Prysm builders launched a autopsy evaluation explaining the December 4 Fusaka mainnet incident that threatened Ethereum community stability.
Abstract
- A Prysm bug after Fusaka brought on validator participation to drop to 75%.
- The community missed 41 epochs and misplaced roughly 382 ETH in proof rewards.
- Ethereum prevented finality loss because of consumer range and fast fixes.
The consensus consumer suffered useful resource exhaustion from costly state recomputation when processing particular attestations, inflicting validators to face extreme operational issues.
The bug surfaced instantly after Fusaka activated at epoch 411392 on December 4, 2025, at 21:49 UTC.
The community missed 41 epochs as validator participation plummeted to 75%, leading to roughly 382 Ethereum (ETH) in misplaced proof rewards. Prysm builders deployed emergency runtime flags earlier than implementing everlasting fixes in variations v7.0.1 and v7.1.0.
Useful resource exhaustion pushed community towards finality loss
The technical failure centered on out of date historic states that created denial-of-service circumstances on affected nodes.
Prysm core developer Terence Tsao defined that “historic state is compute reminiscence heavy, a node could be dosed by massive variety of state replays occurring in parallel.”
Validators operating Prysm, which represented roughly 15% to 22.71% of community validators, confronted crippling efficiency degradation. The participation drop from regular ranges above 95% to 75% pushed Ethereum dangerously near shedding finality.
Had the bug affected a special consensus consumer like Lighthouse as a substitute of Prysm, the community might have misplaced finality fully.
Such an occasion would probably freeze Layer 2 rollup operations and block validator withdrawals till builders resolved the problem.
The Fusaka improve itself launched PeerDAS (Peer Knowledge Availability Sampling) expertise designed to extend blob capability eightfold for Layer 2 scaling.
The improve executed efficiently with zero downtime earlier than the Prysm bug surfaced.
Ten consensus purchasers prevented Ethereum community collapse
Ethereum’s consumer range structure prevented catastrophic failure. Whereas Prysm validators struggled, ten different consensus purchasers together with Lighthouse, Nimbus, and Teku continued validating blocks with out interruption.
The decentralized consumer construction meant that roughly 75% to 85% of validators maintained regular operations all through the disaster. This prevented finality loss and stored the community processing transactions regardless of Prysm’s degraded state.
The Ethereum Basis shortly issued emergency steerage for Prysm operators. Validators utilized the short-term repair whereas Prysm builders constructed everlasting options.
By December 5, community participation recovered to almost 99%, restoring regular operations inside 24 hours of the incident.