Add pprof trace support

Add pprof trace support and --enable-contention-profiling to those
components that don't already have it.

cmd/cloud-controller-manager/app/controllermanager.go

@@ -22,6 +22,7 @@ import (
 	"net/http"
 	"net/http/pprof"
 	"os"
+	goruntime "runtime"
 	"strconv"
 	"time"
 
@@ -111,6 +112,10 @@ func Run(s *options.CloudControllerManagerServer, cloud cloudprovider.Interface)
 			mux.HandleFunc("/debug/pprof/", pprof.Index)
 			mux.HandleFunc("/debug/pprof/profile", pprof.Profile)
 			mux.HandleFunc("/debug/pprof/symbol", pprof.Symbol)
+			mux.HandleFunc("/debug/pprof/trace", pprof.Trace)
+			if s.EnableContentionProfiling {
+				goruntime.SetBlockProfileRate(1)
+			}
 		}
 		configz.InstallHandler(mux)
 		mux.Handle("/metrics", prometheus.Handler())

cmd/cloud-controller-manager/app/options/options.go

@@ -75,6 +75,7 @@ func (s *CloudControllerManagerServer) AddFlags(fs *pflag.FlagSet) {
 	fs.DurationVar(&s.RouteReconciliationPeriod.Duration, "route-reconciliation-period", s.RouteReconciliationPeriod.Duration, "The period for reconciling routes created for Nodes by cloud provider.")
 	fs.BoolVar(&s.ConfigureCloudRoutes, "configure-cloud-routes", true, "Should CIDRs allocated by allocate-node-cidrs be configured on the cloud provider.")
 	fs.BoolVar(&s.EnableProfiling, "profiling", true, "Enable profiling via web interface host:port/debug/pprof/")
+	fs.BoolVar(&s.EnableContentionProfiling, "contention-profiling", false, "Enable lock contention profiling, if profiling is enabled")
 	fs.StringVar(&s.ClusterCIDR, "cluster-cidr", s.ClusterCIDR, "CIDR Range for Pods in cluster.")
 	fs.BoolVar(&s.AllocateNodeCIDRs, "allocate-node-cidrs", false, "Should CIDRs for Pods be allocated and set on the cloud provider.")
 	fs.StringVar(&s.Master, "master", s.Master, "The address of the Kubernetes API server (overrides any value in kubeconfig)")

cmd/kube-controller-manager/app/controllermanager.go

@@ -28,6 +28,7 @@ import (
 	"net/http"
 	"net/http/pprof"
 	"os"
+	goruntime "runtime"
 	"strconv"
 	"time"
 
@@ -135,6 +136,10 @@ func Run(s *options.CMServer) error {
 			mux.HandleFunc("/debug/pprof/", pprof.Index)
 			mux.HandleFunc("/debug/pprof/profile", pprof.Profile)
 			mux.HandleFunc("/debug/pprof/symbol", pprof.Symbol)
+			mux.HandleFunc("/debug/pprof/trace", pprof.Trace)
+			if s.EnableContentionProfiling {
+				goruntime.SetBlockProfileRate(1)
+			}
 		}
 		configz.InstallHandler(mux)
 		mux.Handle("/metrics", prometheus.Handler())

cmd/kube-controller-manager/app/options/options.go

@@ -178,6 +178,7 @@ func (s *CMServer) AddFlags(fs *pflag.FlagSet, allControllers []string, disabled
 	fs.StringVar(&s.ClusterSigningKeyFile, "cluster-signing-key-file", s.ClusterSigningKeyFile, "Filename containing a PEM-encoded RSA or ECDSA private key used to sign cluster-scoped certificates")
 	fs.StringVar(&s.ApproveAllKubeletCSRsForGroup, "insecure-experimental-approve-all-kubelet-csrs-for-group", s.ApproveAllKubeletCSRsForGroup, "The group for which the controller-manager will auto approve all CSRs for kubelet client certificates.")
 	fs.BoolVar(&s.EnableProfiling, "profiling", true, "Enable profiling via web interface host:port/debug/pprof/")
+	fs.BoolVar(&s.EnableContentionProfiling, "contention-profiling", false, "Enable lock contention profiling, if profiling is enabled")
 	fs.StringVar(&s.ClusterName, "cluster-name", s.ClusterName, "The instance prefix for the cluster")
 	fs.StringVar(&s.ClusterCIDR, "cluster-cidr", s.ClusterCIDR, "CIDR Range for Pods in cluster.")
 	fs.StringVar(&s.ServiceCIDR, "service-cluster-ip-range", s.ServiceCIDR, "CIDR Range for Services in cluster.")

