Mass VM Operations Failing Across All Regions Simultaneously

Your team operates a large-scale cloud compute platform where virtual machines are provisioned and managed on behalf of end users. Over a roughly six-hour window, your on-call engineers observe that all VM create, delete, and reimage operations are returning errors across every geographic region simultaneously. Downstream services that depend on ephemeral VM capacity — CI/CD job runners, interactive development environments, automated code-scanning pipelines, and static site build workers — all begin reporting failures. Approximately 100% of new workload requests are unable to acquire compute resources. The platform's underlying cloud provider is reachable and billing is active; no provider-wide outage is declared. Internal dashboards show no recent manual infrastructure changes by the operations team, yet the failure window has a clear start time approximately 5 hours and 53 minutes before service restoration.