The latest issue started with a bad internet route from Teraswitch’s Miami facility that then spread to data centers across Europe and Asia, cutting off validators in London, Amsterdam, Frankfurt, Singapore and Tokyo. North America stayed online. The company fixed the issue in about 10 minutes, and traffic was flowing again by 4:16 a.m. UTC.
One single network operator, identified as AS2032, controlled more than a quarter of all the tokens people had locked up to secure the network, which was more than the Solana-prescribed safety limit. Almost all of those tokens went offline at the same time. Other companies lost another 14 million tokens in the same short period. Most of the affected validators, including the big one called Helius, stayed offline for the full 33 minutes because their backup systems never switched on.
This whole event is a clear warning: if more than one-third of the network’s tokens ever go offline at once, the entire blockchain freezes for every single person holding SOL, and there is no quick way to fix the bigger damage that would follow.
Solana’s network remained operational
Solana Foundation’s VP of Tech Jacob Creech framed the episode as evidence that the network’s design held up under pressure. Put simply, the network came close to the freeze threshold, its performance was unaffected in any way.
