Summary : This proposal is to increase the external slots from 800 to 900.
Background : Currently, there are 800 external slots out of the max 1000 slots of the harmony blockchain. Our goal is to have fully external slots by EOY. So this proposal is to further reduce the number of internal slots.
Motivation : With this proposal, it will change external slots available to external validators from 800 to 900. This will further lower the median of the stakes, and potentially increase the number of unique smaller validators. This is another step towards the fully externalized validators and fully externally owned network.
Supporting this. Does it make sense to do both the 51% upgrade and this one together if it passes the vote? So we minimize the hard forks, if its possible?
We fully agree and support this proposal.
If there would be a higher requirement for machines specs after externalizing more slots, it would be great to announce it prior to the change.
After the events of the spam attack, I will have a hard time supporting HIP-8.
We already have a disproportional distribution of validators across shards due to shard 0 minting blocks at a slower rate and yielding poorer rewards. I worry with the addition of 100 slots and the same hesitancy to be on shard 0, we will see further saturation of the other shards.
Additionally, during the peak of the attack we had validators fleeing shard 0 in masses. With low incentive to be on shard 0, what assurances do we have that we will be able to maintain an adequate number of validators to maintain security and network stability. (I know harmony hold 49% voting power still but the concern remains). This proposal is in theory beneficial for the sake of decentralization, but I take issue.
Ultimately I leave the vote in the hands of my delegates but wanted to take a moment to voice my concerns as I haven’t had the ability following the events of 2 weeks ago. Just 2 cents from a small validator.
Some validators are correlating the slow speed of shard 0 to having an imbalance of slots currently and they are hesitant to support this proposal.
Does the number of slots a shard has have an effect on the shard speed at all? If shard 0 was at 260 slots and the rest were at 180 would the times be lower on shard 0 or would the remain as they are now?
If we can get an answer on these being related or not that would help inform us all better.
I have asked around to confirm, but as long as there is 1 validator in shard 0 the network will keep running. Also this 1 validator will earn all the rewards during that epoch, so likely this will not happen. Times should be the same…
Thanks for your feedback @WellnessOne . The number of total slots has no strong relation with block rewards. The total amount of rewards is the same per block. With more open slots, there could be more validators elected and participate in the network, which is important to keep the network decentralized and more resilient to attack.
Yes, nodes in shard0 will have to do more work as right now, most of the transactions happened on shard0 and shard0 is also the beacon chain shard to handle staking transactions. The expectation on the CPU/network of the nodes on shard0 is a bit higher than other shards at this moment. When the resharding is implemented, it will be the same requirement to all nodes. The number of slots may not necessarily impact the speed of the shard, as the CPU is not the bottleneck so far.
