TLDR
- A routing failure at data center provider Teraswitch knocked nearly 29% of Solana’s staked tokens offline on Wednesday
- The network came within roughly 20 million SOL of the 33.34% threshold that would have caused a full freeze
- The outage originated at Teraswitch’s Miami facility and spread to data centers across Europe and Asia
- One single network operator controlled more than a quarter of all staked SOL, above Solana’s own safety limit
- Solana Foundation said blocks and transactions never stopped, calling it proof of the network’s resilience
Solana came close to a complete network freeze on Wednesday after a routing failure at data center provider Teraswitch knocked nearly 29% of staked tokens offline. The incident raised fresh questions about how concentrated Solana’s infrastructure really is.
Solana Stays Online as Outage Disrupts 102 of 699 Validators
Solana Foundation technology executive Jacob Creech said an infrastructure provider used by some validators suffered an outage overnight, but the Solana network continued producing blocks and processing transactions… pic.twitter.com/lNqpmGYIdX
— Wu Blockchain (@WuBlockchain) August 13, 2026
The failure started at Teraswitch’s Miami facility. A default route was incorrectly advertised and then spread by a route reflector in Amsterdam to locations across Europe and Asia. That left 12 data center locations without a usable network path, including sites in London, Amsterdam, Dublin, Frankfurt, Singapore and Tokyo. North American locations were not affected.
Staking platform Marinade reported that around 90 validators went delinquent during the outage. Together they held 28.83% of total staked SOL. The network came within about 20 million SOL of the 33.34% threshold at which Solana can no longer reach finality.
Finality is the point at which transactions become irreversible. If more than a third of staked tokens go offline at the same time, the entire blockchain stops finalizing transactions for everyone.
Teraswitch identified the problem within around 10 minutes. Traffic was restored by 4:16 a.m. UTC. The 90 affected validators lost a combined 333 SOL in rewards, which Marinade said would be covered by validator bonds.
Stake Concentration Raised as Key Risk
Marinade pointed to stake concentration as the bigger concern. One network operator, identified by the ASN AS20326, held 27.34% of total staked Solana at the time. That was above the 25% limit set under Solana’s own stake distribution rules. Around 94% of the SOL linked to that operator went offline during the incident.
A further 14.1 million SOL went offline across validators hosted by Latitude.sh, Limestone, Butterfly Research and Allnodes. Marinade said it could not confirm whether those outages were connected to the same routing failure.
Marinade also flagged its own stake allocation, noting that four ASNs hold two thirds of the stake it allocates. “Nobody should be comfortable with that, us included,” the platform wrote.
Foundation Says Network Held Up
Solana Foundation VP of Tech Jacob Creech said the episode showed the network’s design working as intended. He noted that 597 of 699 staked validators kept voting throughout the incident. Affected validators recovered within 40 minutes.
“Because Solana validators are distributed across independent infrastructure providers, the failure of a single provider did not interrupt the network,” Creech wrote on X.
Solana has $4.3 billion locked in DeFi protocols and has faced multiple outages in previous years. A February 2024 halt took around five hours to restart. Solana was trading at around $75.79 at the time of reporting.
