Inside Sui's 48-Hour Mainnet Crisis

On May 31, 2026, the Sui Foundation published one of the most transparent post-mortems the crypto industry has seen, detailing how a single protocol upgrade caused its Layer-1 mainnet to grind to a halt three separate times in just over 48 hours . What made the report remarkable wasn't just the technical detail, but a rare admission: the team knowingly shipped an interim fix that had a "recognized probability of causing another halt"—and it did exactly that .

Here is a complete breakdown of what went wrong, how the Foundation recovered, the market fallout, and what changes are coming to prevent a repeat.

The root cause: a cascading failure from v1.72

All three outages stemmed from the v1.72 protocol upgrade, which introduced a new "address balances" feature . This new component interacted unexpectedly with two critical subsystems: the gas charging logic and the consensus mechanism's randomness state .

Outages 1 and 2: the gas charging crash bug

The first mainnet stall began on May 28, lasting roughly 6 hours and 44 minutes . A crash bug in the gas charging logic—specifically in how the new address-balances code handled "gas smashing," the process of combining input coins to pay transaction fees—caused all validators to enter a crash loop . Block production stopped entirely. Approximately $1 billion in on-chain assets were temporarily frozen .

Engineers identified the bug and deployed a patched binary. Once more than two-thirds of staked value upgraded, the network resumed . However, the same root cause triggered a second halt later that day . The interim fix was designed for speed, not completeness, and the known risk materialized .

Outage 3: the randomness-state failure

On May 29, after a more stable fix was deployed for the gas charging bug and validators restarted, a separate latent bug surfaced. The validators failed to properly preserve randomness state across the restart, triggering a fresh consensus halt that lasted over three and a half hours . The team then developed and deployed a second patch specifically addressing this state preservation issue. Validators were re-upgraded across May 29–30 until full participation was restored .

How the Foundation resolved the outages

The recovery process was an iterative scramble across three distinct failures:

• For the first gas-charging crash, engineers developed a hotfix binary and coordinated an emergency upgrade among validators. The network resumed when the upgrade threshold was met .
• For the recurrence, the team iterated on the fix to achieve stability, having knowingly accepted the risk of a repeat failure .
• For the randomness-state stall, a new patch was developed, tested, and deployed to validators over the following day .

Throughout the incident, the Sui Foundation emphasized that no user funds were at risk, no committed transactions were reverted, and wallet security mechanisms remained operational .

Market impact: SUI takes a hit

The outages had an immediate and measurable effect on the SUI token and market sentiment:

• Token price: SUI dropped approximately 6.6% on the first outage day and around 3.27% over the full 36-hour window, with steeper intraday swings .
• Exchange disruptions: Major Korean exchange Upbit suspended SUI deposits and withdrawals, further pressuring liquidity .
• Sentiment shift: Broader risk-off sentiment for Sui-specific positions set in, compounded by whale activity and supply-overhang concerns .

Future safeguards: what the Foundation is changing

In its post-mortem and subsequent announcements, the Foundation outlined several concrete measures to prevent similar incidents :

• Stronger pre-release testing: More rigorous integration testing for new features, with specific focus on gas logic and consensus interactions, before mainnet deployment.
• Validated restart procedures: Updated validator tooling to ensure randomness state is correctly checkpointed and restored across node restarts.
• Automated monitoring and alerting: Enhanced on-chain monitoring designed to detect anomalous validator crash-loop patterns earlier.
• Staged rollout process: A slower, more conservative upgrade rollout with mandatory testnet validation windows to prevent deploying known-risk fixes under time pressure.
• Commitment to transparency: A stated culture shift toward publishing detailed post-mortems with root-cause analysis and measurable action items after any future network disruption.

The bigger picture: speed versus stability

The Sui Foundation's post-mortem was unusually candid. By admitting it knowingly deployed a risky interim patch, the Foundation exposed a tension familiar to many high-stakes engineering teams: the pressure to restore service fast versus the discipline to ensure a fix is complete . For a blockchain positioning itself for institutional-grade use, the incident raises legitimate questions about upgrade governance, testing rigor, and whether the network's reliability posture matches its ambitions .

The safeguards announced are a start, but their effectiveness will only become clear the next time Sui faces a complex upgrade under the public spotlight.