Topic
|
Week
|
|
Presentations are generally based on discussing your hypothetical "company" in the context of required readings of the last week |
Week 2 - 4 |
|
The "Above the Clouds" paper (week 4) talks about a few obstacles to the growth of Cloud Computing. Take three obstacles from there, namely a) Availability of a Service, b) Performance Unpredictability, and c) Scaling Quickly and discuss if Eventual Consistency (second reading) can help in all, some, or none of these for your company |
Week 5 |
|
What exactly is the problem/pain that the "Semantic Approach" paper (week 5) is trying to address? List two things that you found most interesting in the paper (or argue why there is nothing interesting about their problem space and/or proposed solution) |
Week 6 |
|
What are the three most important trends in the online advertising industry from 1994 to 2004, as outlined in the "DoubleClick whitepaper" (week 6)? If you were to write a similar paper today, will you include a) all, b) some (which ones), or c) none of these trends in a 2011 whitepaper? If you could add one important, recent trend (post 2004) to this list, what would that be? |
Week 7 |
|
No presentations (Midterm Paper & Invited Talk) |
Week 8 |
|
Assume the cloud storage startup for media files. What do you think are the two most important design principles for your cloud storage architecture, out of the ones mentioned in "Server-Side Design Principles" paper (week 8) and why? Name one scalability challenge that you think is important, but cannot really be addressed by the design principles listed on the paper. |
Week 9 |
|
Same topic as last week (see above) because of invited lectures. |
Week 10 |
|
No presentations (we'll have a live demo/tutorial instead) |
Lecture of Mon 04/18 |
|
As CTO of a cloud media storage startup you need to decide between using Hadoop and DBMS. Part of your value-addition to enterprise clients (like NY Times) is the analytics you provide on their media content. Will you use Hadoop (MapReduce reading from week 10) or DBMS (Parallel DBMS reading from week 11)? Which argument(s) presented in either paper convinced you? |
Last Two Lectures |
|