cmd/kubelet/app/options/options.go

@@ -150,6 +150,7 @@ func (s *KubeletServer) AddFlags(fs *pflag.FlagSet) {
 	fs.Int32Var(&s.EventBurst, "event-burst", s.EventBurst, "Maximum size of a bursty event records, temporarily allows event records to burst to this number, while still not exceeding event-qps. Only used if --event-qps > 0")
 	fs.BoolVar(&s.RunOnce, "runonce", s.RunOnce, "If true, exit after spawning pods from local manifests or remote urls. Exclusive with --api-servers, and --enable-server")
 	fs.BoolVar(&s.EnableDebuggingHandlers, "enable-debugging-handlers", s.EnableDebuggingHandlers, "Enables server endpoints for log collection and local running of containers and commands")
+	fs.BoolVar(&s.EnableContentionProfiling, "contention-profiling", false, "Enable lock contention profiling, if profiling is enabled")
 	fs.DurationVar(&s.MinimumGCAge.Duration, "minimum-container-ttl-duration", s.MinimumGCAge.Duration, "Minimum age for a finished container before it is garbage collected.  Examples: '300ms', '10s' or '2h45m'")
 	fs.MarkDeprecated("minimum-container-ttl-duration", "Use --eviction-hard or --eviction-soft instead. Will be removed in a future version.")
 	fs.Int32Var(&s.MaxPerPodContainerCount, "maximum-dead-containers-per-container", s.MaxPerPodContainerCount, "Maximum number of old instances to retain per container.  Each container takes up some disk space.  Default: 1.")

cmd/kubelet/app/server.go

@@ -878,7 +878,7 @@ func startKubelet(k kubelet.KubeletBootstrap, podCfg *config.PodConfig, kubeCfg
 	// start the kubelet server
 	if kubeCfg.EnableServer {
 		go wait.Until(func() {
-			k.ListenAndServe(net.ParseIP(kubeCfg.Address), uint(kubeCfg.Port), kubeDeps.TLSOptions, kubeDeps.Auth, kubeCfg.EnableDebuggingHandlers)
+			k.ListenAndServe(net.ParseIP(kubeCfg.Address), uint(kubeCfg.Port), kubeDeps.TLSOptions, kubeDeps.Auth, kubeCfg.EnableDebuggingHandlers, kubeCfg.EnableContentionProfiling)
 		}, 0, wait.NeverStop)
 	}
 	if kubeCfg.ReadOnlyPort > 0 {

federation/cmd/federation-controller-manager/app/controllermanager.go

@@ -24,6 +24,7 @@ import (
 	"net"
 	"net/http"
 	"net/http/pprof"
+	goruntime "runtime"
 	"strconv"
 
 	metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
@@ -131,6 +132,10 @@ func Run(s *options.CMServer) error {
 			mux.HandleFunc("/debug/pprof/", pprof.Index)
 			mux.HandleFunc("/debug/pprof/profile", pprof.Profile)
 			mux.HandleFunc("/debug/pprof/symbol", pprof.Symbol)
+			mux.HandleFunc("/debug/pprof/trace", pprof.Trace)
+			if s.EnableContentionProfiling {
+				goruntime.SetBlockProfileRate(1)
+			}
 		}
 		mux.Handle("/metrics", prometheus.Handler())
 

federation/cmd/federation-controller-manager/app/options/options.go

@@ -64,6 +64,8 @@ type ControllerManagerConfiguration struct {
 	APIServerBurst int `json:"federatedAPIBurst"`
 	// enableProfiling enables profiling via web interface host:port/debug/pprof/
 	EnableProfiling bool `json:"enableProfiling"`
+	// enableContentionProfiling enables lock contention profiling, if enableProfiling is true.
+	EnableContentionProfiling bool `json:"enableContentionProfiling"`
 	// leaderElection defines the configuration of leader election client.
 	LeaderElection componentconfig.LeaderElectionConfiguration `json:"leaderElection"`
 	// contentType is contentType of requests sent to apiserver.
@@ -115,6 +117,7 @@ func (s *CMServer) AddFlags(fs *pflag.FlagSet) {
 	fs.IntVar(&s.ConcurrentReplicaSetSyncs, "concurrent-replicaset-syncs", s.ConcurrentReplicaSetSyncs, "The number of ReplicaSets syncing operations that will be done concurrently. Larger number = faster endpoint updating, but more CPU (and network) load")
 	fs.DurationVar(&s.ClusterMonitorPeriod.Duration, "cluster-monitor-period", s.ClusterMonitorPeriod.Duration, "The period for syncing ClusterStatus in ClusterController.")
 	fs.BoolVar(&s.EnableProfiling, "profiling", true, "Enable profiling via web interface host:port/debug/pprof/")
+	fs.BoolVar(&s.EnableContentionProfiling, "contention-profiling", false, "Enable lock contention profiling, if profiling is enabled")
 	fs.StringVar(&s.Master, "master", s.Master, "The address of the federation API server (overrides any value in kubeconfig)")
 	fs.StringVar(&s.Kubeconfig, "kubeconfig", s.Kubeconfig, "Path to kubeconfig file with authorization and master location information.")
 	fs.StringVar(&s.ContentType, "kube-api-content-type", s.ContentType, "ContentType of requests sent to apiserver. Passing application/vnd.kubernetes.protobuf is an experimental feature now.")

