Dear community,
We have heard your concerns about the number of active validators decreasing over this first week since launch. We are currently at 43 elected validators out of a total of 170 validators created. It is extremely important to us that we address the root causes of this problem so that we can foster a decentralized group of validators for the long term.
Let’s work together to see how we can solve this problem. Here’s an outline of what we see as the underlying issues and potential solutions:
Key problems
- Minimum stake - the amount of stake needed to get elected is going up, leaving behind smaller validators to start getting unelected, and over time churn out.
- Signing issues - Validators (especially smaller & inexperienced ones) are facing signing issues, resulting in low uptime, thereby negatively affecting their future incoming delegation.
Potential Solutions
-
Increase the EPOS median range from 15% to a higher percentage, i.e. 30%
- Reasoning 1: A higher median range will allow bigger validators to take fewer seats at a higher stake per seat, opening more seats for other validators. In addition, the cutoff should be lower as there will be no incentive to bid up the bottom seats above median-30%.
- Reasoning 2: We’ve noticed that the EPOS mechanism actually hurts smaller stakers because they can’t adjust their stake to stay within the EPOS band, whereas larger stakers can. This happens because the larger your stake is, the more evenly divisible it will be into the EPOS band to maximize rewards. While the reverse is true for smaller stakers. Example: I am a small staker with 3M stake, but the median is 2M. If I split to 2 seats, my bids are 1.5M each and I won’t get elected. So I have no choice but to keep my bid at 3M for 1 seat, and receive a low reward.
- Feasibility: this change will need a hard fork.
-
Increase the number of public slots from 320
- Background: Currently, Harmony controls 68% of the voting power and runs 680 of 1000 nodes. We can remove open some of our 680 nodes to the public, but keep our voting power at 68% for safety.
- Reasoning: This increases the number of seats, decreasing the median and allowing smaller stakers to get elected.
- Feasibility: This can be easily done with a rolling upgrade.
-
Helping smaller validators solve signing issues
- Background: Validators (especially smaller & inexperienced ones) are facing signing issues, resulting in low uptime, thereby negatively affecting their future incoming delegation.
- Reasoning: If smaller validators have higher signing rates they will have better uptime and APR, making them more attractive for delegation.
- Feasibility: The long-term solution is to resolve the network layer/p2p signing issues in the pending PR. But in the shorter run, we can support these validators through one-on-one chats
-
Decrease the limit of seats per validator
- Background: The current limit of seats per validator is 1/3rd of the entire network, so a single validator can take up to 106 seats.
- Reasoning: If we reduce this limit, validators will need to create multiple separate validators in order to have the same total number of seats.
- Feasibility: This change will need a hard fork and will be more comestic in nature than solving the root problem as there is an easy work around.
-
Change the staking dashboard design to better promote small validators
- Background: Some have argued that emphasizing ‘Uptime’ rather than ‘APR’ may be beneficial for small validators for example (which we’ve done) and we’ve changed calculation for expected return to use the last epoch so that validators
- Reasoning: The staking dashboard may inadvertently put small validators at a disadvantage due to its design and emphasis on certain metrics
- Feasibility: Does not require hard fork
-
Marketing promotion for smaller validators
- Background: We can help small validators gain visibility through Harmony marketing efforts. We can also run campaigns encouraging delegators to diversify their stake among many validators rather than concentrate them in one
- Reasoning: promoting smaller community validators can help them attract more delegation
Process
We know that time is of the essence so we are moving quickly to arrive at a solution quickly. Here’s the process we’ll be following to decide on the next actions to take:
- Discuss - Continue the discussion here on the forum for the next 48 hours
- Meet - We will host a live meeting on Monday at 1pm PT to go over the points raised
- Vote - If needed, we will hold a vote to decide how and what to implement
- Publish - Write out details of the changes we will be implementing
- Upgrade - We will make the changes and upgrade the network