|
@ -0,0 +1,47 @@ |
|
|
|
|
|
# Proposer selection procedure in Tendermint |
|
|
|
|
|
|
|
|
|
|
|
This document specifies the Proposer Selection Procedure that is used in Tendermint to choose a round proposer. |
|
|
|
|
|
As Tendermint is “leader-based protocol”, the proposer selection is critical for its correct functioning. |
|
|
|
|
|
Let denote with `proposer_p(h,r)` a process returned by the Proposer Selection Procedure at the process p, at height h |
|
|
|
|
|
and round r. Then the Proposer Selection procedure should fulfill the following properties: |
|
|
|
|
|
|
|
|
|
|
|
`Agreement`: Given a validator set V, and two honest validators, |
|
|
|
|
|
p and q, for each height h, and each round r, |
|
|
|
|
|
proposer_p(h,r) = proposer_q(h,r) |
|
|
|
|
|
|
|
|
|
|
|
`Liveness`: In every consecutive sequence of rounds of size K (K is system parameter), at least a |
|
|
|
|
|
single round has an honest proposer. |
|
|
|
|
|
|
|
|
|
|
|
`Fairness`: The proposer selection is proportional to the validator voting power, i.e., a validator with more |
|
|
|
|
|
voting power is selected more frequently, proportional to its power. More precisely, given a set of processes |
|
|
|
|
|
with the total voting power N, during a sequence of rounds of size N, every process is proposer in a number of rounds |
|
|
|
|
|
equal to its voting power. |
|
|
|
|
|
|
|
|
|
|
|
We now look at a few particular cases to understand better how fairness should be implemented. |
|
|
|
|
|
If we have 4 processes with the following voting power distribution (p0,4), (p1, 2), (p2, 3), (p3, 4) at some round r, |
|
|
|
|
|
we have the following sequence of proposer selections in the following rounds: |
|
|
|
|
|
|
|
|
|
|
|
`p0, p1, p2, p3, p0, p0, p1, p2, p3, p0, p0, p1, p2, p3, p0, p0, p1, p2, p3, p0, etc` |
|
|
|
|
|
|
|
|
|
|
|
Let consider now the following scenario where a total voting power of faulty processes is aggregated in a single process |
|
|
|
|
|
p0: (p0,3), (p1, 1), (p2, 1), (p3, 1), (p4, 1), (p5, 1), (p6, 1), (p7, 1). |
|
|
|
|
|
In this case the sequence of proposer selections looks like this: |
|
|
|
|
|
|
|
|
|
|
|
`p0, p1, p2, p3, p0, p4, p5, p6, p7, p0, p0, p1, p2, p3, p0, p4, p5, p6, p7, p0, etc` |
|
|
|
|
|
|
|
|
|
|
|
In this case, we see that a number of rounds coordinated by a faulty process is proportional to its voting power. |
|
|
|
|
|
We consider also the case where we have voting power uniformly distributed among processes, i.e., we have 10 processes |
|
|
|
|
|
each with voting power of 1. And let consider that there are 3 faulty processes with consecutive addresses, |
|
|
|
|
|
for example the first 3 processes are faulty. Then the sequence looks like this: |
|
|
|
|
|
|
|
|
|
|
|
`p0, p1, p2, p3, p4, p5, p6, p7, p8, p9, p0, p1, p2, p3, p4, p5, p6, p7, p8, p9, etc` |
|
|
|
|
|
|
|
|
|
|
|
In this case, we have 3 consecutive rounds with a faulty proposer. |
|
|
|
|
|
One special case we consider is the case where a single honest process p0 has most of the voting power, for example: |
|
|
|
|
|
(p0,100), (p1, 2), (p2, 3), (p3, 4). Then the sequence of proposer selection looks like this: |
|
|
|
|
|
|
|
|
|
|
|
p0, p0, p0, p0, p0, p0, p0, p0, p0, p0, p0, p0, p0, p1, p0, p0, p0, p0, p0, etc |
|
|
|
|
|
|
|
|
|
|
|
This basically means that almost all rounds have the same proposer. But in this case, the process p0 has anyway enough |
|
|
|
|
|
voting power to decide whatever he wants, so the fact that he coordinates almost all rounds seems correct. |
|
|
|
|
|
|