DNN — the Distributed Node Network (formerly USDNN) — is officially live as of September 24, 2026. Three nodes serve attestation, verification, and anti-cheat from coast to coast: no single point of failure, faster responses no matter where you are, and automatic failover if any node ever has issues.
West, East, and our DB node are all online. If one node ever has issues, the others pick it up automatically.
One primary for truth, replicas everywhere for speed. Purple holds the West, green holds the East, and Central rides whichever answers faster.
Purple West (Los Angeles primary) replicates out to both green edges; Central has no box of its own — Texas and the middle of the country ride whichever of West or East answers faster.
Static Labs infrastructure used to run on a single fail-point VPS in one data center — high latency for US-Central and US-East, and a mass of resources we never pushed to their limits while overpaying for them.
A request lands on the nearest healthy node. Reads — project data, ban verdicts, session lookups, dashboard cards — are served from that node's own Postgres replica in single-digit milliseconds.
Writes — bans, project changes, account updates — commit once on the Los Angeles primary, so there is exactly one source of truth and no merge conflicts, ever. All databases are connected to each other and heavily optimized, so everything responds in practically the same time no matter which node you land on. High-volume telemetry (counters, player events, device touches) is written asynchronously in the background, so it never blocks a response.
No region picker, no manual failover. Every client hits the closest healthy node and gets the same answer.
| Player region | Served by | Reads | Writes |
|---|---|---|---|
| US-West | Hillsboro · Los Angeles | Local replica | Forwarded to LA primary |
| US-Central | Fastest of West / East | Local to serving node | Forwarded to LA primary |
| US-East | Vint Hill | Local replica | Forwarded to LA primary |
If the primary ever goes unreachable, edges don't error out — they announce it and keep serving.
Each edge node watches the primary and reacts on its own if it ever becomes unreachable: reads, ban checks, and verification keep serving from that node's local replica, while writes pause instead of failing with a cryptic error.
Game clients keep attesting — verification never depended on the write path. When the primary comes back, nodes resume writes automatically, with nothing to reconcile beyond a brief gap in telemetry counters.
The fleet grows by copying, not by re-architecting: a new region is a replica, a config, and a DNS entry.
Once Kubian has the revenue for it and needs better request handling, we add more fleets across the US — multiples in each location above, plus nodes near Canada — scaling Static Labs LLC infrastructure so everything feels smoother for developers and end-users.
After that comes GDNN, the Globally Distributed Node Network: per-continent edges with local replicas across Europe, Asia, Oceania, and more — the fastest option for international users, with the fastest authentication, verification, and anti-cheat alerts worldwide.
Short answers for the questions studios actually ask about DNN (Distributed Node Network).
Read the integration docs, see how it survives real attacks, or talk to the team about your region.
Status above reflects the September 24, 2026 DNN (Distributed Node Network, formerly USDNN) launch topology (Los Angeles primary + Hillsboro and Vint Hill replicas). For live incident history, contact the team.