nfd-worker: replace --metrics with --port

Use a single port for serving http. In addition to metrics we will have
the healthz endpoint.

Author: marquiz

cmd/nfd-worker/main.go

@@ -108,8 +108,8 @@ func initFlags(flagset *flag.FlagSet) (*worker.Args, *worker.ConfigOverrideArgs)
 		"Kubeconfig to use")
 	flagset.BoolVar(&args.Oneshot, "oneshot", false,
 		"Do not publish feature labels")
-	flagset.IntVar(&args.MetricsPort, "metrics", 8081,
-		"Port on which to expose metrics.")
+	flagset.IntVar(&args.Port, "port", 8080,
+		"Port on which to metrics and healthz endpoints are served")
 	flagset.StringVar(&args.Options, "options", "",
 		"Specify config options from command line. Config options are specified "+
 			"in the same format as in the config file (i.e. json or yaml). These options")

deployment/base/worker-daemonset/worker-daemonset.yaml

@@ -36,5 +36,5 @@ spec:
               cpu: 5m
               memory: 64Mi
           ports:
-            - name: metrics
-              containerPort: 8081
+            - name: http
+              containerPort: 8080

deployment/helm/node-feature-discovery/templates/worker.yaml

@@ -100,13 +100,13 @@ spec:
         {{- range $key, $value := .Values.featureGates }}
         - "-feature-gates={{ $key }}={{ $value }}"
         {{- end }}
-        - "-metrics={{ .Values.worker.metricsPort | default "8081"}}"
+        - "-port={{ .Values.worker.port | default "8080"}}"
         {{- with .Values.worker.extraArgs }}
         {{- toYaml . | nindent 8 }}
         {{- end }}
         ports:
-          - containerPort: {{ .Values.worker.metricsPort | default "8081"}}
-            name: metrics
+          - containerPort: {{ .Values.worker.port | default "8080"}}
+            name: http
         volumeMounts:
         - name: host-boot
           mountPath: "/host-boot"

deployment/helm/node-feature-discovery/values.yaml

@@ -418,7 +418,7 @@ worker:
     #          matchName: {op: In, value: ["SWAP", "X86", "ARM"]}
 ### <NFD-WORKER-CONF-END-DO-NOT-REMOVE>
 
-  metricsPort: 8081
+  port: 8080
   daemonsetAnnotations: {}
   podSecurityContext: {}
     # fsGroup: 2000

docs/deployment/helm.md

@@ -233,7 +233,7 @@ API's you need to install the prometheus operator in your cluster.
 | `worker.*`                                  | dict    |                         | NFD worker daemonset configuration                                                                                                                                                                           |
 | `worker.enable`                             | bool    | true                    | Specifies whether nfd-worker should be deployed                                                                                                                                                              |
 | `worker.hostNetwork`                        | bool    | false                   | Specifies whether to enable or disable running the container in the host's network namespace                                                                                                                 |
-| `worker.metricsPort`                        | int     | 8081                    | Port on which to expose metrics from components to prometheus operator. **DEPRECATED**: will be replaced by `worker.port` in NFD v0.18.                                                                      |
+| `worker.port`                               | int     | 8080                    | Port on which to serve http for metrics and healthz endpoints.                                                                                                                                               |
 | `worker.config`                             | dict    |                         | NFD worker [configuration](../reference/worker-configuration-reference)                                                                                                                                      |
 | `worker.podSecurityContext`                 | dict    | {}                      | [PodSecurityContext](https://kubernetes.io/docs/tasks/configure-pod-container/security-context/#set-the-security-context-for-a-pod) holds pod-level security attributes and common container settins         |
 | `worker.securityContext`                    | dict    | {}                      | Container [security settings](https://kubernetes.io/docs/tasks/configure-pod-container/security-context/#set-the-security-context-for-a-container)                                                           |

pkg/nfd-worker/nfd-worker.go

@@ -19,13 +19,16 @@ package nfdworker
 import (
 	"encoding/json"
 	"fmt"
+	"net/http"
 	"os"
 	"path/filepath"
 	"regexp"
 	"sort"
 	"strings"
 	"time"
 
+	"github.com/prometheus/client_golang/prometheus"
+	"github.com/prometheus/client_golang/prometheus/promhttp"
 	"golang.org/x/exp/maps"
 	"golang.org/x/net/context"
 	"k8s.io/apimachinery/pkg/api/errors"
@@ -94,7 +97,7 @@ type Args struct {
 	Kubeconfig  string
 	Oneshot     bool
 	Options     string
-	MetricsPort int
+	Port        int
 	NoOwnerRefs bool
 
 	Overrides ConfigOverrideArgs
@@ -295,15 +298,13 @@ func (w *nfdWorker) Run() error {
 	labelTrigger.Reset(w.config.Core.SleepInterval.Duration)
 	defer labelTrigger.Stop()
 
+	httpMux := http.NewServeMux()
+
 	// Register to metrics server
-	if w.args.MetricsPort > 0 {
-		m := utils.CreateMetricsServer(w.args.MetricsPort,
-			buildInfo,
-			featureDiscoveryDuration)
-		go m.Run()
-		registerVersion(version.Get())
-		defer m.Stop()
-	}
+	promRegistry := prometheus.NewRegistry()
+	promRegistry.MustRegister(buildInfo, featureDiscoveryDuration)
+	httpMux.Handle("/metrics", promhttp.HandlerFor(promRegistry, promhttp.HandlerOpts{}))
+	registerVersion(version.Get())
 
 	err = w.runFeatureDiscovery()
 	if err != nil {
@@ -317,6 +318,14 @@ func (w *nfdWorker) Run() error {
 
 	// Start readiness probe (at this point we're "ready and live")
 
+	// Start HTTP server
+	httpServer := http.Server{Addr: fmt.Sprintf(":%d", w.args.Port), Handler: httpMux}
+	go func() {
+		klog.InfoS("http server starting", "port", httpServer.Addr)
+		klog.InfoS("http server stopped", "exitCode", httpServer.ListenAndServe())
+	}()
+	defer httpServer.Close()
+
 	for {
 		select {
 		case <-labelTrigger.C: