Architecting Aurora Global Database with Write Forwarding and Low RPO

Act as an AWS Aurora Specialist. Provide a highly comprehensive, enterprise-grade technical guide for provisioning and tuning an Amazon Aurora Global Database configured to support a massive [DB_CLUSTER_SIZE] (e.g., `db.r6g.16xlarge`) workload. The architecture must stretch across multiple [SECONDARY_REGIONS] to provide extreme read-scaling and active-active write capabilities utilizing the Aurora Write Forwarding feature. Explain the precise mechanical configuration of Aurora Global Endpoints to intelligently route traffic. Detail the implementation required to guarantee strict session-level consistency (`read-after-write`) for clients connected to secondary regions. Set up aggressive CloudWatch Alarms specifically monitoring the `AuroraGlobalDBReplicationLag` metric against a strict [REPLICATION_LAG_THRESHOLD] (e.g., < 1000ms). Finally, author an exact, CLI-driven failover runbook detailing how to permanently detach and promote a secondary cluster to primary status with zero data loss during a catastrophic regional outage.