PhD Seminar • Networks and Distributed Systems - Domino: Using Network Measurements to Reduce State Machine Replication Latency in WANsExport this event to calendar

Friday, July 24, 2020 1:30 PM EDT

Please note: This master’s thesis presentation will be given online.

Xinan Yan, PhD candidate
David R. Cheriton School of Computer Science

In this talk, I will introduce Domino, a low-latency State Machine Replication (SMR) protocol for Wide Area Networks (WANs). Domino uses network measurements to predict the expected arrival time of a client request to each of its replicas, and assigns a future timestamp to the request indicating when the last replica from the supermajority quorum should have received the request. With accurate arrival time predictions and in the absence of failures, Domino can always commit a request in a single network roundtrip using a Fast Paxos-like protocol by ordering the requests based on their timestamps.

Additionally, depending on the network geometry between the client and replica servers, a leader-based consensus protocol can have a lower commit latency than Fast Paxos even without conflicting requests. Domino supports both leader-based consensus and Fast Paxos-like consensus in different cycles of the same deployment. Each Domino client can independently choose which to use based on recent network measurement data to minimize the commit latency for its requests. Our experiments on Microsoft Azure show that Domino can achieve significantly lower commit latency than other consensus protocols, such as Mencius, Fast Paxos, and EPaxos.

To join this seminar from WebEx, please go to: https://uwaterloo.webex.com/meet/xinan.yan.

Location 
Online presentation
200 University Avenue West

Waterloo, ON N2L 3G1
Canada

S M T W T F S
31
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
1
2
3
  1. 2024 (96)
    1. April (19)
    2. March (27)
    3. February (25)
    4. January (25)
  2. 2023 (296)
    1. December (20)
    2. November (28)
    3. October (15)
    4. September (25)
    5. August (30)
    6. July (30)
    7. June (22)
    8. May (23)
    9. April (32)
    10. March (31)
    11. February (18)
    12. January (22)
  3. 2022 (245)
  4. 2021 (210)
  5. 2020 (217)
  6. 2019 (255)
  7. 2018 (217)
  8. 2017 (36)
  9. 2016 (21)
  10. 2015 (36)
  11. 2014 (33)
  12. 2013 (23)
  13. 2012 (4)
  14. 2011 (1)
  15. 2010 (1)
  16. 2009 (1)
  17. 2008 (1)