Starknet Reverts 18 Minutes of Transactions After 4-Hour Outage

Second Major Halt in Four Months

Starknet block production froze for four hours on January 5, forcing the network to wipe 18 minutes of transaction history to restore consensus. The outage, detailed in a post-mortem released Monday, marks the second significant liveness failure for the layer-2 since September, raising fresh questions about the operational maturity of zero-knowledge infrastructure.

Validators stopped processing blocks at 09:53 UTC. By the time the network came back online at 14:17 UTC, developers had executed a forced reorganization, reverting the chain to block 5,187,263. Any transaction submitted between 09:24 and 09:42 UTC was effectively erased.

The “Blockifier” Bug

The root cause was not a liquidity drain or external attack, but a logic error in the network’s “blockifier” component. According to StarkWare, the bug caused the sequencer to retain state data from a reverted function during a specific sequence of cross-function calls. This created a discrepancy: the sequencer “thought” a transaction was valid, but the prover, the mechanism responsible for generating cryptographic proofs for Ethereum mainnet, correctly rejected it.

The mismatch triggered the halt. To prevent invalid state roots from settling on Ethereum, the team manually intervened to rewind the chain.

Market Reaction & Trend

Despite the operational failure, spot markets ignored the news. STRK held the $0.09 level (-0.9%) throughout the downtime, with no significant liquidation cascades observed on major venues. This apathy contrasts with the Grinta outage in September 2025, where a nine-hour halt drove a sharper sell-off.

StarkWare engineers have deployed a patch that changes how execution results are verified. Future discrepancies between the sequencer and prover will now trigger an immediate, automated network halt to minimize the window of revertible transactions.