in installing
and configuring the product. Finally, the IBM ISS endpoint
product is exclusively for Windows clients; it cannot be used
to manage servers, Windows or otherwise. Server security is
available only as a completely separate product.
McAfee’s new enterprise management server, ePolicy
Orchestrator (ePO) 4.0, was a great disappointment. The
more you loved the previous versions of McAfee’s ePO, the
more frustrated you will likely be with the new version.
McAfee has completely rewritten its flagship management product with a Web-based GUI, letting admins manage it from any browser
in their enterprise.
ENDPOINTS | Management
The well-laid-out
THE GOOD NEWS Symantec is top-notch from
and quick GUI of eara large enterprise perspective, offering easylier ePO versions has
to-understand GUIs for policy configuration
been replaced with a
and status checking.
complex and bewilderTHE BAD NEWS The more you loved the
ing Web-based interprevious versions of McAfee’s ePO, the
face. The easy dragmore frustrated you will likely become
and-drop features of the
with the new version.
thick-client ePO have
been replaced with countless Web-based drop-down menus
in screens that make it difficult to find what you need.
The difficult-to-use management GUI and default policy are of significant concern. While testing, we accidentally
applied a baseline medium security policy to the McAfee
management server itself. In the complex ePO 4.0 GUI,
such mistakes are frustratingly easy to make—we did it
while being guided step-by-step on the phone by McAfee
D ETE CTI O N
Chart shows the
percentage (lower
percentages are better)
of 8,114 recent malware specimens that
survived on protected
systems after (1) realtime scans of malware
as we attempted to
copy them to the target
system; (2) on-demand
scans of the malware
surviving the real-time
scan; and (3) ondemand scan of all
specimens with realtime scanning disabled.
ANTIMALWARE SCANNING
RESULTS
1
On-demand scan
2real-time scan
following
On-demand scan of
3specimens
all malware
Real-time scan
CA
eEye Digital
Security
IBM ISS
McAfee
Sophos
Symantec
Trend Micro
8.7%
9.3%
100%***
22.3%
100%**
17.6%
8.0%
8.7%
9.3%
58.6%
22.3%*
36.7%
8.3%
7.8%
8.7%
9.3%
34.7%****
100%*
36.7%
8.3%
7.9%
*McAfee’s default action is to alert only during an on-demand scan. Changing this default, 11.6% of malware remained after both real-time/on-demand and pure on-demand scans.
**Sophos real-time scans block only read-actions by default, not write-actions.
***IBM ISS does not block network-writes across Windows shares, even when configured to scan for all write-actions.
****IBM ISS only scored this number after repeated runs from scans that ended prematurely with a messages saying “Successfully Completed”
I N F O R M AT I O N S E C U R I T Y
22
Notably, McAfee blocks all .exe files from a network copy,
even benign test files, due to another default setting. Such
a feature is likely to cause problems in environments
attempting to distribute programs via network file shares,
and is certain to be disabled in some enterprises.
When we tested Sophos, every one of our specimens
survived the initial copy because, by default, Sophos’ realANTIMALWARE SCANNING
time defenses only look at “read” actions, not “write”
o gauge each vendor’s ability to detect and block malactions. Such an approach, possibly done to improve file
ware found in the wild, we ran three tests using 8,114
system performance, prevents the malware from executrecent malware specimens from a private collection
ing, but does not stop infiltration of malware into a file
graciously provided by antispam researcher Bill
system. Sophos does offer an option for changing this
Stearns. Our zoo included a large variety of worms, bots,
default behavior.
backdoors and viruses. For each test, we recorded the perIn the end, both the real-time/on-demand combo test
centage of specimens not eradicated in each round of testing
and the pure on-demand test left 36.7 percent of the
(See “Antimalware Scanning Results,” p. 22).
specimens on the target machine.
Our first test was designed to evaluate each product’s realSophos’ default behavior is to perform “in-place” quartime signature-based defenses by copying the malware from
antine, preventing future access of the file but leaving it in
a hardened machine to a shared directory on the protected
its current location. All the other products move malware
target system. We then recorded the percentage of malware
to a separate quarantine directory or delete it. Sophos says
specimens that made it into the target’s file system, escaping
its approach makes restoration of files misidentified as
detection by the product’s real-time scanning capabilities.
malware easier. If your antivirus tool makes false-positive
We then performed an on-demand scan of all malware
matches on legitimate files, restoring access in their normal
that survived our first test, to assess the combined real-time
locations is a lot easier than scraping them out of a quaranand on-demand scan capabilities for identifying and eraditine directory and finding their homes again. The Sophos
cating malware.
tool can be configured to perform traditional quarantine
Finally, we conducted on-demand scanning indepenor deletion.
dently by disabling real-time scanning, copying all malware
IBM ISS was rated lowest in this series of tests, crashing
to the target file system, and then executing a scan of the
several times and scoring so poorly as to cause us to doubleentire zoo.
check that protection was enabled. IBM ISS leaves signatureTrend Micro, CA and eEye all did very well, generally
based antivirus turned off by default, another indication
detecting and blocking or removing all but about 8 to 9
that this product is typically used to augment another venpercent of the malware we threw at them in all tests.
dor’s antivirus solution. IBM ISS has licensed BitDefender’s
Symantec was close behind, missing 17.6 percent of
antivirus and antispyware functionality in its endpoint suite,
specimens on the real-time scan, but performing on a par
which we activated before starting our test regimen. The
with Trend, CA and eEye in the on-demand scans.
initial real-time test completed without the tool blocking
McAfee was next, with 22.3 percent of our specimens
a single file. According to IBM ISS support personnel, file
eluding the real-time scan. The follow-up on-demand
copies across Windows network shares are not scanned,
scan, however, produced some surprising results: Another
even with the on-write scanning option enabled. This stance
10.7 percent of specimens were detected, but none of
mystifies us, considering that users could copy infected files
those were deleted or quarantined. Likewise, in the pureon a file server back to their clients without any real-time
play on-demand scan, all of the 8,000-plus malware specprotection.
imens survived, despite an avalanche of alerts. That’s
The on-demand scanning was hardly better. The followbecause this new McAfee product has a default action
up on-demand scan started off as expected, but halfway
of alert-only for on-demand scans, in contrast to the
through the scan (according to the progress bar) scanning
competition and a departure from most earlier McAfee
stopped and we were greeted with the message “Successfully
products.
Completed.” However, the same GUI listed “Number of
With the help of McAfee support, we used the
Files Remaining: 4,430” and we
McAfee client to conduct an
still counted 58.6 percent of our
on-demand scan with a delete
ENDPOINTS | Antimalware Scanning
malware in the target machine’s
action, a process that requires
file
system. This stop and start
several rather nonintuitive steps.
THE GOOD NEWS Trend Micro, CA and eEye all did very well,
several times during the
repeated
After that scan, 11.6 percent of
generally detecting and blocking or removing all but about
8 to 9 percent of the malware thrown at them.
scan. We re-ran this test several
our initial specimens remained
times, but 34.7 percent was the
for both the on-demand scan
THE BAD NEWS IBM ISS crashed several times, scoring
best IBM ISS managed in repeatfollowing real-time scanning and
so poorly as to cause us to double-check that the protection
ed on-demand scans.
the pure-play on-demand scan.
was enabled.
support. By simply applying the default security policy to
the management server, ePO killed itself. We were unable to
get access to any of the management capabilities, and had to
reinstall everything from scratch to resume testing.
T
I N F O R M AT I O N S E C U R I T Y
23
EXPLOIT PROTECTION
very vendor in our analysis claims to protect systems
against exploitation using some form of HIPS technology. Different vendors use this term for a variety
of disparate technical defenses (see “HIPS Hydra,” p.
26). Regardless of approach, we wanted to see how each
vendor would fare against exploitation attempts in a series
of three tests. We disabled each product’s firewall component to focus the test exclusively on HIPS functionality.
First, we attempted to exploit client-side software running on the protected hosts, trying to attack Internet
Explorer via the IE CreateObject vulnerability (MS06-014)
and VML flaw (MS06-055). We also tried to exploit the Firefox browser using the Mozilla_CompareTo vulnerability.
Our second test measured how well each product
defended listening services on the protected system, particularly services associated with Windows networking.
We attempted to exploit both the MSRPC DCOM buffer
overflow flaw (MS03-026) and the LSASS buffer overflow
issue (MS04-011). To add some variety to this network service testing, we attempted each exploit with both a standard
command shell payload and the Metasploit Meterpreter
shell, a more sophisticated and often harder-to-detect
attack that provides specialized remote shell access running
from within an exploited process. Our client-side and server-side testing relied on Metasploit Framework version 3.0,
E
using all default settings except for the HTTP port for
browser exploits, which we changed from 80 and 8080 to
another number to simulate an attacker who tricks a client
into clicking on a link with a port number in it.
Our third test was designed to look at how each vendor
could defend against zero-day exploits of third-party applications. We created our own network-listening program
with a buffer overflow flaw, and wrote some code to
exploit it to give remote command-shell access on the target
machine.
Overall, eEye performed best in detecting exploits. It was
the clear leader in identifying client-side attacks, alerting on
all of our tests, but by default did not block; it simply displayed the alert “Application Protection: Suspicious System
Call.” This default behavior could be altered to block such
exploits, as a global switch applied to all such suspicious system-call detection. While blocking is the goal, concern over
false-positive blocks makes eEye’s default setting reasonable.
eEye successfully stopped all service exploits. However,
in the process of blocking the MSRPC DCOM exploit, it
killed the svchost.exe process, which made our Windows
machines reboot themselves within 60 seconds. It is generally considered better to kill an exploited process rather
than run the attacker’s code, but re-booting could result
in loss of valuable data.
eEye detected and alerted on our zero-day exploit; when
A D D I T I O N A L F E AT U R E S
Some vendors provide
additional capabilities
that could be very
helpful in an enterprise
environment. Some
organizations will
certainly desire these
features, while others
will not. For that
reason, we did not
include them in our
overall product ranking,
but still detail them
here for organizations
that will take them
into account in their
buying decisions.
(See “Building Blocks,”
opposite page.)
24
END-TO-END ENDPOINT
SECURITY
Web site reputation analysis helps combat phishing
attacks, spyware downloads and related nastiness. The technique assesses the URLs
accessed by a browser to determine whether
they are associated with a malicious or suspicious Web site. If the endpoint product
detects such a URL, it prevents the browser
from surfing there, either displaying an alert
message or redirecting it to a safe site.
McAfee and Trend Micro deploy an impressive
and vast infrastructure of monitoring software
to discover malicious Web sites and update
their extensive blacklists.
Application execution control enables enterprises
to prohibit the execution of unwanted game
and P2P applications by using a blacklist.
I N F O R M AT I O N S E C U R I T Y
Similarly, application control blacklists can help
limit the outbreak of a fast-spreading worm or
browser exploit by blocking execution of the
malware on the end system. Using whitelist
functionality, an enterprise can severely lock
down a machine, allowing it to run only
required applications. Most endpoint security
solutions can look at the name, file system
location, and/or MD5 hash of a given executable to determine whether the application
should be allowed to run on the protected
machine.
All of the vendor products we tested, except
for Trend Micro, supported some form of application execution control. eEye and Symantec
offer the most flexibility, with custom-designed
whitelists and blacklists based on executable
C O M PA R I S O N
BUILDING BLOCKS
CA
eEye Digital
Security
IBM ISS
McAfee
Sophos
Symantec
Trend Micro
Antivirus
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Antispyware
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Personal firewall
Yes*
Yes
Yes
Yes
Yes
Yes
Yes
HIPS/buffer overflow protection
Yes*
Yes
Yes
Yes
Yes
Yes
Yes
Web site reputation analysis
No
No
No
Yes
No
No
Yes
Application execution control
Yes*
Yes
Yes
Yes
Yes
Yes
No
USB control
Yes*
Yes
No
No
No
Yes
No
Vulnerability scanner
No
Yes
No
No
No
No
Yes
Windows NT,
2000, XP, 2003.
Windows 2000
Pro, XP. No
server support in
endpoint product,
so no Windows
2003**
Windows NT,
2000, XP, 2003,
Linux
Windows
95/98/Me, NT,
2000, XP, 2003,
Linux, Mac OS X
(PowerPC and
Intel)
Windows 2000,
XP, 2003, Vista.
Separate,
unmanaged
clients for Linux
and Mac OS X.
Windows 2000,
XP, 2003, Vista.
Mac OS X
support by
end of year.
Supported operating systems AV/AS: Windows
NT, 2000, XP,
2003, Vista.
HIPS: XP only.
* CA’s endpoint solution requires the purchase of two separate products: its anitvirus/antispyware solution and its HIPS solution. All items marked with a * are available only with the purchase of the HIPS solution.
**Server protection is a separate product.
path, hash, or regular expression matching.
CA’s blacklist and whitelist capabilities were
also impressive, augmented by a graylist function that allows admins to define rules for execution of specific programs only by certain
users and at certain times of day. Sophos has
the most limited application control, supporting
only blacklists for those applications for which
it generates a specific signature.
USB control addresses the concern of the many
enterprises that have been stung by employees
who bring malware into the organization on
USB thumb drives or spirit gigabytes of sensitive data out on them. CA, eEye and Symantec
feature device control capabilities.
eEye’s USB control is an all-or-nothing
feature—all USB tokens are allowed for all
users of a given policy group, or all users
are forbidden from using the devices.
CA and Symantec’s solutions are much
more fine-grained, allowing policies to be
defined for various Windows devices such as
USB, infrared, FireWire, SCSI, and serial or
parallel ports. Their device control policies
can specify certain kinds of files (e.g., Word
documents, executables, etc.) that can be
accessed or denied on specific devices.
Vulnerability scanners assess the local host on
which they are installed to determine its security
stance, looking for unpatched operating system
components or weak security settings. In particular, eEye includes a copy of its comprehensive Retina security scanner on every host;
it’s a powerful suite that can discover several
hundred potential security flaws and propose
fixes. Trend Micro includes a far more limited
scanning tool that can search the system for
applications that the IPS functionality has
signatures to protect.w
—E D S KO U D I S & MAT T CAR P E NTE R
I N F O R M AT I O N S E C U R I T Y
25
ENDPOINTS |
Exploit Protection
THE GOOD NEWS Overall, eEye performed
best in detecting exploits.
we tweaked the configtweaked our test to verify this claim, and Trend Micro did
uration to an action of
detect our attacks on those ports. Administrators can add
“Terminate Process,” it
lists of additional ports for browser and other HTTPTHE BAD NEWS CA fared poorly in detecting
blocked as well.
related defenses. Ideally, an admin would configure the
and blocking client and services exploits.
IBM ISS came next.
endpoint security suite so it monitored for HTTP and
On the client side, it
HTTPS attacks on all ports allowed out through the enterdetected and blocked the VML exploit. However, the alert
prise’s network firewall. In many organizations, unfortumessages for the IE CreateObject and Firefox attacks didn’t
nately, the number of ports allowed outbound are rather
indicate that the product had detected the exploit action,
high and change on a regular basis, making this synchroonly that it identified a Microsoft Windows shell banner
nization of network firewall and endpoint security tool
passing across the network. An attacker could launch such
difficult.
an exploit without creating a banner,
thereby dodging this form of detection.
IBM ISS identified and blocked all
services-based attacks, with an alert that
S E V E N - H E A D E D P R OT E CT I O N
cited the specific exploit we used, the ideal
behavior for the product under these tests.
It allowed our zero-day attack, again
merely alerting to the presence of a Windows shell banner.
Sophos delivered reasonable performance in our client-side testing, alerting
Host-based intrusion prevention system (HIPS) functionality means many
on two exploits as “Buffer Overflow”
different things to the vendors that include such capabilities in their endpoint
behavior, but missing the CreateObject
security suites. The goal, of course, is to prevent the end system from being
exploit. The default action is to alert, but
compromised by an attacker, but the technological approach of the vendors
Sophos can be configured to block the
implementing HIPS varies widely. We interviewed each vendor, asking them to
attacks.
describe their technical approach to blocking exploitation attempts. We wanted
All of our services attacks were detected,
to focus specifically on defenses against buffer overflow and related code exebut by default they were allowed through,
cution exploits. Based on our interviews, we identified seven essential forms of
giving the attacker control of the system.
such exploit detection and prevention:
Sophos neither detected nor blocked our
• System call backtracing analyzes various system API calls to ensure the callzero-day exploit.
ing address exists in a known code segment.
McAfee detected and blocked our VML
and Firefox exploits, but failed to detect
• Spawn blocking limits which programs can run new programs (for example,
our CreateObject exploit. McAfee detectblocking a browser from running a new command shell process).
ed and blocked all of our service exploits.
• Behavior checking monitors system calls for combinations that historically
For zero-day defenses, McAfee requires
have indicated that an attack is under way.
administrators to configure specific appli• DLL loading checking looks for unusual or unexpected DLLs to be loaded
cations to be protected on a machine. By
into
running applications on the machine.
default, nothing other than specific Windows components is protected, so our
• Call verification ensures the return address for the current function is immezero-day attack went undetected. As an
diately preceded by a call instruction.
experiment, we configured McAfee to add
• SEH validation protects against exploits that overwrite exception handlers by
zero-day protection to our custom vulnervalidating the Structured Exception Handler chain.
able application. Unfortunately, our exploit
• Network-based IPS monitors network traffic for known vulnerabilities and
still went undetected.
exploits.
Trend Micro and Symantec came next
in our exploit testing. Neither identified
CA implements spawn blocking, DLL loading checks and network-based
nor blocked a single client exploit. Trend
IPS. eEye relies on system call backtracing, call verification and network-based
Micro support personnel indicated that
IPS. IBM ISS uses system call backtracing and network-based IPS. McAfee
the HIPS protection it licensed from
has created a patented “generic buffer overflow protection,” although it declined
Third Brigade (as well as the protections
to share details with us before press time, as well as network-based IPS.
offered by other vendors) is often conSophos uses system call backtracing. Symantec implements behavior checking
figured by default to look for browser
and network-based IPS. Trend Micro focuses exclusively on network-based IPS.w
exploits only on TCP ports 80 and 8080.
