Ethereum’s 10-year uptime rests on a hidden consensus trade-off

Ethereum researcher Luca Zanolini has explained why the network separates continuous block production from final settlement, arguing that this structure lets Ethereum keep operating through software faults, outages and falling validator participation. 

His account connects Ethereum’s decade-long block production record with current work on faster finality and a cleaner consensus design.

Meanwhile, Ethereum uses one process to keep adding blocks and another to mark older blocks as final. The production layer follows the chain supported by active validators, while the finality layer requires approval from at least two-thirds of the total active stake. If that threshold disappears, finality can pause without stopping new blocks.

That distinction became visible in May 2023, when client faults disrupted finality twice within 24 hours. The first interruption lasted about 25 minutes and the second lasted close to an hour. Blocks continued to arrive, transactions remained available and the network recovered without a coordinated restart.

Why Ethereum Avoids a Full Network Halt

Zanolini said a base-layer halt would freeze more than simple token transfers. Lending platforms could not process liquidations, oracles could not update prices, rollups could not post data or proofs, and bridges could not confirm new state. Risks would continue building while users lacked an on-chain response.

A forced restart would also place recovery in the hands of a small group of developers, operators and validators. They would need to diagnose the fault, agree on a fix and coordinate the network’s return. Ethereum instead aims to keep producing blocks whenever an honest majority of awake validators can communicate.

Slashing and Inactivity Leaks Restore Order

The finality layer protects settled history through signed validator votes. Conflicting blocks or attestations can produce evidence that the protocol can check. “The protocol only punishes what it can prove,” Zanolini wrote. Validators that sign contradictory histories can lose stake through slashing.

Ethereum also uses an inactivity leak when finality remains unavailable for more than four epochs. Offline validators gradually lose effective stake, with penalties rising during a prolonged disruption. This changes the voting balance until participating validators again control enough stake to finalize the chain automatically.

The process does not require a hard fork or manual reboot. Block production continues while inactive stake falls. Zanolini described this recovery path as a core part of Ethereum’s design because the protocol can return to finality without waiting for every offline validator to reconnect.

Client Diversity Limits Software Failure Risk

Ethereum’s model becomes more fragile when one consensus client controls too much stake. A client above one-third can threaten finality during a major fault. Control above one-half can distort fork choice, while a client above two-thirds could help finalize an invalid history before operators can react.

As previously reported by crypto.news, a Prysm fault after the Fusaka upgrade in December 2025 pushed validator participation to about 75%. The network missed 41 epochs and validators lost roughly 382 ETH in rewards, but other clients continued working and Ethereum avoided losing finality.

The Ethereum Foundation’s Protocol Consensus team is now studying ways to separate block production and finality more clearly. A March research proposal suggested using a sampled committee for faster blocks while a separate process finalizes the chain behind it. This would let both systems use different timing and security settings.

A May 11 update said Ethereum’s next consensus work will focus on reducing finality time, which currently takes about two epochs under normal conditions. Separate crypto.news reporting said Vitalik Buterin has backed Minimmit, a proposed one-round finality system. The proposal could settle blocks faster, though its current design accepts lower formal fault tolerance than Casper FFG.

Zanolini’s explanation presents Ethereum’s resilience as a set of linked choices rather than one feature. Continuous blocks preserve access, finality protects settled history, slashing prices provable misconduct, inactivity penalties support automatic recovery and multiple clients reduce the reach of a shared bug across the wider validator network.