CryptoSlate
On Aug. 12, a stale default route on a Miami edge router at Teraswitch, Solana's largest infrastructure provider, spread across twelve European and Asia-Pacific data centers and knocked validators representing nearly 29% of network stake offline. The chain kept producing blocks throughout the roughly 33-minute disruption, but performance deteriorated sharply: independent monitoring firm Metrika measured skipped leader slots rising above 32% and non-vote transaction throughput collapsing from a typical 1,100 to 1,300 per second to under 300. About half an hour passed before newly produced blocks reached finality, leaving any system depending on rapid settlement materially exposed.
Why it matters
Solana's consensus requires more than 66% of stake voting to maintain finality, with a full production halt triggered at roughly 33% offline. The 29% that dropped out on Aug. 12 stayed four to five percentage points under that threshold. The Foundation disclosed that Teraswitch had hosted 38% of network stake in 2025 and that the provider's share had been reduced below 30% only shortly before the incident. That still left a single failure domain capable of placing the chain within a hair of a halt. Many independently operated validators lost connectivity at once because they shared one upstream provider, exposing the gap between on-paper validator decentralization and operational concentration.
Market impact
Settlement was delayed, throughput cratered, and any application depending on rapid Solana finality absorbed the cost. Skipped slots above 32% meant a substantial share of leader windows went unfilled, while non-vote TPS dropped by roughly three-quarters. The backlog finalized as connectivity returned, but the incident lands at a sensitive moment for an ecosystem marketing itself to institutions on uptime and speed. Teraswitch deployed provider-side hardening the same day so an invalid route can no longer block sites from using healthy local edge routers. Additional route monitoring was still in progress at the time of its September postmortem, leaving infrastructure diversity and observability as open resilience questions.
Frequently asked questions
-
What caused Solana's near-finality halt on Aug. 12?
A stale default route on a Miami edge router at Teraswitch spread across twelve European and Asia-Pacific data centers after routine transit-provider maintenance activated it, taking validators holding about 29% of network stake offline.
-
How close did Solana come to a full chain halt?
Roughly 29% of stake went offline, sitting about 4 to 5 percentage points below the ~33% threshold at which Solana would have lost finality and halted block production.
-
What was the user-visible impact during the outage?
Skipped leader slots rose above 32%, non-vote transaction throughput fell below 300 per second from a typical 1,100 to 1,300, and finality stalled for about 30 minutes before the backlog cleared.
-
What has Teraswitch done to prevent a repeat?
Teraswitch deployed provider-side hardening the same day so an invalid route can no longer prevent sites from using healthy local edge routers, with additional route monitoring still in progress at the time of its September postmortem.
-
What does the incident reveal about Solana's validator decentralization?
The Foundation disclosed Teraswitch hosted 38% of network stake in 2025, only trimmed below 30% before the incident. Many separately operated validators share one failure domain, exposing operational concentration that does not show up in on-paper decentralization metrics.
CoinDesk
Crypto News
Lookonchain
TheBlock