TOTEM: Threat Observation, Tracking, and Evaluation Model National Laboratories Information Technology Summit Oak Ridge, TN June 1, 2009.

TOTEM: Threat Observation, Tracking, and Evaluation Model

National Laboratories Information Technology Summit

Oak Ridge, TNJune 1, 2009

2 Managed by UT-Battellefor the Department of Energy

TOTEM: Threat Observation, Tracking, and Evaluation Model

John J. GerberCISSP, GCFA, GCIH, GISP, GSNA

Mark A FloydCISSP, GCFA

“A totem is any supposed entity that watches over or assists a group of people, such as a family, clan, or tribe.”-- Merriam-Webster


TOTEM: Basic Idea

• TOTEM?• Who Are You Guys? • Why Should Anyone Care? • How the ANL Federated IDS Data

Sharing Model Can Help.• Possible Problems.• Existing

Methodologies/Frameworks.• Blended to Create TOTEM.• TOTEM at ORNL.• Screen Shots.• Future Development.


What is TOTEM?

“Totemism : system of belief in which humans are said to have kinship or a mystical relationship with a spirit-being, such as an animal or plant. The entity, or totem, is thought to interact with a given kin group or an individual and to serve as their emblem or symbol.”-- Encyclopædia Britannica

The idea behind TOTEM is simple: • Pick up where the ANL model stops.• Compare threat information from sources such as the federated

model and other watchlists (DShield, Emerging Threats, SenderBase, etc.).

• As new threat information and activity sources are added, a better evaluation can be rendered.

• Use components from the individual site for evaluating risk.• Information is gathered and visualization provided.


Who Are You Guys?

We are like dwarfs standing upon the shoulders of giants, and so able to see more and see farther than the ancients. – Bernard of Chartres Setting an example is not the main means of influencing another, it is the only means. – Albert Einstein


“Danger, Will Robinson!”

According to a May 6th Wall Street Journal article, the Pentagon confirmed that it detected 360 million attempts to penetrate its networks in 2008, which is up from six million in 2006.

The Department of Defense also disclosed that it had spent $100 million in the past six months repairing damage from these cyber attacks.

(04/09/2009) Electricity Grid in U.S. Penetrated By Spies reported in The Wall Street Journal. Under the Bush administration, Congress approved $17 billion in secret funds to protect government networks.

(05/09/2009) FAA's Web Security Audit: 3,857 Vulnerabilities security audit of the Web applications found 763 high risk, 504 medium risk, and 2,590 low risk vulnerabilities.

(04/21/2009) Computer Spies Breach Fighter-Jet Project reported in The Wall Street Journal.Cyber spies have stolen tens of terabytes of design data on the US's most expensive costliest weapons system -- the $300 billion Joint Strike Fighter project.

(05/2009) Inspector General report sent to the FAA - Last year, hackers took control of FAA critical network servers and could have shut them down, which would have seriously disrupted the agency's mission-support network.

(05/20/2009) NARA suffers data breach reported in Federal Computer Week - the missing drive contains 1T of data with "more than 100,000 Social Security numbers (including Al Gore’s daughter), contact information (including addresses) for various Clinton administration officials, Secret Service and White House operating procedures, event logs, social gathering logs, political records and other highly sensitive information.


It is a Dangerous World“IDSs have failed to provide value relative to its costs and will be obsolete by 2005.” -- Richard Stiennon, Gartner Analyst, 06/03

"The worldwide wireless LAN (WLAN) intrusion prevention system (IPS) market is on pace to reach $168 million in 2008, a 41 percent increase from 2007 revenue of $119 million, according to Gartner, Inc." -- Gartner Press Release, 09/18/2008

http://taosecurity.blogspot.com


Detection

Key Points• 4 percent of incidents were detected through event monitoring and

other forms of analytic technologies.• 82 percent of the cases, victim possessed the ability to discover

the breach had they been more diligent in monitoring and analyzing.

• Organizations lack fully proceduralized regimen for collecting, analyzing, and reporting on anomalous log activity.


