Liberating Distributed Consensus - Heidi Howard
Reactive Summit4 HN recommenders · 2019–2020
7K views · talk · evergreen · Uploaded 2018-12-08
Heidi Howard's Reactive Summit 2018 talk on generalizing Paxos by weakening its original requirements. She shows Paxos is one point on a broad spectrum of consensus approaches.
HN recommends it as an illuminating, approachable explanation of distributed consensus.
42:03The Soul of Erlang and Elixir • Sasa Juric • GOTO 2019
GOTO Conferences
40:20"Testing Distributed Systems w/ Deterministic Simulation" by Will Wilson
Strange Loop Conference
49:15Erlang Factory 2014 - That's 'Billion' with a 'B': Scaling to the Next Level at WhatsApp
Erlang Solutions
47:43"Turning the database inside out with Apache Samza" by Martin Kleppmann
Strange Loop Conference
1:02:50The Language of the System - Rich Hickey
ClojureTV
51:27Google I/O 2012 - Go Concurrency Patterns
Google for Developers
51:36Conflict Resolution for Eventual Consistency • Martin Kleppmann • GOTO 2016
GOTO Conferences
41:06"Unison: a new distributed programming language" by Paul Chiusano
Strange Loop Conference
4 comments from Hacker News
Check out Heidi Howard's work for some nice generalisations and variations of Paxos. Her approach using immutable state makes things more tractable. [LINK] [VIDEO]
Heidi Howard is incredible and her work on distributed consensus is illuminating. I think she has actually successfully cracked the cookie of "making consensus easy". [VIDEO]
Tbh I think you got it backwards. CALM is actually very much about eventual consistency. The gist is basically, that distributed systems are only eventually consistent iff they are logically monotonic. Coordination as used in this article basically means "needs to be decided by an authority". But because of CALM we know that in order to elect an authority/leader, you need a consensus algorithm that actually has monotonicity at it's core, like raft. There's a really great talk on this by Heidi Howard: [VIDEO] I actually don't think that database isolation levels are a good way to talk about this stuff, because they are practical guarantees, not theoretical ones. Snapshot isolation for example also requires a leader. From a theoretical point of view there is no in between (at least none that I can see), either you are eventually consistent or you're not, either you are monotonic or you're not.
>A more recent development is Heidi Howard's work Glad someone has mentioned her here. This talk helped me greatly [VIDEO] .