feat: implement decision-based routing with plugin architecture

Signed-off-by: bitliu <[email protected]>

Author: Xunzhuo

e2e/profiles/dynamic-config/profile.go

@@ -241,8 +241,6 @@ func (p *Profile) kubectlApply(ctx context.Context, kubeconfig, manifestPath str
 }
 
 func (p *Profile) verifyEnvironment(ctx context.Context, opts *framework.SetupOptions) error {
-	p.log("Verifying all deployments are healthy...")
-
 	// Create Kubernetes client
 	config, err := clientcmd.BuildConfigFromFlags("", opts.KubeConfig)
 	if err != nil {
@@ -251,9 +249,56 @@ func (p *Profile) verifyEnvironment(ctx context.Context, opts *framework.SetupOp
 
 	client, err := kubernetes.NewForConfig(config)
 	if err != nil {
-		return fmt.Errorf("failed to create kubernetes client: %w", err)
+		return fmt.Errorf("failed to create kube client: %w", err)
+	}
+
+	// Wait for Envoy Gateway service to be ready with retry
+	retryTimeout := 10 * time.Minute
+	retryInterval := 5 * time.Second
+	startTime := time.Now()
+
+	p.log("Waiting for Envoy Gateway service to be ready...")
+
+	// Label selector for the semantic-router gateway service
+	labelSelector := "gateway.envoyproxy.io/owning-gateway-namespace=default,gateway.envoyproxy.io/owning-gateway-name=semantic-router"
+
+	var envoyService string
+	for {
+		// Try to get Envoy service name
+		envoyService, err = helpers.GetEnvoyServiceName(ctx, client, labelSelector, p.verbose)
+		if err == nil {
+			// Verify that the service has exactly 1 pod running with all containers ready
+			podErr := helpers.VerifyServicePodsRunning(ctx, client, "envoy-gateway-system", envoyService, p.verbose)
+			if podErr == nil {
+				p.log("Envoy Gateway service is ready: %s", envoyService)
+				break
+			}
+			if p.verbose {
+				p.log("Envoy service found but pods not ready: %v", podErr)
+			}
+			err = fmt.Errorf("service pods not ready: %w", podErr)
+		}
+
+		if time.Since(startTime) >= retryTimeout {
+			return fmt.Errorf("failed to get Envoy service with running pods after %v: %w", retryTimeout, err)
+		}
+
+		if p.verbose {
+			p.log("Envoy service not ready, retrying in %v... (elapsed: %v)",
+				retryInterval, time.Since(startTime).Round(time.Second))
+		}
+
+		select {
+		case <-ctx.Done():
+			return ctx.Err()
+		case <-time.After(retryInterval):
+			// Continue retry
+		}
 	}
 
+	// Check all deployments are healthy
+	p.log("Verifying all deployments are healthy...")
+
 	// Check semantic-router deployment
 	if err := helpers.CheckDeployment(ctx, client, "vllm-semantic-router-system", "semantic-router", p.verbose); err != nil {
 		return fmt.Errorf("semantic-router deployment not healthy: %w", err)