Solana Network Flirts with Disaster as Routing Error Causes Near-Finality Halt

Solana's network came perilously close to a transaction finality halt due to a routing error in the Teraswitch infrastructure. The issue began when a default route was announced from Miami without necessary specifications, causing edge routers to prefer an invalid path. This resulted in lost valid traffic routing paths in 12 locations worldwide.

The incident led Marinade Finance to analyze the effects of the problem, which revealed that approximately 28.83% of staked SOL entered a non-transactional state at its peak. The critical threshold for transaction finality on Solana is 33.34%, and if another 19.9 million SOL had been deactivated, the network would have experienced a loss of finality.

The event highlighted the risk of overcrowding in Solana's verification infrastructure, with approximately 118.89 million staked SOL on autonomous system number AS20326, representing more than a quarter of the total stake in Solana. The incident also affected validators running on other platforms, such as Latitude.sh and Allnodes.

The problem was identified within about 10 minutes, but it took around 33 minutes for some validators to return to full normal operation. Marinade's analysis noted that concentration measurements based solely on hosting providers might underestimate the true correlation risk.