A multinational SaaS provider hosts a core analytics engine in a central GCP project's VPC network. Multiple independent enterprise customers need private, low-latency TCP connectivity from their own distinct GCP VPC networks to this central analytics engine. The customer VPC networks frequently have overlapping RFC 1918 internal IP address ranges with each other and with the service provider network. Furthermore, security policies require that customers must only have unidirectional access to initiate connections toward the analytics service, preventing any customer from reaching another customer's network or exposing the provider's internal topology. Which VPC networking design should you implement to satisfy these architecture requirements?
- Configure Private Service Connect endpoints in each customer VPC targeting a Service Attachment published in the analytics provider VPC behind an Internal HTTP(S) Load Balancer.Cevap
- BEstablish VPC Network Peering between each customer VPC and the central analytics provider VPC, relying on Cloud NAT in the provider VPC to translate overlapping IP ranges.
- CDeploy High Availability (HA) VPN tunnels between each customer VPC and the provider VPC, establishing dynamic routing via Cloud Router to resolve IP overlaps.
- DPlace all customer projects and the analytics provider project into a single shared VPC Service Controls perimeter and rely on IAM role bindings to enforce private reachability.