Introduction to the Gateway API - Container Service for Kubernetes

The Gateway API is a Kubernetes community standard for traffic management. It supports both Ingress-based north-south traffic management and Service Mesh-based east-west traffic management. This topic describes the core concepts of the Gateway API and north-south traffic solutions available in ACK.

Core concepts

The Gateway API uses a role-oriented, layered design that decouples infrastructure provisioning, cluster O&M, and application routing:

GatewayClass (Infrastructure layer): Like IngressClass, defines the gateway controller type, such as ALB or Envoy Gateway, and general configurations.
Gateway (O&M layer): Defines a gateway instance and its listener rules, such as ports, protocols, and TLS configurations.
HTTPRoute, GRPCRoute, and other route types (Application layer): Define traffic routing rules, such as path matching, header modification, and traffic weighting, attached to backend services.
Policy (Policy layer): Defines configurations such as circuit breaking, rate limiting, and JWT authentication. Can attach to a gateway, route, or backend service.

Gateway API solutions in ACK

Choose a Gateway API solution based on your business needs:

Solution	Gateway with Inference Extension	ALB (Application Load Balancer)
Overview	Gateway with Inference Extension is built on open source Envoy Gateway, optimized for cloud-native and AI inference scenarios. Listens for Gateway API resources to dynamically manage gateways and north-south traffic. Not fully managed—deployed on cluster nodes, requires O&M, and has no cloud product SLA.	ALB Ingress Controller supports the Gateway API from v2.17.0. Configure resources such as Gateway and HTTPRoute to route external traffic to in-cluster workloads. The controller converts Gateway API changes into ALB listener rules, routing rules, and server group configurations in real time.
Scenarios	General traffic and AI inference scenarios	General traffic scenarios
Core advantages	Standard open source Envoy architecture: Tracks the latest Envoy Gateway upstream. Built on Envoy proxy with a mature, high-performance ecosystem for large-scale traffic. Rich traffic management features: Supports rate limiting, circuit breaking, and zone-level routing, plus features such as fault injection, traffic mirroring, compression, and caching. AI inference extension: Provides model-aware load balancing with scheduling policies such as LoRA/KV Cache-aware, Waiting Request Num, and Model Priority. Supports phased model version releases. Flexible extensibility: Integrates with Envoy Gateway extension methods such as EnvoyFilter, Lua, Wasm, and ExtProc for custom traffic management.	Fully managed: Fully managed gateway and components, reducing O&M complexity and costs. High performance and instant elasticity: Built on Alibaba Cloud's network virtualization stack with high QPS throughput and concurrent connection performance. Scales automatically based on traffic loads. Powerful application layer routing: Deeply integrated with container services, supporting advanced routing rules for application-layer traffic. Rolling updates and zero downtime: Dynamically adjust forwarding rules via OpenAPI. Changes take effect in real time without instance restarts, ensuring zero service downtime. Broad scenario support: Handles persistent connections, high concurrent connections, high QPS, traffic spikes, rolling updates, hot upgrades, active zone-redundancy, and active geo-redundancy.
References	Manage general traffic using Gateway with Inference Extension Manage LLM traffic using Gateway with Inference Extension	Manage general traffic using ALB