fix metric scrape port not updated when inference pool target port updated

pkg/epp/backend/fake.go

@@ -31,7 +31,7 @@ type FakePodMetricsClient struct {
 	Res map[types.NamespacedName]*datastore.PodMetrics
 }
 
-func (f *FakePodMetricsClient) FetchMetrics(ctx context.Context, existing *datastore.PodMetrics) (*datastore.PodMetrics, error) {
+func (f *FakePodMetricsClient) FetchMetrics(ctx context.Context, existing *datastore.PodMetrics, port int32) (*datastore.PodMetrics, error) {
 	if err, ok := f.Err[existing.NamespacedName]; ok {
 		return nil, err
 	}

pkg/epp/backend/provider.go

@@ -49,7 +49,7 @@ type Provider struct {
 }
 
 type PodMetricsClient interface {
-	FetchMetrics(ctx context.Context, existing *datastore.PodMetrics) (*datastore.PodMetrics, error)
+	FetchMetrics(ctx context.Context, existing *datastore.PodMetrics, port int32) (*datastore.PodMetrics, error)
 }
 
 func (p *Provider) Init(ctx context.Context, refreshMetricsInterval, refreshPrometheusMetricsInterval time.Duration) error {
@@ -121,7 +121,8 @@ func (p *Provider) refreshMetricsOnce(logger logr.Logger) error {
 		wg.Add(1)
 		go func() {
 			defer wg.Done()
-			updated, err := p.pmc.FetchMetrics(ctx, existing)
+			pool, _ := p.datastore.PoolGet()
+			updated, err := p.pmc.FetchMetrics(ctx, existing, pool.Spec.TargetPortNumber)
 			if err != nil {
 				errCh <- fmt.Errorf("failed to parse metrics from %s: %v", existing.NamespacedName, err)
 				return

pkg/epp/backend/vllm/metrics.go

@@ -55,13 +55,15 @@ type PodMetricsClientImpl struct{}
 func (p *PodMetricsClientImpl) FetchMetrics(
 	ctx context.Context,
 	existing *datastore.PodMetrics,
+	port int32,
 ) (*datastore.PodMetrics, error) {
 	logger := log.FromContext(ctx)
 	loggerDefault := logger.V(logutil.DEFAULT)
 
 	// Currently the metrics endpoint is hard-coded, which works with vLLM.
 	// TODO(https://github.com/kubernetes-sigs/gateway-api-inference-extension/issues/16): Consume this from InferencePool config.
-	url := existing.BuildScrapeEndpoint()
+	url := existing.Address + ":" + strconv.Itoa(int(port))
+
 	req, err := http.NewRequestWithContext(ctx, http.MethodGet, url, nil)
 	if err != nil {
 		loggerDefault.Error(err, "Failed create HTTP request", "method", http.MethodGet, "url", url)

pkg/epp/controller/pod_reconciler_test.go

@@ -35,10 +35,10 @@ import (
 )
 
 var (
-	basePod1  = &datastore.PodMetrics{Pod: datastore.Pod{NamespacedName: types.NamespacedName{Name: "pod1"}, Address: "address-1", ScrapePath: "/metrics", ScrapePort: 8000}}
-	basePod2  = &datastore.PodMetrics{Pod: datastore.Pod{NamespacedName: types.NamespacedName{Name: "pod2"}, Address: "address-2", ScrapePath: "/metrics", ScrapePort: 8000}}
-	basePod3  = &datastore.PodMetrics{Pod: datastore.Pod{NamespacedName: types.NamespacedName{Name: "pod3"}, Address: "address-3", ScrapePath: "/metrics", ScrapePort: 8000}}
-	basePod11 = &datastore.PodMetrics{Pod: datastore.Pod{NamespacedName: types.NamespacedName{Name: "pod1"}, Address: "address-11", ScrapePath: "/metrics", ScrapePort: 8000}}
+	basePod1  = &datastore.PodMetrics{Pod: datastore.Pod{NamespacedName: types.NamespacedName{Name: "pod1"}, Address: "address-1"}}
+	basePod2  = &datastore.PodMetrics{Pod: datastore.Pod{NamespacedName: types.NamespacedName{Name: "pod2"}, Address: "address-2"}}
+	basePod3  = &datastore.PodMetrics{Pod: datastore.Pod{NamespacedName: types.NamespacedName{Name: "pod3"}, Address: "address-3"}}
+	basePod11 = &datastore.PodMetrics{Pod: datastore.Pod{NamespacedName: types.NamespacedName{Name: "pod1"}, Address: "address-11"}}
 )
 
 func TestPodReconciler(t *testing.T) {

pkg/epp/datastore/datastore.go

@@ -265,16 +265,13 @@ func (ds *datastore) PodDelete(namespacedName types.NamespacedName) {
 }
 
 func (ds *datastore) PodUpdateOrAddIfNotExist(pod *corev1.Pod) bool {
-	pool, _ := ds.PoolGet()
 	new := &PodMetrics{
 		Pod: Pod{
 			NamespacedName: types.NamespacedName{
 				Name:      pod.Name,
 				Namespace: pod.Namespace,
 			},
-			Address:    pod.Status.PodIP,
-			ScrapePath: "/metrics",
-			ScrapePort: pool.Spec.TargetPortNumber,
+			Address: pod.Status.PodIP,
 		},
 		Metrics: Metrics{
 			ActiveModels: make(map[string]int),

pkg/epp/datastore/types.go

@@ -26,10 +26,6 @@ import (
 type Pod struct {
 	NamespacedName types.NamespacedName
 	Address        string
-
-	// metrics scrape options
-	ScrapePort int32
-	ScrapePath string
 }
 
 type Metrics struct {
@@ -61,11 +57,10 @@ func (pm *PodMetrics) Clone() *PodMetrics {
 		Pod: Pod{
 			NamespacedName: pm.NamespacedName,
 			Address:        pm.Address,
-			ScrapePort:     pm.ScrapePort,
-			ScrapePath:     pm.ScrapePath,
 		},
 		Metrics: Metrics{
 			ActiveModels:            cm,
+			MaxActiveModels:         pm.MaxActiveModels,
 			RunningQueueSize:        pm.RunningQueueSize,
 			WaitingQueueSize:        pm.WaitingQueueSize,
 			KVCacheUsagePercent:     pm.KVCacheUsagePercent,
@@ -74,7 +69,3 @@ func (pm *PodMetrics) Clone() *PodMetrics {
 	}
 	return clone
 }
-
-func (pm *PodMetrics) BuildScrapeEndpoint() string {
-	return fmt.Sprintf("http://%s:%d%s", pm.Address, pm.ScrapePort, pm.ScrapePath)
-}