Here are the questions for the presentation on Monday:
Dynamo focuses particularily much on being always writeable, they even push conflict resolution to read operations. Why do you think that is? Eirik Iversen Mortensen
Can you elaborate on why it is that Dynamo uses the 99.9th percentile instead of just average? Ida Charlotte Rønningen
The measurements in the paper were done on a live system running a couple hunderd nodes. What challanges do you think they would face if they were to try scaling the system by, an order of magnitude or even more. Morten Grønnesby
Dynamo was not able to meet the SLA requirements of some applications with regard to latency. The paper mentions that one solution would be to set W to 1, this would obviously affect consistency and durability. During peak loads, how do you think this would affect read operations? Why? Adrian Arne Skogvold
Dynamo is said to scale incrementally, what might happen if multiple nodes fails sumultaneously, or if an administrator tries joining or leaving multiple nodes simultaneously? Ruben Andre Mæland
Hi @uit-inf-3203/students ,
Here are the questions for the presentation on Monday:
Bjørn