Most used blockchain averts serious disruption after post-upgrade flaw

Ethereum’s Fusaka upgrade went live on Dec. 4, initially lifting market sentiment as the network pushed through one of its most ambitious technical changes to date. 

Fusaka is designed to make Ethereum faster and cheaper by improving the way Layer-2 networks (L2s) settle transactions back onto the main chain.

These L2s handle transactions at low cost and then submit compressed data to Ethereum. Fusaka upgrades this entire process, allowing the system to eventually handle 5–8× more transactions with lower fees.

Ahead of the activation, developers and data providers pointed to increased L2 activity and higher throughput expectations, but within hours of deployment, attention shifted to a very different issue.

ETH briefly traded near $3,217, according to CoinGecko, before sliding back to $3,131 at press time as an unexpected report sent traders scrambling.

More News:

• 5-star investor issues harsh warning to XRP holders
• Analyst predicts 90% surge for sinking crypto stock
• Andrew Tate gets ‘hyperliquidated’ as Bitcoin crash wipes out his entire balance

A Prysm consensus bug stressed validators

Shortly after Fusaka activated at block 18,200,000, Prysm nodes began experiencing severe performance degradation due to a historical-state generation issue. Prysm nodes are one of the software programs validators use to run the network, began to slow down.

The issue came from something called historical state generation — basically a validator trying to rebuild large chunks of old blockchain data. This process is extremely CPU- and memory-heavy, and under certain conditions, Prysm nodes were hit with too many such tasks at once.

Prysm core developer Terence Tsao explained:

“Historical state generation is compute and memory heavy, and a node can be dos’ed by a large number of state replays happening in parallel,” leading to stale attestations and slower block finalization.

A spike in "off-slot checkpoint targets” forced many validators to reconstruct more historical data than usual, pushing some Prysm operators near failure. Node operators stabilized their systems after enabling an emergency fix, telling the client to skip certain heavy computations until the load drops.

Related: Ethereum edges higher as Wall Street returns on fresh inflows and corporate buying

Ethereum stayed online thanks to client diversity

Despite the stress, the network did not halt.

Client diversity acted as a natural circuit breaker, with Lighthouse, Teku, Nimbus and Lodestar continuing to finalize blocks normally.

Staking protocol Lido confirmed it was “less affected… owing to the distributed and decentralized nature of the validator set,” noting that operations continued without disruption.

Fusaka’s scaling overhaul still lands

Fusaka’s technical changes remain intact. 

The upgrade introduced PeerDAS, a data-availability sampling system that allows nodes to store only a fraction of blob data, enabling up to 8× future throughput increases without hard-fork coordination.

Ethereum co-founder Vitalik Buterin called the feature “significant because it literally is sharding,” adding that:

“Ethereum is coming to consensus on blocks without requiring any single node to see more than a tiny fraction of the data.”

Related: Analyst cuts 60% price target for popular Bitcoin stock