ANL Federated IDS Data Sharing ModelBasic Idea: an incident at one location can be a precursor to an attack on another similar location. Current Members • Argonne National Laboratory (ANL) • National Center for Supercomputing Applications (NCSA)• Los Alamos National Laboratory (LANL) • Lawrence Berkeley National Laboratory (LBNL)• Oak Ridge National Laboratory (ORNL) • U.S. Computer Emergency Readiness Team/DHS

(USCERT) • Thomas Jefferson National Accelerator Facility (JLAB) • Brookhaven National Laboratory (BNL) • Sandia National Laboratories (SNL) • Idaho National Laboratory (INL) • Fermi National Laboratory (FNAL)• National Energy Research Scientific Computing Center

(NERSC)• Pacific Northwest National Laboratory (PNNL)


ANL Federated IDS Data Sharing Model (2)






ANL Federated IDS Data Sharing ModelAdditional Info

"Federated Defenses and Watching Each Others' Backs" by Scott Pinkerton, ANL. Tuesday, 11:00-11:45am.


Violent Felons in Large Urban Counties

A majority (56%) of violent felons had a prior conviction record. Thirty-eight percent had a prior felony conviction and 15% had a previous conviction for a violent felony.


The More Sources the Better?

• DNS-DB Malware Domain Blocklist maintains a list of domains, pulled from various sources, that are known to be used to propagate malware and spyware.

• Global Watchlist pulls the list of suspected malicious IPs/Net ranges from different sources such as SANS DShield, Arbor atlas and so forth, then putting all of them in one place.

• Ninja Chimp Strike Force provides a compiled list of hosts associated with bruteforce attempts, spam, botnets, etc. The list is comprised of data from Arbor Networks, Project Honeypot, Shadowserver, and about 24+ hosts. It is sorted on an hourly basis to keep information current and is consistently changing.


Cooperative Protection Program (CPP)

Purpose

• Define, integrate, deploy and operate sensors to collect high quality, information rich network data

• Data analysis targeted at cyber adversaries and their activities against DOE

• Detect and deter hostile activities directed at the Department’s information assets

• Generate summary and alert information about boundary-crossing Internet traffic at DOE sites


Problems

• An incident at one location can be a precursor to an attack on another similar location.

• Limited ACLs. • False positives.• All sites are not created equal.• Mistakes happen. • Politics.


Trust Management

Nicolas Luhman [1] defines trust management as: a tool allowing our systems to keep working even if assumption of cooperation doesn't hold. Bernard Baber [2] formulates trust as an expectation about the future, citing three fundamental meanings of trust:

1. Expectation of the persistence and fulfillment of the natural and moral social order.

2. Expectation of technically competent role performance from those we interact with in social relationships and systems.

3. Expectation that partners in interaction will carry out their fiduciary obligations and responsibilities (place other's interests before their own).


Trust and Reputation Modeling Techniques

Need: specialized knowledge structures used to predict the reliability of trusting agent's partners in the future interaction using the past experience of interactions with the trustees.

Examples• Feedback mechanisms used by online auction sites

(ex: eBay).• User ranking systems used by social networking.


Dilbert and Albert Einstein


CAMNEP: System ArchitectureSystem developed by Martin Rehak.


CAMNEP: Multi-Source Trustfulness Integration


CAMNEP: Agent Specific Clusters


CAMNEP: Reporting


CAMNEP: Conclusions


Risk

NIST publication SP 800-30: Risk Management Guide for Information Technology Systems. In the text we read:

"Risk is a function of the likelihood of a given threat-source's exercising a particular potential vulnerability, and the resulting impact of that adverse event on the organization. To determine the likelihood of a future adverse event, threats to an IT system must be analyzed in conjunction with the potential vulnerabilities and the controls in place for the IT system.“

"Vulnerability: A flaw or weakness in system security procedures, design, implementation, or internal controls that could be exercised (accidentally triggered or intentionally exploited) and result in a security breach or a violation of the system's security policy."


Topological Vulnerability Analysis (TVA) Approach

Steven Noel, Matthew Elder, Sushil Jajodia, Pramod Kalapa, Scott O'Hare, Kenneth Prole

Basic idea: analyze and visualize vulnerability dependencies and attack paths for understanding overall security posture.

Populate through automated network discovery, asset management, and vulnerability reporting technology.


Operating with Limited DataSeeing the forest through the trees.


Creating TOTEM

Network Capture• Nessus/ISS - VSWeb• NAC• FRAMS• Device Exception System (DES) • Network Registration System • Proxy logs • Splunk/log aggregators• Flow logs• Time Machine

Vulnerability Database• National Vulnerability Database

(NVD)• The Open Source Vulnerability

Database (OSVDB)• Emerging Threat • SANS Internet Storm Center (IC)

Exploit Conditions• IDS/IPS - Snort and Bro

Attack Scenario (Threat)• Federated Model IPs• DNS-DB Malware Domain

Blocklist• Global Watchlist• Ninja Chimp Strike Force


TOTEM: What is the Point?