—E D S KO U D I S & MAT T CAR P E NTE R
Again, independent of our scoring, we
HIPS HYDRA
I N F O R M AT I O N S E C U R I T Y
26
Both Trend Micro and Symantec detected and blocked
all of our services exploits, but neither detected our zeroday attack.
CA fared worst of the seven products in this series of
tests, failing on most. It didn’t detect or block any of the
client exploits with its default security policy. Although not
part of the scoring, we experimented with its “Restrictive
Policy,” which did block all of the exploits, but also prevented Firefox from accessing the network.
The next set of results were, if anything, poorer, as it
did not alert or block our services exploits, even when we
applied Restrictive Policy.
The one success was that CA detected and blocked our
zero-day exploit under default policy.
REPORTING
e evaluated each product’s reporting functionality,
used to pull information such as long-term attack
and infection trends, policy compliance information, and lists of the most problematic groups of
machines. In particular, we looked at comprehensiveness,
flexibility and ease of use
Though McAfee’s management GUI was disappointing,
ePO’s reporting features are excellent, including more than
70 different reports that break down all aspects of the enterprise. The point-and-click custom report creation tool is
stellar, making it easy for people who are not database experts
to massage the information into highly useful reports.
Symantec is also solid, offering more than 70 reports,
with impressive performance. Symantec’s custom reporting
capabilities are focused on defining filters for its existing
reports to create useful subsets, a valuable capability but
somewhat less flexible than McAfee.
The IBM ISS reporting tool provided good coverage,
addressing long-term trends and top attacked and infected
machines. However, getting at the report files is a little
obscure. Admins have to remember where they were
generated in the file system to open the report from
within the management GUI. Further, to open a report,
you have to right-click on it and go to “Properties,” a
bizarre GUI twist that takes some getting used to.
Trend Micro’s reporting is handled by a separate product,
Trend Micro Control Manager, which is not tightly
bundled into the existing management GUI, making a
little more work for installation and use. On the positive
side, this separate reporting tool applies to all
ENDPOINTS | Reporting
Trend Micro enterprise
products, including gateTHE GOOD NEWS McAfee ePO’s reporting
way security appliances,
features are excellent, including more
than 70 different reports that break
antispam products, etc. It’s
down all aspects of the enterprise.
included in the purchase
of the endpoint suite, and
THE BAD NEWS Sophos’ reporting
provides a full complement
capabilities are quite skimpy. Only about
a dozen reports are available.
of well-laid-out reports.
W
27
I N F O R M AT I O N S E C U R I T Y
eEye’s built-in reporting features are decent and offer
some features for creating custom queries in its published
database schema. However, building custom or tweaked
report queries is a complicated process, even using the builtin templates.
CA’s reporting for antivirus and antispyware is stellar,
with more than 70 reports available. Unfortunately, CA’s
HIPS and firewall features offer very little reporting, with
only about a dozen high-level reports providing much less
visibility into these important aspects.
Sophos’ reporting capabilities are quite skimpy. Only
about a dozen reports are available. They don’t include Top
10 style reports of most infected systems, users or groups.
The look and feel of the reporting engine makes the product appear better suited for small and medium businesses,
rather than large enterprises. However, Sophos publishes
its database schema for customers to use with third-party
reporting tools, such as Crystal Reports.
INTEGRATION OF COMPONENTS
ndpoint security suites should integrate disparate
components into a coherent, manageable whole. Most
of the vendors have worked hard to integrate various
aspects of their solution, with high marks going to
eEye, McAfee, Sophos and Symantec.
The IBM ISS product
has good integration, but
ENDPOINTS | Integration
often looked like it was
THE GOOD NEWS Most of the vendors have
packaged around compleworked hard to integrate various aspects of
menting another vendor’s
their solution, with high marks going to
antivirus and antispyeEye, McAfee, Sophos and Symantec.
ware, rather than providTHE BAD NEWS CA was never one for
ing a whole solution. This
deep integration of components in its
approach makes sense
antimalware solutions.
given that IBM ISS HIPS
capability has been an
added defensive layer to complement traditional
antivirus/antispyware products. IBM ISS has licensed
BitDefender’s antivirus/antispyware technology, but the
management GUI still appears as though it is merely
grafted in.
Trend Micro’s antivirus/antispyware integration is decent,
but integrating the personal firewall and HIPS, licensed
from Third Brigade, needs some work. These components
are a plug-in inside the management GUI, with a separate
set of configuration screens that don’t have the same look
and feel of the configuration of antivirus and antispyware.
Further, on the client side, antivirus and antispyware is a
completely separate program from the HIPS software.
CA was never one for deep integration of components
in its antimalware solutions. Its endpoint security product
continues to separate management of antivirus and antispyware on different screens, but at least they both are available in one GUI application. CA’s HIPS, on the other hand,
is a separately purchased product. It is installed and man-
E
R E P O RT CAR D
MAKING THE GRADE
CA
Threat Manager
8.1, Host-Based
Intrusion Prevention System 8
www.ca.com
eEye Digital
Security
Blink Enterprise
Edition (Blink
Professional 3.1,
REM Security
Management
Console 3.5)
www.eeye.com
IBM Internet
Security Systems
Proventia Desktop
Endpoint Security
9.0
www.iss.net
McAfee*
Total Protection
for Enterprise
www.mcafee.com
Sophos
Endpoint Security
and Control 7.0
www.sophos.com
Symantec*
Endpoint
Protection 11.0
www.symantec.
com
Trend Micro*
OfficeScan
version 8.0
www.trendmicro.
com
Enterprise
management 30%
B
B
C
D
A-
A
B+
Antimalware scanning 20%
A-
A-
D
C
C-
B+
A-
Exploit protection 20%
D
A-
B
B-
B-
C
C
Reporting 20%
B-
B
B+
A
C
A-
B+
Integration 10%
C
A
A-
A
A
A
B
VERDICT
B-
B+
C+
C+
B-
B+
B
Lacks integration
of separate
products.
Pros: Good antimalware scanning
Cons: Weak
exploit protection
and integration
A worthy new
entry with a
comprehensive
solution.
Pros: Excellent
protection
with decent
management
Cons: Client GUI
better than
Enterprise
Management GUI
Product much
more geared
around augmenting an antivirus/
antispyware
solution, not
replacing it.
Pros: Decent
exploit protection
Cons: Malware
scanning default
settings weak and
prone to failures
A disappointing
new management
GUI could make
transition difficult.
Pros: Excellent
reporting
Cons: New
management
interface complex
and difficult
to use
A moderate
performer in all
categories.
Pros: Very
straightforward
management
capabilities and
integration
Cons: Skimpy
reports that can
be augmented
with third-party
tool
Solid product for
the enterprise.
Pros: Excellent
enterprise
management
Cons: Exploit
protection
relatively weak
Decent product,
but HIPS integration needs work.
Pros: Good
antimalware
scanning and new
plug-in architecture for expanded
functionality
Cons: Client-side
exploit protection
tied to port
number
*Beta. Symantec Endpoint Protection 11.0 was made available Sept. 27; all tested components of McAfee Total Protection for Enterprise have been available since Oct. 11; Trend Micro OfficeScan 8.0 will be available Nov. 15.
aged using its own GUI and a separate client package is
installed on protected workstations.
In the End, Take It Slow
ymantec’s new offering looks very solid, and eEye
is a worthy new competitor in the endpoint security
space. Trend Micro has a decent solution and a promising plug-in architecture for future expansion. CA and
Sophos did reasonably well, but neither shined consistently.
Finally, we were very disappointed with the numerous
glitches, unfortunate design decisions and poor performance of McAfee and IBM ISS.
Regardless of which vendor you choose, keep in mind
that the endpoint security market is relatively immature—
witness our beta testing of three major vendors—and the
complexity of any of these products warrants a carefully
planned deployment strategy. We urge you to experiment
with the products on your own laboratory test systems with
S
images from your production environment to make sure
they don’t have any adverse consequences on your particular
application mix.
Double check default policy settings to make sure they
offer reasonable protection, and if not, adjust them for your
environment and risk profile. And, finally, have your support staff become familiar with the various quirks of these
management GUIs before production roll-out.w
Ed Skoudis is a co-founder of Intelguardians, an information
security research and consulting company. He is also a fellow
with the SANS Institute, where he teaches the Hacker
Techniques, Exploits and Incident Handling course.
Matt Carpenter is a senior security analyst for Intelguardians
with expertise in hacker attacks, defenses and security vulnerability
research. He has released several open source security research
tools. Send comments on this article to [email protected]
mag.com.
I N F O R M AT I O N S E C U R I T Y
28
P RODUC T
Reviews
example, the provisioning process requires an “authoritative list,” in a specifically formatted comma delimited
text file.
Further, IMAG wouldn’t recognize user formats used
by key applications such as SQL Server and Project
Server for AD accounts. As a result, these IDs must be
manually matched to their account on the user store.
IDENTITY MANAG EM ENT
IMAG 500
R R EVI EWE D BY B RAD CAU S EY
Apere
www.apere.com
Price: $15,000 for a single device
Policy Control
Apere’s IMAG 500 appliance aims
to simplify the complex maze of
identity management through the
automatic discovery of distributed
identity stores, consolidation, reconciliation and provisioning of all
user accounts, and access to practically all network resources through
a single control point.
It’s an attractive proposition, but
we found the implementation rough going.
Configuration/Management
D
Setup was difficult, ultimately requiring the company’s
technical support to remotely access the device to assist
with the configuration process. Although Apere claims
that the device can learn the location and type of applications, each application required manual configuration.
Basic device configuration is like any other appliance:
You set up DNS servers, email servers, log servers and
VLANs.
Things got sticky when we tried to add authentication
resources and user stores, particularly for Web applications. Because user accounts don’t generally live on the
Web server itself, IMAG doesn’t have a way to tie users
to the resource they need to access. In the absence of an
API, you have to enumerate users from each identity
store and reassign the resource that IMAG associated
with the users. You then have to get the user information,
reformat it and redirect the allowed resources for those
users to point to the Web server.
Rather than use an Active Directory domain, for
D
Access control entries are based on user ID, VLAN and IP
address, but IMAG lacks a proper grouping mechanism,
so it can assign only one or all users, for example, to a
resource. Anything in between will require extensive
work to set up.
Policies provide the ability to control which users or
groups of users can access particular applications, but
are only usable when the device has been deployed in
an in-band configuration. When deployed out of band,
IMAG can still serve as a central ID consolidation and
reconciliation point, but any other benefits are lost.
Effectiveness
D
IMAG has some major challenges to overcome. An ID
management solution needs to be able to effectively link
and manage identities from stores of user accounts.
Tasks such as importing users from identity stores and
consolidating them proved extremely difficult. Some of
the most basic applications, such as SQL Server and
LDAP, were a major challenge.
Proper user provisioning requires that you manually
create an account in each application and assign that
user to each resource via the user interface.
Reporting
B
Reporting is handled through an easy-to-use Web interface. Each report is customizable, and canned reports are
available for applications, authentication stores and user
provisioning.
You can research orphaned users, unmanaged
resources and user stores that have not been reconciled.
Each report can be filtered based on criteria such as
specific users or resources.
Verdict
Testing methodology: Our lab included a single Active Directory
domain and a single LDAP tree. User accounts were enumerated from
various sources such as MySQL, SQL Server, Web applications and
various client-server applications. User roles such as administrators,
power users and end users were set up to test access controls.
Apere says many of the issues we encountered
are addressed in its next release, but mid-enterprise
businesses may not have the tolerance for a product
with so many features missing or unfinished.w
I N F O R M AT I O N S E C U R I T Y
29
P RODUC T
Reviews
P R I V I LE G E MANAG E M E NT
HotPick
TION
R MAIT Y ®
O
F
IN
UR
SEC
BeyondTrust
Privilege Manager 3.0
R EVI EWE D BY B RAD CAU S EY
BeyondTrust
www.beyondtrust.com
Price: $30 per seat
downloaded free from Microsoft’s Web site.)
Once installed, the Privilege Manager settings are
available by simply opening the Group Policy Object Editor. It gives you a single interface to manage the custom
add-ons as well as the default GPO settings, simplifying
management and reducing administrative overhead.
Each managed computer requires client software to
capture and manage permissions for processes and programs; it can be installed through standard software
deployment or via Group Policy. The client reads the
custom GPO settings and modifies the security token on
programs as they launch, giving the user elevated privileges as defined by Privilege Manager rules.
Policy Control
The least privilege security model is
the de facto standard for reducing the
risks of elevated user privileges. This
can be a challenge in Windows environments. You don’t want your end
users to have general admin rights,
but they may need them to run the
applications required to do their jobs.
There’s no easy way to manage this,
so companies wind up letting users
have excessive privileges, leaving their
desktops, user accounts and software
vulnerable to attack.
BeyondTrust’s Privilege Manager
3.0 solves this dilemma through a Group Policy extension that allows organizations to control permissions for
selected processes and applications. BeyondTrust has also
introduced a new technology, called ShatterProof process
isolation, that prevents shatter attacks, a complex privilege escalation technique.
Configuration/Management
A
Privilege Manager adds GPO extensions that integrate
with Internet Explorer and Microsoft’s Group Policy
Management Console, so admins can work directly
through a customized Active Directory interface. The
installation was very easy and fast, consisting of an MSI
with few requirements, chiefly the .NET framework and
AD’s Group Policy Management Console. (These can be
A
Creating policies for application privileges is simple and
intuitive. Each rule allows you to identify a target process
or executable name. This can be done by a number of different methods, including MSI GUID, hash, path, folder
or ActiveX rules, giving you tremendous flexibility.
For each rule, you define what action will be taken,
including modifying privileges and permissions for target
applications. Defined privileges dictate what components
of the system will be accessible when the program or
process is initiated and for the duration of its run time.
These rules can be configured with filters that restrict
what settings apply to what group. For example, you can
disable the policy for a specified application based on a
wide range of criteria, such as subnet, computer name,
user, security group or organizational unit. In addition,
you can modify Internet Explorer behavior and ActiveX
security through a custom administrative template.
Effectiveness
A
Privilege Manager provides an extremely effective framework for implementing least privilege policies. The overall concept of least privilege in an enterprise environment
is plagued with difficulties. Often, developers have to get
involved, code has to be changed, and massive amounts
of time will be spent during implementation and dealing
with unknowns. Because Privilege Manager integrates
with Group Policy, it will significantly simplify the management of application privileges and permissions.
Verdict
Testing methodology: Clients in our AD domain consisted of several
Windows 2000 and Windows XP computers with various service
packs. A variety of applications were tested including Web sites with
ActiveX requirements, DOS-based applications, network-based
applications and locally installed programs.
Privilege Manager will prove invaluable for implementing and managing a least privileges program. Although
long-term management of each application will be
complex, it helps cut the job down to size.w
I N F O R M AT I O N S E C U R I T Y
30
P RODUC T
Reviews
PASSW ORD SECU RITY
Enterprise
Password Vault 4.0
R E V I E W E D BY TO M B OW E R S
Cyber-Ark Software
www.cyber-ark.com
Price: EPV server,
$25,000; user pricing
starting at $220
per user
Privileged users hold the
keys to your kingdom:
passwords that control
administrative access to
devices and applications
across your enterprise.
Cyber-Ark’s Enterprise Password Vault (EPV) is among
a handful of specialized products designed to securely
manage these sensitive passwords, controlling privileged
accounts across a wide range of client/server and mainframe OSes, switches, databases, etc.
It provides the privileged account controls mandated
by regulations, and its central repository makes it an
ideal addition to identity/access management projects.
Installation/Configuration
C+
Overall, this was a tedious installation/configuration
process. EPV is in serious need of an installation wizard
and graphics-filled documentation to help users understand the purpose of each of its components and where
it sits in the architecture. The documentation, while
voluminous, is disjointed and difficult to follow.
The expectation is that the four components be
distributed on at least two Windows 2003 servers, and
we sorely missed an overall diagram to reference the
separate installations.
We were somewhat vexed, for example, when we
Testing methodology: EPV was tested on multiple fully patched
and hardened Windows 2003 servers and Windows XP workstations.
We used a sample database of users and passwords, and scanned
the system for weaknesses using standard penetration testing tools
and forensic analysis software.
installed the last component, Password Vault Web Access.
We belatedly realized that you need IIS installed on the
second server—something the documentation didn’t
mention until then.
A
Effectiveness
The EPV experience is superb once the system is
installed. Operationally, the end user password management system is an intuitive, wizard-driven interface,
requiring little to no training.
The system is organized around the vault, which contains multiple safes. Each safe is independent and may be
connected to one user or group, or many of both. A person in one group or safe cannot see the existence of other
safes nor access them without explicit permission. Each
safe also has an owner or owners that control access. Via
the safe, passwords are synchronized with the end products, such as routers, switches and servers; changing the
password in the safe also changes it on them.
Essentially, the EPV takes control of the admin logon
function. For example, an admin logs on to the EPV Web
interface to access the password object associated with a
switch they wish to manage. This object gives them the
new password, they log on to the switch and conduct
their maintenance. Passwords can be generated based on
internal policies and/or regulations such as FFIEC or the
Family Educational Rights and Privacy Act.
The architecture is very secure. That’s obviously a
critical point, but we don’t see it often enough in enterprise security products. We encountered no way for password information to leak, either through the vault or the
browser-based interfaces. A firewall on the PrivateArk
server protects the host, opening a single port that allows
only Cyber-Ark’s proprietary protocol.
B
Reporting
Reporting is very well executed, but lacks a cohesive
export mechanism.
Reports are clear and concise. A nice dashboard presents reports and graphs that provide good auditing
capabilities to help meet regulatory requirements.
The exporting mechanism is smooth yet somewhat
disappointing. Reports can be exported only to Microsoft
Access and Excel, or via CSV format.
Verdict
EPV is a valuable tool and a maturing product
that performs its privileged password management
function very well.w
I N F O R M AT I O N S E C U R I T Y
31
P RODUC T
Reviews
able to register a fingerprint in less than 10 seconds. The
workstation software automatically detects any DigitalPersona Pro servers on the local network.
