search

lkuper / CSE232-2020-10

A graduate course on distributed systems
https://decomposition.al/CSE232-2020-10/
11 stars 1 forks source link

11/9 Fundamentals of Fault-Tolerant Distributed Computing in Asynchronous Environments #15

Open hmadkhali opened 3 years ago

hmadkhali commented 3 years ago

The Scribes are @abhay3010
@Rajhi1988m @lakshmikrish-97 @versey-sherry @maxwellrbradley

lakshmikrish-97 commented 3 years ago

Hi @lkuper Lindsey, the first draft is ready for your review - thank you!

lkuper commented 3 years ago

A few comments:

Q1: looks OK.

Q2: OK, but I'm not sure what the discussion of UDP adds: what kind of fault tolerance is it meant to be an example of / what part of the question is it answering?

Q3: Example 1 is indeed something that was brought up in class, but it's not really in the spirit of what the question is about. When we speak of different processes observing the same computation in this question, we don't mean the same operation being executed twice, once at each of two nodes (which is what is happening in Example 1). What we mean is the execution of the entire system, which takes place only once, but different processes may make different observations of that single execution, such that a certain property is true in one process's observation, but false in another. There's also something to say here about the role of snapshot algorithms and how they can help if the property we want to detect is stable, but not if it's unstable.

versey-sherry commented 3 years ago

@lkuper Hey Lindsey, we deleted the UDP part in Q2 because we have decided that it was not relevant. We rewrote Q3 and Abhay prepared a diagram for the example. Please review the updated version, thank you!