Sui Recovers From 6 Hour Halt After Safety Mechanism Stops Consensus

TLDR

• A consensus bug caused validators to produce conflicting checkpoints, halting Sui for 6 hours.
• No certified transactions were rolled back and no user funds were at risk during the halt.
• The error was unrelated to traffic, usage spikes, or external attacks, Sui confirmed.
• Recovery began after validators fixed logic and safely replayed consensus from divergence.

---

Sui has confirmed a six-hour mainnet disruption on January 14, 2026, caused by an internal consensus processing bug. While the network temporarily stopped processing transactions and certifying checkpoints, no user funds were lost, and no data was rolled back. The issue triggered Sui’s safety-first system to halt activity rather than risk inconsistent state finalization.

Network Activity Halted After Validator Divergence

On January 14, 2026, Sui’s mainnet experienced a six-hour disruption caused by a bug in the validator consensus processing logic. The problem emerged from a rare edge case in the way conflicting transactions were handled under certain garbage collection conditions.

This led to different validators computing different outputs from the same consensus data. Consequently, validators began signing conflicting checkpoint digests. Since over one-third of stake supported different versions, no checkpoint could be certified, and progress stopped.

The Sui team in a blog post stated that the disruption was not related to network usage or external interference. Throughout the halt, read-only Remote Procedure Call (RPC) queries continued to serve the last certified state to users.

Safety Systems Activated to Prevent Forks or Rollbacks

The network’s architecture is designed to halt if checkpoint inconsistency is detected. During the incident, validators could not reach agreement, and this automatic fail-safe mechanism was triggered.

According to Sui, “No certified state forks occurred and no certified transactions were rolled back.” Funds remained safe and untouched. Although no transactions were processed during the outage, previously confirmed state was available for reads.

Sui emphasized that the system functioned as intended under stress, halting transaction processing to prevent inconsistent state finalization. This decision prioritized safety over uptime.

Bug Fix and Validator Recovery Steps

The problem was traced to a bug in the consensus commit logic. A fix was quickly developed to remove incorrect consensus data and restore correct validator behavior. Validators from Mysten Labs first deployed a canary version of the fix to verify checkpoint generation.

Once confirmed, the wider validator set adopted the updated binary. They replayed consensus commits from the divergence point and resumed signing matching checkpoint digests. 

Once a quorum was achieved, normal network activity resumed. The entire recovery took approximately six hours from the time the issue began. During this period, around $1 billion in on-chain value remained temporarily inactive, but untouched.

Planned Improvements to Shorten Future Recovery Time

Following the event, Sui announced improvements to reduce recovery time if similar issues occur. These include better tools to detect checkpoint inconsistencies early and pause consensus faster.

The team also plans to improve automated recovery tools to clean up internal state safely. In addition, testing frameworks have been updated to simulate such consensus edge cases and ensure fixes are validated before mainnet deployment.

Sui confirmed this was the second major outage since its 2023 launch, following a brief issue in 2024. However, the market response remained stable, with minimal impact on the SUI token price. Sui concluded that the outage, while disruptive, proved the network’s focus on protecting state consistency over continuous uptime.

The post Sui Recovers From 6 Hour Halt After Safety Mechanism Stops Consensus appeared first on CoinCentral.