Recall that answers to reading questions should be sent via email to remzi AT cs before 9am on the day of the class.
Tip #1: Don't spend time regurgitating obvious stuff from the papers! The point of these questions is to think , and then to write down what it is that you thought about. I read 20 to 30 write-ups, so the more interesting you are, the better! If you find yourself just repeating a lot of details from the paper, you are going down the wrong path.
Tip #2: Don't spend time criticizing how the authors wrote the paper. That is, I don't need to know whether you thought the paper was well written or not. We'll have plenty of time to talk about those types of things later in the class. Focus on technical aspects for these questions.
Questions for the Week (04/21)
Q15, 04/25: What trade-offs does GFS make to enable a simpler system design?
Questions for the Week (04/14)
Q14, 04/18: Write a small program that accesses files: it should perform much better on NFS than AFS. Now write another small program that accesses files: it should perform much better on AFS than NFS. What key properties of each system are you exploiting?
Questions for the Week (04/07)
Q13, 04/11: What is your biggest critique of the RPC system described in the paper?
Questions for the Week (03/31)
Q12, 04/02: What do the performance figures teach us in the scheduler activations paper?
Questions for the Week (03/24)
No Question for 3/28. Just read!
No Question for 3/26. Just read!
Questions for the Week (03/10)
Q12, 03/12: What are workloads that would perform poorly when using the Superpages approach described in the first paper? What are workloads that would cause ARC to have difficulty in the second paper?
Questions for the Week (03/03)
Q11, 03/05: What feature of Multics is found in modern systems?
Questions for the Week (02/24)
Q10, 02/28: What techniques from AutoRAID solve problems that are described in the original RAID paper? Describe.
Q9, 02/26: Why is OptFS faster than typical journaling file systems?
Questions for the Week (02/17)
Q8, 02/19: What types of workloads perform well on LFS? What types of workloads perform poorly? How could you design a workload to perform better on FFS than LFS?
Questions for the Week (02/10)
Q7, 02/14: No question today! Happy Valentine's day, and see you in class.
Q6, 02/12: The exokernel is different from a VMM (such as Disco) in some ways, and different in others. What are the important similarities? What are the important differences?
Questions for the Week (02/03)
Q5, 02/07: Disco is not able to run IRIX unmodified. What are some reasons? How could you solve these problems to enable Disco to run IRIX without recompilation?
Q4, 02/05: Mach is an early micro-kernel, whereas Nooks is a modern kind of micro-kernel. What aspects of Mach are found in Nooks? What aspects are missing? Do you think the micro-kernel approach to system design is important?
Questions for the Week (01/27)
Q2, 01/29: Part 1: What aspects of Atlas are still found in modern computing systems? Which aspects are not? Part 2: VMS memory management combines segmentating and paging in a novel way; what problems does this solve? What problems does this create?
Q3, 01/31: Part 1: Which policies and mechanisms within FFS make sense today? Which policies and mechanisms do not?
Questions for the Week (01/20)
Q1, 01/24: Part 1: THE and Nucleus present different philosophies for system design. What are they, and which is better? Part 2: why did UNIX become the dominant OS of its era?