Walk me through how you would diagnose a sudden spike in latency within a distributed Kubernetes cluster for Global Payment Holding.
