Yesterday, Ethereum’s Beacon Chain went through a seven-block reorganization (reorg), ahead of the Merge, tentatively scheduled for August.
Data from Beacon Scan shows that seven blocks numbered 3,887.075 to 3,887.081 were removed from the Beacon Chain on May 25, between 08:55.23 and 08:56.35 UTC.
Reorg is a term that refers to an event where a block which was part of the canonical chains, such as the Beacon Chain gets knocked off because of a competing block.
This can happen due to malicious attacks by a miner or bug with high resources. These incidents can cause the chain to unintentionally fork and duplicate.
Developers believe the issue is due more to circumstance than anything serious like a security problem or fundamental flaw. A “proposer boostfork” was highlighted in this instance. This refers to a system in which certain proposers have priority in selecting the next block in a blockchain.
Preston Van Loon, Core Ethereum developer, suggested that the reorg was caused by a “nontrivial segmentation” of old and new client node software and wasn’t necessarily something malicious. Vitalik Buterin, co-founder of Ethereum, called the theory a “good hypotheses.”
Block reorg: Beacon Scan
Martin Koppelmann, co-founder of EVM compatible Gnosis chains, was among the first to tweet about the event yesterday morning. He noted that it “shows how the current attestation strategy for nodes should be reviewed to hopefully result in more stable chains!” (Proposals already exist).
Van Loon responded to Koppelmann by suggesting that the reorg be attributed to the proposer boost fork, which had not been fully implemented yet.
“We suspect that this is due to Proposer Boost not being fully rolled out to network. This is not a sign of a poor fork choice but rather a non-trivial separation of out-of-date client software.
Once we have high levels of confidence about the root cause, all details will be public. He said, “We can expect a post-mortem by the client development community.”
This could be because Proposer Boost’s fork choice implementation has not been fully rolled out to all networks. This is not a sign of a poor fork choice but rather a non-trivial separation of out of date and updated client software.
— prestonvanloon.eth (@preston_vanloon) May 25, 2022
Terence Tsao, another developer, echoed the hypothesis earlier today to his 11,900 followers on Twitter. He noted that the reorg was caused by “boosted and non-boosted nodes in network” and the timing of a late arriving block.
“Given that the proposer boost is a non-consensus-breaking change. The roll-out was gradual due to the insynchronicity of client releases. The proposer boost was not updated by all nodes simultaneously.
Related: The Merge is fast approaching and OpenEthereum support ends
Van Loon, who spoke at the Permissionless conference, said that the Merge (and switch to Proof-of-Stake) could occur in August “if all goes according to plan.”
Van Loon and other developers are yet to say if the reorganization will have any effect, although it is certain that there will be questions about the timeline.