Base Says Sequencer Failure Caused Block Production Halt of 33 Minutes

Block production on Coinbase’s (COIN) Base network halted for 33 minutes early Tuesday following a sequencer failover that didn’t recover as expected, developers said in a post-mortem report Wednesday.

The outage began at 06:07 UTC on Aug. 5, when the active sequencer fell behind due to congestion from on-chain activity. While Base’s Conductor module — a core component of the OP Stack designed to maintain uptime — correctly attempted to shift leadership to a standby sequencer, the new instance had not been fully provisioned and was unable to produce blocks.

A sequencer organizes transactions for placing into blocks. The system correctly tried to hand off responsibility to a backup sequencer, but that backup wasn’t fully ready yet and couldn’t produce blocks.

Since it couldn’t automatically switch again, production stalled until engineers manually fixed the issue. The network was fully recovered by 06:40, according to the report.

To avoid reorganization risks — that is, when a blockchain temporarily rewrites history by replacing confirmed blocks with alternative ones — the team paused Conductor and coordinated a controlled leadership transition. This process contributed to the length of the outage.

As such, the outage highlighted a key operational risk in layer-2 rollup networks that rely on centralized sequencers to order and submit transactions. These systems remain dependent on rapid failover mechanisms and complete provisioning, and a single-point gap in this chain can lead to full network stalls.

Base said it will implement infrastructure changes to ensure all sequencers added to the cluster are Conductor-ready, even before election, and will prioritize improved test coverage to validate this logic.

The incident follows previous halts on other OP Stack chains and comes just as Base has seen record traffic from new token launches and NFT mints tied to on-chain social apps like Farcaster and Zora.

Read more: Base Network Suffers 1st Downtime Since 2023, Halts Operations for 29 Minutes