Base Freezes Up Again — Coinbase's Layer-2 Stalls for the Second Time in 48 Hours

Base, the Coinbase-incubated layer-2, stalled out again — the second time in roughly 48 hours. Block production was interrupted following an alert at 15:33 UTC and resumed by 16:11 UTC, according to status updates tracking the chain. For a network that processes a huge share of low-cost Ethereum activity, two halts in two days is not a great look, even if funds were never at risk.

What Happened

The first stall hit the day before, on June 25, when Base's block production went 'unhealthy' around 16:03 UTC. A consensus issue caused an invalid block to be sequenced, halting all new blocks after height 47,806,542. That outage ran nearly two hours; node operators had to restart their Base mainnet nodes to resume syncing before the chain came back online June 26 at 16:11 UTC. Then, almost immediately, a second alert fired at 15:33 UTC and block production stalled again before recovering shortly after.

The timing is rough. Base had scheduled its second major network upgrade — Beryl — for June 25 activation. Beryl introduces the B20 native token standard and trims Ethereum withdrawal delays from seven days down to five. Instead of a clean upgrade-day story, the team spent it firefighting consensus failures.

Why It Matters

Base markets itself on reliability and Coinbase's institutional backing, so repeat consensus stalls cut against the pitch. The chain's most recent prior outage was a 33-minute halt in August 2025 caused by a sequencer handoff issue — meaning Base has now logged multiple block-production interruptions tied to its centralized sequencing model. Coinbase and Base developers identified the issue, shipped a fix, and confirmed all user funds remained safe throughout. But the incidents revive the long-running debate about how decentralized — and how resilient — these Coinbase-run rollups actually are when a single bad block can freeze the whole network.

What's Next

Expect a post-mortem from the Base team detailing the root cause of the back-to-back stalls and what's changing in the sequencer setup. With Beryl now live and withdrawal windows shortening, the bigger question is whether the upgrade hardened the chain or introduced the very fragility that knocked it offline twice in two days.

☕₿