280 likes | 388 Views
The Use of Honeynets to Detect Exploited Systems Across Large Enterprise Networks. John Levine*, Richard LaBella***, Henry Owen*, Didier Contis*, Brian Culver** *School of Electrical and Computer Engineering **Office of Information Technology Georgia Institute of Technology
E N D
The Use of Honeynets to Detect Exploited Systems Across Large Enterprise Networks John Levine*, Richard LaBella***, Henry Owen*, Didier Contis*, Brian Culver** *School of Electrical and Computer Engineering **Office of Information Technology Georgia Institute of Technology *** South Florida Honeynet Project
Agenda • Introduction • Establishment of the Honeynet on the Georgia Tech Campus • Exploitations Detected on the Georgia Tech Network • Lessons Learned • Conclusions and Further Recommendations
Introduction • Definition of a Honeynet • Concept of Data Capture and Data Control • Generation I vs. Generation II Honeynets • Description of the Georgia Tech Campus Network
Introduction • Current Vulnerabilities on the Internet • Current Tools to Protect Networks • Firewalls • Intrusion Detection Systems (IDS)
Shortcomings Associated with Firewalls 1. The firewall cannot protect against attacks that bypass it, such as a dial–in or dial-out capability. 2. The firewall at the network interface does not protect against internal threats. 3. The firewall cannot protect against the transfer of virus–laden files and programs
Shortcomings Associated with Intrusion Detection Systems (IDS) • Increase Complexity of Security Management of Network • High Level of False Positive and False Negative Alerts • Must Know Signature or Anomoly Detection Pattern
Definition of a Honeynet • Network Established Behind a Reverse Firewall • Captures All In-Bound and Out-Bound Traffic • Any Type of System • Network is Intended To Be Compromised • All Honeynet traffic is suspicious
Data Capture and Data Control • Data Capture • Collect all information entering and leaving the Honeynet covertly for future analysis • Data Control • Covertly protect other networks from being attacked and compromised by computers on the Honeynet
Generation I vs. Generation II • GEN I Honeynet • Simple Methodology, Limited Capability • Highly effective at detecting automated attacks • Use Reverse Firewall for Data Control • Can be fingerprinted by a skilled hacker • Runs at OSI Layer 3 • GEN II Honeynet • More Complex to Deploy and Maintain • Examine Outbound Data and make determination to block, pass, or modify data • Runs at OSI Layer 2
Georgia Tech Campus Network • 15000 Students, 5000 Staff, 69 Departments • 30000-35000 networked computers on campus • Average data throughput 600Mbps/4 terabytes per day • NO FIREWALL BETWEEN CAMPUS & INTERNET! • Why? Requirement for Academic Freedom, high throughput • However, individual enclaves within Georgia Tech use firewalls • IDS is run at campus gateway • Out of band monitoring and follow-on investigation
Establishment of the Honeynet on the Georgia Tech Campus • Established in Summer of 2002 • Uses Open Source Software • Initially Established As One Honeynet Machine behind the firewall • IP Address Range Provided by Georgia Tech Office of Information Technology (OIT)
Hardware and Software • No Requirement for State of the Art Equipment (Surplus Equipment) • No Production Systems • Minimum Traffic • Use Open Source Software (SNORT, Ethereal, MySQL DB, ACID) • Use Reverse Firewall Script Developed by Honeynet.org
Intrusion Detection System (IDS) • SNORT • Open Source • Signature-Based, with Anomaly-Based Plug-in Available • Can Write Customized Signatures • Run Two Separate SNORT Sessions • One Session to Check Against Signature Database • One Session to Capture All Inbound/Outbound Traffic
Logging and Review of Data • Honeynet Data is stored in two separate locations • Alert Data is stored in SQL database • Packet Capture Data is stored in a daily archive file • Data Analysis is a time consuming process In my Experience: • One hour/day to analyze traffic • One hour of attack traffic can result up to one week of analysis
Exploitations Detected on the Georgia Tech Honeynet • 36 possible exploited machines have been detected at Georgia Tech in previous 9 months (through June 2003) • A report is made to OIT on each suspected compromise
Identification of a System with a Compromised Password • Previously Compromised Honeynet Computer Continued to Operate as Warez Server • Another Georgia Tech Computer Connected to the Warez Server • Investigation Revealed that Password had been Compromised on Second Georgia Tech Computer
Detection of Worm Type Exploits • GEN I Honeynet Well-Suited to Detect Worm Type Exploits • Repeated Scans targeting specific ports • Analyze captured data for time lapses • Ability to Deploy Specific Operating System on Honeynet
Exploitation Pattern of Typical Internet Worm • Target Vulnerabilities on Specific Operating Systems • Localized Scanning to Propagate (Code Red) • 3/8 of time within same Class B (/16 network) • 1/2 of time within same Class A (/8 network) • 1/8 of time random address • Allows for Quick Infection Within Internal Networks with High Concentration of Vulnerable Hosts
Lessons Learned from Gen I Honeynet • Start Small • Maintain Good Relations with Enterprise Administrators • Focus on Attacks and Exploits Originating within your Enterprise Network • Don’t Publish Honeynet Addresses • Don’t Underestimate Time Requirement • Powerful Machines are NOT Necessary
Initial Observations of Gen II Honeynet • Configuration is more complex than Gen I • Must use variants of Linux 2.4 kernel in order to run Sebek keystroke logger capability • Data must continue to be monitored on a daily basis • Lessons Learned from Gen I Honeynet still apply
Conclusions and Further Recommendations • Honeynet Assists in Maintaining Network Security • Provides Platform for Research in Information Assurance and Intrusion Detection • Deployment of a Distributed Honeynet
Honeynet Portscan Activity • Date Public: 7/24/02 Date Attack: 1/25/03
Honeynet Portscan Activity • Date Public: 7/16/03 Date Attack: 8/11/03
Honeynet Portscan Activity • Date Public: 8/15/2003 Date Attack: 8/22/03