The client of the Ethereum Prysm consensus said that validators lost 382 Ether, which is equivalent to more than one million dollars, after a software bug caused disruptions in the network immediately following the recent Fusaka upgrade.
A post-incident report titled "Prysm Incident on the Fusaka Mainnet" clarified that the incident arose due to resource exhaustion that affected almost all Prysm nodes and led to the loss of blocks and attestations.
What caused the Prysm outage?
Prysm developer Offchain Labs mentioned that the issue arose on December 4 when a previously introduced software bug delayed validator requests.
This delay led to the loss of blocks and attestations at the network level.
The project clarified that the Prysm beacon contract received attestations from contracts that may have been out of sync with the network, and those attestations referred to a block root from the previous cycle.
This disruption resulted in the loss of 41 cycles, with 248 blocks lost out of 1,344 available stakes. This represented a stake loss rate of 18.5% and reduced overall network participation to 75% during the incident.
Offchain Labs stated that the software bug responsible for this behavior was included and deployed on test networks about a month ago, before being activated on the main network after the Fosaka upgrade.
Although a temporary mitigation measure reduced the immediate impact, Prysm clarified that it has since implemented permanent changes to the verification logic to prevent the issue from recurring.
Client diversity in Ethereum
At the same time, this outage renewed scrutiny around the concentration of Ethereum clients and the risks posed by software monoculture.
Offchain Labs stated that this disruption could have led to more serious consequences if Prysm had controlled a larger percentage of the Ethereum validator base. The company pointed to the diversity of Ethereum clients as a key factor in preventing widespread network failure.
The company stated that if a client controls more than 1/3 of the network, it will lead to a temporary loss in the final form and the loss of more blocks, while a software client with a bug that controls more than 2/3 can make the chain invalid for final confirmation.
Despite that precautionary measure, the incident underscored the need to enhance client diversity.
Mega Labs data indicates that Lighthouse remains the dominant client for Ethereum in terms of consensus, accounting for 51.39% of validators, with Prysm in second place at 19.06%, followed by Teku at 13.71% and Nimbus at 9.25%.
Set the Lighthouse stake about 15% points away from the threshold that some researchers consider a systemic risk.
Developers and ecosystem participants once again urged validators to consider switching to alternative clients to reduce the likelihood that a single software bug could disrupt the core operations of the blockchain.
