Solana: Building, verifying and earning trust in an open environment

Foundations of Trust for Infrastructure Providers: Transparency, Learning, and Open Build

Infrastructure providers that support today's Internet, cloud services, and financial systems have earned our trust through years of transparency, learning from mistakes, and Open Build. Solana must win trust in the same way.

Since its launch in 2020, the network has experienced early outages, unpredictable surge in demand, and the most intensive period of activity in the cryptocurrency industry. Each test allows us to learn valuable lessons, and the network has become stronger after each challenge, establishing its status as a tough and field-tested network.

Proof of the most recent stress test

The latest proof appeared on August 12, 2026, when Solana's largest infrastructure provider experienced a routing failure, causing nearly 29% of network pledges to go offline. Despite this challenge, blocks continue to be generated and transactions continue to be implemented. Users are almost unaware. The infrastructure provider achieved a full recovery in just over 30 minutes.

This does not mean that this incident is irrelevant, but rather that the network has sufficient redundancy to absorb the impact in the event of infrastructure failure without disrupting service. In contrast, when AWS's us-east-1 area failed last October, most Internet functions were paralyzed for most of the day. However, when Solana's largest infrastructure provider encountered routing failures, the underlying network remained stable, reliable and resilient.

I believe this resilience is no accident. It is the result of years of honestly facing failures, quickly fixing problems, and always working to make Solana the most reliable link.

Solana: Can stand the test of practice, not a fluke

Like almost all critical infrastructure, Solana has been tested repeatedly and openly under pressure. What makes Solana different is the way we respond. We and our developers see tough times as opportunities to make the network stronger. We publish public disclosure and diagnostic reports and introduce fixes that go beyond fixing immediate problems to prevent similar problems from happening again.

Evolving from lessons

For example, in the early days of networks, bobot-driven traffic exposed the challenges of validators in handling incoming traffic. To improve transaction intake, we increased the connection capacity of the verifier and introduced pledge-based Quality of Service (SWQoS) to better manage traffic.

Another lesson comes from the transaction scheduler, which once rewarded the behavior of sending the same transaction repeatedly. We completely redesigned this to make repeating transactions meaningless.

This goes back to the recent routing failure. Last year, our team found that TeraSwitch hosted 38% of the network's pledges. This concentration poses a risk, so we took proactive measures to reduce it below 30%. It was because of these preventive efforts that the network itself was able to continue to operate when the routing failure occurred on August 12. Even DevNet, which lost about three-quarters of its pledge amount in the same incident, can recover on its own without a coordinated restart. The network has self-healed.

In my opinion, this is a model of a resilient decentralized network working as expected. While we cannot completely avoid failures, we can design systems so that failure of individual components does not drag down the entire network. In addition, the implementation of all the characteristics we have accumulated in production practices has together created Solana's resilience today.

Solana Uptime: Maintain 100% for two and a half years

Organizations typically use "nine" to measure reliability. Four nines (99.99%) mean approximately 53 minutes of downtime per year. Five nines mean only about five minutes of downtime per year.

AWS, the world's largest cloud provider, will provide credit compensation if it fails to meet the four nines (99.99%) standards. In fact, AWS does fall below that standard-in the past two years, the AWS us-east-1 region has experienced approximately 15 hours of major outages, or approximately 99.91% availability. GitHub, the critical infrastructure for most software teams around the world, promises three nines (99.9%) availability, which means they incur nine hours of downtime per year.

However, since February 2024, Solana has maintained 100% uptime. During this period, it carried more daily transactions than all other major blockchains combined and processed record single-day transactions. As usage increases, reliability becomes more meaningful. Networks that carry small amounts of activity may seem reliable, but networks that process millions of transactions per second, manage billions of dollars in assets, and support real financial activity must prove reliable under high loads.

Solana did this, giving financial institutions and businesses the confidence they needed to build on this network.

Cumulative effect of confidence

Resilience is not only about surviving the worst days, but also about achieving future ambitions. Because the network has proven its strength, Solana engineers can now focus on driving new improvements.

• Block time has been shortened to 300 milliseconds (the target next step is 200 milliseconds).
• Storage costs (rent) have been reduced by 90%.
• Now supports larger transactions, enabling greater certification and better privacy protection.
• Alpenglow is expected to be released on the main network later this year, and the final confirmation time will be shortened to 150 milliseconds.
• Recently, just after the TeraSwitch event, the network achieved 5,000 Sustainable User Transactions (TPS) per second for the first time.

In short, we have made the network faster, cheaper and more efficient. Only on a trustworthy network can developers focus on such improvements.

Community is resilience

Solana's resilience is not only rooted in the network, but also deeply rooted in our culture. For years, Solana's developers ignored ups and downs and noise and focused on delivering results. They built-and are still building-the infrastructure that supports $17.5 billion in stablecoins, more than $4 billion in real-world assets, and growing remittance flows.

This culture also means constantly improving standards. Earlier this year, the Solana Foundation and Asymmetric Research launched STRIDE, a comprehensive Solana DeFi security initiative that provides independent security assessments, publicly discovered issues, and ongoing operational security and threat monitoring for eligible protocols. This is another example of ecosystems learning from real-world challenges and turning them into stronger practices.

Solana: A network that can stand the test of combat

A network that has never seen a problem is a network that has never been proven. Solana has been criticized many times, but despite this, the ecosystem is still being built and improved.

For our users-including developers, founders, traders and large companies like Western Union, Franklin Templeton, MoneyGram, SoFi, BlackRock and others-this dedication and persistence create a chain that proves its commitment and resilience. Solana has become the fastest, most widely used and highest-paying blockchain. We are expected to become the financial infrastructure for the Internet's 5.5 billion users, fueling the ongoing token super cycle.

We know that the next stress test will come. But when it comes, we will continue to do as usual: publish diagnoses, roll out fixes, and continue to build and become stronger.