Both the server and workstation software can be
purchased with or without DigitalPersona’s fingerprint
reader. The latest version of DigitalPersona Pro offers wide
support for third-party readers, such as those becoming
popular in new business-class laptops. The DigitalPersona
optical reader is quite good; we found it to be accurate,
with few false negatives and no false positives.
B I O M E TR I C A UTH E NTI C ATI O N
DigitalPersona
Workstation Pro
and Server 4.0
Workstation (Single User)
R E V I E W E D BY B R E N T H U STO N
DigitalPersona
www.digitalpersona.com
Price: Server, $1,499, plus $50 authentication
license per user; Workstation without reader, $60,
with DigitalPersona U.are.U 4000B reader, $149
Biometric authentication has met
considerable market resistance, mostly because of integration issues, accuracy and cost. With improved technology and the introduction of laptops equipped with fingerprint
readers, biometrics may be starting to move into the mainstream.
DigitalPersona Pro is a robust single sign-on
(SSO) software suite that allows an enterprise to replace
passwords with biometric fingerprint readers or provide
dual-factor authentication.
Installation and Setup
B+
There are two pieces to the suite: DigitalPersona Pro
Workstation software for individual systems and the
server component, which integrates with Active
Directory on your domain controller. While the workstation software can function by itself, the server provides domain-wide SSO.
Installation is straightforward. The server installation
requires a few more steps to integrate with Active
Directory, but it’s all detailed in the manual. After installation, the workstation software starts a wizard, which
records your fingerprint. After a few repetitions, we were
Testing methodology: DigitalPersona Pro Workstation was tested as
a standalone product on Windows XP desktops, and in an AD environment with the server component on Windows Server 2003.
B-
The workstation software, in standalone mode, is rather
simple. It integrates with Windows logon and also provides an SSO function that seems to be geared toward
home users. The SSO feature provides an automatic wizard that will detect the login fields in many applications.
Unfortunately, we found there are some apps it does not
support (such as terminals, like Putty). It also supports
only Internet Explorer, a problem considering the growing popularity of Firefox. However, it is very easy to use,
fast and accurate with applications it supports.
Server (Centralized Environment) B+
The server software is much more robust. The SSO wizard allows manual creation of login templates to support
applications that the automatic wizard can’t detect; these
templates are pushed out to desktops via GPOs.
Creating a template is fairly easy. You need to make
sure the window title is accurately reflected within the
SSO administration tool. You then enter the actions
required for login—e.g., entering keystrokes into a field,
time delays, or x-y coordinates of a window. Templates
can also be created for password-change forms, which
can be used to automatically generate passwords. The
created templates can be either pushed out to workstations via GPO or copied manually.
The server centrally manages fingerprint data for all
users, with tight Active Directory integration. It also provides event logs for fingerprint logins to help with regulatory compliance, but lacks strong reporting capabilities.
It also provides a handy query tool to easily discover who
has registered fingerprints.
Verdict
Enterprises looking for a biometric single sign-on solution won’t be disappointed with what DigitalPersona
Pro offers. The software is easy to use, and can function with single- and two-factor authentication.w
I N F O R M AT I O N S E C U R I T Y
32
P RODUC T
Reviews
Security managers can also require that certain
remote connection requests be approved by one or more
designated persons. Connection requests and approvals
can be sent to a ticketing system.
ACCESS CONTROL
eGuardPost
Configuration/Management
R EVI EWE D BY STEVE N WE I L
e-DMZ Security
www.e-dmzsecurity.com
Price: Starts at $12,500 for five concurrent
sessions
Secure remote vendor
and system administrator access to information systems is a critical business
requirement for many organizations, but it
can be a challenge to manage and audit. While VPNs
are fine for most users, they can require client software
and don’t offer the level of audit and forensic capabilities
demanded by regulatory requirements and high-security
environments. eGuardPost is a hardened appliance that
can be used to secure, manage and audit these sensitive
connections.
Policy Control
B+
eGuardPost allows security managers to apply granular
access controls to remote connections. The appliance
comes bundled with Security’s Password Auto Repository
(PAR), e-DMZ’s flagship product, which securely stores
and manages administrative passwords.
We were able to successfully create multiple users and
enforce a variety of access controls on them.
Once users log in via HTTPS and are authenticated
via RSA Security’s SecurID, Secure Computing’s SafeWord or LDAP (or against user accounts created and
stored on eGuardPost), eGuardPost determines what
type of remote access they are allowed and which systems
they can connect to. Security managers can assign specific roles (e.g., requester, approver, auditor and administrator) to remote users.
eGuardPost can be configured to automatically log in
specific users; it retrieves the necessary password from
the local or a remote PAR. The password is never shown
to, or known by, the remote user.
Testing methodology: Our test network included a Windows XP
laptop, an unmanaged switch and three Windows 2003 Web, FTP
and domain controller servers.
B
Configuration is straightforward and easy thanks to
excellent documentation. The appliance is managed via
HTTPS. The management interface is well designed and
mostly easy to navigate.
Systems to be managed are defined, users are created,
and the security manager determines which users have
what type of remote access to which systems. You can
even limit access to specified time periods, which will be
very useful for vendors and contractors, as well as admins
assigned to particular tasks. Systems and users can be
placed into and managed as groups.
Users do not need to install any software; eGuardPost
proxies all remote connections. It can establish connections to systems via Telnet, Windows Terminal Server,
SSH, VNC and X5250.
B
Reporting
eGuardPost’s forensics capabilities are unique, offering
VCR-like recording and playback of every mouse, keyboard and screen action during a remote session. We
conducted multiple remote sessions via eGuardPost then
watched their recordings; each was flawlessly presented.
eGuardPost can automatically move recorded sessions to
designated archives.
eGuardPost can produce detailed reports of user
rights and activities, security alerts, firewall events, database events and Web server events. Reports can only be
exported to Excel and some of them are a bit cryptic. The
appliance supports SNMP and syslog.
B+
Effectiveness
We found eGuardPost to be a very effective product, correctly and efficiently managing and auditing all of the
many remote connections we sent through it.
eGuardPost is carefully hardened, with an embedded
firewall and hard drive encrypted with 256-bit AES. Our
security scans of the appliance found no vulnerabilities.
Verdict
eGuardPost is a well-designed and highly capable
product that meets an important need. It has strong
security and great forensics capabilities.w
I N F O R M AT I O N S E C U R I T Y
33
P RODUC T
Reviews
we entered the domain name, service account name and
password, and to LDAP using the service account
domain name, password, IP address and port number.
We could create fall-through rules across multiple directory services for a variety of situations (for example,
check AD first to authenticate a VPN user, then LDAP).
IDE NTITY MANAG EM ENT
Identity Engines
Ignition Server
Policy Control
R E V I W E D B Y S A N D R A K AY M I L L E R
Identity Engines
www.idengines.com
Price: Starts at $33,500
Identity Engines’ Ignition
Server manages access
controls across disparate
directory services platforms (Active
Directory, LDAP, eDirectory) by consolidating them into a single user store. Deployed as
an alternative to RADIUS, the appliance includes a
comprehensive policy engine to use with multiple
access control devices (wireless access points, switches,
firewalls, VPNs) throughout a heterogeneous enterprise.
Configuration/Management
B+
Because of well-written documentation, we completed
basic network installation in minutes. But that’s where
simplicity ends. Users must have an extensive knowledge
of authentication protocols, directory structures, virtual
provisioning and certificate management to take full
advantage of the Ignition Server’s features.
There are three major aspects of the Ignition Server:
networked devices (authenticators), user stores (directory
services) and policies.
Authenticators—devices attached to the network—
can be bundled by subnet to facilitate large installations.
They can be managed according to several attributes,
including service categories—groups of authenticators
to which policies are applied. Adding authenticators
was the same as with RADIUS: Provide a name, IP and
shared secret. Service category, device type (wired,
wireless, VPN) and vendor are added the same way.
Ignition Server automatically connected to AD once
A
The Ignition Server is really a policy engine that speaks
RADIUS. It does everything a RADIUS server would do,
but it’s the policy engine that sets it apart. We liked how
multiple authenticators are tied together into a single service category to which three different policies—authentication, identity routing and authorization—can be easily
configured and applied.
Authentication policy determines the tunnel protocols,
credentials and ciphers for communication between the
supplicant, Ignition Server and directory services.
An identity routing policy traverses directory services
during authentication, determining which user store to
apply based on the user’s network domain or what device
is making the authentication request.
The authorization policy controls access according to
the user account.
Effectiveness
A
We authenticated users to specific devices, such as wireless access points, and assigned a common policy using
credentials from two directory services (AD, LDAP).
Ignition Server supports strong authentication, such
as RSA SecurID and Secure Computing’s SafeWord.
Security is solid. Built on a 64-bit hardened appliance running a stripped-down version of BSD, security
features include onboard IDS, 256-bit AES encrypted
file system, and protection against physical tampering.
Reporting
C
This is Ignition Server’s biggest shortcoming. While
real-time statistics and logging are available, the logs
could only be exported hourly, daily or weekly—nothing customized or on-demand. We’d welcome the ability to export the statistics displayed in the individual
tabs.
Verdict
Testing methodology: Ignition Server was deployed in place of the
RADIUS server in our simulated enterprise network. It provided AAA
services for our wired and wireless network access, as well as for a VPN.
Organizations that need a unified policy engine to control network access using multiple authentication systems will be able to justify Ignition Server’s price tag.w
I N F O R M AT I O N S E C U R I T Y
34
P RODUC T
Reviews
HotPick
Sun Java System
Identity Manager 7.0
IDE NTITY MANAG EM ENT
TION
R MAIT Y ®
O
F
IN
UR
SEC
R EVI EWE D BY B RAD CAU S EY
Policy Control
Sun Microsystems
www.sun.com
Price: $50 per user
Sun Microsytems’ latest ID
management solution unifies its user provisioning
and auditing products, providing an impressive level
of integration and functionality in a single package. Sun Java System
Identity Manager 7.0 is
a complete solution that
allows an enterprise to use a single console for a multitude of ID management tasks, including
role delegation, password synchronization, automated
provisioning and compliance auditing.
Configuration/Management
for nearly any type of integration, including Web applications, which generally present a huge challenge because
of their distributed nature.
Most of the common primary identity stores, such as
Active Directory, require that at least one Sun Identity
Manager Gateway be installed. The Gateways make
Identity Manager very scalable; you add as many Gateway
servers as you need.
B
Setup was somewhat lengthy, although not difficult; a
minimum installation required a dedicated server, JRE,
JDK, Tomcat and MySQL. Large enterprises will need
robust hardware and software components (all major
databases and application servers are supported). The
documentation is thorough and well written.
User data sources are added via agentless connectors.
Among the supported sources, which Sun calls resources,
are RSA products, BlackBerry Enterprise Server, Remedy,
PeopleSoft, Siebel and all database servers. Supported
resources can be added in a few simple steps, and others
can be accessed through generic connectors, or custom
built through the API. Sun has integrated SPML to allow
Testing methodology: Our lab included two Active Directory domains
and one OpenLDAP tree. User accounts were enumerated from various
sources, including MySQL and SQL Server, Web applications, and
various client-server applications. User roles such as administrators,
power users and end users were created to test access controls.
A
Policy and audit is where Identity Manager really shines.
By integrating fully functional auditing capabilities into
the standard interface, it allows you to provision a new
user for Active Directory, RACF and Oracle, and compare
the access given to current policies. If there are any violations, provisioning is automatically escalated for approval
based on a process you define. You can even periodically
audit existing identities for policy violations.
Delegation of duties reduces cumbersome management overhead.
Effectiveness
A
Sun has done an impressive job in furnishing a comprehensive ID management solution for the large enterprise, providing fast and effective linking of users to
identities. In addition to the great administration features, it handles user interaction very well. Users can easily log in to Identity Manager to handle password resets
and requests for resource access. Automatic resource discovery allows a simpler approach to adding and configuring identity stores, while ID consolidation helps link
various user accounts throughout the enterprise.
Information from ID stores can be reconciled, eliminating inconsistencies and reducing errors.
Reporting
A
Identity Manager handles all major reporting functions—
getting the data, formatting and moving it—remarkably
well. Clicking on the reports tab in the management interface provides access to canned reports, and you can also
easily create very flexible custom reports.
Reports can be scheduled, cloned, downloaded or
emailed in PDF or CSV format, or viewed in real time in
a custom-built dashboard.
Verdict
Sun Java System Identity Manager excels with agentless connectors, scalability and amazing auditing.w
I N F O R M AT I O N S E C U R I T Y
35
P RODUC T
Reviews
k
c
i
P
t
o
H
Symark PowerBroker
AC C E SS C O NTR O L
TION
R MAIT Y ®
O
F
N
I
UR
SEC
R E V I E W E D B Y S A N D R A K AY M I L L E R
Symark
PowerBroker 5.0
www.symark.com
Price: Starts at $1,000 per server
Symark PowerBroker solves
the dilemma of providing
root access privileges to
multiple users on Unixbased systems without
compromising security.
It delivers comprehensive security controls
through granular policies, and exhaustive auditing for
rock-solid regulatory compliance.
The client/server-based software resides at the shell
level, making no changes to the kernel. PowerBroker
supports 30 different types of encryption—AES 256 is
the default—to secure network traffic, logs and configuration files.
Configuration/Management
A
Installation requires moderate expertise in Unix environments and an understanding of basic shell scripting.
We used a simple batch file to disseminate the necessary
files to client systems.
PowerBroker works with HP-UX, Linux, Solaris,
SCO and AIX and integrates well with existing infrastructure such as routers and firewalls.
PowerBroker can be configured and managed by
command line or its well-designed Web GUI, which
can easily be used by someone with minimum knowledge of Unix. We used the GUI to quickly set up privileges, create and assign policies, create alerts, manage
encryption, and generate and view audits, logs and
reports.
Policy Control
A
PowerBroker’s policy control is extremely granular, based
on a programmable scripting language.
By assigning root-level privileges based upon on
role, the actual root password is never revealed. Policies
can also be assigned based upon user authentication
through centralized repositories such as LDAP and SSO
systems.
The new access control lists allow those unfamiliar
with programming or shell scripts to write policies that
control privileges through global categories such as user,
system, command, time of day and day of week.
Reporting
A+
PowerBroker’s greatest capability is logging and reporting. Ad hoc and custom reports are easily set up and run
from the Web-based report utility, drawing from massive amounts of information in the encrypted log files.
The Entitlement Report will satisfy auditors, presenting a quick overview of who can run what, and under
what circumstances.
The I/O logging option records all screens and keystrokes, storing them in an encrypted file that can be used
for forensic analysis or to meet rigorous regulatory requirements. It can also be used for real-time monitoring.
Data is logged in syslog format, so it can be ported to
SIM/SEM products, or exported in CSV and text formats.
Effectiveness
A
Everything the shell touches can be controlled through
PowerBroker. Instead of logging in through bin/bash or
csh, PowerBroker offers two transparent secured Kornand Bourne-based shells. When we logged in through
the PowerBroker shell, we did not have to type pbrun in
front of every request we wanted to run as root.
We were impressed by the control that can be
assigned to users based on role and circumstance. For
example, we elevated privileges of users so they could
access a particular system, such as a Web server, as root,
while denying similar root privileges to a mail server.
Security features include blocking predefined keystrokes,
automatic termination of idle root sessions, and checksum comparisons to identify potential malicious code.
Verdict
Testing methodology: Symark PowerBroker was deployed in a Linuxbased environment with a variety of servers requiring root privileges,
including a Web server and mail server.
PowerBroker is a scalable solution that effectively delegates root privileges securely and provides excellent
audit trails for regulatory compliance.w
I N F O R M AT I O N S E C U R I T Y
36
P RODUC T
Reviews
INTRUSION PREV ENTION
ISG 2000 with IDP
R E V I E W E D B Y P H O R A M M E H TA
Juniper Networks
www.juniper.net
Price: Starts at $42,500
The marriage of firewalls and intrusion
prevention makes good sense, as IPS
technology matures and gets serious
enterprise interest. Juniper Networks’
ISG 2000 appliance combines firewall,
VPN and its latest intrusion detection
and prevention software in an effective,
high-performance package.
Installation/Configuration
B+
The ISG 2000 is a multigigabit integrated firewall/VPN
system with a modular architecture, enabling high scalability and flexibility. To add IDP, the organization has to
get an advanced license, possibly buy extra memory and
purchase up to three security modules, depending on
their usage and throughput requirement.
ISG with IDP tightly integrates the software available
on standalone IDP products with ScreenOS 5.4.0r2, a
security-specific operating system with the capacity to
handle high-speed, high-volume traffic inspection.
Although the appliance offers a console for configuration, the best way is to use the Netscreen Security
Manager (NSM), a dedicated Red Hat Linux or Solaris
console for managing Juniper security products. The
user interface or the management client is the final
component that is installed on an administrator’s
machine (Windows or Linux) to configure the ISG and
any other ScreenOS-based devices in the network.
The user interface is designed well but still complex
because of the number of settings and features available.
When the device is added, NSM automatically detects
the OS and the installed license, and enables/disables
appropriate features accordingly. Adding IDP rules is
Testing methodology: We set up a lab with Windows and Linux
PCs sending legitimate as well as malicious traffic back and forth
through ISG 2000.
easy and similar to adding firewall/VPN rules. Juniper
provides a rich database of checks that can be used to
match and drop, or just log the attack traffic between
specified sources and destinations.
A
Effectiveness/Performance
Juniper Networks’ Multi-Method Detection (MMD)
technology uses up to eight different intrusion detection methods, including stateful signature, protocol
and traffic anomaly detection, and backdoor detection.
We tried—without success—to dupe the ISG 2000
using a variety of detection-evasion techniques such as
splicing and fragmentation, while executing DoS and
OS exploit attacks. We were amazed to see how little all
those attacks affected the performance of this beast,
which leverages a fourth-generation security ASIC, the
GigaScreen3, along with high-speed processors.
NSM lets you view the code of the current checks and
create your own checks within the IDP database.
B
Administration
Like any access control system, it is imperative that the
IDP rules be verified and updated on regular intervals
on the basis of the normal traffic flow. It’s easy to set
up daily updates and many other tasks, such as importing updated configurations and rebooting devices. The
management interface can be used to specify actions like
SNMP, syslog or email alerts when specified criteria
are matched. Because NSM stores all the information
required on the server, you can take care of device and
log backups like any other system.
