A role-based network structure organizes clusters and their dedicated VPCs around operational roles such as production, nonproduction, preview, development, quality assurance, and product management. The role can span multiple clusters and regions, so network capacity and naming must support growth, replacement, and lookup. section 7.2.1
Apply the pattern by:
- Reserving IP space large enough for every network that a role may eventually require. section 7.2.1
- Naming each VPC after its Kubernetes cluster, using role, business line, instance, cloud provider, and region. section 7.2.1
- Separate platform development and preview from user-facing nonproduction and production so changes can be tested progressively. section 7.2.1
Role-and-region naming
Names such as sbx-i01-aws-us-east-1 and prod-i01-aws-us-east-2 identify role, instance, provider, and region while giving Terraform a predictable VPC lookup key. section 7.2.1
sbx-i01-aws-us-east-1 prod-i01-aws-us-east-2
Use the naming examples to inspect how role and region are encoded in cluster and VPC identifiers. section 7.2.1
