- EIP-7514 is designed to slow down growth of the active validator set
- Limiting growth in the validator set size gives Ethereum community time to design solutions for validator set size
- Limiting validator set size is important for the roadmap (see single slot finality, SSF, for instance) and for P2P network load
- Ethereum is healthy and continues to function as intended
- There will be no impact to those currently running validators, likely negligible impact on potential validators (i.e., those wanting to enter the active validator set)
- This proposal represents an abundance of care and could be argued this is not currently necessary as Ethereum would continue to run as intended
What is EIP-7514?
The number of validators that can be activated during a particular epoch is restricted by the churn limit (see get_validator_churn_limit in specs here).
Currently, this limit changes as the number of active validators changes – it increases as more validators enter the active set and decreases as the number of active validators decreases. Specifically, the number steps up (and down) in multiples of CHURN_LIMIT_QUOTIENT, or 65,536 validators.
Today, with about 806,416 validators, the current churn limit per epoch is: 806,416 / 65,536 = 12.305 or 12 validators per epoch. This means that 2700 validators can be added (or exited) from the active set per day (225 epochs per day * 12 = 2700). When demand to enter the active set exceeds 12 validators per epoch a queue forms.
Currently the queue to enter the active validator set is about 7,795.
EIP-7514: Add Max Epoch Churn Limit proposes setting the max churn limit to 8. This has two effects: first, it reduces the number of validators that can be added to the active set and second, it no longer increases as more validators are added to the set.
Importantly, the change is asymmetric with respect to activations and exits. Currently, the same churn limit applies to both activations and exits. This proposal would just limit the activation side. In other words, inclusion of this proposal will introduce one churn limit for activations and another for exits. The churn limit used on the exit side will continue to fluctuate with the number of active validators.
This proposal will be included in the Deneb upgrade – likely to happen in Q4 of 2023.
Why is EIP7514 important?
Since the Shapella upgrade earlier this year we’ve seen the active set increase from about 558,606 at the end of March to over 800,000 today. This trend could very well continue.
A larger validator set is a good thing as it generally means more staked ETH, which implies greater security, as the cost to attack the network increases. However, a larger number of validators presents some potential challenges.
One such issue is that some validator duties become more onerous, like aggregating attestations. There is some evidence that Ethereum is still far away from facing serious challenges on this front. For example, the Holesky testnet will be launched with 1.46M validators and (so far) testing has been positive with this validator set size.
Beyond this, some aspects of the Ethereum roadmap such as single slot finality (SSF), are not practical at very large validator set sizes (>1M validators).
An incorrect take is that this somehow implies Ethereum is facing some sort of emergency situation or that this represents “hitting the panic button.” The core community (Ethereum Foundation and core developers), have decided to implement this change to slow down the rate of growth – effectively getting ahead of the issue, giving them time to incorporate changes to alleviate network load and make implementing aspects of the roadmap easier.
For more about the rationale of limiting the number of validators on Ethereum see Mike Neuder of the Ethereum Foundation’s proposal to increase MAX_EFFECTIVE_BALANCE, which is another proposal to help alleviate some of the issues above. Read our breakdown on the proposal to increase MAX_EFFECTIVE_BALANCE here.
What do Stakers & Potential Stakers Need to Know?
For those currently running validators on Ethereum this has no direct impact and will not affect the time required to exit the active set.
For those thinking about staking this might have an impact, but it could be negligible. How much of an impact this has will depend on how much demand we see to run a validator on Ethereum. As mentioned, there is currently a queue to activate validators on Ethereum. If this change were to be incorporated today it would slow the rate of activations by about 33%. While the activation queue has been decreasing for a while, it’s possible that the current queue will diminish to near zero in the next couple of weeks. If this is the case, then the proposal will have no effect.
The reality is that the effect will likely be somewhere in between. This should be viewed as a relatively minor change that will be incorporated in Deneb to manage the risk that the active set grows too quickly. The Ethereum core community is getting in front of a (potential) future issue. This represents an abundance of care that should be encouraged by the community.