Reporting
B+
NSM’s reporting module is a powerful and intuitive tool,
with multiple predefined reports grouped by type of
data, including firewall/VPN, IDP and administration.
Each grouping includes many report templates for top
attacks, attackers and targets, giving comprehensive
information with graphs. You can also create custom
report queries and run them automatically. Reports can
be exported only in HTML format.
Verdict
ISG 2000 with IDP is an excellent appliance that
offers a powerful combination of effectiveness and
performance, flexibility and manageability, and low
cost of ownership.w
I N F O R M AT I O N S E C U R I T Y
37
P RODUC T
Reviews
INTRUSION PREV ENTION
SGI-2000S IPS
R E V I E W E D B Y P H O R A M M E H TA
Stonesoft
www.stonesoft.com
Price: SGI-2000S, $31,900; SGI-200ANZ, $8,950
The implementation of
intrusion detection/prevention systems has
increased considerably,
in part because of
improved effectiveness
and the need to comply with federal and industry regulations. Stonesoft offers a strong new entry into this
crowded market with its StoneGate IPS products.
tion, while minimizing the number of false positives.
Attacks are detected by using context-sensitive fingerprinting defined with regular expressions. For example,
this expression matches any of the following patterns
in the traffic: “/bin/{ash|bash|csh|ksh| sh|tcsh}.” You can
make it context-sensitive by defining situation elements
that generate an alert only when the above expression is
detected for traffic originating from untrusted sources.
StoneGate detects zero-day attacks through protocol
and statistical anomaly detection techniques.
We tried multiple vulnerability scanners (Nessus,
WebInspect, AppScan) and penetration-testing techniques—denial of service, gain remote shell, overflows,
etc.—accompanied with evasion techniques, such as
time delay and fragmentation. Most were detected without affecting the normal traffic when run in inline mode.
In sniffer mode, the sensors can respond to selected
threats by sending TCP resets directly to the communicating parties or by giving a blacklisting command to a
StoneGate firewall, if one exists on the network.
Administration/Management
Installation/Configuration
B+
StoneGate’s security platform is highly flexible and scalable, featuring a three-tier architecture—user interface,
management and IPS (and firewall if you own it as well).
Organizations can deploy clusters with up to 16 nodes.
We tested the SGI-2000S IPS sensor appliance and
SGI-200ANZ analyzer device (for event correlation).
Since the appliances come with IPS engines installed,
we only needed to install the three management components, which we put on a single Windows server. (Linux
and Solaris OS versions are available.)
Wizard-driven installation and configuration of the
management server, which can manage all Stonesoft
products, is fairly simple.
There’s no auto-update capability, so we recommend
you download and install the latest IPS signature updates
(released about once per week) as regularly as possible.
Effectiveness/Performance
B+
Designed for gigabit networks, these beasts are equipped
with an IPS engine that uses various techniques for misuse and anomaly detection, for effective intrusion detecTesting methodology: In a typical lab setup with multiple Windows
and Linux machines, we sent legitimate as well as malicious traffic
back and forth between the machines and the Internet through the
SGI-2000S IPS.
B+
The use of regular expressions makes creating new rules
and custom checks easy. For example, if installing a
certain product is prohibited, administrators can create
a custom situation element and add an HTTP request
URI with a regular expression containing the address
the forbidden software uses to send or receive data.
The UI offers various situation context elements to
help users write intelligent context-sensitive regular
expressions to detect malicious traffic. Users can create
custom situations, category tags and even workflow for
certain events or sets of events
B
Reporting
The management server provides extensive reporting
tools for generating reports on the logged firewall and
IPS events. You can create reports on log, alert and audit
entries as well as statistical monitoring information. A
variety of report designs are ready for use in the software
and new reports can be designed and customized as
needed. Reports can be exported into PDF or text.
Verdict
Stonesoft has delivered another strong product (see
review of StoneGate SG-4000 firewall appliance,
February 2006) that thwarts attacks and monitors
traffic on internal networks without noticeable
degradation of bandwidth.w
I N F O R M AT I O N S E C U R I T Y
38
P RODUC T
Reviews
items display according to the level of privileges
granted to a specific user.
LO G MANAG E M E NT
LogLogic’s LX
R E V I E W E D B Y P H O R A M M E H TA
LogLogic
www.loglogic.com
Price: Starts at $50,000
Although device logs can contain a wide variety of information, little attention was given
to the review and management of these logs until regulations
like SOX, HIPAA and PCI made it mandatory. Now
companies are finding there is far more to gain from
reviewing and auditing logs than just compliance.
LogLogic offers enterprise-class appliances for analyzing and archiving log data that enable organizations to
achieve compliance, while offering decision support and
improved availability. We reviewed the LX 2010, one of
the LX family of appliances for real-time log data collection and analysis. (The ST series interfaces with NAS and
WORM devices for mass storage.)
Installation/Setup
B+
Configuration
B+
LX 2010 is a beast of an appliance with a 2 TB RAW (1
TB in RAID 10) storage capacity, dual 2.4 AMD Opteron
processors and 4 GB of memory. The setup is as simple
as it gets, supported by a hardened Linux kernel, MySQL
database, Apache Web server and Java.
The only thing left for the user is mounting the hard
drive in the slots in allotted order and changing the
default IP address on the Ethernet interface used to access
the Web-based management console, which is easily done
through the GUI or CLI.
An appliance can be used to manage multiple LX
and/or ST deployments. The access control feature
allows you to restrict network access based on source
IP address and destination port, similar to access lists
used by routers or firewalls. The GUI provides controls to access different parts of the system, and menu
Testing methodology: Logs were obtained from Windows and UNIX
servers, Cisco routers, Check Point firewalls and other networking
devices generating logs in syslog format.
LogLogic supports most of the widely deployed devices
in the industry. At a sustainable rate of 4,000 messages
per second, the LX 2010 can become the syslog and/or
SNMP server for all servers and devices in the network.
Logs can also be imported via HTTP, HTTPS, SCP, FTP
or SFTP. Multiple log formats covering virtually all types
of devices are supported—but not all log types. For
instance, for firewall/VPN products with proprietary log
formats, only Check Point Software Technologies, Cisco
Systems, Juniper Networks and Nortel are supported.
Email (Exchange) and database (Oracle) server support
is also limited.
Configuring log sources is straightforward. Adding
devices requires configuration changes on the source
devices as well. The documentation provides step-bystep instructions for setting up the log transfer rules and
frequency. We configured a few syslog devices, Windows
servers using LogLogic’s own open-source Lasso tool, a
couple of Cisco routers and a Check Point firewall. Since
most of the configuration happens on the log sources
themselves, adding and setting up devices on LX 2010
usually takes less than a minute.
B+
Reporting
Reporting is the most important component of this
product. Two excellent status dashboard screens show
the current mps rate, alerts, system performance and
total message counters. Another screen shows all added
devices and their message counters. The Real-Time
Viewer tab shows log messages as they are received.
LogLogic offers many built-in real-time reports for
access control, connectivity, database event logs, IBM
i5/OS, IDS, email and Web activity.
Administrators can create keyword or regular
expression searches to produce custom reports to
monitor network security and health. The ability to
replay old log data should prove very useful for incident response.
Verdict
LogLogic’s LX 2010 offers much-needed help to
companies in the areas of log review, analysis and
archiving. It can help organizations not only with
compliance but also with detection and prevention
of dangerous events.w
I N F O R M AT I O N S E C U R I T Y
39
P RODUC T
Reviews
U N I F I E D TH R E AT M A N A G E M E NT
FortiGate 3600A
R E V I E W E D B Y S A N D R A K AY M I L L E R
Fortinet
www.fortinet.com
Price: Starts at $32,995
The FortiGate 3600A rolls eight
homegrown security services—firewall, IPsec and SSL VPNs, IPS, traffic shaping, antivirus, antispam and Web filtering—into
one high-availability appliance built for speed.
It’s a good choice for publicly facing data centers and
managed service providers, with enterprise-class features
such as redundant power supplies, dual-core processing,
the new FortiASIC Content Processor-6, AMC network
adapter expansion slots, two accelerated gigabit SFP
ports and eight 10/100/1000 copper gigabit ports.
Configuration/Management
B
Policy Control
B-
The 3600A can be deployed as a gateway between the
Internet and private network (NAT/route mode), or on
a single subnet invisible to the rest of the network (transparent mode). We chose NAT mode in order to include
multiple subnets.
Using the quick-start guide, we planned our network
configuration and connected to the Web-based manager
in minutes. The interface is pleasantly clean and easy to
navigate. Thanks to the expandable menu tree, moving
through the initial setup was fairly intuitive.
For example, the VPN option expands to provide
instant access to IPsec, PPTP, SSL and certificate administration. We created firewall rules, applied policies for
content filtering and set up VPN tunnel associations.
Fortigate supports RADIUS, LDAP and Active
Directory authentication.
Our only significant frustration was with the client
software, which provides endpoint security and IPsec
VPN connectivity. It was extremely slow to install and
created instabilities in several instances.
Testing methodology: The FortiGate 3600A was deployed in
NAT/route mode between the Internet and a simulated enterprise
network. Threats specific to individual components were executed,
along with behaviors and traffic denied by policy.
Creating customized firewall rules, IPS signatures and
adding URLs to the Web filter was straightforward.
Working primarily through the firewall, we quickly
assigned numerous policies relating to network settings,
logging, traffic shaping and restricting client network
access based on policy compliance, such as up-to-date
antivirus and IPS signatures.
However, given the extensive hardware support for
high throughput (the pair of SFP connectors for optical
networks), we were dismayed there was little standard
policy control for VoIP. Also, there are only four IM services listed in the IM/P2P policy tab (MSN, Yahoo!, AIM
and ICQ); we would have liked to seen more choices,
given the explosive growth of IM clients.
A
Effectiveness
We were impressed with the quality of security services
on a single appliance, as well as flexibility for deployment and ease of administration. For instance, the IPS is
signature- and anomaly-based, and multiple VPN technologies are included. Automatic updates and system
backup and restore for multiple security services simplify
life for admins and reduce the chance of human error.
Each layer of security functioned effectively when
faced with common threats such as syn floods, malware,
port scans and spam. Prohibited Skype traffic and potentially hazardous URLs and sites containing blacklisted
keywords were blocked.
B-
Reporting
Logging is outstanding. The 3600A provides three
avenues for logging: local, syslog and through the
FortiAnalyzer, an additional dedicated appliance for data
collection and analysis from multiple FortiGate devices.
The exhaustive logging was easily parsed using singleclick column filtering.
Using check boxes, we set up custom email alerts for
more than a dozen different events, to be sent at defined
intervals. The event log is also highly customizable.
Unfortunately, there are few onboard reporting
features unless the data is sent to a FortiAnalyzer, which
was not included in our testing.
Verdict
Considering the costs and IT resources for managing
individual products, the FortiGate 3600A offers an
affordable and manageable enterprise solution.w
I N F O R M AT I O N S E C U R I T Y
40
P RODUC T
Reviews
U N I F I E D TH R E AT M A N A G E M E NT
Sidewinder 2150 v7
R E V I E W E D B Y D AV I D S T R O M
Secure Computing
www.securecomputing.com
Price: Varies; as tested, $35,900
In its June issue, Information Security tested six Unified
Threat Management (UTM) boxes; this month we
review Secure Computing Sidewinder UTM, specifically
the large-enterprise 2150 appliance. The new Sidewinder
release was too late for the comparative evaluation, but
would stack up in the middle of the pack.
Enterprise Management/Control B-
Sidewinder has a collection of different software management tools. Unlike most other UTM products, it
does not have a built-in Web server but uses a Windowsbased tool that doesn’t run on Vista yet. One big drawback is that the product doesn’t include a DHCP server
for the local network; you’ll need to supply your own.
It took about an hour to set up.
Sidewinder doesn’t allow multiple concurrent administrators to save configurations, although they can view
configuration and monitor operations. It also comes with
dual power supplies, which is handy if one fails. It also has
two available add-in slots in the model we tested.
Daily Operations
B+
We followed the same testing procedure as for the comparative review and tested how these products would
work on a daily basis. While Sidewinder’s IDS/IPS is
wired to live inside its firewall module, it has a very flexible IPS coverage and can scan for attack signatures and
behaviors. It can also explicitly detect outbound attack
signatures. Sidewinder has a very useful front-page dashboard that shows alerts, CPU and memory usage, and
other summary statistics in one convenient place. It is
also easy to set up and change security policies.
Sidewinder At-a-Glance
Ethernet Ports
Max attachment size for antivirus scans
AV supplier
Content filter supplier
IDS: Patterns or behavior
IDS in or out of firewall?
VPN types
Other ports scanned
Web App firewall
Authentication options
Multiple concurrent admins?
8*
User selected < 1 GB
Sophos
Own
Both
Inside only
IPsec
IM, P2P, VOIP, SQL
Extensive
Radius, LDAP, AD
No
*Includes two gigabit Ethernet ports that came installed on our test unit
Authentication & Security
B
Sidewinder sets up most of it security policies for each
network interface, but has separate controls for content
filtering, antivirus and antispam modules that are applied
across these interfaces. Sidewinder offers connections
to a variety of authentication servers, including Radius,
LDAP, Active Directory and iPlanet servers. It includes an
IPsec VPN only—no SSL.
Feature Module Integration
B
Sidewinder uses Sophos antivirus scanning but also has
its own SmartFilter content filtering engine. However,
SmartFilter requires a separate Windows-based administration and configuration tool and its own obscure
setup with nested sub-menus. This is because Secure
Computing sells this as a separate product that can be
run on other vendors’ firewalls. We’d like to see it completely integrated into the main console. One nice feature is the ability to run several antivirus scanners in
parallel on the same box to balance the processing load.
A maximum 1 GB file attachment can be scanned.
Although Sidewinder was able to easily block Skype
with its default settings, it doesn’t have explicit protection rules for other IM/P2P protocols.
It does extensive port scanning, including ports used
for VOIP, IM, P2P, SQL server and Citrix applications. It
also protects against common Web server attacks, such
as SQL injection and cross-site scripting.
Verdict
Testing methodology: We connected the Sidewinder box on a test
network with Windows XP, Vista and Apple Macintosh clients and a
Windows 2003 Enterprise Server, and ran tests using Skype, AOL and
Google Talk IM clients, and various security penetration techniques.
Sidewinder offers solid security features and is easy to
set up and manage. Its strengths are extensive IDS/IPS
and antivirus scanning features; its biggest weakness
is its separate content filtering module.w
I N F O R M AT I O N S E C U R I T Y
41
P RODUC T
Reviews
FI R E WALL
SonicWALL
TZ 180W
R EVI EWE D BY J O E L S NYD E R
SonicWALL
www.sonicwall.com
Price: $700 for 10-user version with standard O/S
SonicWALL has been a major player
in the SME firewall market for as
long as there has been a market, helping to define what SME firewalls should
look like, cost and do. The SonicWALL TZ
180W UTM continues SonicWALL’s tradition of
products designed and sized for the small business.
Ease of Use
A-
Small-office firewalls are often difficult to use, as developers jam more and more features into poorly designed
user interfaces. The TZ 180W, from opening the box to
final deployment, has a sophisticated and refined feel.
Although we ran into a bug in the initial deployment
wizard and some misdirected online help, every other
part of the system was easy to use.
The TZ 180W with SonicWALL’s SonicOS Standard is
stripped down from an enterprise firewall, which means
that some features, such as NAT, come as “one size fits
all.” However, SonicWALL has chosen an excellent subset
of features—more than most network managers will
need—for the small office version.
Security Features
B
SonicWALL has set a very attractive price for its Comprehensive Gateway Security Suite, a UTM add-on service
for the TZ 180W that includes software support along
with content filtering, antivirus, antispyware and intruTesting methodology: We installed the SonicWALL TZ 180W on
a production network and used both Windows and Mac laptops to
validate security services. To test performance, we used Spirent’s
WebAvalanche and WebReflector testing devices to deliver a typical
mix of Internet traffic sizes and pages.
sion prevention subscription services. Consider this
service a must-have for any TZ 180W, as it unlocks the
full potential of the product.
We found content filtering, antivirus and antispyware effective. Intrusion prevention was less effective.
SonicWALL has loaded what looks like a subset of
Snort’s signatures into the IPS, but not all the intelligence. For example, the TZ 180W alerted on a possible
port scan, which was actually the box communicating
with SonicWALL’s own Web servers. It missed all but one
of our outbound attacks.
Having all those IPS signatures configured to protect
against inbound attacks against servers on a device primarily designed to protect end users doesn’t make a lot
of sense. A better approach is signatures focused on endusers, such as malware protection or browser-focused
overflow attacks.
A
Performance
The TZ 180W provides significantly greater performance than the earlier TZ series. Because we had a 10user unit, we were unable to push the firewall to its
advertised limit of 90 Mbps using typical Internet traffic mix. Our system ran out of CPU at about 27 Mbps.
When we turned on all the TZ 180W’s security services,
goodput was about 9.7 Mbps, close to SonicWALL’s
advertised speed of 10 Mbps.
It should keep up with DSL and cable modem connections, but don’t be tempted to run LAN backups or
file sharing through it with security services turned on.
B
Wireless
The TZ 180W has a built-in, dual-antenna 802.11b/g
wireless access point. Although this only adds $95 to the
price, it was disappointing that we couldn’t get 802.11a
or 802.11n on a new system. Wireless security is limited,
but we were impressed that we were able to set up WPA2
with RADIUS authentication in seconds.
However, you can’t easily set up the wireless so that
insiders and guests can use it, so it’s best to pick one set
of users. With a cool set of features aimed specifically at
guest users, it will probably fit best to give guests secure
temporary Internet access without allowing them access
to your internal network.
Verdict
The TZ 180W is an outstanding small-office and
home-office UTM firewall, offering good value and
a broad suite of gateway security services.w
I N F O R M AT I O N S E C U R I T Y
42
P RODUC T
Reviews
U N I F I E D TH R E AT M A N A G E M E NT
Firebox X 1250e
R E V I E W E D BY B R E N T H U STO N
WatchGuard Technologies
www.watchguard.com
Price: Ranges from $2,290 (plus $4,420 for UTM
bundle for the Firebox X550e) to $3,790 (plus
$7,400 for the UTM bundle for the 1250e)
WatchGuard’s unified threat management
(UTM) appliances are a onestop shop for border security
needs, especially for a small- to medium-sized business.
