Solana Stays Online as 102 of 699 Validators Stop Voting
Solana remained online on Aug. 12 even after a routing failure temporarily disrupted a large slice of its validator network. According to Solana Foundation technology executive Jacob Creech, 597 of 699 staked validators kept voting while blocks and transactions continued processing normally.
What Happened
The disruption came from an infrastructure issue at Teraswitch, which reported a malformed default route from its Miami facility that spread through its routing system and affected sites in Europe and Asia. The incident briefly made some validators unreachable, but it did not stop Solana’s mainnet from producing blocks. Affected operators recovered within about 40 minutes.
Creech described the event as a proof point for Solana’s resiliency, saying the network stayed live even as a significant portion of validators stopped voting. The official Solana status page also recorded no mainnet incident and showed 100% Mainnet Beta cluster uptime across the prior 90 days.
Validator Impact
At the peak of the disruption, 102 of 699 staked validators were not voting, which means roughly 597 validators continued supporting consensus. Marinade Finance separately estimated that around 28.83% of staked SOL became delinquent for about 33 minutes. That figure is important because Solana needs more than two-thirds of stake to participate for transaction finality to continue normally.
The delinquent stake level came close to the 33.34% threshold where finality would stop entirely. Even so, the network remained operational and continued processing transactions. Marinade said the event showed how close the system came to a critical state without actually crossing that line.
Infrastructure Failure
Teraswitch said the problem started with a malformed route from its MIA1 facility in Miami, which was then propagated by a route reflector in Amsterdam into parts of Europe and Asia Pacific. That caused connectivity loss at 12 sites, including London, Amsterdam, Dublin, Frankfurt, Singapore, and Tokyo. North American sites were not affected.
Engineers identified the issue within about 10 minutes and removed Miami from the private backbone. Service was restored at 04:16:15 UTC, and Teraswitch later pushed a global configuration change to reduce the chance of a similar routing fault blocking traffic in the future. The company said its root-cause investigation is still ongoing.
What It Means
The episode highlights that blockchain resilience depends not only on software, but also on the physical infrastructure behind validators. Even if validator software keeps running, a concentration of stake on shared hosting or routing systems can still expose the network to large coordinated disruptions. Marinade noted that one autonomous system held about 118.9 million SOL, or more than one quarter of all staked SOL, and that roughly 94% of that stake went offline together.
This also contrasts with Solana’s February 2024 halt, when the network stopped producing blocks and required a coordinated restart. In the Aug. 12 incident, no mainnet restart was needed, and transaction flow continued throughout. Solana’s growing use of multiple validator clients, including Firedancer, adds another layer of resilience on the software side, but this event tested infrastructure diversity instead.
Why It Matters for Investors
For SOL holders and network watchers, the key takeaway is that Solana did not fail at the consensus layer, but it did show concentration risk in validator hosting. That kind of risk can matter just as much as software reliability because it affects finality, uptime perception, and confidence in the network’s decentralization. The fact that the network stayed online is positive, but the margin was narrower than a simple validator count suggests.
The event may also push operators to improve redundancy across autonomous systems and data centers. Marinade said it plans to review concentration limits and automatic failover arrangements, which could lead to stricter operational standards for validators. For Solana, the real test is whether those changes reduce the chance that a single routing failure can threaten such a large share of staked voting power.
FAQ
Did Solana go offline?
No. Solana continued producing blocks and processing transactions during the incident.
How many validators were affected?
Creech said 102 of 699 staked validators stopped voting at one point, while 597 kept voting. Marinade’s analysis suggested about 90 validators were directly affected by the routing failure.
How long did the disruption last?
Affected operators recovered within about 40 minutes, and Marinade estimated the delinquent stake remained elevated for about 33 minutes.
Did the outage hit Solana’s status page?
No mainnet incident was recorded, and the official status page showed 100% Mainnet Beta cluster uptime over the prior 90 days.
Was finality at risk?
Yes, but only briefly. Marinade said the network reached 28.83% delinquent stake, close to the 33.34% point where transaction finality would stop.
What caused the issue?
Teraswitch traced it to a malformed default route that spread from Miami through its routing infrastructure and affected multiple regions.
