Base Network Halted Twice in 24 Hours: A Single Bug, a Delayed Hard Fork, and the Single-Sequencer Debate

Coinbase's Layer 2 network Base suffered two block production outages within 24 hours on June 25–26, 2026. Both were caused by the same underlying bug in the block builder logic . The incidents stalled the Beryl hard fork, drew a sharp rebuke from creator Jesse Pollak, and revived criticism of Base's single-sequencer design as a single point of failure.

What caused the halts? The same bug, twice

Both outages were triggered by a single bug in the sequencer's block-building logic . The official Base team account on X confirmed the root cause was identified, fixed, and shared with other OP Stack chains as feedback . In a detailed post-mortem, Base engineers explained that an invalid transaction was received by the block builder and failed during execution as expected, but the system erroneously did not clear the journal state containing the accounts and storage slots that had been accessed . When a valid transaction was executed immediately afterward, it used the stale journal state, leading to incorrect gas calculations and producing an invalid state-transition block . That invalid block jammed the sequencing pipeline, halting all block production.

The same sequence—failure to clear state followed by incorrect gas calculations—caused both outages . No user funds were at risk during either incident .

Outage timeline

• June 25, ~16:03 UTC: Block production stopped at height 47,806,542 after the sequencer produced an invalid block, triggering a consensus failure . The chain stalled for roughly two hours. Engineers identified the issue at 16:52 UTC, resumed preliminary sequencing at 17:51 UTC, and confirmed normal block production by 17:58 UTC .
• June 26, ~15:33 UTC: A second halt occurred, lasting approximately 14 minutes before service was restored .

Jesse Pollak: 'A halt is not okay'

Base creator Jesse Pollak responded publicly on X, acknowledging both incidents. He stated that user funds were safe throughout but added bluntly that "a halt is not okay" for a platform that aims to support global finance . He described the outages as unacceptable for Base's ambitions and said the team would use the incidents as an opportunity to improve the platform's stability and reliability .

Beryl hard fork delayed by one day

The Beryl hard fork—Base's second independent network upgrade—was originally scheduled for June 25 at 18:00 UTC . After the first outage, Base postponed the upgrade to June 26 at 18:00 UTC to ensure the B20 Activation Registry was fully operational before activation . The team stated that the outage and the upgrade were unrelated .

What Beryl introduces

Beryl bundles three major changes :

• B20 token standard: Base's native token standard for stablecoins, real-world assets (RWAs), and long-tail token issuers, implemented as an ERC-20-compatible Rust precompile built directly into the node software .
• Shorter withdrawal window: Single-proof withdrawal finality reduced from 7 days to 5 days .

Renewed criticism of the single-sequencer architecture

The outages revived long-standing criticism that Base's single-sequencer design creates a single point of failure . Because Base runs a single sequencer operated by Coinbase under the OP Stack, there were no backup sequencers or validation nodes that could take over when the invalid block jammed the pipeline . Multiple analysts and community members pointed to this architecture as the underlying systemic risk, arguing that a single-sequencer model is inherently fragile for a network aspiring to support global financial infrastructure .

What comes next

Base has published a full post-mortem on the June 25 outage, and the team stated it has communicated the findings to other OP Stack chains as feedback . The Beryl hard fork was successfully activated on June 26, but the incidents have left a lingering question: can a chain built on a single sequencer credibly claim to support global finance?