How does one effectively distinguish false positives from actual threats?

The answer may only be visible by looking at multiple sources with different levels of trust and doing a little aggregation and anomaly detection. Our goal is to create attack road maps with weights/prioritizations in order to manage the possible risks.


TOTEM Analysis

Trust model defined• Past and current traffic• Traffic patterns to hosts• Traffic volume to hosts

Evaluation Engine• Traffic acquisition and data

processing layer• Cooperative threat

detection layer• Operator and analyst

interface layer

Data is processed in stages• Anomaly detection• Trust update• Collective trust estimation• Method integration• History integration


Creating TOTEM: Federated Model The devil is in the details

Classic LAMP System• Linux• Apache• MySQL• Perl

Additional Software• GPG• GeoIP • Graphviz• Request Tracker• ModSecurity


Information Shared by the Federated IDS Data Sharing Model

• Strictly unclassified information• Information on (usually external) IP addresses that

was malicious enough to warrant a site response (blocking or other)o IP address:tcp/udp port #o Time of attacko Type of attacko Exploit attemptedo Severity of attacko Previous history of offending IP at that site

(corporate memory)o We could periodically share watch lists

• Information presented in a standardized exchanged formato Small XML fileo Using IETD standards for cyber data exchange


Other Blacklists Provide Information

# watchlist.security.org.my, contact [email protected]# ip/net, source, comment, name, last update (GMT+8)202.99.11.99, http://www.dshield.org/ipsascii.html, Dshield: Top IPs, dshield-top-ips, 2009/05/1395.215.76.0/22, www.spamhaus.org/drop/drop.lasso, Spamhaus Block List, spamhaus, 2009/05/13114.80.67.30, www.emergingthreats.net/rules/bleeding-rbn.rules, ET RBN, rbn, 2009/05/13 122.1.21.148, www.emergingthreats.net/rules/bleeding-compromised.rules, ET, compromised,

# domain type original_reference-why_it_was_listed note--pound sign=comment# notice notice duplication is not permitted00.devoid.us malware www.cyber-ta.org/malware-analysis/DNS.Cumulative.Summary 20090321scan4lux.info fake_antivirus www.malwaredomainlist.com/update.php 20090505junglemix.in phishing isc.sans.org/diary.html?storyid=6328 20090505

Wed May 13 07:59:03 CDT 2009

99.254.50.13999.248.26.17799.245.29.3899.234.219.183


Other Blacklists Provide Information (2)

Top 10 Blacklist Providers

Using 266 IPs from malware.Using 235 IPs from rbn.Using 172 IPs from coolwebsearch and spamhaus.Using 55 IPs from rogue.Using 23 IPs from malspam.Using 20 IPs from dshield-top-blocks.Using 15 IPs from exploit and sql_injection.Using 13 IPs from spyware and trojan.Using 11 IPs from rogue_antivirus.Using 10 IPs from botnet.

Total Blacklisted IPs Downloaded: 1214

Blacklisted IPs Added Today: 39


Sample Reports: Blacklist

1. Denotes IPs that are blacklisted by the Internet community more recent than 2009-05-11 17:12:07.

3. Denotes IPs that was blocked by the DOE Federated Community more recent than 2009-05-11 17:12:07.

4. Denotes IPs that was blocked by the DOE Federated Community prior to 2009-05-11 17:12:07.


Sample Reports: Blacklist (2)


Signature Based Information Can be Useful

In respect to Snort, we have been looking at trend information for awhile.


Sample Reports: Blacklist (3)


Sample Reports: ORNL Shuns

1 Denotes IPs that are blacklisted by the Internet community more recent than 2009-05-11 18:02:07.

4 Denotes IPs that was blocked by the DOE Federated Community prior to 2009-05-11 18:02:07.


Sample Reports: ORNL Shuns (2)






There is a great deal of work yet to be done. Some key areas to develop will be:

• Additional work on the evaluation engine.• Improved visualization.• CPP.• ICSI Bro.• ICSI Time Machine. • Integration with Request Tracker (RT).


Comments

Seriously, we would appreciate any comments. After the presentation, please feel free to contact us.

Mark Floyd John Gerber [email protected] [email protected]

Mark [email protected]

John [email protected]

Thank you for the opportunity to discuss TOTEM.

TOTEM: Threat Observation, Tracking, and Evaluation Model National Laboratories Information Technology Summit Oak Ridge, TN June 1, 2009.

Documents

department of energytotem

department of energywhat

department of energydanger

department of energywho

department of defense

data breach

new threat information

threat observation