VPCs

Virtual private cloud (VPC) is a logically isolated virtual network within a public cloud that a customer can control as if it were their own data center, while benefiting from cloud services. A VPC provides a private IP address space, usually from a chosen CIDR range, and enables creation of subnets, route tables, and gateways. Resources such as virtual machines and managed services run in subnets that can be designated as public or private. Routing policies control traffic between subnets and to external networks, including the internet.

Connectivity options include an internet gateway for inbound and outbound internet traffic, NAT gateways or instances for private subnets to access the internet, and VPN connections or dedicated links to on‑premises networks (for example Direct Connect or ExpressRoute). Security is enforced with security groups (stateful, instance‑level) and network access control lists (stateless, subnet‑level), along with identity and access management controls.

VPCs support isolation and multi‑tenancy, allowing separate VPCs for development, testing, production, or different teams. They can be connected using VPC peering or a transit gateway, and can span regions or accounts in some clouds. Design considerations include IP planning, subnet sizing, routing complexity, gateway costs, and data transfer charges. While VPCs provide strong isolation, they rely on shared cloud infrastructure and may require additional configuration for cross‑VPC or cross‑region communication. Notable examples include AWS VPC, Google Cloud VPC, and Microsoft Azure VNet.