Proposal to increase the maximum num of validators

Proposer: Do Kwon / Founder of Terraform Labs
Date: 05/12/2021

I propose we increase the maximum num of the active validator set on the Terra Network from 100 to 130.

Reasons:

  • Having more validators is value accretive to the network.
  • Several of the new validators have projects tied to them (Orion, Angel, Flipside) and these need validators to be in the active validator set to redirect commission revenue as public sources of funding
  • The Cosmos Hub has been running with a comparable no of validators in the active set for many months without issues

This proposal will be submitted on-chain next week barring any major issues.

26 Likes

What are the downsides to doing so? Would it impact throughput significantly?

Hi Do,
Is there any plan from you to increase the number of bLuna / Anchor-side validators, or is it Lido’s responsibility to select or add more? It would be good for it to be broadened as the platform grows as only having 6 right now will make it a bit centralized.

1 Like

This would be lido’s domain - please feel free to reach out to them via lido.finance website

1 Like

No - Cosmos Hub validator set scaling without issues

It’s good to see that we’re going to see more validators

I can understand 130 is being proposed because Cosmos Hub has proven that it can be done, but is there any known limit or risk to increasing it even further? The terra ecosystem seems to be growing at a fast rate, which is nice, but with that in mind, I wonder if adding a bigger buffer wouldn’t just be better for agility.

That said, I’m not aware of the risks or potential limitations, so hence my question.

More validators should bring more decentralization and related security, but at the cost of a slower and expensive network but those effects of slowness and price are not perceptible with such a change. If network activity did a 10000x and validators in the 1000s maybe we notice something?

I might be in the minority here (and yes I run a validator) but I think this will affect the community negatively.

It takes about ~$12k/yr to run the hardware, even more if you take into account labor… sure it could be $5k … but it’s in that ball park.

For every thousand luna delegated you get you earn about $5/year per commission point.

so to even break even today, you need about 2million @ 1% commission or 100k @ 20% commission. There are already 30% of the validators who don’t have 100k delegates, so they are already doing it at a loss.

now … if there are more choices, then people will have more choice, and it means less delegations per validator… Their only choice is to for them to either leave, cut their costs, or raise their rates…

What we really need, in my opinion, is to find ways to decrease the divide between the top validator (who has 28 million luna delegated) and the bottom one (who has about 2,000), and find ways that validators can earn a living by doing this…

And I’m all for the new protocols having validators… As we saw with Orion, great ideas get luna staked to them, regardless of how many validators there are.

7 Likes

I think we can increase # of validators while also solving top-heavy consolidation issues by creating an upper limit bound on how much can be delegated to a given validator (e.g. cap at 5M LUNA). We also need validators to charge some % of commission, as this “0 commission race to the bottom” trap is nonsense for everyone. Validators undercutting other validators so everyone takes an L? We can do this better.

1 Like

Isn’t this all part of a free market? Sure 0% fees suck for validators and to many validators will dilute the staked Luna but all of that will balance out eventually, the network should do the changes that benefit the network, it’s not the network responsability to make every validator profitable, unless it really needs those validators.

All that said, an upper cap on staked Luna, carefully calibrated, could be a change for the better to promote further decentralization.