GitHub adds Azure capacity but October outages expose migration and regional-recovery risks

By November 9, 140 of 151 production services and 32 of 50 MySQL primaries were running in Azure Central US, whose Kubernetes capacity more than doubled. October incidents still disrupted Actions, other GitHub services, and enterprise migrations; a regional network event affected all three Central US zones and at least 34,128 signed-in users. GitHub says East US 3 is about a week behind plan, with its first validated traffic ramp planned for the first week of December.

The added Central US capacity is already absorbing migrated-service growth, but the October outages show that spare compute alone does not prevent failures when databases, routing, or regional dependencies are shared. Builders should treat migration load limits, dependency isolation, and tested regional failover as capacity-planning requirements; the report does not yet establish how much the planned East US 3 ramp will improve recovery.