A number of Cosmos validators joined the dYdX mainnet with the hope of attracting delegations from major stakeholders in the ecosystem. For many of these validators, this required both an investment on tokens (acquiring DYDX tokens in the market) and time/money in infrastructure.
The reality as of today is:
Zero support have been offered so far from majority token stakeholders.
Anonymous and previously unknown validators have materialized. It is impossible to tell if these are operated by the same entity and are thus introducing risk at the consensus layer.
Many of the aforementioned Cosmos validators will be pushed out of the validator set, or have been already.
For our own part, we have now exited the active set and are contemplating either:
Cutting our losses and liquidating the tokens (if possible)
Trying to support other Cosmos validators that are at risk of exiting the set.
Raising the issue here, hopefully resulting in a DAO-sponsored proposal to increase # of validator slots.
Snapshot December 6, 2023 showing chain can be halted by two anonymous validators. Increasing the number of active slots would also not affect the current block production as this is already controlled by the top 4 validators (2/3+ VP).
Totally supporting that proposal. Due to the mentioned risk of chain halting, increasing the number of validators in active set should solve that problem. This could enhance network security and provide a more inclusive environment for validators
We appreciate your dedication to the Cosmos ecosystem and hope for a constructive resolution.
It seems likely that DAO / treasury will need to come up with a delegation program to fix the VP skew. It is negative affecting block production, network reliability and the viability of operating a dYdX validator for many teams.
So short term, the proposal it will help retain low VP validators who will otherwise not be around to see this change.
the risk of chain halting is more to do with the vote concentration at the top.
increasing the number of validators won’t help with the vote concentration at the top (or halting/security).
the major concern (to me) is the lack of capital being staked (someone quoted ~3%).
having another 40 validators each with 20k tokens staked won’t really help.
Exactly my stance and speaking as a validator at the bottom of the set. In fact, it will:
Slow down the chain.
Increase costs for all validators in terms of storage and resources long term.
Not do anything to solve the VP concentration and low stake amount.
We are all closely monitoring the early days of the dYdX Chain.
A few weeks after launch, as a community, we can be proud of:
140+ validators (active & inactive) around the chain.
High performance of the chain.
Soon celebrating 1B trading volume
Reaching USDC 200k of trading fees distributed to validators and stakers.
? Can a newborn chain reach a similar state of validators distribution as older ones? No.
? Is the current state of the chain ideal? Not yet
? Do we have enough data around the ideal number of validators for optimal performance of the chain and optimal user experience? Not yet.
? Is the current state of the chain rapidly evolving? Yes
I am of the personal view that decentralization doesn’t need to be “all-or-nothing”.
With proper community planning, diligence and patience, we will decentralize further and to a pragmatic level over time.
I wouldn’t be surprised if the “top 25 validators” counts in the coming months centralized exchanges and large web3 wallets. “Ledger by Meria” illustrates this rising trend, jumping from inactive to ~ 20th rank in a matter of hours after launch. More to come, undoubtedly.
The foundation published its staking principles and will progressively deploy part of its treasury in the coming days/weeks. It will be a humble but opinionated contribution to our validator’s community. We plan to rotate our staking allocations too.
Liquid Staking conversations progress, and so are community treasury deployment.
? If the community have experience and ideas on how to attract additional delegations to good & responsible validators, we shall discuss and plan accordingly.
In the meantime, the dydX Foundation relentlessly communicates globally about the dYdX chain, its benefits and how to onboard & contribute. Our team guides custodians, exchanges, wallets on numerous integration programs. A healthy validator set is something we think deeply about and put lots of effort into.
Further readings, some dashboards are being built by grantees, validators and community members:
As a few people have mentioned already, there’s a pretty good chance that increasing the size of the valset (especially by this many validators) will result in slower block times which, for a chain like dYdX v4, could lead to issues with performance.
Because of this, there need to be some pretty compelling decentralization gains to justify a valset increase, imo. Notably, valset increases very infrequently (if ever) lead to such improvements at the top of the valset.
Here’s a thread I made analyzing this same question on Osmosis last year.
The findings were pretty compelling that (at least on Osmosis) validator set increases have diminishing returns on decentralization. Jumping straight from 60 validators to 100 is almost certainly not going to justify the additional TTC.
Would still encourage caution with any validator set increases, as it’s a bit of a slippery slope. When the set becomes too competitive again after the increase, we’ll likely see additional valset increase proposals. This continues over and over again, leading to degraded performance each time. The Cosmos Hub, for example, is now up to 180 validators, and there’s already been talk of increasing the size of the set again.
Before doing any valset size increases, I’d suggest establishing a framework for when and how this can be done, and ideally one that sets a hard upward cap on the number of validators that will be able to validate the chain, otherwise dYdX may end up in the same position in terms of ceaseless valset increase proposals!
Thanks for this Charles. I would urge the foundation to request from staking UIs to consider better staking interfaces that don’t focus on total VP. Stargaze chain has done this with extremely positive results and there have been no issues with validators going down (common argument to not having random listing’s). Reference: Stargaze
I also believe the draconian 30 day unbonding has an effect on staking, costing thousands in missed commissions plus a month of being locked out of using your assets. We should move to the standard 14 days and will make a thread to discuss it further.
