3. $ cat disclaimer.txt
“The opinions expressed in this presentation
are those of the speaker and do not reflect
those of past, present or future employers,
partners or customers”
5. Are You Ready?
• Most organizations are NOT prepared to
deal with security incidents
• If anything can go wrong, it will!
(Murphy’s law)
• Assigned internal resources?
6. Technical Issues
• Networks are complex
• Some components/knowledge are
outsourced
• Millions of daily events
• Lot of console/tools
• Lot of protocols/applications
7. Find the Differences
Aug 27 14:33:01 macosx ipfw: 12190 Deny TCP
192.168.13.1:2060 192.168.13.104:5000 in via en1
%PIX-3-313001: Denied ICMP type=11, code=0 from
192.168.30.2 on interface 2
8. Economic Issues
• “Time is money”
• Real-time operations
• Downtime has a huge financial impact
• Reduced staff & budget
• Happy shareholders
9. Legal Issues
• Compliance requirements
• Big names
• Initiated by the group or business
• Local laws
• Due diligence & due care
10. Belgian Example: CBFA
From a document published in April 2009:
“Tout établissement qui connecte son
infrastructure sur Internet dispose d’une politique
de sécurité qui tient compte de:
...
la création, l’archivage de fichier “historique
d’évènements” techniques adaptés à leur
analyse, leur suivi et leur reporting.”
13. Raw Material
• Your logs are belong to you
• If not stored internally (cloud,
outsourcing), claim access to them
• All applications/devices generate events
• Developers, you MUST generate GOOD
events
14. 3rd Party Sources
• Vulnerabilities Databases
• Blacklists (IP addresses, ASNs)
• “Physical” Data
• Geolocalization
• Badge readers
21. Correlation
• Generation of new events based on the
way other events occurred (based on their
logic, their time or recurrence)
• Correlation will be successful only of the
other layers are properly working
• Is a step to incident management
24. Let’s Kill Some Myths
• Big players do not always provide the best
solutions. A Formula-1 is touchy to drive!
• Why pay $$$ and use <10% of the
features? (the “Microsoft Office” effect)
• But even free softwares have costs!
• False sense of security
25. LM vs. SIEM
• A LM (“Log Management”) addresses the
lowest layers from the collection to
reporting.
• A SIEM (“Security Information & Event
Management”) adds the correlation layer
(and incidents management tools)
26. Grocery Shopping
• Compliance
• Suspicious activity
• Web applications monitoring
• Correlation
• Supported devices
• Buying a SIEM is a very specific project
28. Syslog Daemons
• Syslog is well implemented
• Lot of forked implementations
• syslogd, rsyslogd, syslog-ng
• Multiple sources
• Supports TLS, TCP
• Several tools exists to export to Syslog
(ex: SNARE)
29. SEC
• “Simple Event Correlation”
• Performs correlation of logs based on Perl
regex
• Produces new events, triggers scripts,
writes to files
32. Personal Researches
• Examples based on OSSEC!
• MySQL integrity audit
• USB stick detection in Windows
environments
• Detecting rogue access
• Mapping data on Google Maps
35. Conclusions
• The raw material is already yours!
• The amount of data cannot be reviewed
manually.
• Suspicious activity occurs below the radar.
• Stick to your requirements!
• It costs $$$ and HH:MM
• Make your logs more valuable via external
sources