Topics & Readings

Midterms Returned
Final Exam Question Brainstorming

Class Notes

Format

Expect the final exam to include some of the (possibly tweaked) questions from the midterm along with some of the questions below.
A copy of the midterm questions is here.

Possible Questions

Describe m attacks and n defences. For each, say when they would be "suitable". (Potentially give a list of each to choose from.)
Describe a real-life example of functionality versus security. How did security compromise functionality? What attempt was made to minimize the impact of security? What alternatives would there be to reconcile this conflict?
Compare and contrast two similar security technologies (attacks or defences), emphasizing their relative strengths and weakness in the context of a specific thereat model. Be sure to describe the threat model and focus on specific implementations.
Explain the evolution of the reference monitor from Multics to current-day operating systems.
If you would design your own secure OS, what mechanisms would you include and why.
Alternately, what security mechanism would you add to the primary OS that you use on a daily basis. What threats(s) would this mechanism address and how effective would this mechanism be in thwarting those threats, given your usage patterns?
In the far, far dystopian future, users who fail to authenticate successfully to their home security system are fatally electrocuted. Describe possible attacks, defences, and how it would impact the end user.
Choose an offensive network security tool. Explain its functionality briefly and analyze its effectiveness in the context of a specific defensive scenario.
You are the new security officer of a law firm that has a number of clients who require great levels of discretion. As part of your initial audit, you discover that the authentication database was compromised some tie ago. The passwords were hashed "well" (e.g., bcrypt with many rounds). You run a password cracker (e.g., Hashcat) using the default settings for a week using a cluster of 1000 top of the line systems. You crack none of the passwords. You verify that the cracker is working properly. What do you tell management and why? Be sure to make your answer understandable to your bosses...otherwise you're going ot get fired.
Outline an attack strategy for a particular target. Be sure to consider all phases of the attack (getting in, accomplishing your goal, covering your tracks). Then, outline how you would defend against that attack strategy.
Aleph One outlined a classic buffer overflow atttack. What is one mechanism that prevents such naive attacks from succeeding on most current systems? How does it stop the attack? What is a way to circumvent that protection (and thus make the attack work again)?