Consensus Node

Rewards

A total of 238.5 million tokens will be awarded in the first year, distributed evenly among all eras within the year.

The annual rewards will decrease in a stepped manner, with an annual decay rate of 0.841 (0.5 raised to the power of 4), halving every four years. For each era (24 hours in CESS), validators are rewarded proportionally based on their earned era points.

Era points are reward points earned through certain behaviors, such as producing non-uncle blocks, producing references to previously unreferenced uncle blocks, and producing referenced uncle blocks.

An uncle block is a relay chain block that is valid in all aspects but did not become the main block. This can occur when two or more validators produce blocks in the same slot, and one validator's block arrives at the next block producer before the others. We refer to the delayed blocks as uncle blocks.

Rewards are distributed at the end of each era. Regardless of their staked amount, all validators roughly share block production rewards. However, the rewards of specific validators may differ based on era points. While earning era points has a probabilistic component and may be slightly influenced by factors such as network connectivity, well-performing validators should have a similar sum of era points over a large number of eras.

Validators can also receive "tips" from transaction senders as an incentive for including transactions in their produced blocks. Validators receive 100% of these tips directly.

Penalties

Non-Responsive

If a validator fails to produce any blocks and does not send heartbeat signals during an era, it will be reported as non-responsive. Depending on the number of repeated violations and the non-responsiveness or offline status of other validators during the era, reduction penalties may be applied.

Validators should have robust network infrastructure to ensure node operation and minimize the risk of reduction or cooling penalties. High availability setups and backup nodes are recommended, and the backup node should be launched only when the primary node is verified as offline (to avoid double-signing and potential reduction penalties as described below).

Let x = offenders, n = total number of validators in the active set

Non-responsiveness reduction = min((3 * (x - (n / 10 + 1))) / n, 1) * 0.07

Ambiguity

GRANDPA Ambiguity

Validators sign two or more votes for different chains in the same round.

BABE Ambiguity

Validators produce two or more blocks in the same slot.

Reduction penalties

Let x = offenders, n = total number of validators in the active set

min( (3 * x / n )^2, 1)

Both GRANDPA and BABE ambiguities use the same formula to calculate reduction penalties

Validators can run nodes on multiple computers to ensure continued validation work in case one node fails. However, validators should exercise caution while setting up these nodes. Ambiguities may arise if they do not coordinate well in managing the signing machines, and the reduction rate for ambiguous offenses is higher than that for similar offline offenses.

If a validator is reported for any offense, it will be removed (cool-down) from the validator set and will not receive rewards during its absence. It will be immediately marked as inactive and needs to redeclare its validation intent.