Skip to content
Nantian Gateway

Quick Start

This guide walks through installing Nantian Gateway on a local Kind cluster. After completing the steps below, the control plane, data plane, and dashboard will be operational.

Expected time to complete: approximately five minutes.

Before you begin

Section titled “Before you begin”

Complete the prerequisites before proceeding:

  • A running Kubernetes cluster (Kind recommended for local development)
  • kubectl and helm installed and available in PATH
  • Gateway API CRDs installed on the cluster

Step 1: Install the Helm chart

Section titled “Step 1: Install the Helm chart”

Nantian Gateway is distributed as a Helm chart. The chart is available via the Helm repository as nantian-gw/nantian-gw.

For local development, install directly from the chart directory:

Terminal window
helm install nantian-gw nantian-gw/nantian-gw

To review the default configuration before installing:

Terminal window
helm show values nantian-gw/nantian-gw

Step 2: Verify the installation

Section titled “Step 2: Verify the installation”

After Helm completes, verify that all pods are running:

Terminal window
kubectl get pods -n nantian-gw

Expected output:

NAME                                    READY   STATUS    RESTARTS   AGE
nantian-gw-controlplane-xxxxxxxxx-xxxxx  1/1     Running   0          30s
nantian-gw-controlplane-xxxxxxxxx-xxxxx  1/1     Running   0          20s
nantian-gw-dataplane-xxxxxxxxx-xxxxx     1/1     Running   0          25s
nantian-gw-dataplane-xxxxxxxxx-xxxxx     1/1     Running   0          15s
nantian-gw-dashboard-xxxxxxxxx-xxxxx     1/1     Running   0          10s

The deployment should show two control plane replicas, two data plane replicas, and one dashboard pod. Each pod should display 1/1 in the READY column and Running in the STATUS column.

Verify that the GatewayClass resource was created:

Terminal window
kubectl get gatewayclass

Expected output:

NAME               CONTROLLER                                    ACCEPTED   AGE
nantian-gateway    gateway.nantian.io/gateway-controller         True       1m

Step 3: Deploy an example application

Section titled “Step 3: Deploy an example application”

The chart includes a sample application that demonstrates the Gateway API routing model. Deploy it to verify end-to-end connectivity:

Terminal window
kubectl apply -f gateway/examples/httpbin.yaml

This manifest creates a deployment, service, and HTTPRoute for the httpbin testing application.

Verify the route was created:

Terminal window
kubectl get httproute

Expected output:

NAME                   HOSTNAMES   AGE
httpbin-route                               1m

Step 4: Test the gateway

Section titled “Step 4: Test the gateway”

Send a request through the gateway to the sample application:

Terminal window
curl -s http://localhost:18080/get

Expected response:

{
  "args": {},
  "headers": {
    "Accept": "*/*",
    "Host": "httpbin.example.com",
    "User-Agent": "curl/8.0.0"
  },
  "origin": "127.0.0.1",
  "url": "http://httpbin.example.com/get"
}

The response confirms that the request passed through the data plane and reached the backend service. The Host header was set by the gateway, and the origin reflects the data plane’s IP address.

What was deployed

Section titled “What was deployed”

The Helm installation created the following components:

  • Control plane (2 replicas) — watches Kubernetes resources, translates configuration, and pushes it to data plane instances via gRPC xDS
  • Data plane (2 replicas) — handles client traffic, terminates TLS, performs routing, and forwards requests to backend services
  • Dashboard (1 replica) — serves the web-based admin interface for monitoring and configuration
  • GatewayClass — defines the gateway controller implementation
  • Gateway — configures listeners for HTTP and HTTPS traffic

Next steps

Section titled “Next steps”