Intrusion detection 2001
- 1. Intrusion Detection CTO Forum November 9, 2001 Tom Casey Tcasey@pec.com 703.679.4900
- 2. 2 Agenda • Risks Associated with E-business • Elements of an Intrusion Detection Strategy • Misuse and Anomaly Detection • Application, Host, and Network Based Tools • Active and Passive Response • Intrusion Detection System Architecture • Technical and Legal Issues • Commercial and Open Source ID systems
- 3. 3 Reported Incidents Increasing Number of Incidents Reported 0 5000 10000 15000 20000 25000 30000 35000 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 *2001 Years NumberofIncidents *Q1-Q3 2001Cert.org Statistics October 15, 2001 34,754 21,756 2,412 132
- 4. 4 Risks Associated with E-business • Defaced Websites • Denial of Service/DDOS • Theft of Company Proprietary Information • Theft of Customer Information • Downtime = Loss of Revenue • Negative Press = Negative Public Image • Internal and External Threats
- 5. 5 History of Intrusion Detection • Intrusion Detection (ID) defined: – Process of monitoring computer networks and systems for violations of security policy • First ID System--manual “system audits” • 1980, ID was born – First document need for automated audit trail review to support security goals • Growth of Internet
- 6. 6 The Importance of Intrusion Detection • A perfectly secure system is a myth – Firewalls and filtering routers aren’t enough to protect electronic assets • Effective audit information analysis required a tool • An IDS is one of many components supporting a robust security architecture-”Defense in Depth” – Firewalls, VPN, Virus Protection, Vulnerability Assessments etc. • Protect valuable information resources from internal and external threats
- 7. 7 An IDS can accomplish the following • Prevents and/or mitigates the damage resulting from intrusion • Identifies a precursor of more serious activity • Identifies perpetrators • Discovers new attack patterns
- 8. 8 Elements of a Complete Intrusion Detection Strategy • Policy! – Policy is living, constantly evolving – ID configuration/design must support policy • Intrusion Detection System (IDS) architecture • Institutionalized Incident Response – Responses map to policy – Working with law enforcement – CERTs • Trained security personnel • Awareness Programs - Support from Users
- 9. 9 Time Line of an Attack Probing: •Port Sweeps •Address sweeps •Doorknob Ratting Break-in: •Operating System Bugs •Sniffed Passwords •Social Engineering •Back Door Malicious Actions: •Steal Data or Programs •Hop to other systems •Install Back Door •Setup Sniffer •Steal CPU time
- 10. 10 Misuse Detection • Misuse, signature/pattern-matching • Reliably detecting “known” use patterns • Detects only known intrusions • Difficult handling large volumes of data • Does not handle uncertainty
- 11. 11 Anomaly Detection • Anomaly Detection • Establish profile of “normal” user behavior • Patterns of abnormality, rare, unusual behavior • Accommodate adaptations to changes in user behavior • Statistical and Quantitative analysis • Assumes users exhibit predictable, consistent patterns of system usage
- 12. 12 Anomaly Detection (con’t.) User Normal Behavior Anomaly in User Behavior System Administrator Secretary Programmer •Log in as root •Edit user’s access permissions •Run system configuration/ monitoring tools •Logged in locally during company working hours •Uses office automation software (word processing, etc) •Reads and sends emails •Logged in from early morning to late night •Uses software development tools •Browses Internet more often in the evening then the daytime •Becomes a programmer •Accesses Software Development tools •Accesses Software project sources •Logs in from a remote host •Assumes the role of a manager •Logs in as a human resources manager •Gains access to personnel database
- 13. 13 Intrusion Detection Tools • Application-based – Collects information and detects intrusion at the application layer – Placement: E-commerce Server, WebServer • Host-based – Agent software on host – Monitors: event logs, critical system files, registry settings, etc – Alerts management console, reacts actively and/or passively depending upon policy • Network-based – Operates at the network level – Detects DOS or dangerous payloads before the reach destination – Dedicated host, two interfaces: Management and Stealth
- 14. 14 Active Responses • User driven • Automatic Responses • System takes action to block the progress of attack – Closing holes, shutting down services, logging an intruder – Block IP address(es) • Collect more information (honey pots)
- 15. 15 Passive Responses • System logs and reports problem • Alarms and notification – visual, audible, email paper • SNMP traps • Archiving and reporting
- 16. 16 IDS Architecture Recommendations • Network based – At Internet connection points – Key internal network segments – In the DMZ – Just inside the Firewall (Intranet) – Behind WAP server, WAN router, modem pool • Host-based – Servers containing critical data – Domain servers • Optimum Architecture: Combine misuse and anomaly detection
- 17. 17 Sample IDS Architecture Firewall Internet Router Web Server(S) DMZ Services Email Relay Border Directory Host IDS Agent Domain Controller Personnel Database User Workstations User Workstations IDS Central Management Console Network Sensor Network Sensor Network Sensor User WorkstationsStealth Mode Customer Database Corporate Private Network Web Server(s) File and Print Server
- 18. 18 Technical Issues • Scalability – Scaling over space as the network grows • Management – Network Management – Sensor Controls – Investigative Support – Performance Loads – User Interface • Reliability – Quality of analysis engines – Response mechanisms
- 19. 19 Technical Issues (con’t) • Analysis – Difficulties categorizing attacks/threats – False positives/negatives (tuning anomaly detection engines) – Trend analysis, event correlation, data mining • Interoperability – Tools to collect information from: multiple abstraction layers, hardware, software – Audit trail standards • Integration – Intrusion detection in a Switched Environment – Intrusion detection in a Crypto Environment
- 20. 20 Legal Issues • Legislation – Computer fraud and abuse statutes – Electronic Communications Privacy Act Sec 2510 • System logs are circumstantial evidence – Requires proof of authenticity – Testimony of responsible parties – Expert to explain log file contents – Maintaining redundant event log records • Electronic Monitoring – System admin monitoring vs. Law enforcement monitoring • Cyber Forensics
- 21. 21 Commercial and Open Source • Leading Commercial Vendors – Internet Security Systems (ISS): RealSecure – NetworkICE: BlackICE – Enterasys System: Dragon – Cisco Secure Systems: IDS – NFR: Network Intrusion Detection • Open Source – Snort.org • Managed Security Providers (MSPs) – Leverage the MSPs’ security expertise – Ideal for Small/Mid-sized business – Leverage MSP experience with other customers – Focus your staff and resources on your core business activities – 24X7X365 Monitoring and Notification
- 22. 22 Current and Future Trends in IDS • Protocol Scanners • “Meta” Detection – Interoperability – Centralized Administration, Management, and Reporting • IDS Appliances – No general purpose OSes to configure and maintain – No patches/Drivers to install – Facilitates: accuracy, speed, and remote management – 100 Gigabit Detection
- 23. 23 References • Internet Security Systems: www.iss.net • Enterasys Networks: www.enterasys.com • Cisco Systems: www.cisco.com • Snort: www.snort.org • NFR Security www.nfr.com • CERT @ Carnegie Melon: www.cert.org • Sans Institute: “The Twenty Most Critical Internet Security Vulnerabilities” http://www.sans.org/top20.htm • Computer Security Institute: "2001 Computer Crime and Security Survey" http://www.gocsi.com/prelea/000321.html