feat: add dedicated health check endpoints for flightdeck

flightdeck/http/server.go

@@ -34,6 +34,8 @@ func (s *Server) setupRoutes() {
 	Home(s.mux, s.siteClient, s.config)
 	Config(s.mux, s.siteClient, s.config)
 	Help(s.mux, s.config)
+	Health(s.mux)
+	Ready(s.mux, s.siteClient, s.config)
 }
 
 func (s *Server) Start() error {
@@ -65,8 +67,9 @@ func requestLoggingMiddleware(next http.Handler) http.Handler {
 
 		duration := time.Since(start)
 
-		// Skip logging for static assets at debug level
-		if len(r.URL.Path) > 7 && r.URL.Path[:8] == "/static/" {
+		// Skip logging for static assets and health check endpoints at debug level
+		if (len(r.URL.Path) > 7 && r.URL.Path[:8] == "/static/") ||
+			r.URL.Path == "/healthz" || r.URL.Path == "/readyz" {
 			slog.Debug("request",
 				"method", r.Method,
 				"path", r.URL.Path,
@@ -125,3 +128,37 @@ func Config(mux *http.ServeMux, getter internal.SiteInterface, config *internal.
 func Static(mux *http.ServeMux) {
 	mux.Handle("GET /static/", http.StripPrefix("/static/", http.FileServer(http.Dir("public"))))
 }
+
+// Health handles liveness probe at /healthz
+// This endpoint only checks that the process is alive and can serve HTTP requests
+// It does not check any external dependencies
+func Health(mux *http.ServeMux) {
+	mux.HandleFunc("GET /healthz", func(w http.ResponseWriter, r *http.Request) {
+		slog.Debug("health check requested")
+		w.WriteHeader(http.StatusOK)
+		w.Write([]byte("OK"))
+	})
+}
+
+// Ready handles readiness probe at /readyz
+// This endpoint checks that the k8s API is reachable and Site CR can be fetched (if Sites exist)
+func Ready(mux *http.ServeMux, getter internal.SiteInterface, config *internal.ServerConfig) {
+	mux.HandleFunc("GET /readyz", func(w http.ResponseWriter, r *http.Request) {
+		slog.Debug("readiness check requested")
+
+		// Try to fetch the Site CR to verify k8s API connectivity
+		// Note: If Sites CRD doesn't exist, the Get method returns an empty Site and no error
+		_, err := getter.Get(config.SiteName, config.Namespace, metav1.GetOptions{}, r.Context())
+		if err != nil {
+			// Only fail readiness if there's a real error (not just missing CRD)
+			slog.Warn("readiness check failed", "error", err)
+			w.WriteHeader(http.StatusServiceUnavailable)
+			w.Write([]byte("Not Ready"))
+			return
+		}
+
+		// If we got here, either the Site exists or the CRD doesn't exist (which is OK)
+		w.WriteHeader(http.StatusOK)
+		w.Write([]byte("Ready"))
+	})
+}

flightdeck/http/server_test.go

@@ -0,0 +1,94 @@
+package http
+
+import (
+	"context"
+	"errors"
+	"net/http"
+	"net/http/httptest"
+	"testing"
+
+	"github.com/posit-dev/team-operator/api/core/v1beta1"
+	"github.com/posit-dev/team-operator/flightdeck/internal"
+	"github.com/stretchr/testify/assert"
+	metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
+)
+
+// Mock implementation of SiteInterface
+type mockSiteClient struct {
+	site *v1beta1.Site
+	err  error
+}
+
+func (m *mockSiteClient) Get(name string, namespace string, options metav1.GetOptions, ctx context.Context) (*v1beta1.Site, error) {
+	if m.err != nil {
+		return nil, m.err
+	}
+	return m.site, nil
+}
+
+func TestHealthEndpoint(t *testing.T) {
+	// Health endpoint should always return 200 OK
+	mux := http.NewServeMux()
+	Health(mux)
+
+	req := httptest.NewRequest("GET", "/healthz", nil)
+	w := httptest.NewRecorder()
+
+	mux.ServeHTTP(w, req)
+
+	assert.Equal(t, http.StatusOK, w.Code)
+	assert.Equal(t, "OK", w.Body.String())
+}
+
+func TestReadyEndpoint_Success(t *testing.T) {
+	// Ready endpoint should return 200 when Site CR can be fetched
+	config := &internal.ServerConfig{
+		SiteName:  "test-site",
+		Namespace: "test-namespace",
+	}
+
+	mockClient := &mockSiteClient{
+		site: &v1beta1.Site{
+			ObjectMeta: metav1.ObjectMeta{
+				Name:      "test-site",
+				Namespace: "test-namespace",
+			},
+		},
+		err: nil,
+	}
+
+	mux := http.NewServeMux()
+	Ready(mux, mockClient, config)
+
+	req := httptest.NewRequest("GET", "/readyz", nil)
+	w := httptest.NewRecorder()
+
+	mux.ServeHTTP(w, req)
+
+	assert.Equal(t, http.StatusOK, w.Code)
+	assert.Equal(t, "Ready", w.Body.String())
+}
+
+func TestReadyEndpoint_Failure(t *testing.T) {
+	// Ready endpoint should return 503 when Site CR cannot be fetched
+	config := &internal.ServerConfig{
+		SiteName:  "test-site",
+		Namespace: "test-namespace",
+	}
+
+	mockClient := &mockSiteClient{
+		site: nil,
+		err:  errors.New("failed to connect to k8s API"),
+	}
+
+	mux := http.NewServeMux()
+	Ready(mux, mockClient, config)
+
+	req := httptest.NewRequest("GET", "/readyz", nil)
+	w := httptest.NewRecorder()
+
+	mux.ServeHTTP(w, req)
+
+	assert.Equal(t, http.StatusServiceUnavailable, w.Code)
+	assert.Equal(t, "Not Ready", w.Body.String())
+}

internal/controller/core/flightdeck_controller.go

@@ -295,7 +295,7 @@ func (r *FlightdeckReconciler) reconcileFlightdeckResources(
 							LivenessProbe: &corev1.Probe{
 								ProbeHandler: corev1.ProbeHandler{
 									HTTPGet: &corev1.HTTPGetAction{
-										Path: "/",
+										Path: "/healthz",
 										Port: intstr.FromString("http"),
 									},
 								},
@@ -308,7 +308,7 @@ func (r *FlightdeckReconciler) reconcileFlightdeckResources(
 							ReadinessProbe: &corev1.Probe{
 								ProbeHandler: corev1.ProbeHandler{
 									HTTPGet: &corev1.HTTPGetAction{
-										Path: "/",
+										Path: "/readyz",
 										Port: intstr.FromString("http"),
 									},
 								},

internal/controller/core/flightdeck_test.go

@@ -196,12 +196,12 @@ func TestFlightdeckReconciler_DeploymentHasProbes(t *testing.T) {
 
 	// Verify LivenessProbe exists
 	assert.NotNil(t, container.LivenessProbe)
-	assert.Equal(t, "/", container.LivenessProbe.HTTPGet.Path)
+	assert.Equal(t, "/healthz", container.LivenessProbe.HTTPGet.Path)
 	assert.Equal(t, int32(10), container.LivenessProbe.InitialDelaySeconds)
 
 	// Verify ReadinessProbe exists
 	assert.NotNil(t, container.ReadinessProbe)
-	assert.Equal(t, "/", container.ReadinessProbe.HTTPGet.Path)
+	assert.Equal(t, "/readyz", container.ReadinessProbe.HTTPGet.Path)
 	assert.Equal(t, int32(3), container.ReadinessProbe.InitialDelaySeconds)
 }