A service in Cloud-native API Gateway represents a backend upstream that the gateway routes traffic to. Each service maps to a real backend, such as a Kubernetes workload, a Nacos-registered microservice, a Function Compute function, or a fixed IP endpoint. Before a gateway instance can route requests, you must add at least one service and register it as a route destination.
This topic covers the full service lifecycle: creating services, updating them, managing service versions for canary releases, configuring traffic rules, and setting up active health checks.
Create a service
Cloud-native API Gateway supports the following service sources:
Container Service for Kubernetes (ACK) clusters
Microservices Engine (MSE) Nacos instances
Fixed endpoints (IP addresses or hostnames)
Domain Name System (DNS) domain names
Function Compute
Add a service to a gateway instance and register it as a route destination so that routes can forward traffic to it. For more information, see Create a service.
Manage services
For services backed by fixed endpoints, the gateway cannot automatically detect when a backend node's IP address or port changes — fixed-endpoint services are not registered with a dynamic service discovery source. When a backend node's endpoint changes, manually update the service configuration so the gateway routes traffic to the correct address. For more information, see Manage services.
Manage traffic rules
Traffic rules control how the gateway transfers traffic from a gateway instance to a backend service. Two categories of rules are available:
Certificate encryption — configure TLS certificates to encrypt traffic between the gateway and the backend
Load balancing policy — distribute requests across backend nodes according to a specified algorithm
For more information, see Manage traffic rules.
Configure active health checks for services
Active health checks let the gateway periodically probe each backend node and automatically remove unhealthy nodes from the load balancing pool. When a node recovers, the gateway re-adds it automatically. This improves route availability for services running in multi-replica deployments.
Configure active health checks for a service.
Troubleshoot health check failures.
For more information, see Configure active health checks for services.