We evaluated the Firebox X1250e,
which features eight 10/100 interfaces, stateful
packet inspection, application proxies, remote-user
and site-to-site VPN, and optional modules for gateway
antivirus, antispyware and antispam protection, plus
URL filtering.
Configuration/Management
A
Setup is straightforward. We followed the included quickstart guide to get the device working in less than an hour.
The management interface is one of the best we’ve
seen. The rules setup is logical and does not require
knowing any cryptic languages. The proxies and other
features are well integrated, and can be configured and
enabled/disabled easily for each rule.
Effectiveness
B+
The firewall immediately stood out on its own, thanks to
the ease of setting up rules. Rules are granular, and you
don’t have to worry about putting them in the correct
order—Firebox takes care of that for you.
Application proxies for HTTP, FTP, SMTP and DNS,
and a generic TCP proxy allow the firewall to inspect
traffic and deny or allow the request based on your
policy. For example, we set up a rule in the FTP proxy
Testing methodology: We tested the Firebox X 1250e protecting
two internal networks and a DMZ that included a Web server, FTP
server, SMPT and POP server.
to deny “get” requests. The rule worked as intended and
wouldn’t allow any file downloads. The controls are
granular; you can, for example, block the download of
certain extensions, and block or allow HTTP requests or
content types in the HTTP proxy.
Firebox’s IPS capabilities are strong. By default, it
will block anyone trying to port-scan or send suspicious packets through the device; our port scans got us
quickly blacklisted. We set up a Web site behind the
Firebox and attacked it using Metasploit, but all our
attacks were stopped.
The antivirus module is based on open-source
ClamAV, which we’ve found to be a competent antivirus.
One issue here is that you can only use the antivirus
through the HTTP and SMTP proxies, so, for instance,
there is no way to scan files going through the FTP proxy.
The VPN uses IPSec and PPTP, supporting remote
user and branch connections. Back-end authentication
can be implemented through Firebox itself, RADIUS,
Active Directory, LDAP or RSA Security’s SecurID.
The VPN client only works with Windows—a restriction for some shops, which can use the less secure PPTP
option.
The antispam filtering, provided by Commtouch,
picked up spam that even our tuned SpamAssassin filter
missed.
While Firebox’s URL filtering module features many
categories and blacklisted sites, it was possible to get
around some by using the IP address.
B+
Reporting
Reporting capabilities are good, but you can only export
the results in HTML and NetIQ formats (but it derives
the reports from XML data, so importing it elsewhere is
not out of the question).
However, the reporting gives you an excellent breakdown of device statistics, traffic stats, and IPS alerts, and
a report of hits on any rules you have in place (such as
users trying to visit blocked Web sites.
There are also extensive real-time monitoring capabilities including traffic and bandwidth monitors, device
statistics (memory usage, processes running) and a list of
authenticated users.
Verdict
Despite some minor flaws, the Firebox X series is an
excellent UTM deal, with its low entry price, terrific
firewall and routing capabilities, and top-notch
filtering services.w
I N F O R M AT I O N S E C U R I T Y
43
P RODUC T
Reviews
SSL VPN
Objects can be defined by a solitary IP address or a
network segment.
Setting up individual users and groups was equally
effortless. The 4000 supports user authentication via
LDAP, Active Directory, NT and RADIUS.
SonicWALL
SSL-VPN 4000
R EVI EWE D BY SAN D RA K. M I LLE R
SonicWALL
www.sonicwall.com
Price: Starts at $6,995
SonicWALL steps up
to the enterprise market
with the affordable SSL-VPN
4000 appliance, offering secure
clientless remote access to files, shares
and applications.
Support for port-forwarding applications,
such as Citrix, puts SonicWALL in league with industry stalwarts. Moreover, SonicWALL has no per-tunnel
licensing fees or restrictions on concurrent users, making
it even more attractive to growing organizations.
The SSL-VPN 4000 supports up to 200 concurrent
connections and includes support for two-factor
authentication, such as RSA Security tokens.
Configuration/Management
A
Policy Control
A
Using the administrator’s guide, we were able to log on
to the appliance within minutes. All major browsers and
OSes are supported.
SonicWALL’s familiar easy-to-use Web-based console
gave us instant access to major features, each offering a
subset of functionalities.
For example, the network tab includes access to the
interfaces, DNS, network paths, host resolution and
network objects—all straightforward. After basic network settings, we quickly set up services to which we
would provide secure remote access: HTTP, HTTPS,
terminal services (Java and ActiveX), VNC, FTP, Telnet,
SSH (versions 1 and 2), file shares and Citrix Portal.
Testing methodology: We tested SonicWALL SSL-VPN 4000
on a simulated Windows-based enterprise network behind a third-party
firewall. Remote access was tested from a variety of laptops and
remote machines, running an assortment of operating systems and
Web browsers.
We were impressed with the granular policy control,
which let us assign access privileges at the user, group
and global levels.
We were able to delineate authentication to our
AD server, so that domain members were automatically
assigned the policies and access privileges of their associated group.
Policies are granular and highly accessible. A single
window enabled us to assign general settings, such as
enabling single sign-on using SSL VPN credentials, creating individual policies for network objects, IP addresses and ranges, and server paths, such as for Citrix. In
addition, we were able to set up detailed login policies,
such as one-time passwords sent via email and logins
from specific IP addresses or defined browsers.
A
Effectiveness
We were extremely satisfied with SonicWALL’s interoperability, including the product’s Web access to email,
files and Web-based applications.
Additionally, the NetExtender thin client can be automatically downloaded and installed to provide access to
email using client software installed on remote machines
and non-Web-based applications, such as CRM systems
and proprietary software.
We simulated a variety of scenarios that tested the
granularity of policy features, such as allowing global
access to email while limiting access to specific file shares
and applications.
C+
Reporting
The VPN lacks a comprehensive view. Interface statuses
are under the system tab, active user sessions are under
the users tab, and viewing events requires going to the
log tab.
Logging is very basic, although it supports syslog and
can email logs and alerts to a single address.
Verdict
SonicWALL SSL-VPN 4000 is an affordable and
capable appliance for mid-sized enterprises.w
I N F O R M AT I O N S E C U R I T Y
44
P RODUC T
Reviews
HotPick
FaceTime Internet
Security Edition
INTE RNET SECU RITY
TION
R MAIT Y ®
O
F
IN
UR
SEC
R E V I E W E D B Y S A N D R A K AY M I L L E R
FaceTime Communications
www.facetime.com
Price: Starts at $7,125
Policy Control
FaceTime’s Internet Security
Edition tackles the greynet
challenge of sifting Web
traffic to differentiate between
legitimate and unauthorized use of real-time
communications applications such as instant messaging,
Web browsing and VoIP.
The combination of RTGuardian (RTG), a hardened
Linux rack-mounted inline appliance, and Greynet
Enterprise Manager (GEM), a Windows-based server,
delivers security at the perimeter and endpoint by identifying malware, spyware, adware and unauthorized traffic.
Configuration/Management
B
Despite well-written documentation, setup took extensive preparation and knowledge of Active Directory and
domain credentials.
RTG enforces policies and ties into GEM, which provides centralized management and reporting through a
secure Web interface.
GEM automatically discovers endpoints by querying
the primary domain controller. Administrators can also
specify a range of IP addresses and discover endpoints
through ping and Windows Management Instrumentation. In both cases, GEM failed to detect several Windows desktops and all our non-Microsoft machines.
Effectiveness
tering and managing greynet applications. What really
impressed us is FaceTime’s approach to protecting desktops against malware, spyware and adware.
When RTG identifies malicious behavior, it feeds the
data to GEM, which deploys a temporary client to clean
the machine and scan for additional infection. It inoculates the machine, using ActiveX kill bits and Windows
software restriction policies, which prevent the code
from executing again. This feature stopped spyware cold,
despite our repeated attempts to reinfect the machines.
A
RTG controls traffic at the gateway, performing URL fil-
A
We were able to set global and granular policies for Web
browsing, Web mail, IM, file sharing, streaming media
and VoIP.
For example, we allowed the use of certain public IM
clients while prohibiting others. No one on our network
was permitted to use P2P applications, and Skype was
accessible only to the sales group. We could schedule
automatic scans, spyware removal and inoculation. Policies can be assigned according to multiple criteria such
as IP address, host, user, domain and operating system.
Comprehensive URL filtering categories let us turn
off access to generally prohibited and productivitydraining sites (porn, gambling, shopping, news, travel).
Custom policies can be set by users, groups, location, file
extension and content.
Reporting
A
GEM provides centralized reporting and logging; the realtime reporting dashboard includes activity blocked by
RTG as well as GEM activity tracking infected computers.
In addition to providing statistical analysis for everything from infections to policy violations, FaceTime
offers a variety of executive and auditing reports. Administrators can quickly see the rate of spyware infections
and spot trends as to what users/systems were most vulnerable and often infected, while auditors have access to
detailed information about data transferred via various
Web-based channels.
IM reports can be split into events and usage, providing detailed, critical information, such as transferred
files. Reports can be sent via email or exported to an
FTP server for automated distribution.
Verdict
Testing methodology: The RTGuardian appliance was deployed
on the span port of a DMZ switch; GEM Server was installed on a
Windows 2003 Server. Numerous applications were tested, using
malware including spyware and adware.
FaceTime offers an effective and affordable solution
to manage, control, secure and provide policy
compliance around Web-based applications.w
I N F O R M AT I O N S E C U R I T Y
45
P R ODUCT
R EVIE W
From USB drives to MP3
players to DVDs, portable
storage media are an end
user’s dream and a security
manager’s nightmare.
Gone in
We evaluated six device
control tools that can
help you rest easier.
BY SANDRA KAY MILLER
Did your business just walk out the door?
Our mobile workforce can steal or lose sensitive
data quickly and without detection, from a software
developer sneaking out gigabytes of valuable source
code on his iPod to an executive’s wireless-enabled
laptop being sniffed at the local coffee shop.
Think about all the ways we move and store data
on mobile devices: USB ports, which support a
multitude of portable storage devices, including
flash drives, portable hard drives, printers, and
music and video players; FireWire, PCMCIA, serial
and parallel ports, CDs/DVDs, tape drives and even
the lowly floppy drive. Add unprotected WiFi,
Bluetooth and Infrared (IrDA) connections, and
you have a real security nightmare on your hands.
46
I N F O R M AT I O N S E C U R I T Y
Photographs by JOHN KUCZALA
a Flash
I N F O R M AT I O N S E C U R I T Y
47
TE ST B E D
It wasn’t long ago that security administrators controlled access to USB ports
with epoxy or caulk and physically disabled onboard wireless. Now, however,
instead of trying to ban use of portable
storage devices and wireless connections,
organizations can select from a fairly new
but effective group of products that give
them granular policy-based control over
their use. Device control products can
help balance productivity with security
by allowing administrators to centrally
authorize and monitor endpoint devices.
In a head-to-head review, Information
Security examined six device control products, all of which provide centrally managed granular control over ports, interfaces and storage devices: DeviceLock 6.0
from SmartLine, Sanctuary Device Control 4.0 from SecureWave, Endpoint Access
Manager 3.0 from ControlGuard, DeviceWall 4.5 from Centennial Software, Safend
Protector 3.1 from Safend and Protect
Mobile from Workshare.
Each product was graded based on its
ease of installation and configuration, policy, tampering resistance, port and device
control, encryption support, performance,
and monitoring, alerting and reporting.
Overall, we found all the products performed as advertised, but there are enough
differences to consider when choosing a
portable endpoint data control solution
(see “Making the Grade,” p. 51).
Getting Started
About this
review
Centennial Software’s
DeviceWall 4.5
ControlGuard’s Endpoint
Access Manager 3.0
Safend’s Safend Protector 3.1
SecureWave’s Sanctuary
Device Control 4.0
SmartLine’s DeviceLock 6.0
Workshare’s Protect Mobile
I
nformation Security deployed six portable storage device control
products in our test lab.
All the products were tested in a Microsoft Windows environment with Active Directory, although some supported Novell. All the products utilized either an embedded or external version of a SQL database.
Our testing environment included wired and wireless network connectivity
with both desktops and laptops, supporting an array of portable storage
devices including USB flash drives, FireWire external hard drives, CD-RW
drives and floppy disk drives. Our testing also included PDAs and serial
docking stations, smart phones with Bluetooth connections, PCMCIA wireless adapters and multifunction printer/scanner/fax/copier machines on both
USB and parallel ports.
Concentrating on real-world scenarios, we blocked devices such as
portable music players and storage devices (flash drives, FireWire drives)
while allowing legitimate peripherals including keyboards, mice, printers, faxes
and scanners. Drilling down into the granular policies, we set CD/DVD drives
to read-only and disabled Bluetooth and IrDA connections, while allowing
WiFi use.
Multiple attempts to introduce devices contrary to policy were performed
using a variety of devices and connections, including portable storage devices
infected with known malware including worms, Trojans and keyloggers.w
ll the products we tested have
similar architectures—server, console and client/agent—based on a
Windows platform, although each included support for Novell’s directory services.
We deployed each product on an identical
simulated enterprise network (see “About This Review,”
above) using numerous desktops and laptops, supporting
multiple ports and removable storage devices.
Centennial’s DeviceWall was the easiest product to
install, since it requires only two components—the Control
Center and the Client Service. With a half-dozen ways to
roll out the client, DeviceWall got our top vote for installation and configuration.
ControlGuard and DeviceLock have similar setups,
consisting of a server, client and multiple Windows-based
ways to administer the product (Active Directory, MMS,
SMS, GPO). With two different client agents—active and
passive—ControlGuard gave us more to consider during
setup. We deployed both types of agents and concluded
A
48
I N F O R M AT I O N S E C U R I T Y
—SAN D RA K AY M I LLE R
that this aspect of ControlGuard should be simplified with
a single agent that could perform in either or both modes.
DeviceWall and DeviceLock had easy install wizards that
walked us through setting up initial permissions and policies. DeviceLock’s wizard allowed us to set permissions
for ports and devices, getting us running quickly.
Installing and configuring Workshare Protect Mobile
took the most effort, because it is part of an enterprise
suite of three components. It delivered
comprehensive endpoint protection, but
didn’t provide the depth of granularity or
functionality as the other products.
The installation of SecureWave was the
most difficult, because of its four compo-
nents—a database server, an application server
with two subcomponents, the management console and the client. We also encountered several
client deployment issues that required extra time
reconfiguring our firewall.
Policy Configuration/Enforcement
ltimately, everything boils down to policy
and enforcement and performance. Policy
granularity is a driving factor in each of
these six products. For portable storage devices, our
testing revealed nearly identical features, including
monitoring and control over reading, writing and
blocking.
Policies were determined by device types and
classes, ports, connections, machines and users. With
all the products, we could set up who could use what
device/port/connection and when.
The policy options available are so plentiful, it’s
easy to get overwhelmed and confused. We found it
was easier to start with our global policies and work
to more detailed policies, such as those for individual users. We were also able to set different policies
for the same user/computer determined by online/offline
status. That means when a mobile user returns to the office
and logs in to the domain, wireless interfaces can be turned
off, and corporate asset protection, such as file filtering,
engaged.
All the products allowed very fine-grained policy,
mainly through whitelists—the more granular the policies
a product supports, the better the controls. DeviceLock
provided the most detailed assignment of authorized
devices. For example, we were able to allow a single FireWire portable hard drive based on its serial number. The
exceptions can also work in reverse; for example, you can
shut down access for terminated employees or limit devices
to read-only.
We liked how SecureWave’s Sanctuary Device Control
comes out of the box with a default deny-all policy. No
data was allowed to be transferred to external storage
devices until we set up authorization. Allowing only what
you authorize—instead of trying to blacklist what you
don’t—is sound security policy.
SecureWave has a number of ways to keep tabs on traffic,
including data transfer throttling and file type filtering.
U
For example, we set policies that limited file types to
Microsoft Office files no larger than 5 MB. Regardless of
how we tried to save CAD files—both less than and in
excess of our size limit—to flash drives, portable hard drives
or write to CD, we were unable to do so.
ControlGuard earns kudos for recognizing that many
mobile workers also connect directly to the corporate network. We easily set up two distinctly different policies,
offline and online. We simulated a common problem that
occurs when mobile workers connect their WiFi-enabled
laptops directly to the corporate network—they still have
a live wireless connection. For our testing purposes, when
laptop users logged on to the domain, their WiFi adapters
were disabled.
ControlGuard addresses another real-world scenario,
exercising control over multiple users logging on to a
single machine or a single user with access to multiple
machines. This is where a firm understanding of policy
hierarchy is required. For example, a user having rights to
a USB port on one machine doesn’t necessarily mean he
has the same rights on another.
For organizations that want to further enforce policy
through enterprise-class management systems, ControlGuard’s Endpoint Access Manager is designed to integrate with third-party products like HP OpenView and
CA Unicenter.
Safend offers similar policy control through rolebased access and prohibiting simultaneously enabling
multiple networking protocols. One feature that
really caught our attention was the ability to easily
print a summary of our entire policy anytime. This
means corporate policy can be posted or viewed
by management, which doesn’t have access to
I N F O R M AT I O N S E C U R I T Y
49
KEY ISSU E
Encryption Gains
ground
D
ata encryption has long been a strong security technology, but its
use has been generally limited because of the complexity of implementing and maintaining it, as enterprises wrestle with thorny issues
such as a key management and security.
That’s all changing because of regulatory requirements and the exposure of
data through Internet-facing applications. Nowhere has this become more evident than with the ubiquitous use of high-capacity portable storage media. Five
out of six solutions Information Security tested for this review have integrated
automatic forced encryption capabilities into their products.
Although the use of automated encryption for portable storage media is
solving numerous security issues, there are still significant challenges to
address.
Nate Lawson, senior researcher at Cryptography Research (www.crypto
graphy.com), a security consulting and technology licensing firm, points out
that there is plenty of room for improvement with the widespread use of
encryption for storing information.