pkg/apis/componentconfig/types.go

@@ -210,6 +210,8 @@ type KubeletConfiguration struct {
 	// enableDebuggingHandlers enables server endpoints for log collection
 	// and local running of containers and commands
 	EnableDebuggingHandlers bool
+	// enableContentionProfiling enables lock contention profiling, if enableDebuggingHandlers is true.
+	EnableContentionProfiling bool
 	// minimumGCAge is the minimum age for a finished container before it is
 	// garbage collected.
 	MinimumGCAge metav1.Duration
@@ -757,6 +759,8 @@ type KubeControllerManagerConfiguration struct {
 	ApproveAllKubeletCSRsForGroup string
 	// enableProfiling enables profiling via web interface host:port/debug/pprof/
 	EnableProfiling bool
+	// enableContentionProfiling enables lock contention profiling, if enableProfiling is true.
+	EnableContentionProfiling bool
 	// clusterName is the instance prefix for the cluster.
 	ClusterName string
 	// clusterCIDR is CIDR Range for Pods in cluster.

pkg/apis/componentconfig/v1alpha1/types.go

@@ -264,6 +264,8 @@ type KubeletConfiguration struct {
 	// enableDebuggingHandlers enables server endpoints for log collection
 	// and local running of containers and commands
 	EnableDebuggingHandlers *bool `json:"enableDebuggingHandlers"`
+	// enableContentionProfiling enables lock contention profiling, if enableDebuggingHandlers is true.
+	EnableContentionProfiling bool `json:"enableContentionProfiling"`
 	// minimumGCAge is the minimum age for a finished container before it is
 	// garbage collected.
 	MinimumGCAge metav1.Duration `json:"minimumGCAge"`

pkg/kubelet/kubelet.go

@@ -180,7 +180,7 @@ type KubeletBootstrap interface {
 	GetConfiguration() componentconfig.KubeletConfiguration
 	BirthCry()
 	StartGarbageCollection()
-	ListenAndServe(address net.IP, port uint, tlsOptions *server.TLSOptions, auth server.AuthInterface, enableDebuggingHandlers bool)
+	ListenAndServe(address net.IP, port uint, tlsOptions *server.TLSOptions, auth server.AuthInterface, enableDebuggingHandlers, enableContentionProfiling bool)
 	ListenAndServeReadOnly(address net.IP, port uint)
 	Run(<-chan kubetypes.PodUpdate)
 	RunOnce(<-chan kubetypes.PodUpdate) ([]RunPodResult, error)
@@ -2120,8 +2120,8 @@ func (kl *Kubelet) ResyncInterval() time.Duration {
 }
 
 // ListenAndServe runs the kubelet HTTP server.
-func (kl *Kubelet) ListenAndServe(address net.IP, port uint, tlsOptions *server.TLSOptions, auth server.AuthInterface, enableDebuggingHandlers bool) {
-	server.ListenAndServeKubeletServer(kl, kl.resourceAnalyzer, address, port, tlsOptions, auth, enableDebuggingHandlers, kl.containerRuntime, kl.criHandler)
+func (kl *Kubelet) ListenAndServe(address net.IP, port uint, tlsOptions *server.TLSOptions, auth server.AuthInterface, enableDebuggingHandlers, enableContentionProfiling bool) {
+	server.ListenAndServeKubeletServer(kl, kl.resourceAnalyzer, address, port, tlsOptions, auth, enableDebuggingHandlers, enableContentionProfiling, kl.containerRuntime, kl.criHandler)
 }
 
 // ListenAndServeReadOnly runs the kubelet HTTP server in read-only mode.

