Discussion: reducing BLS keys

IMO shard 0 needs to be incentivized until the resharding. Maybe just a slight bump in rewards to keep keys from leaving for shards 1-3. IDK how difficult that would be to do.

3 Likes

sorry to side track, but may I ask why are validators avoiding shard 0?

1 Like

Not able to attend sorry! But would like to view the recording if possible. Or recap. Thanks!

If I understood correctly, the shard 0 is slowing overtime, and by doing so, the other shards are generating more rewards, since the epoch time is defined by the shard 0, and the other shards are still generating rewards at the default time of 2 blocks per sec. More blocks = more rewards. Shard 0 being slower on block generation, means, the other shards generate more blocks to be signed in the same time.

Someone correct me plz if my understanding was wrong.

1 Like

Either that or secure resharding is implemented before decentralized nodes.

These two roadmap items are being tackled the wrong way round for me, surely security should come first?

2 Likes

I do agree, I wouldn’t mind having the resharding implemented while we still have some harmony nodes running as a failsafe.

For those who missed the yesterday Validator Twitter Space
VDAO Twitter Space #1 // AMA & HIP-16 and Future Proposals

6 Likes

I like this idea of a 10% per shard key limit which tackles the shard takeover risk @leo described head on. This would also address the problem of kucoin being all on shard 0. I suspect they are contributing to slow block times by running underpowered nodes.

5 Likes

So many great thoughts and insights here!

Thank you everyone!! @Jimbo_JCR.one @DKValidator @ValidatorONE @ben2k_Stakeridoo @Maffaz @Valyd8Chain @Manumix +++
Where is the document referenced in the recording from the TwitterSpaces?

What are the next steps?
Is there enough consensus for what a “best option” would be to move forward?

Seems like another Talk poll is needed, perhaps several because there are several options (Correct?)…

I was curious what would be the effect, per option, for the end-users on the network?

How would changing the BLS keys requirements effect someone staking with a pool which had to use fewer BLS keys?

Would it basically result in larger validators APRs lowering until stake-holders moved their stake to a pool with the correct amount of keys?

IMO, It’s absolutely vital that we support our smaller Validator Nodes to ensure the network is operated by and for the community.

Thank you everyone!

1 Like

Think this is the document, with all the ideas from the first HIP-16-thread:


Yes exactly that.

4 Likes

I am quite new to the validator system but I was thinking that we currently have 900 external slots taken by ~155 validators. We have ~74 unelected validators. So instead of reducing BLS keys, why can’t we implement a system that will consider 1 slot for all validators in the network and those validators with multiple slots will be fighting among themselves, just as they are right now?

4 Likes

That is a great idea… This discussion was already made into a proposal here:

Feel free to create your own thread as well…

1 Like

Anyone familiar with the history of this thread care to take a shot of a TL:DR on this topic to help those catch up. Prefereably someone at the very top of the thread that can just provide an overview of the discussion points made so help those of us who are just too lazy to read every comment on the way down?

Non-big Validators are trying hard to push for more decentralization with more or less radical measures, big Validators mostly want none of that and have all the voting power, so that’s it.

5 Likes
1 Like

Going to keep it short. I see so much drama in the harmony discord about new validators getting completely upset about not being able to get delegations and get elected. I understand everyone wants to be a validator for harmony and its great! There has to be some type of level of engagement from everyone who wants to get involved also. I dont see many out there asking to delegate and the few that do get delegations and eventually get elected.
Limiting the amount of ONE someone can delegate is not going to work in my opinion. If I want to delegate all my wallet into self delegation I should be allowed to. Limiting BLS keys is the only way and it will naturally fix the need to talk about this. If larger validators have millions staked and only allowed 5 keys or 10 keys it will lower their APY and force some delegations to go to smaller validators with higher APYs . There bid will also be alot higher. There has been some shady validators bidding lower by taking more keys to big under 4 million and kicking several new validators out. Lets just keep it fair. You have 20 million you dont need 9 keys. Be fair and honest and keep decentralization working the way its supposed to. my opinion only. I will vote yes on any BLS key limitations that sound proper and fair to veteran validators , I mostly likely dont agree with limited stake from single wallets.

2 Likes

Your way of thinking is ok. But the problem with the high delegation from a single wallet is the corporative addresses. You can notice that even when the validator is marked as unelected, those wallets do not undelegate. So even with low rewards, the stake would remain there. So the BLS KEY limitation will have the efficiency dismissed. So the best scenario is to have BOTH proposals running. The more hips focusing decentralization running and approved, the better for the network.

I think it’s inaccurate to say only a few unelected validators are out there trying to drum up support for themselves, and that those who do market themselves “eventually get elected”. It is a struggle to become elected and then remain elected. It shouldn’t be this hard. And there are ways to go about making it easier. What would be wrong with having more validators and furthering the network’s decentralization? Keep in mind, the front page of Harmony’s website says “1,000 validators” is a feature of the protocol. Currently that’s an inaccurate statement. But we can make it a truthful one

I also think turning delegating/validators into a popularity/marketing “contest” is the wrong way to go. I think validators can do that stuff if they wish, but I don’t think their election should be dependent upon it

If it wasn’t thanks to the work of “wallet 18/19”, OR1ON, and Harmoforce, I think there would have been very few new elected validators over the last several months

4 Likes

there is nothing wrong with having more validators. I am all for that. But making it max stake per wallet per delegator isn’t going to do anything. You will just have large wallets create multiple validators or 50 wallets. Maybe implementing something like avax. You can only validator for 10x what your self staking. The system in my opinion is decent now but the BLS keys management is just out of control when validators with minimal delegations take multiple keys and bid really low just to boost rewards .5 % . It unelects everyone under that low bid 90% of the time. And to your point it is a popularity contest at the moment. It shouldn’t be.

If they want to go through the trouble to split their balance through 50 wallets or so…it’s up to them. But the path of less resistance is almost always followed. If they said wallets set up another validator, would contribute to decentralization as well. More validators, more decentralization. :man_shrugging:.

3 Likes