Minimun Validators self-Stake
The recent proposal to update staking parameters (#162) was related to increasing the limits for the nominators. The reason that the community has opted to do this so far, is that the supply and demand of the number of nominator slots that Polkadot can serve are not aligned, and there are many more people who wish to nominate than the number of spots available.
The same limitation has always existed on the set of validator candidates of the network as well. Let us briefly recap: Having the intention to validate is fully permissionless and anyone is allowed to do so. Nonetheless, a limited (or better put: finite) number of spots are available for validator candidates to enter the NPoS election, and potentially become active validators of the network (and author blocks). The following figure illustrates this 3 step process:
To make sure this process happens smoothly, we propose introducing a minimum bond for validators as well. This bond will ensure that those validator spots (first column of the previous figure) cannot be filled trivially. Moreover, if they do become filled, the network’s token holders and governance can respond by increasing the aforementioned minimum validator bond, combined with the permissionless ‘chill_other’ transaction that can remove stakers who are below the corresponding stake limit.
Speaking of chilling, the important implication of this proposal is that if and when enacted, validator candidates who do not meet the minimum validation bond will also be susceptible to the chill_other transaction. Therefore, we highly recommend validators to react timely, and move a reasonable amount of funds to their self-stake.
Similarly, we recommend the network governance to make sure the majority of the validators have updated their stake, or at least have had the chance to be informed and react.
Lastly, as for the amount of the minimum bond, we propose a starting amount of 500 DOTs. Two notes should be emphasized about the proposed value:
- The Polkadot community has had multiple rounds of discussion about a fair minimum validators stake from an economical and incentives point of view. Nonetheless, we want to assert that the limit proposed here should be mostly interpreted from a technical point of view, and our primary concern for this proposal is to defend against potential sybil attacks on the validator candidacy process.
- This value is merely a starting point, to set a precedence. Keeping the previous technical aspect in mind, the governance of the chain can always decide to increase or decrease it as it sees fit, based on the demand and supply.
This is an open discussion for the entire community to participate. Particularly, I hope the validator community comes back to feedback before this is moved forward to actually being proposed to the governance of Polkadot.
Comments (9)
There are several "good intention" validators in the 1KV program that today don't fulfil the 500DOT requirement. Excluding them from the 1KV program could hurt decentralization of the network and the program reputation. I personally could see how the 500DOT requirement could be construed as an exclusion of lower income economies. I can therefore not support this proposal.
Hi @kianneigma ,
I am in agreement with the proposal and the minimum of 500 DOT. I would however like to clarify what happens when an active validator is chilled. Does their validator host machine still perform its duties to the chain in the background for the rest of the era? If that's the case then a chill is fine, if not, then we need to rethink the solution. I'm also a little concerned if a validator is chilled in epoch 6 for instance and he is elected to be an active validator in the next era.
Here's my thoughts, providing that chill_other is used on a validator:
If we are in era X, the validator will be chilled as at session 1 era X+1. If they are already selected as an active validator for era X+1, then allow them to validate for era X+1 and then be chilled as at session 1 in era X+2. Overall if they are due to be chilled as at session 1 of the next era then they should not be taken into consideration for the next era's election.