BlockBeats News, June 28th. According to official sources, the root cause of the consecutive interruptions on Coinbase's Layer2 network Base last week has been identified as a software bug in the Sequencer block building logic.

The Base engineering team stated that after a failed transaction, the corresponding temporary state (Journal State) should have been cleared. However, due to a program defect, the old state was incorrectly retained, ultimately causing the Sequencer and validation nodes to be unable to continue generating new blocks. On June 26th, the Base mainnet experienced two block interruptions, with the first one lasting 116 minutes and the second one lasting 20 minutes.

The team mentioned that although the Sequencer state update issue was fixed through a patch, after the system restart, a "Race Condition" occurred, causing the Sequencer to fail to resynchronize promptly, leading to a second crash. In addition, some infrastructure issues prolonged the overall recovery time.

Base stated that they will enhance protocol fuzz testing to increase vulnerability discovery capabilities in exceptional input scenarios and optimize the network recovery mechanism to prevent similar events in the future that would require manual rebooting of validation nodes.

It is worth noting that this is not the first time Base has experienced a failure due to Sequencer-related issues. It previously had network interruptions of approximately 17 minutes in September 2024 and 30 minutes in August 2025. Currently, Base remains the second-largest Ethereum Layer2 network by Total Value Locked (TVL).