Base reviews the network outage incident: user funds are safe, the root cause is a vulnerability in the sequencer block construction logic.

BlockBeats news, June 27, the Base engineering team released a postmortem of the block production outage on June 25, stating that the Base mainnet experienced two block production interruptions on June 25 and June 26. The first event occurred at 11:47 ET on June 25 and lasted 116 minutes; the second occurred at 11:28 on June 26 and lasted 20 minutes. Both events had the same root cause. Chain integrity was not affected, and all funds on Base are safe. Block production safely resumed after the team mitigated the incidents.

The root cause was a vulnerability in the sequencer's block building logic, which caused the expired journal state to not be cleared after a transaction validation failure. An invalid transaction was received by the block builder and executed as expected to fail, but the system incorrectly retained the journal state containing the accessed accounts and storage slots. A subsequent valid transaction was executed before the journal state was correctly cleared, resulting in a mismatch in the gas fee amount, and thus generating a sequenced block with an invalid state transition. Other nodes could not accept this block, causing the chain to come to a complete halt.

During the incident, Base was unable to produce new L2 blocks. The sequencer and validator nodes could not progress past the invalid block, and new transactions could not be included on-chain. The team fixed the outage by applying a patch to the sequencer to ensure that the journal state is correctly updated during execution. Base stated that it will strengthen its protocol fuzz testing and load testing tools to more easily identify adversarial transaction patterns that might expose similar vulnerabilities, and continue to improve its operational and monitoring capabilities.