Early yesterday morning, the Ethereum network experienced a harrowing moment. 382 ETH (worth over a million dollars) vanished instantly after a routine software update, but this was neither a hacker attack nor a market bloodbath—rather, a long-standing code vulnerability was activated, triggering a system-level disaster.

The incident dates back to a month ago. At that time, a fatal flaw had quietly entered the testing environment but went unnoticed. It wasn't until Fusaka's update went live on December 4th that this ticking time bomb truly detonated.

Problems came quickly. The Prysm consensus client began experiencing strange delays when processing validator requests, followed by the loss of blocks and invalidation of on-chain proofs. A domino effect was set in motion—within just 41 epochs, 248 out of 1344 slot blocks completely disappeared, with a loss rate of 18.5%. Network activity also plummeted, with participation dropping sharply from normal levels to 75%.

According to the incident report, the root cause was traced to resource exhaustion leading to system crashes. Prysm's developers admitted that the beacon node received proof data from nodes that might not have been fully synchronized, pointing to a block root from an earlier epoch—this inconsistency triggered a chain reaction of failures. Nearly all nodes running the Prysm client were affected, underscoring the severity of this incident.