The Railway team experienced an outage on Google Cloud that lasted approximately 10 minutes, affecting inbound traffic on Edge servers. The outage was caused by a change to IP allow-list rules, which were later reverted. Private networking and outbound requests were unaffected. The incident highlighted the need for more efficient communication and testing processes, with improvements such as automatically applying "page override" for merging changes and skipping the "plan" phase during reverts in Terraform. Railway's goal is to build resilient infrastructure by supporting multiple cloud providers, including Google Cloud, Amazon Web Services (AWS), and metal datacenters, to mitigate similar incidents in the future.