Live data from Hacker News

Consensus Is Harder Than It Looks (2020)

brooker.co.za

31–32 of 32 posts

Re: Consensus Is Harder Than It Looks (2020)

#31

If we can't make machines reach a consensus, what hope is there for human beings?

One of the things about Raft that bugs me is that it appears that the nominee counts the votes. And the leader is the only machine that approves transactions. If our favorite protocol is susceptible to first coups and then despots, then we have a long long way to go to model humans.

Re: Consensus Is Harder Than It Looks (2020)

#32
post #28

Earlier quoted context omitted.

Happy to mildly disagree! "RAFT uses a protocol that looks awfully like Paxos for election of a single leader" Agreed. "so it has nothing to do with multi-Paxos (as I understand it)" Except that Raft is not only a protocol for single leader election, it's also a replication protocol (see the "AppendEntries" message), and that's why it is Multi-Paxos. Multi-Paxos is just the category or classification for a family of…

I guess my issue here is calling all these "paxos" when 2 (VR and VS) predate Paxos. Anyway, did you check out Egalitarian Paxos? I think this protocol deserves more love, specially in WAN contexts (there is a ref. impl. on github): https://www.usenix.org/system/files/nsdip13-paper14.pdf

> I guess my issue here is calling all these "paxos" when 2 (VR and VS) predate Paxos.

Definitely. All VR needs now is a paper called "VR Made Famous" because it's already been made simple, easy, fast and understandable! And, after all, it was first.

Thanks for the reference to Egalitarian Paxos, that's a great example of active replication in the Paxos (and not Multi-Paxos) sense.

Post reply on HN