pkg/kubelet/server/server.go

@@ -25,6 +25,7 @@ import (
 	"net/http/pprof"
 	"net/url"
 	"reflect"
+	goruntime "runtime"
 	"strconv"
 	"strings"
 	"time"
@@ -119,11 +120,12 @@ func ListenAndServeKubeletServer(
 	port uint,
 	tlsOptions *TLSOptions,
 	auth AuthInterface,
-	enableDebuggingHandlers bool,
+	enableDebuggingHandlers,
+	enableContentionProfiling bool,
 	runtime kubecontainer.Runtime,
 	criHandler http.Handler) {
 	glog.Infof("Starting to listen on %s:%d", address, port)
-	handler := NewServer(host, resourceAnalyzer, auth, enableDebuggingHandlers, runtime, criHandler)
+	handler := NewServer(host, resourceAnalyzer, auth, enableDebuggingHandlers, enableContentionProfiling, runtime, criHandler)
 	s := &http.Server{
 		Addr:           net.JoinHostPort(address.String(), strconv.FormatUint(uint64(port), 10)),
 		Handler:        &handler,
@@ -140,7 +142,7 @@ func ListenAndServeKubeletServer(
 // ListenAndServeKubeletReadOnlyServer initializes a server to respond to HTTP network requests on the Kubelet.
 func ListenAndServeKubeletReadOnlyServer(host HostInterface, resourceAnalyzer stats.ResourceAnalyzer, address net.IP, port uint, runtime kubecontainer.Runtime) {
 	glog.V(1).Infof("Starting to listen read-only on %s:%d", address, port)
-	s := NewServer(host, resourceAnalyzer, nil, false, runtime, nil)
+	s := NewServer(host, resourceAnalyzer, nil, false, false, runtime, nil)
 
 	server := &http.Server{
 		Addr:           net.JoinHostPort(address.String(), strconv.FormatUint(uint64(port), 10)),
@@ -192,7 +194,8 @@ func NewServer(
 	host HostInterface,
 	resourceAnalyzer stats.ResourceAnalyzer,
 	auth AuthInterface,
-	enableDebuggingHandlers bool,
+	enableDebuggingHandlers,
+	enableContentionProfiling bool,
 	runtime kubecontainer.Runtime,
 	criHandler http.Handler) Server {
 	server := Server{
@@ -208,6 +211,9 @@ func NewServer(
 	server.InstallDefaultHandlers()
 	if enableDebuggingHandlers {
 		server.InstallDebuggingHandlers(criHandler)
+		if enableContentionProfiling {
+			goruntime.SetBlockProfileRate(1)
+		}
 	}
 	return server
 }
@@ -381,6 +387,8 @@ func (s *Server) InstallDebuggingHandlers(criHandler http.Handler) {
 			pprof.Symbol(resp, req.Request)
 		case "cmdline":
 			pprof.Cmdline(resp, req.Request)
+		case "trace":
+			pprof.Trace(resp, req.Request)
 		default:
 			pprof.Index(resp, req.Request)
 		}

pkg/kubelet/server/server_test.go

@@ -240,6 +240,7 @@ func newServerTest() *serverTestFramework {
 		stats.NewResourceAnalyzer(fw.fakeKubelet, time.Minute, &kubecontainertesting.FakeRuntime{}),
 		fw.fakeAuth,
 		true,
+		false,
 		&kubecontainertesting.Mock{},
 		fw.criHandler)
 	fw.serverUnderTest = &server

plugin/cmd/kube-scheduler/app/server.go

@@ -146,6 +146,7 @@ func startHTTP(s *options.SchedulerServer) {
 		mux.HandleFunc("/debug/pprof/", pprof.Index)
 		mux.HandleFunc("/debug/pprof/profile", pprof.Profile)
 		mux.HandleFunc("/debug/pprof/symbol", pprof.Symbol)
+		mux.HandleFunc("/debug/pprof/trace", pprof.Trace)
 		if s.EnableContentionProfiling {
 			goruntime.SetBlockProfileRate(1)
 		}

staging/src/k8s.io/apiserver/pkg/server/options/feature.go

@@ -42,7 +42,7 @@ func (o *FeatureOptions) AddFlags(fs *pflag.FlagSet) {
 	fs.BoolVar(&o.EnableProfiling, "profiling", o.EnableProfiling,
 		"Enable profiling via web interface host:port/debug/pprof/")
 	fs.BoolVar(&o.EnableContentionProfiling, "contention-profiling", o.EnableContentionProfiling,
-		"Enable contention profiling. Requires --profiling to be set to work.")
+		"Enable lock contention profiling, if profiling is enabled")
 	fs.BoolVar(&o.EnableSwaggerUI, "enable-swagger-ui", o.EnableSwaggerUI,
 		"Enables swagger ui on the apiserver at /swagger-ui")
 }

staging/src/k8s.io/apiserver/pkg/server/routes/profiling.go

@@ -30,4 +30,5 @@ func (d Profiling) Install(c *mux.APIContainer) {
 	c.UnlistedRoutes.HandleFunc("/debug/pprof/", pprof.Index)
 	c.UnlistedRoutes.HandleFunc("/debug/pprof/profile", pprof.Profile)
 	c.UnlistedRoutes.HandleFunc("/debug/pprof/symbol", pprof.Symbol)
+	c.UnlistedRoutes.HandleFunc("/debug/pprof/trace", pprof.Trace)
 }