HIP32 signifies a major step toward full decentralization, with our Harmony internal node losing its complete voting power. However, we're currently confronted with a challenge tied to crosslinks and the full externalization process
Crosslinks play the role of transmitting staking data from shard 1 validators to shard 0, facilitating the assessment of signature performance. If crosslinks become stuck for an entire epoch, shard 1 validators may find themselves with zero signing activity by the epoch's end :
"blocks": {
"signed": 0,
"to-sign": 0
},
This poses a considerable risk, as, at the conclusion of the epoch, the staking protocol automatically excludes all external validators. Consequently, the shard is left without either internal or external validators, effectively bringing the shard consensus down. Shard 1 devnet and testnet are currently in that situation.
We will have to work on updating the staking logic to avoid such situation to happen and will delay the HIP32 mainnet hardfork implementation for now
HIP32 signifies a major step toward full decentralization, with our Harmony internal node losing its complete voting power. However, we're currently confronted with a challenge tied to crosslinks and the full externalization process
Crosslinks play the role of transmitting staking data from shard 1 validators to shard 0, facilitating the assessment of signature performance. If crosslinks become stuck for an entire epoch, shard 1 validators may find themselves with zero signing activity by the epoch's end :
This poses a considerable risk, as, at the conclusion of the epoch, the staking protocol automatically excludes all external validators. Consequently, the shard is left without either internal or external validators, effectively bringing the shard consensus down. Shard 1 devnet and testnet are currently in that situation.
We will have to work on updating the staking logic to avoid such situation to happen and will delay the HIP32 mainnet hardfork implementation for now