Base recovers from mainnet transaction delays caused by a batcher backlog

What happened

A transaction-volume spike created a Base batcher backlog. The sequencer temporarily reduced block sizes while pending data was submitted to Ethereum, causing some user transactions to be delayed or dropped and some nodes to fall behind.

What changed

For approximately three hours, effective mainnet transaction capacity fell to roughly one-third of its target and users could need to resubmit dropped transactions. The incident exposed batcher throughput and node synchronization as operational dependencies for timely Base bridge use.

What did not change

Base reported that user funds were never at risk, normal processing recovered, and no bridge custody, withdrawal contract, dispute-game authority, pause role, or finality rule changed.

Confirmed

• The incident began on April 9, 2025 at approximately 01:15 UTC.
• The sequencer reduced block sizes to about 20 Mgas versus a 60 Mgas target for roughly three hours.
• Some transactions were delayed or dropped from the mempool, and some node operators experienced synchronization problems.
• Base reported that the backlog cleared, normal processing resumed, and the incident was fully resolved.

Still open

• Whether the planned batching and mempool changes were subsequently deployed and independently verified.
• The recurrence risk under a comparable transaction-volume or L1-submission backlog.
• Whether any bridge-specific transactions experienced materially longer completion than the broader transaction-pool impact described.

What it means for an advisor

• Review the Base bridge memo's operational assumptions for sequencer, batcher, RPC, and node availability.
• Preserve transaction resubmission and alternate-RPC procedures in the bridge operating runbook; the resolved event does not independently require restricting the approved bridge.