Base, Coinbase’s Ethereum layer-2 network, has resumed normal operation after a consensus-related issue temporarily halted block production for nearly two hours on Thursday. The incident triggered “unhealthy” block building, leaving new blocks unable to be created until the network isolated and corrected the underlying problem.

Base said blocks were later being produced normally and that the broader ecosystem infrastructure had recovered and synced. The outage was unusual for a chain that has become a go-to venue for activity on Ethereum’s scaling roadmap, highlighting how sensitive rollup operations can be when consensus and sequencing logic fail.

Key takeaways

Base went offline briefly due to a consensus problem that resulted in an invalid block being sequenced and prevented new block creation.
Base reported recovery of “healthy blockbuilding” and confirmed ecosystem-wide infrastructure synchronization.
The outage occurred around an on-chain upgrade window, with a Base upgrade (“Beryl”) scheduled for 6 pm UTC.
Network creators emphasized that user funds remained safe, while reiterating that a block-production halt is not acceptable.
Thursday’s incident joins a small set of notable recent outages affecting major L2 ecosystems.

How the outage unfolded

Base’s status page said it began investigating “unhealthy” block production at 4:03 pm UTC on Thursday. In a subsequent update at 5:21 pm UTC, the Base team explained that it had “isolated a consensus problem” which caused an invalid block to be sequenced.

According to the status updates, that invalid sequencing stopped progress at the protocol level: “This prevented new blocks from being created.” In other words, the issue was not presented as a simple infrastructure hiccup; it was tied to the chain’s ability to agree on what should be built next.

Recovery confirmed, post-mortem expected

Base later posted an operational recovery update just before 6 pm UTC. It said the network had “recovered healthy blockbuilding,” and that ecosystem-wide infrastructure was able to sync again. Base also indicated that it had identified the issue and would continue investigating the root cause, promising a full post-mortem.

Separately, Base’s creator Jesse Pollack used X to reassure users that funds on the network are safe. Pollack also framed the halt as a solvable operational setback, saying it would be used to “level up” Base as a platform aimed at supporting continuous, global finance.

An uncommon downtime event for a leading L2

The episode stands out as a rare downtime event for Base. The report notes that Base is widely considered among the most-used Ethereum layer-2 networks, and it cites the chain’s previous major outage in August 2025, when Base reportedly went down for 33 minutes. That earlier incident is referenced via Base’s status page.

Such disruptions matter for users and developers because L2 block production is the backbone for transaction inclusion, sequencing, and timely settlement flows. Even if funds remain safe, a halt can translate into delays for withdrawals, reduced reliability for dApps, and friction for systems that assume steady block cadence.

Upgrade timing raises questions for reliability planning

Base downtime appeared to occur separately and shortly ahead of a scheduled network upgrade dubbed “Beryl,” planned for 6 pm UTC. The described goal of the Beryl upgrade was to reduce delays on withdrawals and introduce a new token standard intended for real-world assets and stablecoins.

That timing is important for operators and integrators: when an outage overlaps with a planned change, teams typically have to ensure that the network can recover cleanly and continue the upgrade without compounding issues. Base did not state that the outage directly affected the Beryl rollout, but the proximity means builders watching Base would likely want to see post-recovery monitoring closely through the upgrade window.

The incident also comes amid broader reminders across the L2 landscape. The report points to Sui experiencing two periods of downtime on back-to-back days in May, each involving temporary stops in block production. In that case, Sui later attributed the downtime to a network update it said it knew had a low probability of causing a halt—an example of how even planned changes can create operational risk.

What to watch next

Base has said it will share a detailed post-mortem and identified the consensus problem responsible for blocking new block creation. Investors, traders, and developers should watch for that report, plus confirmation that the Beryl upgrade proceeds smoothly with stable block production and no renewed consensus symptoms around the new token standard changes.