Base’s state update system went down and nobody noticed

Blocks reportedly continued to be produced and the network appeared to function normally. But behind the scenes, a key component which aws responsible for updating Base’s state on Ether stalled for more than 30 hours. This event came to notice after developers flagged that the state updates and withdrawals to Ether were stopped.

Base’s 30-hour glitch raised layer-2 concerns

Developer donnoh.eth addressed the issue in an X post. He noted that the outage went unnoticed because Base withdrawals already require a seven-day challenge period. 

He stated that “It’s kind of crazy that Base state updates have been down for over 30 hours now because of some bug related to the recent upgrade and no one even noticed just because withdrawals take seven days anyway.”

According to Base’s status page, the problem was traced to the network’s Trusted Execution Environment (TEE) enclave. The malfunction prevented the proposal system from generating the state updates needed to anchor Base’s activity back to Ethereum.

The chain kept itself processing and was moving transactions normally. Meanwhile, Base’s state effectively stopped updating until the issue was resolved.

In the case of rollup like Base, transitions are executed on L2 before compressed state commitments are periodically posted back to Ether. The TEE helps to generate cryptographic attestations. This helps in proving that state transitions were computed correctly.

This suggests that when that system stops working, users can continue transacting on Layer 2. Meanwhile, the settlement pipeline connecting the network back to Ethereum can grind to a halt.

No funds were lost and the outage did not expose user assets to theft. But still, it did temporarily freeze one of the most important pieces of infrastructure supporting the rollup. It stands crucial that, as it happened just a few days after Base deployed its Azul upgrade. It was designed to improve scalability and reportedly increase throughput to as much as 5,000 transactions per second.

Despite this, the network found itself struggling. 

Base and Sui face different failure modes

Earlier this year, Base saw periods of transaction delays during heavy network activity. However, those issues never halted settlements. Yet, they exposed capacity constraints as usage continued growing.

Base is not alone that faced this problem. Sui reported a consensus failure that disrupted transaction processing for roughly six hours in January. The network experienced multiple outages tied to software bugs introduced during protocol upgrades. It temporarily freezes transfers, DeFi activity, and NFT transactions.

The tech involved behind those incidents are very different. Base saw outage that involved a TEE-assisted proving mechanism. On the other side, Sui’s problems emerged from validator consensus and gas-accounting logic.