Base stalled block production for a second time in less than 24 hours on Friday morning, describing the incident as showing ‘similar symptoms’ to Thursday’s roughly two-hour outage, with the event arriving less than three hours before the B20 Activation Registry was anticipated to go live.
Base, the Ethereum Layer 2 network incubated by Coinbase, halted block production for the second time in two days on Friday, arriving hours before a scheduled activation of its new B20 token standard on mainnet.
The second stall began at 15:33 UTC Friday when Base’s status page flagged block production as unhealthy. The team identified “similar symptoms” to Thursday’s outage within one minute and restored production by 15:47 UTC, a roughly 14-minute halt. Thursday’s incident ran for approximately two hours after an invalid block at position 47,806,542 choked the sequencer and interrupted withdrawals. The Defiant covered that first halt as it resolved Thursday evening.
Beryl and B20
Thursday’s outage had already forced a schedule shift for Base’s Beryl hardfork. Teams had pushed the upgrade window to allow the B20 Activation Registry to complete its initialization process, a sequence that can take up to 60 minutes after the hard fork activates.
The B20 standard, a Rust precompile embedded directly in Base’s node software, introduces native token management built for stablecoins and tokenized real-world assets. It adds role-based minting controls, transfer restrictions, and freeze capabilities while maintaining full ERC-20 compatibility.
The registry at contract address stores feature flags that govern whether the B20 standard is live. Until the flags are flipped by a designated admin, any call to create a B20 token reverts with a “FeatureNotActivated” error. Beryl completed on Thursday at 20:00 UTC. A dev relations representative confirmed the B20 registry activation was scheduled for Friday at 18:00 UTC, placing that window roughly two hours after the second stall resolved.
Same Bug Returns
Friday’s status update explicitly described the halt as exhibiting “similar symptoms” to Thursday’s, indicating the root cause had not been fully remediated before the second event struck. After the Thursday stall, Base stated it had found the root cause and was “verifying a fix to ensure it cannot recur,” with a full postmortem promised. No postmortem had been published as of Friday afternoon, and the second incident remained in “monitoring” status as of 16:11 UTC.
Node operators were required to restart their Base Mainnet nodes to resume syncing after both incidents. Base’s 90-day block-production uptime stood at 98.72% as of Friday, per the status page. No public statement on the B20 registry activation timeline had been issued as of publication.