Sui is down again! Official: Network congestion, the mainnet is currently unable to process transactions.
The well-known public blockchain Sui in Taiwan experienced a network stall late at night on the 14th Taiwan time. The official statement indicated that the mainnet was temporarily unable to process transactions properly. Some decentralized applications (dApps) and blockchain explorer services, including Slush and SuiScan, may have experienced connection issues or transaction delays. The Sui Core team immediately intervened to address the problem and promised to release further updates once the issue is clarified.
Sui Network Stalls
Late at night on the 14th Taiwan time, the Sui mainnet experienced a network stall. The official statement pointed out that the mainnet was once unable to process transactions normally. Some dApps and blockchain explorer services, including Slush and SuiScan, may have experienced temporary disconnections or transaction delays. The Sui Core team promptly intervened and committed to providing further updates after the issue is resolved.
It is worth noting that this is not the first time Sui’s mainnet has experienced a complete outage. Looking back to November 21, 2024, the Sui mainnet completely stopped operating between approximately 1:15 and 3:45 a.m. Pacific Time, when all validator nodes simultaneously entered a crash loop, causing the entire network to be unable to process any transactions. This incident also sparked discussions in the market about “whether high-performance public chains sacrifice system stability in pursuit of throughput.”
(Sui’s first block production halt after launch: developers said the issue was minor, and the next day Franklin Templeton announced a partnership)
Recap of the previous outage cause: congestion control code triggered validator crashes
According to official technical explanations, the November 2024 outage was caused by an assert! check in the congestion control module of Sui. When certain conditions are met simultaneously, it directly causes all validator nodes to crash, leading to network stagnation.
Trigger conditions include:
- Congestion control mechanism enabled in TotalGasBudgetWithCap mode
- A transaction received by the network contains a “variable shared object as input”
- The transaction does not include any MoveCall instructions
Under these combined conditions, validators encounter errors during cost calculation, resulting in synchronization crashes.
What is congestion control? A necessary component of Sui’s high-performance design
Sui adopts an object-centric ledger model that allows for high concurrency of transactions, which is fundamental to its high throughput. However, if multiple transactions attempt to write to the same shared object simultaneously, they must be processed sequentially, which can create performance bottlenecks.
Therefore, Sui introduced a congestion control mechanism to limit the number of transactions processed for a single shared object within a given time frame, preventing the system from being slowed down by a few high-frequency shared objects. Previously, the Sui Foundation explained in an offline reading session with XueDAO that its core logic involves grouping and batching transactions with causal relationships.
Recently, Sui upgraded this mechanism by adding the TotalGasBudgetWithCap mode, which allows for more precise assessment of transaction computational costs and complexity. However, a logical vulnerability in this mode was the key factor leading to the recent mainnet outage.
After confirming the issue, the Sui team quickly submitted a patch (PR #20365) and released mainnet v1.37.4 and testnet v1.38.1 updates. The official statement noted that after the patch was released, the validator community cooperated highly in upgrading. From the patch release to full network recovery, it only took about 15 minutes, demonstrating very efficient collaboration.
This article: Sui crashed again! Official: Network stall, mainnet currently unable to process transactions. Originally published on Chain News ABMedia.