![Sun [New]](https://s2.coinmarketcap.com/static/img/coins/64x64/10529.png)
![BitTorrent [New]](https://s2.coinmarketcap.com/static/img/coins/64x64/16086.png)
This article was first published on TurkishNY Radio.
Telemetry shared by Solana ecosystem operators says the network absorbed a distributed denial-of-service wave that peaked around 6 terabits per second, yet block production stayed steady with no reported downtime and no meaningful increase in user fees.
What “healthy” looked like under pressure
Monitoring circulated during the incident reported a median confirmation around 450 milliseconds, p90 under 700 milliseconds, and slot latency holding roughly 0 to 1 slot. These indicators are where stress tends to surface first, because they capture both user experience and validator timing.
Solana Labs co-founder and president Raj Gokal echoed the takeaway, saying the campaign had “zero effect on performance,” aligning the narrative with the published telemetry.
Why this matters for credibility
Solana’s reliability narrative has been shaped by earlier outages tied to bugs and spam floods that behaved like DDoS events. A recent recap highlighted multiple incidents across 2020 through 2024, which is why the ecosystem still gets judged on resilience, not raw throughput.
Separate reporting cited Solana’s status history and noted the last recorded outage was February 6, 2024, framing the current uptime streak as a meaningful shift in operating reality.
Scale, optics, and the money argument
Reporting framed the peak as one of the larger internet-scale attacks, with Solana’s reported traffic sitting in the same conversation as cloud-provider incidents.
Anatoly Yakovenko leaned into the economics, writing that “someone is spending as much as the chain makes in revenue to send it.” The point is not that attacks are good; it is that surviving them without user pain raises the cost of disruption.
The defenses that likely made the difference
Reporting around the event pointed to QUIC for transport and local fee markets that can drop or price spam nearer to ingress, limiting how much hostile traffic reaches the parts of the network that users compete over.
Michael Hubbard, interim CEO of Sol Strategies, described an “incredible load” and credited high-availability tooling that helps validator clusters fail over cleanly instead of requiring the sort of manual recovery Solana became known for in earlier years.
Consolidation, and the fine print
The same reporting said active validator operators were down by more than 35% in 2025. That can heighten centralization concerns, but it can also mean the remaining set is better provisioned to absorb internet-scale pressure.
An ecosystem contributor also argued the 6 Tbps figure may have been the peak of a short burst rather than a sustained constant rate, a useful reminder that peak numbers are not the same as sustained load.
Market context
One report said SOL fell below $130 to a seven-month low during the broader selloff even as the network held steady under attack. During this reporting window, SOL traded around $127.90.
Conclusion
One report noted it could not independently verify the full scale of the attack, so caution is warranted. Still, named confirmations and specific telemetry, paired with a lack of visible disruption, make this episode a cleaner data point for Solana’s “uptime era” than any marketing promise. For developers and market makers, predictability beats peak speed, because applications can plan risk when the chain behaves similarly in calm and chaos.
FAQ
Did Solana go down?
Reporting described no downtime and stable confirmations.
What does p90 under 700 milliseconds mean?
It means 90% of observed confirmations were faster than that threshold.
Why do local fee markets matter?
They can contain congestion to hotspots instead of spreading higher fees network-wide.
Glossary of key terms
DDoS: A coordinated traffic flood intended to degrade or interrupt service.
QUIC: A transport protocol used to manage and filter traffic more effectively.
Slot latency: A measure of delay in block production relative to expected slot timing.
References
