Google Cloud moved Cloud Run service health to general availability, a feature that automates cross-region failover and failback for multi-region Cloud Run services using per-instance readiness probes, configurable through a two-click setup rather than hand-rolled health-check and DNS-failover plumbing. The timing is what makes this worth noting: the GA release landed just six days after a power interruption at one of Google's Netherlands data centers had already tested the underlying failure mode in production. Before this feature, running a genuinely resilient multi-region Cloud Run deployment meant building your own external health-checking and traffic-shifting logic, typically bolted on with Cloud Load Balancing configuration and custom monitoring, which is exactly the kind of undifferentiated infrastructure work that serverless platforms are supposed to eliminate. With service health, Cloud Run itself monitors instance-level readiness across regions and automatically redirects both internal and external traffic away from an unhealthy region and back once it recovers, without an operator needing to write or maintain that logic. For teams already running latency-sensitive or availability-critical services on Cloud Run, this closes one of the more common gaps between "serverless is supposed to just work" and the reality that multi-region resilience has historically required real engineering effort even on managed platforms. It's also a useful data point on how cloud providers are increasingly shipping resilience features in direct response to their own well-publicized incidents, which is worth factoring in when evaluating how mature a given managed service's failure-handling story actually is before you need it.