A GKE web app must handle tens of thousands of users with latency below a threshold. To validate current deployment before launch, what should you do?

The most effective way to validate a GKE web app's ability to handle high user loads and maintain low latency before launch is through load testing. A load-testing tool can accurately simulate the expected concurrent users and request volume, providing realistic performance metrics like latency under stress. While autoscaling is crucial for production, simply sending curl requests won't simulate a realistic load or provide comprehensive latency analysis. Replicating across regions and using a global load balancer is a good strategy for high availability and global reach, but it doesn't directly validate the current deployment's performance under load. Cloud Debugger is for debugging code issues, not for assessing application performance under high user traffic.