“How do I make sure I never lose or destroy that key, because if I do, it’s
like losing the entire set of data,” Lawson says. “I won’t be able to decrypt
it again.”
While there are lots of standards for encryption (AES, DES, 3DES, etc.)
and protocols (SSL), there is little standardization for key backup.
Lawson sees this as a potential pain point, especially in M&A scenarios
and because of the speed at which technology becomes obsolete.
Before organizations begin routinely encrypting portable storage media
through solutions such as the ones we tested, they need to examine the lifecycle of the data being stored and ensure access to decryption tools, such
as keys and software.w
computer resources, there are bound to
be exceptions to the rules. DeviceWall
was our favorite product for bending the
rules. It let us assign temporary access for
up to three devices either for the current
Windows session or by start time and
duration. Even when we weren’t connected to the network or Internet to push out
a change in policy, DeviceWall gave us the
option to generate a key that could be
verbally exchanged or sent via text message over a mobile phone to provide temporary access to the restricted port or
device.
DeviceLock’s exception to policy functions similarly to DeviceWall’s temporary
access, but lacked the granularity to assign
any length of time, giving only the option
to use the restricted resource during that
particular Windows Session.
Safend allows for the temporary suspension of the client, even when the computer is offline.
Tamper-proofing
e attempted to circumvent
our installed clients through
a variety of methods. Often,
installed components can be sidestepped
by local users who have administrative
rights to their machine. Even with local
admin rights, we were unable to modify
—SAN D RA K AY M I LLE R
or remove any of the installed clients.
Tens of millions of USB flash drives are
sold every year, and you can bet some are
the console, but needs access
going to be lost or stolen, sometimes with sensitive data.
to security policies.
DeviceWall was our pick for the lost flash drive scenario.
Overall, DeviceWall’s policy
When we inserted a USB flash drive into a bare-bones lapconfiguration was the least intutop running no device control client, we received the mesitive of the products tested,
sage that our drive was not formatted and asked if we
although the Master Policy tree
would like to format it. Had the drive contained confidenaccessed through the Control
tial information, the cost associated with losing the data
Center provided a clean interface
to the wrong entities could be devastating, but thanks to
for configuring 16 different device
DeviceWall, less than a minute after plugging in the unconcategories, including digital camtrolled drive, it was wiped clean.
eras, scanners, smart phones, and
We also addressed the issue of theft, loss and tampering
BlackBerry, Palm OS and Windows
of removable storage devices and media through the prodMobile devices. We would have liked
ucts’ use of encryption. DeviceLock, which was generally
to have seen all the individual cateoutstanding in other areas, was the only product in our
gories for mobile handheld devices
testing that did not support any type of encryption, which
under a single high-level branch on
brought down its overall grade.
the tree, instead of each given its own.
SecureWave set the bar with two different types of
It would make the Control Center
encryption—centralized, which allows administrators
interface much less cluttered.
to set the requirements, and decentralized, meaning an
When building complex policies
authorized user can decide when to encrypt. Additionally,
that limit or deny the availability of
you can export keys to a file or to the portable device for
50
I N F O R M AT I O N S E C U R I T Y
W
R E P O RT CAR D
Making the
grade
Vendor
ControlGuard
Endpoint Access
Manager 3.0
Starts at $10 per seat
www.controlguard.com
SmartLine
DeviceLock 6.0
Starts at $35 per seat
www.devicelock.com
Centennial
Software
DeviceWall 4.5
Starts at $25 per seat
www.devicewall.com
Safend
Protector 3.1
Starts at $32 per seat
www.safend.com
SecureWave
Sanctuary Device
Control 4.0*
Starts at $45 per seat
www.securewave.com
WorkShare
Protect Mobile
Starts at $10 per seat
www.workshare.com
Installation & Configuration
How easy was the product
to install and set up?
10%
B
B
A
B
C
B
Policy Configuration
Level of granularity and
ease of deployment.
20%
A
A
A
A
A
A
Loss, theft, tampering
Client and portable
storage security.
10%
A
A
A
A
A
B
Port & Device Control
What does the solution cover?
10%
A
A
A
A
A
B
Encryption
How well is it supported?
10%
B
N/A
C
B
A+
C
Performance
How did the product stack
up against testing?
20%
B
B
B
B
B
B
Monitoring, Alerting &
Reporting
20%
A
A
B
A
A
B
AComprehensive
control through
client-based agents;
provides complete
control over all I/O
devices including
PCI, ISA and
optical devices
B+
Powerful shadowing
feature lets companies know exactly
how their critical
information is moving and being stored.
Would like to have
seen support for
encryption.
B+
Easy to use, great
reporting; includes
encryption but
doesn’t provide the
policy granularity or
alerting features of
competing products.
ADelivers comprehensive visibility and
control over endpoints, devices and
network interfaces.
AOffers complete I/O
control over all
devices on Microsoft
Windows and Novell
networks.
B
Focuses on mobile
devices such as
laptops and PDAs.
The Verdict
*SecureWave’s Sanctuary Device Control 4.0 is part of the Sanctuary Suite.
access to encrypted media offline, although we felt that this
compromised the security of the portable storage device.
SecureWave offers the strongest encryption, with AES 256.
DeviceWall offers two different ciphers—AES and
Blowfish—in both global and individual user key models.
For instance, a company might require its HR employees
to automatically encrypt all data transmitted via WiFi
or saved to portable media. However, encryption is only
available for use with USB flash drives. On the plus side,
DeviceWall allows you to easily back up the Global Key,
so data can be retrieved if the key is lost.
ControlGuard also provides encryption for secured USB
drives. We liked its “self-destruct” feature, which limits the
lifecycle of the data accessible on the drive.
Workshare Protect Mobile provides the most flexible
client-side encryption through PGP based upon content.
Once files have been identified as requiring additional
security, they are automatically encrypted.
Safend’s encryption is the most transparent to users. We
were able to use the same encrypted USB drive on all the
machines on our network with the Safend client installed
without ever realizing the device had been encrypted. Of
I N F O R M AT I O N S E C U R I T Y
51
course, when we attempted to use the drive in a nonSafend computer, we were unable to access the drive.
One big worry with encrypted files on portable media
is the decryption software won’t be available when needed. Safend had the forethought for just such a scenario and
includes a Home Decryption Utility that allows authorized users to access information on encrypted devices
when the Protector Client is not present.
Wireless Control
ireless covers a lot of territory on today’s
mobile devices. All the products we looked at
included comprehensive control over WiFi,
Bluetooth and IrDA interfaces.
Since its introduction, there has been a lot of handwringing over WiFi connections. Administrators disable
onboard wireless, but still have to worry about an employee
using their own inexpensive PCMCIA wireless adapter so
they can hook up at home or a hotspot.
Administrators are just catching up to smart phones
and PDAs, which are increasingly taking advantage of
Bluetooth technology for file transfer and synchronization
with laptops. An inexpensive USB Bluetooth adapter can
quickly connect a PC to a Pocket PC.
And let’s not forget about IrDA. Not as powerful or
popular as WiFi or Bluetooth, infrared personal area network connectivity still presents a vulnerability.
Safend clocked in with the best control for WiFi, based
upon MAC addresses, SSID and network security levels.
The remaining products didn’t provide as much control as Safend, but they all provided basic permit/deny
wireless interface blocking functionalities that identified
all wireless interfaces regardless of type. For instance, we
set policy to deny all WiFi with a laptop containing an
onboard wireless adapter. As we added PCMCIA and USB
wireless adapters, they too were disabled despite those
ports not having any deny policy assigned to them.
W
Auditing and Reporting
hile the majority of our testing was devoted to
the verification of security features, in today’s
regulatory environment, a robust auditing feature can be just as critical as security.
The most comprehensive monitoring feature for this
purpose is shadowing, which is the ability to record all data
transferred to and/or from a device or port. DeviceLock
and SecureWave both support shadowing.
During our testing, shadowing allowed us to capture all
data sent to specific devices, including our printer/scanner/copier/fax machine. How many companies actually
monitor the information sent over a fax or documents that
have been scanned? Low-tech crimes are often overlooked.
The only drawback we could see with data shadowing
was, ultimately, data storage. A large enterprise could generate an enormous amount of data.
W
Safend uses file logging; while not as robust as shadowing, it
lets administrators track what files are being accessed,
moved, deleted, created and modified.
Safend took the honors for the most useful logs, with
excellent information for forensic investigations.
We also liked DeviceWall’s detailed Policy Change Logs,
which record all the policy changes made and provide
comprehensive connection reporting. On the other hand,
we found the graphical Audit Log Reports and Acceptable
Risk meter of little use to a security professional.
Spotlight on the Endpoint
obility and portability make data protection a
far more complicated problem than it once was.
They’ve given this product market traction it
wouldn’t have seen just a couple of years ago. As the products mature, they feature improved reporting and central
management capabilities, and, in most cases, combine encryption with device control for stronger endpoint security.
We’re starting to see similar capabilities in more comprehensive security and data protection products. Expect
to see these products extend their feature sets or get folded into broader products, as larger companies continue
their pattern of acquiring key new technologies.
Time and experience have taught the security community that going to the root of a problem will often save
time and money. Priced on a per-user, annual basis, for
larger enterprises, these products can become pricey on
top of existing security solutions licensed on an annual
basis, such as antivirus/antispyware. Nevertheless, endpoint
security has clearly moved center stage, and many corporations are going to do what it takes to protect their data as it
moves out into the world.w
M
Technical editor Sandra Kay Miller is a frequent contributor
to Information Security. Send comments on this article to
[email protected]
I N F O R M AT I O N S E C U R I T Y
52
P R ODUCT
R EVIE W
UNIVERSAL
IS YOUR BUSINESS
READY TO ROLL
NETWORK SECURITY
INTO A SINGLE
PLATFORM?
WE EVALUATE
SIX LEADING UTM
APPLIANCES TO
HELP YOU PUSH
THE RIGHT BUTTONS.
REVIEWED BY DAVID STROM
53
I N F O R M AT I O N S E C U R I T Y
Photograph by JOHN KUCZALA
CONTROL
UNIFIED THREAT MANAGEMENT IS A GROWING,
competitive field, with more than a dozen vendors.
The idea is to consolidate your security appliances into
a single box and manage an integrated protection profile for your corporate network. While especially appealing for companies with several branch offices without
any resident IT or security staff, the implementation
isn’t perfect for a corporate-wide deployment, primarily because of limits on how you are allowed to administer the component security applications.
We asked vendors to deliver a product that could act
as a firewall and virtual private network gateway, and
protect our test network against attack with a minimum of four defense mechanisms—antivirus, Web
content filtering, intrusion prevention and antispam.
I N F O R M AT I O N S E C U R I T Y
54
TE ST B E D
ABOUT THIS
We reviewed six UTM appliances in a
head-to-head evaluation: Astaro Internet
Security’s Astaro Security Gateway 320;
Check Point Software’s UTM-1 2050;
Fortinet’s FortiGate-1000A; IBM Internet
Security Systems’ Proventia Network
Multifunction Security MX5010; Juniper
Networks’ SSG 550 and SonicWALL’s
SonicWALL Pro 5060c.
Astaro Security Gateway 320
Check Point Software’s
Fortinet’s FortiGate-1000A
We examined log files and configuraUTM-1 2050
tion reports to determine how each appliance stacked up in enterprise management
and control, daily operation, authentication and policies, and feature integration.
All of the products sell for between
IBM Internet Security Systems’
Juniper Networks’ SSG 550
SonicWALL Pro 5060c
$12,000 and $18,500. But getting specific
Proventia NMS MX5010
price configurations isn’t easy, as each
product has a complex range of user and
feature licenses. Further confounding the
e connected each UTM box on a test network with Windows XP,
pricing issue is that you will need to match
Vista and Apple Macintosh clients and a Windows 2003 Enterprise
the capacity of the product with the
Server running Microsoft’s IIS Web server.
expected network traffic it will protect. We
Each UTM box was configured with two interfaces—a local network with
a DHCP server enabled, and an external network connecting to our DSL
tried to compare appliances that had a simmodem. We set up firewall and intrusion rule sets, ran Outlook Express POP
ilar number of network ports and capacity
email clients, and used Skype, GoogleTalk and AIM messaging sessions.
for a 1 Gbps external network connection.
We also connected to a WebDAV server to share files over the Internet.
We asked vendors to send us the boxes
We
connected to each product’s built-in Web management server using
with the highest throughput possible and
both
Firefox v2 and Internet Explorer 6 and 7. We also used SSH to perform
geared toward the largest networks. When
command-line
configuration tasks when necessary.w
we did our tests, we turned on all of the
—DAVI D STR O M
security modules—in the real world, this
will severely limit their overall performance and is something to consider when
Web servers. Check Point actually has three configuration
deploying these products. However, we did not test perforinterfaces—command line, a Web-based initial configumance. This is because testing performance is fraught with
ration tool for basic tasks, and its SmartView Monitor
all sorts of issues. Either you test with synthetic clients to
Windows-based administration tool (See Check Point screen
generate phony traffic so you can compare how different
shot, below). Unfortunately, you’ll need to be familiar with
products respond on the “same” artificial lab network,
all three. For example, you have to go to the command line
or you do your tests on a live network and hope that the
interface to set up a DHCP server on an internal network.
insights gained with your actual conditions are worth the
Some of the Web interfaces are more logically designed
loss of having the comparable traffic data. As a potential
purchaser, you should match throughput specs with what
you ultimately need on your network.
REVIEW
W
1
ENTERPRISE MANAGEMENT
AND CONTROL
e examined how each product is managed, typically with a Web browser, and the various administrative roles that can be performed concurrently.
We also looked at how particular functions are
licensed, and how threat signatures are updated. Because
these products handle a variety of security tasks, ease of
setup is important, and being able to delegate and divide
administrative roles is also critical.
CONFIGURATION. All of the products, except Check Point,
are primarily configured by connecting to their built-in
W
55
I N F O R M AT I O N S E C U R I T Y
Check Point’s SmartView
Monitor for administering its operations.
than others. For example, Astaro, IBM ISS and Fortinet
separate the functional modules—separate menu trees for
antivirus and IDS, for example—and logically lay them
out. Juniper has the poorest interface of the six, because its
commands and controls are buried several levels down or
require operators to visit multiple pages to set up even the
simplest procedures, such as changing one of the antivirus
settings. SonicWALL’s interface is just a little better than
Juniper, hiding many of its UTM features under a single
“security services” menu tab.
Setting up the IBM ISS box took about an hour, and
Check Point took several hours. The others were somewhere
in between. While this may not be terribly important if you’re
installing a single box, it will add up for large deployments.
IBM ISS stood out from the pack with superior defaults,
such as setting up internal network routes and activating
features at the click of a button. This default-driven approach
could be a bit problematic if your tastes run to doing something more sophisticated. For example, most of the other
UTM appliances could handle connections to a WebDAV
server for sharing files; with IBM ISS, we needed to set up
a special firewall policy to allow this traffic. Nevertheless,
this was a minor inconvenience—not enough to keep IBM
ISS from getting the clear top grade in overall enterprise
management.
LICENSING, UPDATING. Each product has intricate licensing and signature file update issues, mainly because customers will purchase varying configurations, feature sets
and user counts. None of the products did a particularly
good job troubleshooting licensing errors; Check Point
and Juniper had the most complex and unintuitive licensing procedures. In fact, we had trouble with our Check
Point licenses even after its engineer spent several hours on
site setting up our box that turned up a bug. The other
products make installing and upgrading licenses, and
updating threat signatures, far easier.
IBM ISS makes this process a snap; it consolidates all of
its updates for antivirus, IDS and firmware in a single
screen. You can set it to check for updates automatically on
a schedule. The others are more complex; you will have to
visit multiple screens or do more than just push a single
button to update everything.
ADMINISTRATION. Consolidated security administration
is a key value proposition for UTM. However, getting to
this consolidation won’t be easy. Because these products
cover a wide range of protection methods, they need to have
the flexibility to be operated by multiple administrators.
Fortinet, Juniper and Astaro can handle multiple concurrent administrators and immediately post any configuration changes to their boxes in a “last one wins” scenario:
This means that any intermediate changes will be ignored,
which isn’t ideal and means one person needs to have
ultimate authority over all UTM appliances. Check Point,
SonicWALL and IBM ISS only allow for a single administrator to be connected at any one time to avoid conflicts.
Check Point has the most complex and useful approach,
providing great flexibility across a large deployment.
Multiple administrators can run its SmartDashboard in
read-only mode to view, but not change, the configuration. And it has other tools, such as the separately priced
Provider-1, which can segregate roles between, say, a desktop department to handle antivirus configuration and a
network group to manage the firewall setup. Juniper has
something similar with its separately priced NetScreen
Security Manager for managing role-based administration.
(SonicWALL is coming out with a new version of its management software that will allow multiple concurrent admin
users, but this wasn’t available for our tests.)
2
DAILY OPERATION
critical value UTM products offer is the ability
to quickly determine if your network has been
breached or if you need to adjust the various protective mechanisms, since you have access to firewall,
IDS and VPNs all in the same place. This means that if
you mistakenly open a firewall port for the VPN, you can
receive alerts to fix it without having to compare logs from
two different places.
We used a typical scenario in which we ran the box for
several days, examined the reports based on an initial firewall and protection rule set, and then adjusted our rules
based on two situations—places where we wanted to eliminate false positives, and places where we needed to tighten
down the box to prevent typical security weaknesses. Part
of this exercise was to examine how reports would be created and examined and how threats will be evaluated and
acted upon by the device.
Overall, Fortinet has the best set of tools to handle the
day-to-day life of a security administrator, and Juniper
scored lowest with its quirky main menu that scatters controls in almost random order (See Juniper screen shot, below).
Juniper also requires that you visit several places to examine
reports and other screens to change its protection rules.
A
Juniper’s quirky main menu (down
the left side of this screen) presents control
settings in almost random order.
