Sui Network's mainnet stalled for approximately two hours on May 28, 2026, marking the third major outage for the Layer 1 blockchain since its May 2023 launch. The disruption — caused by what the project described as a "network stall" — temporarily halted transaction processing across all dApps a...
"Uptime matters. Ethereum: flawless uptime since launch. Solana: 18 months since last outage. Sui: 9 months since last outage. Choose wisely." — Onchain Foundation, via X (May 2026)
Sui Network's mainnet stalled for approximately two hours on May 28, 2026, marking the third major outage for the Layer 1 blockchain since its May 2023 launch. The disruption — caused by what the project described as a "network stall" — temporarily halted transaction processing across all dApps and DeFi protocols on the chain. SUI token fell 8-9% in the 24 hours surrounding the event, trading near $0.91 with market capitalization declining to approximately $3.6 billion.
The incident arrives five months after a six-hour consensus outage in January 2026 that froze an estimated $1 billion in on-chain assets, and 18 months after a two-hour scheduling bug crash in November 2024. Cumulatively, Sui has logged approximately 10 hours of total downtime across three incidents in 19 months of operation — a rate that places it between Solana's historically troubled early years and Ethereum's unblemished record.
This report benchmarks the reliability track records of four major Layer 1 networks — Ethereum, Solana, Sui, and Aptos — and examines the economic costs of outages for users, DeFi protocols, and institutional adopters evaluating infrastructure risk.
At approximately 07:15 PDT on May 28, 2026, Sui's official status page flagged a major outage for mainnet validators. Block production ceased and no new checkpoints appeared on explorers including SuiScan. The Sui Core team issued a statement: "Sui Mainnet is currently experiencing a network stall. The Sui Core team is actively working on a solution. Be aware that transactions may be paused at this time."
By 07:36 PDT, engineers reported identifying the issue and began deploying a fix. Public RPC nodes remained operational, but settlement and validator coordination were interrupted. The network was restored after approximately two hours.
At the time of reporting, Mysten Labs had not disclosed the specific technical cause. The SUI token, already under pressure from a broader crypto selloff triggered by U.S.-Iran military escalation, traded at approximately $0.91 — down roughly 9% over 24 hours. The broader crypto market saw $928 million in liquidations the same day, complicating attribution of SUI's decline to the outage alone.
Sui currently holds approximately $567 million in total value locked across DeFi protocols, according to DefiLlama — down from a peak of $2.6 billion in October 2025.
Three major outages in 19 months of mainnet operation:
| Date | Duration | Cause | Assets Affected | |------|----------|-------|-----------------| | Nov 21, 2024 | ~2 hours | Transaction scheduling bug; validators crashed | SUI dropped ~10% to $3.42 | | Jan 14-15, 2026 | ~5 hrs 52 min | Consensus divergence; 68% of validators experienced clock sync failures | ~$1B in assets frozen | | May 28, 2026 | ~2 hours | "Network stall"; cause undisclosed at time of reporting | ~$567M TVL inaccessible |
The January 2026 post-mortem, published January 28, attributed the failure to "checkpoint certification divergence" — a condition where validators could not reach supermajority agreement on block validity due to timing discrepancies in internal clocks during a routine protocol update. Preventive measures announced included redundant time synchronization services, enhanced checkpoint validation logic, and automated alerts for validator timing drift.
The pattern is notable: each incident has involved a different failure mode. The November 2024 crash stemmed from a scheduling bug. The January 2026 outage was a consensus-layer timing failure. The May 2026 cause remains undisclosed. This suggests the issues are not recurrences of the same bug, but rather emergent failures in different subsystems of a relatively young codebase.
A comparative view of outage histories for four major Layer 1 networks:
Ethereum
Solana
Sui
Aptos
The data illustrates a reliability spectrum. Ethereum occupies one extreme with zero downtime in over a decade of operation, enabled by massive validator diversity and multiple independent client implementations. Aptos, despite sharing Move-language roots with Sui, has maintained a clean record since its sole outage 31 months ago. Solana's trajectory is instructive: after accumulating 68.5 hours of downtime across eight incidents in its first four years, the network has been outage-free for over 15 months.
