Increase the Validator Set by 5-10 slots

General
1

This discussion suggests increasing the validator set by 5 positions to a total of 105, allowing for additional participants to join the network and contribute to its security and resilience.

The existing validator set of 100 has proven its effectiveness in securing the Celestia network, but concerns have arisen regarding its accessibility. The high entry requirement of ±371k $TIA tokens, which at the time of writing is approx. ±6,250,000 USD, poses a significant hurdle for potential validators, limiting the diversity and decentralization of the chain.

By expanding the validator set, the minimum $TIA required for entry would decrease, promoting broader participation with easy entry into the active set. Notably, this opens the door for smaller validators enabling them to contribute actively to the chain’s development.

3 Likes
2

With 12-second block times, there is no reason not to.

3 Likes
3

Why not increase the validator set by 10, 50, or 100?

5 seems arbitrary and would make a precedent case to keep increasing it over time, where performance would degrade.

Another point is that adding k Amount of Validators increases the proving time for Tendermint X, so I would be careful with adding more unless necessary and the tradeoff justified.

4 Likes
4

Additionally, larger block might fill up the 12 second block intervals easily; i.e. comet block propagation (celestia-core is essentially comet v0.34.x) might take longer than 12 seconds with a larger validator set.

My two cents: proposals like these should be accompanied with a very clear analysis and an as close to real-world as feasible simulation around block intervals and validator set sizes.

5 Likes
5

yes, it indeed may be increased by 10 slots, and it shouldn’t the proving time much, will look into it closer!

1 Like
6

as stated above it shouldn’t effect the block intervals much, might need to have dev thoughts on this

1 Like
7

Let’s do it. Seems like a good move.
More decentralization.

2 Likes
8

Agreeing with your rationality for increasing the validator set. The current validator set is beyond reach for most and the block times indicate that additional validators can be supported. Celestia community is one of the frequent user of Smart Stake analytics and being part of the active set will allow Smart Stake to continue to contribute further.

PS: given the high interest in Celestia, an increase of 5-10 will likely lead to almost a similar entry point in a matter of days.

2 Likes
9

Need to 100x node the count to be taken seriously and get ETH/Solidity developer buy in to Celestia in a big way (lots of other reasons to do this), which means this community needs to start designing for such scale or shutdown imo. 10K nodes next step design goal or shut down?

That puts the cost of entry as a validator in the range of US $ 6,250.00 at current Market prices and makes the project hugely viable to all SC Developers.

Go big or go home.

1 Like
10

@ismail good news on that, I’ve actually attempted this on several chains. I’m going to beg forgiveness for being a tad hand wavy here, but both QS and Omniflix were able to get approx 9 second blocks with 700-1000 vaidators.

IMO, the 12 second blocks are 100% safe for a validator set of let’s say, 300. But actually I bet the network could swing 400.

note this is hand-wavy, experiential data, and we would probably want to do a testnet to confirm this, but it is very high confidence hand-wavy experiential data.

1 Like
11

Ethereum is currently planning to shrink its node count by allowing more to be staked to each validator.

Celestia does not need to 100x node count.

But I actually do feel a bit bullish about 2-3x.

This is one of the advantages that Celestia has gained for itself by being super minimal, I think that it could very credibly handle a 200-300 node network, and that it would benefit from that.

2 Likes
12

I’m aware of Omniflix etc but what were the block sizes? The questions really are: could the network support >100MB blocks with consistent block times with the current number of active validator slots? If it could, how much wiggle room would there be to further increase the block size? How does increasing the validator set impact larger block sizes?

IMHO, the more important property for Celestia is that the blocks and their availability can still be verified by really low resourced machines, aka light clients. This verifiability is way more relevant for decentralisation than a large set of active consensus / block producing nodes.

2 Likes
13

In the case of 100mb blocks, my opinion shifts to no, and maintaining the current set size.

With that said, I do find this an interesting research topic.

Being able to sustain that kind of load would be fantastic.

2 Likes
14

I think the reasoning of increasing the validator set to increase the decentralisation of a network is a fallacy. Looking at the current distribution, increasing the validator set by 10 or even 20 nodes may cause a redistribution of around 1%. Far more critical is the distribution of power within the current set with the top 6 validators capable of halting the network.

It would be far more valuable if we could find methods of validator selection for the community that naturally leads to a more even distribution of power.

Another thing to note: Celestia’s approach to security is to not rely solely on validators but provide the means for everyone to verify in a lightweight manner. The idea being, if you want to increase the security of the network, run more light nodes.

2 Likes