I N F O R M AT I O N S E C U R I T Y
56
C O M PA R I S O N
UTM FEATURES
Astaro
Internet Security
ASG 320
Check Point
Software
UTM-1 2050
Fortinet
FortiGate-1000A
IBM Internet
Security Systems
Proventia MX5010
Juniper Networks
SSG 550
SonicWALL
Pro 5060c
Ethernet ports
8
8
10
10
4*
6
Maximum attachment
size for antivirus scans
User selected
User selected
User selected
< 139 MB
User selected
< 1 GB
User selected
< 24 MB
Unlimited
Antivirus supplier
Authentium, Clam
AntiVirus, hardware-based
CA
Own
Sophos, own
Kaspersky
Own
Content filter supplier
SurfControl**
SurfControl
Own
Own
SurfControl
Own
IDS: Patterns or behavior
Both
Both
Both
Pattern recognition
Both
Behavior
IDS in or out of firewall
Either
Either
Either
Inside only
Inside only
Inside only
VPN types
SSL, IPSec
SSL, IPSec
SSL, IPsec
IPsec
IPsec
IPsec
Other ports
scanned
IM, P2P, VOIP
IM, P2P, VOIP
and many more
IM, P2P, VOIP
IM, P2P
IM, P2P, VOIP
VOIP, IM
Web application firewall
Minimal
Extensive
Minimal
Minimal
Extensive
Extensive
Authentication options
eDirectory, AD,
RADIUS, LDAP
RADIUS
RADIUS, LDAP, AD
RADIUS
RADIUS, LDAP,
SecurID
RADIUS, LDAP
Multiple concurrent admins
Yes
No***
Yes
No
Yes
No****
* The Juniper SSG box that we tested has room for six add-in cards
and came with two additional gigabit Ethernet NICs installed.
** Websense recently announced it will acquire SurfControl.
The other products are capable and about equal in this area.
Fortinet’s front page gives you just enough details to
monitor its overall operations. You can quickly find attack
summaries in its menus, and the policy definitions are easy
to set, and more importantly, easy to change when you have
done something wrong.
FIREWALL-IDS. Part of the usefulness of a UTM appliance is how its firewall and IDS work together, and flexibility
in terms of where it can be used across different configurations of an enterprise network. In other words, some products can position the IDS module outside of the firewall to
repel attacks and reject this traffic before it is processed any
further, or to work with an existing firewall infrastructure
at a headquarters network.
Fortinet and Astaro can also examine incoming encrypted packet streams and act on this analysis before passing
these streams through other modules, thereby saving on
processing power.
Check Point, Juniper, Fortinet and Astaro IDSes scan for
both attack signatures and attack behaviors. SonicWALL
only analyzes behaviors and IBM ISS only signatures. The
IDS modules of both IBM ISS and SonicWALL UTMs can
also explicitly detect outbound attack signatures.
The SonicWALL, IBM ISS and Juniper IDSes are hardwired to “live inside” the firewall, meaning that all network
packets from the outside world go first to the firewall and
then to the IDS for inspection. The advantage is that packets
57
I N F O R M AT I O N S E C U R I T Y
*** Available with separate Provider-1 product
**** Planned for version 4
are filtered out by the firewall, reducing the inspection burden on the IDS. However, you do lose some insights because
having the IDS outside the firewall can help you identify
attack vectors early. This may be fine for organizations
that manage both with the same administrative group, but
problematic if the administrative roles are split.
REPORTING. The products have varying methods for
producing reports, with different levels of details. All of
the vendors except Astaro sell separate reporting tools (not
evaluated for this review) that work across their larger security product lines. This assumes that you have more than
just UTM boxes from these vendors and want to consolidate reports so that all firewall information is in one place,
all IDS alerts are in another, and so forth. This may not
work for all usage scenarios, and could be cumbersome if
you have multiple vendors’ products in your data center.
Having to purchase add-on reporting tools somewhat
undercuts the purpose of having an integrated appliance.
Astaro includes reports as part of the Web administrative
interface and produces an “executive report,” which doesn’t
do much more than show some nice graphs of traffic flows.
LIVE MONITORING. We examined several critical pieces of
information available from the Web interface: real-time
CPU and memory load, current alerts of potential network
attacks, antivirus-related messages, and system health messages that required immediate attention.
This is helpful to see if your UTM box is overloaded or
IBM ISS’ antivirus status screen
shows protocols protected and traffic statistics.
mismatched with the particular network traffic and inspection loads.
All of the UTM products except Check Point and IBM
show the current CPU load and, in some cases, memory
consumption on the home page of their Web interface, so
it is easy to find and easier still to track. IBM ISS buries
its status screen, while you have to visit Check Point’s
SmartView Monitor (a separate piece of software that comes
as part of the UTM package) to get this information.
The three most useful front pages were from Astaro,
SonicWALL and Fortinet, which offer all sorts of helpful
summary information in one convenient place. Fortinet also
includes a secure command-line console window within its
Web interface, while the others require an SSH client to connect to their box if you need access to the command line.
SonicWALL also tells you if you have set up the box with a
known security weakness, such as allowing management
from the WAN interface.
Check Point uses Windows software for its management,
which means an admin must carry around a laptop with the
software installed, rather than simply logging in through
a browser. IBM ISS and Astaro can’t be managed through
Macintosh-based Firefox browsers, and we found some bugs
when we administered SonicWALL with Firefox on a Mac.
Antivirus statistics are very important, since few things
light up the help desk lines like email problems. IBM ISS
has a simple-to-understand antivirus status screen (See IBM
ISS screen shot, above), showing messages blocked, signatures, and which ports are being blocked or scanned. Astaro
also has a good summary display of its email traffic, but
tweaking the protection results requires visiting several different sub-menus. Check Point and Fortinet put this information on summary screens; Juniper and SonicWALL have
separate screens that summarize the virus penetrations.
SonicWALL and Fortinet clearly lead the pack in this
regard with the others scoring equally behind. Even if you
don’t activate all of the security modules, both vendors’
approach is easy to understand and provides just enough
feedback so as to not overwhelm an administrator.
There are two basic approaches to how security policies
are created:
• Integrated policy that applies to particular users or
network interfaces. This has its advantages if your UTM box
sits on several different network segments and you want to
deploy different policies by segment or by user group (for
example, one with servers on it, or one with engineering
users). With this method, an administrator sets one policy
that cuts across all of the individual security modules, with
specifics for antivirus, IDS and so forth. Call this the traditional firewall approach, and each policy can enable different security modules for particular situations.
Fortinet and Check Point use this approach; Fortinet
does a better job, setting up a series of four default protection policies that gives you a great starting point and examples that make it easy to modify them for your specific
needs (See Fortinet screen shot, below).
• Separate policies that are module-specific. This means
there will be one policy for antivirus, another for general
firewall tasks, and more for IDS actions. IBM ISS uses this
approach; while it also has chosen lots of defaults to get you
started, making modifications isn’t as easy as with Fortinet,
because you must make them in several places. Juniper also
sets up security policies by module.
The appropriateness for your company depends largely
on how you have structured your support staff. If you have
an antivirus person on staff, and you have a box that requires
adjusting antivirus policies in several different places, you
have a lot more maintenance work than with a box where
you can set these policies in a single place. However, your
security staff may wear a lot of different hats and thus this
might not be as much of an issue. It is really a matter of
taste and organizational structure.
3
AUTHENTICATION AND POLICIES
etting up and tuning security policies for the various
modules is at the core of these products. Ideally, you
would want an appliance that makes it easy to figure
out how to keep your network protected, but still allows
users room to get actual work done, all the while providing
feedback when you have too strong or too weak a policy.
S
Fortinet protection profiles
provide a good base that can be modified
for particular requirements.
I N F O R M AT I O N S E C U R I T Y
58
SonicWALL zones offer modulespecific protection policies.
SonicWALL and Astaro mix both approaches. Astaro
has policies that are based on application-layer protocols
(Web, email, IM and so forth) and has separate policies for
network layer events. This means that to make changes in
the UTM operations, you need to touch screens in both the
protocol section and the network interfaces. If you forget
one or the other, you will have configuration problems or,
worse yet, think you are protected when you aren’t.
SonicWALL policies are module-specific, and are applied
to particular network routes. That has a lot of appeal, and is
why we give it top marks here. All of its protection rules are
organized in a single section, and it is easy to apply them to
the appropriate interface (See SonicWALL screen shot, above).
Authentication capabilities are relevant if they are used
for remote VPN connections. For most site-to-site VPNs,
this isn’t important unless you want to do some rudimentary endpoint protection or create policies based on particular user groups or roles.
All of the products support RADIUS authentication;
Astaro and Fortinet can connect directly with Active Directory user store; Astaro also supports authenticating to
Novell’s eDirectory. Juniper can integrate with RSA’s SecurID
tokens directly.
All of the products offer IPsec VPNs, and Astaro, Check
Point and Fortinet support SSL VPN terminations. None
of the SSL modules has anywhere close to the level of
features that a standalone SSL VPN box would provide.
4
interface. The other products fall somewhere in between in
terms of complexity.
Each product uses different combinations of homegrown and third-party security services to round out its
UTM coverage. Astaro, Check Point and Juniper use
SurfControl for Web content filtering, while the others have
developed their own content-filtering capabilities. Astaro
uses Snort, while the others have their own IDS engines.
Astaro supplies three virus scanners—a proprietary one
using the Authentium antivirus engine, another based on
open-source Clam AntiVirus, and a PCI hardware-based
antivirus capability from Sensory Networks. Juniper uses
Kaspersky, and Check Point uses CA. IBM ISS uses Sophos,
along with a second scanning algorithm that examines network behavior. SonicWALL and Fortinet have their own
antivirus scanners.
The six products differ on how big a file attachment they
will scan through their antivirus engines. SonicWALL claims
an unlimited file size because it scans while streaming the
packets, while the others are more limiting because they
have to cache the files first. If performance bogs down, an
administrator can automatically block files beyond a certain
size. IBM ISS hides this setting in its advanced settings, while
the others make it easier to adjust the maximum limit.
All of the products can at least monitor IM traffic (See
Astaro screen shot, below, for example), and some have rudimentary mechanisms to (sometimes) block particular IM
protocols. SonicWALL was the only vendor that can completely block Google Talk and Skype conversations. Fortinet’s
IM protection is somewhat obscure. You have to go to two
different places, one to handle policies for individual users
and one to monitor or block the specific IM protocols.
Any solid defense against IM use will require combining
Web filters to block access to particular sites as well as using
the IM modules’ features.
Check Point also does some very extensive port scanning, including ports that are used for VOIP, IM and P2P
applications.
Web application scanning is absolutely essential if your
company’s Web servers are in remote locations or if you
FEATURE INTEGRATION
ur final series of tests looked at how the various
functional modules work together. We also determined the third-party suppliers for these modules
and what noteworthy features one product has that
the others do not.
We gave SonicWALL the top grade because of its superior
antivirus features, protection rule flexibility and implementation of IM protection across all of its security modules.
Juniper and IBM ISS scored lowest because of the difficulty in making changes to their protection rules. For example,
in order to implement protection or blocking of a specific
protocol, you have to hunt down the rules that apply to that
protocol and make adjustments in several places in the user
O
59
I N F O R M AT I O N S E C U R I T Y
Astaro presents a wide
variety of choices to allow, monitor
or block IM sessions.
R E P O RT CAR D
MAKING THE GRADE
Company, product
Astaro Internet
Security
Astaro Security
Gateway 320
www.astaro.com
Check Point
Software
UTM-1 2050
www.checkpoint.com
Fortinet
FortiGate-1000A
www.fortinet.com
IBM Internet
Security
Systems
Proventia Network
Multifunction
Security MX5010
www.iss.net
Price as configured
$12,465
$15,500
$14,995
$14,890
$18,375
$10,995
Enterprise management, control
B
B-
B
A
C
B-
Daily operations, reports
B
B
A
B
B-
B
Authentication and security
B
B
A
B
B
A
Feature integration
B
B+
B
C
C
A
Verdict
B
B
A-
B
C+
B+
Feature-rich, nice
Web interface,
relatively simple
to set up.
Pros: Numerous
antispam and
antivirus features;
authenticates to both
AD and eDirectory
Cons: Confusing
method of entering
security policies
Licensing is its
big weakness
Pros: Flexible and
extensive firewall
and IDS
Cons: Too many
individual pieces
of management
software
Flexible and capable
product
Pros: Solid
reporting and
security features
Cons: IM protection obscure
A complex product
made more so with
a bad interface
Pros: Feature-rich
Cons: Miserable
management interface and messy
menus will confound
configuration
Solid feature integration, reports and
security features
plan to set up a new Web server on an unprotected network
such as at a branch office. Check Point, SonicWALL and
Juniper offer protective mechanisms for preventing common Web application attacks such as SQL injection and
cross-site scripting. We didn’t find policy setting particularly straightforward for any of them.
The others just give lip service here, or require you to
spend your days writing firewall rule sets.
For additional features, we liked Check Point’s safe
upgrade, requiring an administrator to complete a successful login within a specified (and user-selected) period of
time; otherwise the box will roll back to a previous version.
SonicWALL allows management of its wireless access points
from its UTM device.
Not All Things to All People
UTM is one of those concepts that sounds great in theory,
but is messy in practice. The six products tested all had
their quirks, and we would have found show-stopping
issues on all of the boxes if we didn’t have a lot of support
from each vendor.
While Fortinet and SonicWALL clearly have the best
collection of features and Juniper the weakest, the others all
Easiest to set up
and manage; nice
arrangement of
commands and
policies
Pros: Default
settings are a
big help
Cons: Feature
integration needs
some work
Juniper
Networks
SSG 550
www.juniper.net
SonicWALL
SonicWALL Pro
5060c
www.sonicwall.com
Pros: Web application firewall features,
unlimited file attachment size for scans
Cons: Complex
configuration menus
and a tired user
interface
had their good points, and the differences among each of
the products is more a matter of taste and judgment than
anything else.
Weigh the ability for multiple people to manage these
boxes with how you organize your security staff. If you
have separate groups managing firewalls and antivirus, for
example, you might be better off choosing the products
that separate their security policies.
You will also want to examine how a UTM deployment
for your branch offices—which makes a lot of sense and
can reduce your overall support burden—will be balanced
with the products that you use or will use on your headquarters network. While Check Point and Juniper have solid
solutions for the headquarters, they have less satisfying and
less mature UTM product lines. Think carefully about what
functions and modules you want to consolidate, and how
you will go about managing the appliances before you invest
heavily in any solution.w
David Strom is a freelance writer, speaker and former
editor-in-chief of Tom’s Hardware and Network Computing
magazines. Send your comments on this article to
[email protected]
I N F O R M AT I O N S E C U R I T Y
60
P RODUC T
Reviews
FO R E N S I C S
Management and Features
P2 Enterprise
Shuttle
R E V I E W E D BY B R E N T H U STO N
Paraben
www.paraben.com
Price: $6,995
Paraben’s P2 Enterprise
Shuttle is a remote digital
forensic suite, allowing you
to remotely conduct undetected forensic tests on
Windows machines in your
network without taking the
machines offline.
This can be useful to
acquire the data without
raising suspicion of the target. It may also be used to monitor infected systems in
real time.
Installation
C+
Installation was pretty straightforward—a CD guide
walks you through it—but we did encounter some issues
after the installation. We were unable to get the proxy
functioning due to a misconfiguration.
The installation automatically filled in the IP address
to be used by the proxy and server with the hostname,
which did not seem to work. The proxy would not start
and did not really give a reason. We corrected the issue
by editing the config files and changing the hostname
to be the actual IP address.
The client agents can be installed directly or through
the Captain, which controls agents and acquires and
analyzes data from systems.
The latter allows you to place the agent without alerting the user or install agents on multiple machines.
Testing methodology: Server, Proxy and Captain were all installed
on the same system. Agents were installed on a variety of Windows
XP SP2 and Windows 2000 machines.
B+
There are four major parts to the enterprise suite: the
Agent, Captain, Proxy and Server. These modules interact with each other over a 128-bit encrypted channel.
The Paraben Agent is invisible to the user, although a
savvy user may suspect something by the increased CPU
load and network activity during acquisition. We were
also able to see it with a rootkit detector.
The GUI-based Captain has a tabbed and framed
design. Navigation is smooth, and buttons are easy to
figure out with contextual help.
The Paraben Proxy, naturally, acts as an encrypted
proxy between all of the components. It’s installed on a
system with an Internet connection
The Server is the main module, performing all
authentication and acting as the central repository for
acquired data. It verifies access permission for any actions
initiated by the Captain and Agent to provide increased
security. The Server should be installed on an isolated
and secured system with no direct Internet connection.
You will spend most of your time with the Captain,
which has quite a few tools to analyze clients. You can do
a forensic dump of data, copying over each file or directory, or perform deep system inspections while the system is running. You can view running processes, what
files those processes are accessing, and which registry keys
they have open. Other capabilities include capturing
screenshots, viewing the registry, processes, drivers and
network sessions, as well as viewing the files on the system.
You can create a full snapshot and save it to the database.
Reporting
B
Reporting functions are fairly simple and to the point.
Reports can be generated for module access such as
server/proxy connecting, login and logout of the server,
and agent connections. Each event is assigned a priority
(fatal, error, warning and information); you can filter
based on the event and the priority. Reports are generated
in a table inside of Captain GUI. There are no charts or
fancy graphics, but they’re definitely not needed here.
Reports can be saved to text, HTML or XML formats.
Verdict
Paraben’s P2 Enterprise Shuttle is a good offering if
you are looking for a remote forensics tool to use in a
Windows environment. It provides all the tools necessary
for a complete forensic analysis of a system, as well as
the security to ensure the integrity of the acquired data.w
I N F O R M AT I O N S E C U R I T Y
61
P RODUC T
Reviews
k
c
i
P
t
o
H
SECURITY TESTI NG
Core Impact 6.0
TIO®N
R MA
I N F OCU R IT Y
SE
R EVI EWE D BY M I K E P O O R
Core Security Technologies
www.coresecurity.com
Price: $25,000 for annual subscription
Prior to Core Impact, the vast majority of security penetration testers
would use “off-the-Web” exploit code,
after scouring an application code for
backdoors and covert channels. Core
Impact changed the security landscape
by providing stable, tested and trustworthy exploits for ethical hacking.
