Linode interfaces

Quake AI concept: Ports

• Linode does not expose a standalone Neutron-style port object; the documented network interface model is per-Linode, and interfaces are managed at `GET/POST /linode/instances/{linodeId}/interfaces` with individual interface IDs under `/interfaces/{interfaceId}`. The canonical API summary lists Linode interfaces, not a top-level port resource analogous to OpenStack Neutron ports (Linode API summary).
• A Linode can have at most one VPC interface, plus at most one public interface and multiple VLAN interfaces on the same instance. That is a fixed per-instance interface model rather than OpenStack's decoupled port inventory where many ports can be created and attached independently (Manage Linode interfaces).
• Interface changes are documented as instance-level operations that generally require the Linode to be powered off, and changes take effect after powering the Linode back on. The docs explicitly say to power off before adding or editing interfaces and then power on for changes to take effect (Manage Linode interfaces).
• The VPC interface model is IP-centric rather than port-centric: a VPC interface can auto-assign or manually set a VPC IPv4 address, and can optionally enable public IPv4 access via 1:1 NAT. That differs from OpenStack ports, which are first-class L2/L3 attachment points with fixed network and subnet binding semantics (Manage Linode interfaces).