An Ingress is a Kubernetes resource that defines how external requests reach Services by path. Pods hold application workloads, Services expose those pods as endpoints, and an Ingress maps a request path to a Service. An Ingress controller makes those rules effective for incoming traffic. section 8.2.3 section 8.2.7
For several microservices, a LoadBalancer Service per application can mean a load balancer for each one. An NGINX Ingress controller instead provides a single load balancer and routes requests according to Ingress resources. When an Ingress changes, the controller refreshes its NGINX configuration with the new paths. section 8.2.7 section 8.3.4
apiVersion: networking.k8s.io/v1
kind: Ingress
metadata:
annotations:
kubernetes.io/ingress.class: nginx
nginx.ingress.kubernetes.io/backend-protocol: GRPC #1
nginx.ingress.kubernetes.io/ssl-redirect: "true" #2
cert-manager.io/cluster-issuer: selfsigned-issuer #3
name: order
spec:
rules:
- http:
paths:
- path: /Order #4
pathType: Prefix
backend:
service:
name: order
port:
number: 8080
tls:
- hosts:
- ingress.local #5
The Order service manifest maps the /Order path to the order Service on port 8080. Its annotations select NGINX, specify a gRPC backend, enable SSL redirection, and name a certificate issuer; the TLS rule names ingress.local. section 8.2.7
The controller can be installed with Helm before applying Ingress resources. For local TLS testing, the Order configuration uses a self-signed certificate and the ingress.local hostname; that certificate is for local development and is not automatically trusted by clients. Client-side certificate handling is therefore a separate step when calling the gRPC service through Ingress. section 8.2.7 section 8.3.3
