BlockBeats reported on June 27th that the Base engineering team released a recap of the June 25th block production outage, stating that the Base mainnet experienced two block production outages on June 25th and 26th. The first incident occurred at 11:47 AM Eastern Time on June 25th, lasting 116 minutes; the second occurred at 11:28 AM on June 26th, lasting 20 minutes. The root cause of both incidents was the same. Chain integrity was unaffected, and all funds on Base are safe. Block production resumed safely after the team mitigated the incidents.

The root cause is a vulnerability in the sorter's block construction logic, which prevents expired log states from being cleared after a transaction verification failure. An invalid transaction is received by the block builder and fails to execute as expected, but the system incorrectly retains the log state containing accessed accounts and storage slots. Subsequently, a valid transaction is executed before the log state is properly cleared, resulting in a gas billing mismatch. This leads to the generation of a sorted block containing invalid state transitions, which other nodes cannot accept, causing the chain to stop completely.

During the incident, Base was unable to produce new L2 blocks, the orderer and validator nodes were unable to proceed beyond invalid blocks, and new transactions could not be added to the chain. The team fixed the outage by applying a patch to the orderer to ensure that the log state was correctly updated during execution. Base stated that it will strengthen its protocol fuzzing and load testing tools to more easily identify adversarial transaction patterns that may expose similar vulnerabilities, and will continue to improve its operational and monitoring capabilities.