The latest release of this automated,
commercial-grade penetration-testing
software platform is an invaluable tool for professional
penetration testers and corporate security engineers.
Configuration/Management
A
Installing Core Impact 6.0 was a breeze—download,
double-click and enter the long string to decrypt the
installation executable.
There are two main workflows: The rapid penetration
test guides the user through the phases of reconnaissance, exploitation and reporting via a series of menudriven wizards. You can choose the type of exploits to
test, as well as the levels of risk to take (e.g., whether or
not to run exploits that might crash or DoS the service).
The second workflow is conducted via modules. In
this more granular mode, you choose from this version’s
plethora of available exploits.
Effectiveness
A
We were able to run multiple exploits, test and compromise machines in minutes, giving the attacker complete
command and control over our target systems and armTesting methodology: We used VMware virtualization software to
install a fully patched Windows XP Pro system to host Core Impact;
and a Windows 2000 Advanced Server system with a few service
packs missing to play the victim.
ing us with detailed information.
We first ran a rapid penetration test in which Core
Impact walks you through a simple set of questions to
identify the target systems. It scanned the network and
identified live hosts, listening ports and OS versions,
allowing us to choose the exploit modules most likely to
compromise the target.
We chose remote exploits first, attacking the Microsoft
Windows Plug and Play services umpnpmgr.dll vulnerability (Microsoft bulletin MS05-039). One click brings up
a quick description of the exploit and the vulnerability,
including links to patches and remediation information.
We then tested client-side exploits, switching to the
“Modules View” to select individual attacks. We ran the
IE IFRAME buffer-overflow exploit—which automatically sets up a Web server on the attack system, with
a Web page serving up the exploit—and browsed the
attack site to compromise our target system. Compromising a target using remote and client-side exploits
demonstrates the need to patch the vulnerable software.
One of the biggest benefits of running exploits against
real systems is gaining insight into how credible the threat
posed by vulnerabilities is in our environment.
At the end, Core Impact removes all the agents from
the target system, a step often neglected by inexperienced
penetration testers. Test activity is logged for review.
Reporting
B
Core Impact comes with a number of report generation
options, from a simple executive summary to a detailed
vulnerability report. The reports are simple and straightforward. The vulnerability report includes the number
of systems compromised, along with detailed information regarding the vulnerabilities exploited. Administrators could use this report to remediate the exposures
by following the remediation information and links to
patches.
Core Impact uses Crystal Reports, with an XMLbased generation system that can be altered and customized to meet your requirements. To take advantage
of this feature, however, you have to get under the hood
and change the XML templates.
Verdict
Core Impact 6.0 is an amazing tool to validate your
security posture. We highly recommend it to security
engineers to verify the vulnerability of their networks,
or confirm test results from third-party consultants.w
I N F O R M AT I O N S E C U R I T Y
62
P RODUC T
Reviews
P E N E TR ATI O N TE STI N G
Metasploit
Framework 3.0
R E V I E W E D B Y P E T E R G I A N N A C O P O U LO S
Metasploit LLC
www.metasploit.com
Price: Free
The Metasploit Framework is a platform for developing,
testing and executing exploit code for all popular Unix,
Linux and Windows platforms; it’s an essential tool for
the serious penetration tester or security professional.
The latest release only serves to further cement its formidable capabilities.
Installation is a breeze: download the appropriate
package (Windows or Linux) and execute. The Metasploit Framework can either be used via the familiar and
capable console interface or the much improved Web
interface. For a free product, it contains surprisingly
good documentation for users and developers, so it’s
fairly easy to get productive quickly.
Metasploit has a searchable database of more than
180 exploits, targeting multiple processor architectures
Testing methodology: We installed the Metasploit Framework console on a Windows XP SP2 and SUSE Linux 9.3 hosts with no hitches
and used both platforms to successfully exploit vulnerable versions of
Windows, Red Hat, SUSE and Fedora hosts.
and operating systems, with more than 70 payloads that
can be delivered to exploitable systems. Using the product is ridiculously simple: select the exploit via the Web
or console GUI, specify target, payload and options, and
run the exploit. It really is as easy as “point, click, own.”
The payloads range from simply binding a reverse
shell to injecting DLLs (like a VNC server) into the target’s memory space to uploading and executing scripts
or apps on the target. As if this isn’t enough, there are
also tools for building your own exploits, such as developing a NOOP sled to exploit a buffer overflow. Building
new exploits is essentially writing code, so you’ll need
to have Ruby development skills (some C experience
wouldn’t hurt either). This shouldn’t be a problem,
since almost all of Metasploit’s target audience will
have some ability in this area or work with someone
who does.
Exploits can be delivered either directly to the target
host, or via a chain of proxies, which are nice for obfuscating attacks. Additionally, various browser hijacking
routines will let you load malicious ActiveX controls
(either your own or some that are bundled with
Metasploit) to vulnerable Internet Explorer versions.
One way or another, you will be able to gain a foothold
in a vulnerable system and leverage it for greater access.
Determining whether or not an exploit succeeds
depends on the payload chosen. For example, if you
elect to bind a shell, Metasploit will open a console session and connect back to the host via the specified port
number.
Metasploit can continually update itself with the
latest exploits and payloads developed by its sizable
user community. Even if you don’t possess the deep
programming knowledge to make full use of its exploit
development capabilities, you’ll benefit from the work of
others and stay current as new exploits come online and
old ones are addressed by patches.
Metasploit isn’t a shrinkwrap port scan or vulnerability assessment tool for the casual user. It’s best to
think of the product as a development environment
akin to Visual Studio, but with a laser focus on developing usable exploit code. It is a serious pen tester’s delight,
but it’s also the sort of tool that gives security officers
nightmares, reinforcing the need for aggressive patching,
layered defense and encryption of data at rest.
Verdict
Metasploit Framework is a mandatory tool for every
security professional. This brief overview offers a
glimpse of its capabilities.w
I N F O R M AT I O N S E C U R I T Y
63
P RODUC T
Reviews
RISK/ POLICY MANAGE ME NT
Elemental
Security Platform
R E V I E W E D BY B R E N T H U STO N
Elemental Security
www.elementalsecurity.com
Price: Management server, $35,000; desktop agent,
$60; server agent, $600
The Elemental Security Platform (ESP) is a powerful tool
for monitoring and enforcing
system compliance, and provides effective asset management, asset-centric access
controls and risk management.
Since we reviewed Elemental Security’s version
1.1, then called Elemental Compliance System
(August 2005), the
product has matured
and extended its capabilities,
with support for new client OSes, risk management, support for ticketing systems and better LDAP
integration.
Setup
B+
Our ESP server was preconfigured, but Elemental typically sends an engineer on-site to install the device and
provide a rundown on features and usage. We were
impressed with the ease of client agent installation, and
getting the clients/servers running.
Agent installation simply requires giving it the
address of the ESP server and answering one or two
other questions, depending on the platform.
The client connects securely to the server, reports
gathered information and downloads relevant policies.
Testing methodology: The system we received was preconfigured
for our environment. In our tests we used a variety of OSes, including
Windows, Mac OS X and Red Hat Linux.
The server automatically gathers data about open ports
and services to categorize hosts, and places them in
groups that can be defined manually or imported from
LDAP.
As a key to risk assessment, the ESP server assigns a
value to the system, depending on what services it’s running. These values can be overwritten.
B+
Effectiveness
The user interface is clean and functional. Pages are uniform, with all dropdown menus on the left side, navigational buttons for selecting your page at the top, and
relevant page information, such as reports, or policies
you are creating, in the middle.
But it can still use a bit more tweaking. For instance,
during policy creation, if you click on rules for a closer
look, they open on the same page, so there’s no facility to
backtrack to where you were. So we had to hit the backspace button, which erased any rules we had already configured. You can right-click on the link and open a new
window to bypass that inconvenience.
ESP can be used as a basic asset inventory tool or a
granular asset-centric access control solution, depending
on policy. Policies can contain a variety of rules, from
packet filters, to whether the user can install a piece of
software, to rules that check for compliance with baselines (such as CIS, or HIPAA security requirements).
We defined some simple policies, such as denying
access to secured hosts by unsecured hosts (hosts not
running the agent), by naming the policy and adding
rules. Some rules require additional configuration, such
as ports for the network filters.
B+
Reporting
Reports can be created for any aspect of ESP for managers, and viewed on-demand or scheduled. You can view
reports for each policy, as well as specific host groups
under a policy.
The reports are easy to read and feature a variety of
graphs and charts to effectively represent the information. Data can be exported to a variety of formats,
including CSV and PDF.
Verdict
We are as impressed with the latest release of
Elemental Security’s tool for monitoring and continuously assessing the security posture of large, heterogeneous enterprises as we were with its early version.w
I N F O R M AT I O N S E C U R I T Y
64
P RODUC T
Reviews
P OLICY MANAG E M E NT
nCircle Configuration
Compliance Manager
R E V I E W E D BY B R E N T H U STO N
nCircle www.ncircle.com
Price: Management server, $20,000, plus $199
per monitored server, $199 per network device
and $49 per other assets
Riding herd on the integrity of
your security infrastructure is
not an option—it’s a necessity.
But keeping your IT assets in
compliance in a large, complex
environment is daunting without automated tools. nCircle
Configuration Compliance
Manager brings policy compliance and security management together into
one centralized suite, helping cut the job down to size.
Configuration/Installation
B+
Installation is straightforward, either on a single server
or multiple systems. nCircle’s agentless technology simplifies the process.
Vulnerability scanning is handled by third-party scanners, such as Nessus, IBM ISS Internet Scanner and
QualysGuard, allowing you to integrate existing tools.
nCircle pulls vulnerability data directly into its database
(we used MSDE, but larger organizations will want to use
SQL Server). Integrating Nessus in our lab was a snap.
Features and Interface
A
nCircle is a deep product with a lot of features. It finds
assets on your network, either actively or using its passive
detection technology, identifying hosts and determining
basic information, such as OS and open ports.
The efficient management console presents three
primary tabs to a paned interface. The changes pane
contains the aggregate of the latest alerts. The inventory
section lists all network assets, broken down by subnet
range. The compliance view is similar to inventory, but
adds columns for compliance with applicable policies,
both pass/fail and by percentage.
B+
Effectiveness
nCircle Configuration Compliance Manager assesses
systems against predefined policy, vulnerability assessments and several other options, such as spider-like
scans of Web servers, in response to events or at predefined times. It monitors files for activity, and tracks file
attributes, runs MD5 checks for integrity and issues
alerts when changes are detected.
We were impressed with nCircle’s asset inventory capability, identifying and collecting detailed information on
each system. For example, on Windows machines, nCircle
reported every piece of installed software, users, groups,
services running, shares available and updates installed.
For instance, it can check for the latest AV version or
unauthorized apps.
We configured nCircle Configuration Compliance
Manager to reactively scan a host with Nessus, and issue
a policy compliance check whenever target files were
modified. Many other events can trigger tasks, such as
finding new assets, or starting a task when a previous one
has finished. You can also schedule scans and set tasks by
single host or by group.
Compliance starts with predefined policies for various OSes. Creating policies from scratch can be daunting, but nCircle can automatically create policies from
“gold standard” machine configurations.
nCircle has included a PCI compliance policy, and
said it was planning to include HIPAA and SOX policies
in upcoming versions.
B+
Reporting
Reporting is critical for compliance tools, and nCircle’s is
thorough and easy to generate. Reports can include asset
and file changes, vulnerability reports, installed software,
risk trends, etc. They are easy to read and have colorful
charts as well as technical breakdowns. Reports can be
exported in .pdf, Excel, Crystal Reports, Word and RTF
formats. nCircle also integrates with the Remedy ticketing system.
Verdict
Testing methodology: We tested nCircle Configuration Compliance
Manager in our lab, including Windows (XP, Server 2003), Linux and
Solaris systems, and Cisco networking devices.
nCircle Configuration Compliance Manager is a
compelling package that rolls policy compliance
and vulnerability detection into one usable package.w
I N F O R M AT I O N S E C U R I T Y
65
P RODUC T
Reviews
ically according to standards you create for admindefined or platform-based groups of devices. When a
patch becomes available, you may create a rollout schedule based on your specific needs.
As noted above, tighter AD integration would allow
enterprises to leverage the time they’ve already invested
in creating an OU structure to apply policies to different
parts of the organization.
We were impressed with PatchLink’s ability to create
custom patch packages and deploy them to the enterprise or specific groups on a scheduled basis. These
packages can also be used to change configuration settings, install software and run automated scripts.
We also like the flexibility to grant end users varying
degrees of control over the PatchLink agent. For example, administrators may choose to allow users to delay
patch deployments and system reboots.
VU LN ERABILITY SCAN NI NG
PatchLink
Update 6.3
R E V I E W E D BY TO M B OW E R S
PatchLink
www.patchlink.com
Price: $1,495 per update server,
plus annual subscription fee
starting at $18 per client
Keeping systems patched can be a nightmare
for enterprises. Medium- and large-sized
enterprises will find this automated patching tool from PatchLink an excellent, costeffective solution.
Configuration/Management
B
PatchLink’s wizard-based process walks you through
installation, including MSDE if it (or SQL Server) isn’t
already present.
PatchLink Update provides a Web-based manual
installation process for one-off installs. It’s also possible
to automate installation by using a script or Group Policy
Object with the standalone installer. Agents can be automatically deployed through the console using Active
Directory/LDAP lookup or IP/DNS scanning.
One of the product’s major shortcomings is weak
integration with AD. Although you can deploy the agent
through AD, it’s not possible to manage devices through
AD organizational units (OUs) or import OU membership information into PatchLink’s group-based management structure.
Policy Control
B+
PatchLink Update grants administrators a great deal of
policy control, including the ability to schedule scans
and deployments.
PatchLink will enforce minimum baselines automat-
Effectiveness
A-
PatchLink Update is a robust and flexible automated
patch tool that will go a long way toward taking some of
the pain out of Patch Tuesday.
The Windows-centric product provides a baseline
level of support for other OSes, including Mac OS X and
several Unix/Linux variants. It also supports many
Windows applications, but only a handful of Mac apps
and none for *nix.
Determining which systems are unpatched and verifying successful deployments are major pain points for
enterprises. PatchLink uses digital signatures for each
patch and scans the host system to determine patch level.
If the initial patch fails, it attempts to redeploy it up to
three times.
Reporting
B
PatchLink has a wide range of reports on the status
of agents, enterprise-wide package compliance, patch
deployment status for systems/groups and the current
mandatory baseline.
Customization and filtering are limited. For example, you can filter your report based on devices or device
groups, but not on complex criteria such as creating a
report listing devices missing patches for a certain period
of time.
Verdict
Testing methodology: PatchLink Update was tested in a Windows
Server 2003 and Windows XP environment within VMware
Workstation. We ran the PatchLink Web server on IIS 6.0.
PatchLink Update 6.3 is a solid solution to the enterprise patch management problem and demonstrates
its true power in a Windows environment.w
I N F O R M AT I O N S E C U R I T Y
66
P RODUC T
Reviews
RISK MANAGEM ENT
RedSeal Security
Risk Manager
R EVI EWE D BY ADAM H O STETLE R
RedSeal Systems
www.redseal.net
Price: Starts at $25,000
Your network produces a flood of information that
could tell you where your business is at greatest risk. But
how do you sort through it all and determine exactly
how your critical assets are threatened?
RedSeal’s Security Risk Manager (SRM) enables security administrators to model and manage threats to
those corporate assets and network infrastructure. The
appliance transforms network device configurations,
vulnerability data and system value ratings into a graphical view that shows how systems can be compromised.
Setup and Configuration
B
Setup is fairly easy. We hooked up the serial cable, ran a
few configuration commands and installed the SRM Java
administrative application through our browser (only
Windows is supported).
SRM generates risk and threat maps based on
imported device configurations and vulnerability data.
SRM supports popular network devices out of the box,
including Cisco IOS, Cisco PIX5/6/7, Juniper ScreenOS
and Check Point Firewall-1/VPN-1 NGX, as well as vulnerability sources such as Nessus and QualysGuard.
Other devices can be imported with the help of RedSeal,
or by creating an XML schema. Device and vulnerability
data can be imported manually, or SRM can retrieve it
directly from the devices or a central repository through
a variety of means (FTP, SSH, HTTP/S, Telnet, CVS).
Effectiveness
A
information and makes it easy to import data or edit
device and system values. This clean interface carries
over into the network map, which appears quite haphazard at first, with systems appearing out of order and all
over the map. This is quickly alleviated by the autoarrangers, which are a great feature for larger networks.
The map shows connectivity between devices and networks, accounting for traffic flow restricted and allowed
by ACLs.
You can assign values (from 1-100) to systems to help
determine where your company is at greatest risk. SRM
uses this data to generate risk and threat maps.
The threat map is similar to the inventory map, but
includes threat calculations based on exposure and business value, modeling how an attacker might get to a system, and through which vulnerabilities. You can pick any
point in your network to see which systems can talk to
this system, or what systems your selected system can
see. A “heat box” style risk map shows which systems are
at greatest risks and establishes mitigation priorities.
The threat map showed us systems at risk, such as firewalls allowing improper traffic, or systems that had severe
vulnerabilities. After correcting the issues and reloading
the data, we could regenerate the maps and see that the
issues were mitigated. For instance, SRM showed that a
high-value internal database server could be attacked
from an FTP vulnerability on an external server. After the
issue with the FTP server was mitigated, SRM showed
that the database server was no longer threatened.
Reporting
B
Reports can be generated for a number of different categories, including inventory, network device configuration errors, exposure to vulnerabilities and performance
data of the appliance. The network configuration checks
are quite useful, as they compare your network devices
against a built-in rule set to check for common configuration errors. Some reports contain colorful heat bars, or
expanding bubbles to show threats, others just text.
SRM’s clean tabbed interface nicely displays available
Verdict
Testing methodology: We tested the RedSeal SRM appliance using
RedSeal-provided data that modeled a network containing a mixture
of network devices, and vulnerability data, in addition to data generated
in our lab.
RedSeal SRM is a very good tool to provide an overall
view of network threats and risks, and will help you
prioritize mitigation measures.w
I N F O R M AT I O N S E C U R I T Y
67
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement