Incident Timeline and Impact
Base mainnet experienced two block production outages on June 25 and June 26, 2026, lasting approximately 116 minutes and 20 minutes respectively. According to the official postmortem published by the Base engineering team, both incidents did not compromise on-chain asset security, and user funds remained accessible at all times. However, during the outages, the network completely halted block production, transactions could not be included on-chain, and the mempool became congested. Users submitting eth_sendRawTransaction requests received persistent error responses.
Root Cause: Sequencer State Management Bug
The postmortem identified that the core issue was a defect in the sequencer's block construction logic. Specifically, when a transaction execution failed, the system failed to properly clean up the historical journal state. This caused subsequent legitimate transactions to encounter gas calculation anomalies during execution, leading to the generation of invalid state transfer blocks. Consequently, the entire L2 network stopped producing blocks. This bug is classified as a sequencer internal state management issue and has no relation to the underlying chain's security mechanisms.
After the first outage, the team deployed a fix via PR #3806, which addressed the journal cleanup flaw and resumed block production. However, during the restart process of the sequencer cluster, an engine reset race condition occurred, hindering sync recovery. This race condition indirectly caused the second, shorter outage the following day (20 minutes). The second incident was primarily a result of incomplete restart procedures.
Remediation and Future Improvements
To prevent similar failures, Base outlined several improvement measures. First, protocol-level fuzz testing and stress testing will be significantly enhanced to detect abnormal transaction paths earlier. Second, the monitoring and operations system will be upgraded to improve awareness of anomalous states. Third, the team plans to introduce a more graceful recovery mechanism that enables the network to automatically recover faster during similar failures, minimizing user impact.
This incident serves as a reminder for L2 network operators that the sequencer, as a centralized component, requires robust state consistency handling and restart fault tolerance. As a major L2 chain incubated by Coinbase, Base's experience in fault resolution is valuable for the entire Rollup ecosystem, highlighting the importance of rigorous testing and resilient recovery design in production environments.