Blockchain outages carry costs that extend beyond the duration of the halt itself:
Direct costs during outage:
Post-outage costs:
For context, during the January 2026 outage, approximately $1 billion in on-chain assets were frozen for nearly six hours. While the Sui Foundation confirmed no funds were at risk and no network fork occurred, the inability to transact for six hours in a market that operates 24/7 represents a concrete economic disadvantage relative to chains with continuous uptime.
Research on DeFi security events more broadly shows that 55% of disruption events lead to significant negative price impacts averaging approximately 14%, with indirect economic losses in related ecosystem market capitalization exceeding direct costs, according to academic studies on DAO economic impacts.
The structural differences between these networks' validator architectures help explain their divergent reliability records.
Ethereum runs on a massive decentralized validator set (700,000+) with multiple independent client implementations (Prysm, Lighthouse, Teku, Nimbus, Lodestar). A bug in any single client affects only a fraction of validators. This client diversity is widely considered the primary reason Ethereum has never experienced a full network halt.
Solana operated with a single client (Solana Labs) for most of its outage-prone period. The introduction of the Firedancer client by Jump Crypto and the Agave client brought client diversity to Solana for the first time, coinciding with its extended outage-free period. The network runs approximately 1,800 validators.
Sui operates with 100 validators — the smallest set among major Layer 1s — running a single client implementation from Mysten Labs. The consensus mechanism is a modified Byzantine Fault Tolerance protocol optimized for parallel transaction processing. The January 2026 post-mortem revealed that 68% of validators experienced simultaneous clock synchronization failures, a failure mode enabled by the homogeneous client environment. With only 100 validators, the coordination required for recovery is simpler, but the risk of correlated failures is higher.
Aptos uses a similar Move-based architecture with AptosBFT consensus and approximately 100 validators. Despite architectural similarities to Sui, Aptos has maintained a clean record since October 2023, suggesting that implementation specifics — not the underlying language or consensus family — drive reliability outcomes.
Solana's history offers a relevant case study. Between December 2020 and February 2024, the network suffered eight major outages totaling 68.5 hours of downtime. Causes ranged from bot-driven spam floods (the September 2021 Grape Protocol incident caused a 17-hour outage) to low-level runtime bugs (the June 2022 durable nonce bug) to infrastructure failures (the February 2023 Turbine block that caused a 19-hour outage).
Solana's path to stability involved several structural changes:
The network survived a DDoS attack peaking near 6 Tbps in December 2025 without downtime — among the largest ever recorded against a distributed system.
For Sui, the Solana experience suggests that client diversity — currently absent — may be the most impactful structural improvement available. With a single Mysten Labs client serving all 100 validators, any software bug has the potential to halt the entire network simultaneously, as demonstrated in all three outage incidents.
The May 28, 2026 outage is not, by itself, a catastrophic event. Two hours of downtime on a two-year-old blockchain is within the range of growing pains experienced by every high-performance Layer 1. The concern is the pattern: three incidents in 19 months, each with a different root cause, on a network with no client diversity and the smallest validator set among its peers.
The economic value framework for evaluating blockchain networks must account for reliability as a first-order variable, not a secondary feature. A chain that cannot guarantee continuous operation imposes hidden costs on every user, protocol, and institution that builds on it — costs that compound with each additional outage.
Solana's trajectory provides evidence that these problems are solvable. But the solutions — client diversity, expanded validator sets, enhanced testing regimes — require sustained engineering investment and, often, years of iteration. Whether Mysten Labs pursues this path, and at what pace, will determine whether Sui's reliability record converges toward Ethereum's standard or remains in a pattern of periodic disruption.
The data, for now, speaks for itself: 10 hours of cumulative downtime across three incidents, $567 million in TVL exposed to outage risk, and a validator architecture that structurally favors correlated failures. Institutions building infrastructure selection frameworks should weight these figures accordingly.