Tell how monitoring and alerting are configured — what metrics, what does the metrics collection pipeline look like, and how are developers notified of critical situations?