Railway Platform-Wide Outage After Google Cloud Incorrectly Suspends Production Account — 8 Hours Down

Railway Blog·high signal

Railway suffered a platform-wide outage from 22:20 UTC May 19 to ~06:14 UTC May 20 after Google Cloud's automated systems incorrectly suspended their production account. The suspension took down Railway's API, control plane, databases, and all GCP-hosted compute, cascading to affect all workloads as cached network routes expired. The incident hit 486 points on Hacker News and exposed a hard dependency on GCP for workload discoverability.