[DRC] Reduce Active Validator Set to 50

[mev_council]

Summary

This proposal seeks to reduce the size of the dYdX active validator set to 50.

Abstract

Due to several validators winding down their operations in recent months, those ranked #51 and below in the active set now carry minimal stakeweight, collectively representing only around 0.02% of the total stakeweight. We have seen concerns raised about the inactivity and apparent disengagement of some of these validators from the dYdX protocol.

A previous proposal by Nethermind to reduce the active set from 60 to 30 validators was discussed extensively within the community. While it ultimately did not proceed to an on-chain vote due to the scale of the change, it sparked meaningful dialogue about validator performance and network efficiency.

Motivation

Validators ranked #51 and below collectively contribute just 0.14% of the total stakeweight. Each of them has a significantly lower stakeweight than the average (4.6M) and median (3.4M) stakeweight of active validators. Most of these validators also appear inactive or disengaged from the network.

With fewer validators, block and transaction propagation become faster, as messages need to reach fewer nodes, reducing overall network latency. Consensus finality also improves since Tendermint-based chains like dYdX require ⅔+ validators to prevote and precommit. Fewer validators mean quicker collection of these votes, resulting in more consistent and faster block finalization.

A smaller validator set reduces communication overhead, lowering the complexity and bandwidth demands of validator-to-validator messaging. This leads to smoother and more efficient network operations. For a trading-focused chain like dYdX, fewer validators also means fewer hops for orders to reach the proposer, accelerating order routing and improving the time-to-finality for trades, an essential factor in delivering a CEX-competitive trading experience. Finally, under periods of high load or congestion, a smaller, more coordinated validator set contributes to greater network stability, helping maintain predictable latency and throughput when it matters most.

Reducing the active validator set also enhances network security by removing inactive or poorly maintained nodes that may pose operational or attack risks. A smaller, more engaged validator group improves responsiveness during upgrades or incidents, reduces the attack surface, and simplifies coordination in critical situations.

Specification

We propose to reduce the active set to 50 by changing the max_validator parameter to 50.

Next Steps

We invite validators to provide feedback on this post, if there’s no significant objection, we will submit the on-chain proposal on Thursday, May 29, 2025.

[Govmos]

This seems like a logical step for the chain to reduce attack surface. We actually encourage the chain to consider further reduction in the near-term to better align with the chain’s economy. On behalf of the PRO Delegators’ community, we expect to cast a YES vote on this proposal and will support renewed discussions on the topic.

[ttt_vn]

Hello dYdX community.
Very reasonable. Validators ranked from 51 to 60 have little to no incentive to stay active and should be removed from the validator set to help reduce network latency.

Additionally, removing inactive or poorly maintained nodes enhances network security by reducing operational and attack risks. A more engaged and responsive validator set can react faster during upgrades or incidents, reduce the attack surface, and simplify coordination in critical situations.

[Winfred]

Stakecito supports this proposal.

Many of the validators ranked 51–60 show signs of operational disengagement: reflected in low stake weight, poor uptime, and limited participation in governance. Some are showing uptimes as low as 80%. There’s also a clear degradation in performance metrics from validator #51 onward, including a marked increase in missed blocks and reduced reliability.

Reducing the validator set to 50 improves the quality of the set, as well as performance and finality times across the network, and it also helps remove underperforming or inactive nodes that present unnecessary operational risk.

[cryptoplaza]

Dragon Stake supports this proposal.
Over the past year, the competitiveness to join the active validator set has significantly declined, largely due to the communicated changes in staking and the subsequent drop in validator profitability. As a result, we have observed a gradual natural reduction in active participation, particularly among the lowest-ranked validators.

As we anticipated, one of the consequences of these changes has been a lower demand for the token, which has negatively impacted the asset’s price. This underscores the importance of aligning validator incentives with the broader health of the ecosystem.

At this point, we believe it makes complete sense to reduce the size of the active set. Validators in the lower ranks hold only marginal stake, and maintaining their presence adds operational complexity without meaningful benefit to the protocol.

Reducing the validator set improves network latency and overall efficiency — a critical factor for a trading-focused chain like dYdX. A leaner, more engaged validator group will contribute to a more robust and responsive network, enhancing performance during both normal and high-load conditions.

We see this change as a logical and necessary step forward for the continued optimization of the protocol.

— Dragon Stake Team