Solana comes close to halting after 29% of staked SOL goes offline

Solana comes close to halting after 29% of staked SOL goes offline

A routing failure at infrastructure provider Teraswitch sent roughly 90 validators offline and pushed delinquent stake close to Solana's 33.34% finality threshold.

Solana came close to losing transaction finality on Wednesday after a routing failure at infrastructure provider Teraswitch caused 28.83% of the network’s staked SOL to become delinquent.

The disruption brought Solana to roughly 86% of the 33.34% delinquent stake threshold at which the network would no longer have the two thirds supermajority required to finalize transactions. The network continued operating during the incident.

Around 90 validators were affected, collectively missing an estimated 333 SOL in staking rewards. The missed rewards are expected to be covered by validator bonds at the end of the epoch.

The incident originated from a routing fault at Teraswitch. According to the provider’s account of the event, a default route from its Miami site was advertised without the expected attributes. 

Advertisement

A route reflector in Amsterdam then distributed the route across Teraswitch locations in Europe and Asia Pacific.

Edge routers at the affected locations preferred the incorrect route while the underlying core network rejected it as invalid, leaving 12 sites without a valid forwarding route. Locations affected included sites in London, Amsterdam, Dublin, Frankfurt, Singapore, and Tokyo. North American locations were not affected.

Teraswitch identified the issue within around 10 minutes, with traffic restored by 04:16:15 UTC. Some validators remained offline longer, with certain operators experiencing around 33 minutes of downtime.

The incident also exposed the concentration of Solana validator infrastructure among individual network providers. Autonomous system AS20326, operated by Teraswitch, hosted approximately 118.9 million SOL in stake, with about 94% of that stake going offline simultaneously during the disruption.

Additional validators hosted through Latitude.sh, Limestone, Butterfly Research, and Allnodes also became delinquent during the same period, accounting for roughly another 14.1 million SOL, according to the analysis.

Marinade said 59 validators representing roughly 80.2 million SOL recovered within the same narrow period as routing service returned across Amsterdam, Frankfurt, and Tokyo, indicating that the affected validators had remained offline until network connectivity was restored rather than switching to alternative infrastructure.

The incident occurred just days after Solana reached 30 consecutive months without a network wide outage. Its last full halt occurred in February 2024, when validators had to coordinate a manual restart that took close to five hours.

Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.
Solana comes close to halting after 29% of staked SOL goes offline
Solana comes close to halting after 29% of staked SOL goes offline

A routing failure at infrastructure provider Teraswitch sent roughly 90 validators offline and pushed delinquent stake close to Solana's 33.34% finality threshold.

Share

Add us on Google

Solana came close to losing transaction finality on Wednesday after a routing failure at infrastructure provider Teraswitch caused 28.83% of the network’s staked SOL to become delinquent.

The disruption brought Solana to roughly 86% of the 33.34% delinquent stake threshold at which the network would no longer have the two thirds supermajority required to finalize transactions. The network continued operating during the incident.

Around 90 validators were affected, collectively missing an estimated 333 SOL in staking rewards. The missed rewards are expected to be covered by validator bonds at the end of the epoch.

The incident originated from a routing fault at Teraswitch. According to the provider’s account of the event, a default route from its Miami site was advertised without the expected attributes. 

Advertisement

A route reflector in Amsterdam then distributed the route across Teraswitch locations in Europe and Asia Pacific.

Edge routers at the affected locations preferred the incorrect route while the underlying core network rejected it as invalid, leaving 12 sites without a valid forwarding route. Locations affected included sites in London, Amsterdam, Dublin, Frankfurt, Singapore, and Tokyo. North American locations were not affected.

Teraswitch identified the issue within around 10 minutes, with traffic restored by 04:16:15 UTC. Some validators remained offline longer, with certain operators experiencing around 33 minutes of downtime.

The incident also exposed the concentration of Solana validator infrastructure among individual network providers. Autonomous system AS20326, operated by Teraswitch, hosted approximately 118.9 million SOL in stake, with about 94% of that stake going offline simultaneously during the disruption.

Additional validators hosted through Latitude.sh, Limestone, Butterfly Research, and Allnodes also became delinquent during the same period, accounting for roughly another 14.1 million SOL, according to the analysis.

Marinade said 59 validators representing roughly 80.2 million SOL recovered within the same narrow period as routing service returned across Amsterdam, Frankfurt, and Tokyo, indicating that the affected validators had remained offline until network connectivity was restored rather than switching to alternative infrastructure.

The incident occurred just days after Solana reached 30 consecutive months without a network wide outage. Its last full halt occurred in February 2024, when validators had to coordinate a manual restart that took close to five hours.

Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.