At Storj, resilience isn't just a design principle — it's something we live and prove every day. Earlier today, I saw it firsthand when a key facility supporting our network went offline — and our customers never missed a beat.
During a scheduled maintenance window, a key facility providing storage capacity for our network — responsible for 40 Gbps of traffic — experienced a complete outage. In an instant, traffic dropped from 40 Gbps to zero. The facility was offline for several minutes.
Our systems detected the failure immediately. Traffic was automatically rerouted. While our monitoring alerted us and a few brief upload errors were logged, reads saw no impact. For a major enterprise customer, the real-world impact was limited to just 4% upload errors during that short window — a minor blip.
Now imagine that same incident happening at a single-datacenter provider.
They wouldn’t have seen a 4% error rate.
They would have seen a 100% outage.
Built Different by Design
This wasn’t luck. It was the result of deliberate engineering:
- Triple-redundant metadata ensures there’s never a single point of failure.
- Edge locations are designed to seamlessly failover if any facility goes down.
- Distributed storage spreads data across independent nodes worldwide, not trapped in a single building.
We don’t patch over failures. We design them out.
We've weathered edge failures, server deaths, drive failures, and ISP disruptions — all without customer downtime. But today’s event was a reminder of how well our architecture performs when it matters most.
Where Others Invest in Redundancy, We Invest in Resilience
Traditional providers pour money into hardware and facility redundancy, trying to patch over single points of failure. At Storj, we took a different path: software redundancy and distributed design.
The truth is simple:
Competitors build patchworks of hardware and hope for the best. We deliver global resilience by design — not a workaround, but a platform built for scale and reliability.
That's why while others lose customer data, trust, and confidence — Storj customers experience something different: real resilience, built-in.
Closing Thoughts
Today’s 40 Gbps outage showed — once again — why architecture matters.
At Storj, resilience isn’t an afterthought. It’s fundamental.
And when the unexpected happens, it’s not about if you fail over — it’s about whether your customers even notice.
— Jacob Willoughby, CTO at Storj
Want to dig deeper into how Storj is engineered for resilience?
Download the Storj CTO’s Guide to the Distributed Cloud for an inside look at the architecture that keeps your data available—even when the unexpected happens.
