Incident Report: October 15th, 2025
Blog post from Railway
Railway experienced a major outage on October 15, 2025, affecting its dashboard and deployment pipeline from 20:49 UTC to 22:15 UTC, during which users faced difficulties deploying new services and accessing the platform's control systems, though existing services continued to operate. The outage was triggered by a large deletion operation that caused a surge in internal requests, leading to elevated error rates in backend systems and an unexpected overload on network management systems. The on-call team responded promptly, pausing deployments and implementing a fix by 21:45 UTC, with the platform beginning to recover shortly after and full access restored by 22:15 UTC. Post-incident measures include improved rate limiting of internal APIs, enhanced visibility and alerting for request patterns, and increased resilience of the network control plane to prevent similar issues in the future. Railway has expressed its commitment to maintaining a top-tier cloud experience and apologizes for any inconvenience caused by the incident.