kubelet manages /etc/hosts file

pkg/kubelet/dockertools/manager.go

@@ -312,22 +312,6 @@ type containerStatusResult struct {
 
 const podIPDownwardAPISelector = "status.podIP"
 
-// podDependsOnIP returns whether any containers in a pod depend on using the pod IP via
-// the downward API.
-func podDependsOnPodIP(pod *api.Pod) bool {
-	for _, container := range pod.Spec.Containers {
-		for _, env := range container.Env {
-			if env.ValueFrom != nil &&
-				env.ValueFrom.FieldRef != nil &&
-				env.ValueFrom.FieldRef.FieldPath == podIPDownwardAPISelector {
-				return true
-			}
-		}
-	}
-
-	return false
-}
-
 // determineContainerIP determines the IP address of the given container.  It is expected
 // that the container passed is the infrastructure container of a pod and the responsibility
 // of the caller to ensure that the correct container is passed.
@@ -1886,12 +1870,10 @@ func (dm *DockerManager) SyncPod(pod *api.Pod, runningPod kubecontainer.Pod, pod
 				glog.Warningf("Hairpin setup failed for pod %q: %v", podFullName, err)
 			}
 		}
-		if podDependsOnPodIP(pod) {
-			// Find the pod IP after starting the infra container in order to expose
-			// it safely via the downward API without a race.
-			pod.Status.PodIP = dm.determineContainerIP(pod.Name, pod.Namespace, podInfraContainer)
-		}
 
+		// Find the pod IP after starting the infra container in order to expose
+		// it safely via the downward API without a race and be able to use podIP in kubelet-managed /etc/hosts file.
+		pod.Status.PodIP = dm.determineContainerIP(pod.Name, pod.Namespace, podInfraContainer)
 	}
 
 	// Start everything

pkg/kubelet/dockertools/manager_test.go

@@ -2059,61 +2059,3 @@ func TestGetIPCMode(t *testing.T) {
 		t.Errorf("expected host ipc mode for pod but got %v", ipcMode)
 	}
 }
-
-func TestPodDependsOnPodIP(t *testing.T) {
-	tests := []struct {
-		name     string
-		expected bool
-		env      api.EnvVar
-	}{
-		{
-			name:     "depends on pod IP",
-			expected: true,
-			env: api.EnvVar{
-				Name: "POD_IP",
-				ValueFrom: &api.EnvVarSource{
-					FieldRef: &api.ObjectFieldSelector{
-						APIVersion: testapi.Default.Version(),
-						FieldPath:  "status.podIP",
-					},
-				},
-			},
-		},
-		{
-			name:     "literal value",
-			expected: false,
-			env: api.EnvVar{
-				Name:  "SOME_VAR",
-				Value: "foo",
-			},
-		},
-		{
-			name:     "other downward api field",
-			expected: false,
-			env: api.EnvVar{
-				Name: "POD_NAME",
-				ValueFrom: &api.EnvVarSource{
-					FieldRef: &api.ObjectFieldSelector{
-						APIVersion: testapi.Default.Version(),
-						FieldPath:  "metadata.name",
-					},
-				},
-			},
-		},
-	}
-
-	for _, tc := range tests {
-		pod := &api.Pod{
-			Spec: api.PodSpec{
-				Containers: []api.Container{
-					{Env: []api.EnvVar{tc.env}},
-				},
-			},
-		}
-
-		result := podDependsOnPodIP(pod)
-		if e, a := tc.expected, result; e != a {
-			t.Errorf("%v: Unexpected result; expected %v, got %v", tc.name, e, a)
-		}
-	}
-}

pkg/kubelet/kubelet.go

@@ -20,6 +20,7 @@ package kubelet
 // contrib/mesos/pkg/executor/.
 
 import (
+	"bytes"
 	"errors"
 	"fmt"
 	"io"
@@ -102,6 +103,8 @@ const (
 	// Minimum period for performing global cleanup tasks, i.e., housekeeping
 	// will not be performed more than once per housekeepingMinimumPeriod.
 	housekeepingMinimumPeriod = time.Second * 2
+
+	etcHostsPath = "/etc/hosts"
 )
 
 var (
@@ -955,8 +958,17 @@ func (kl *Kubelet) syncNodeStatus() {
 	}
 }
 
-func makeMounts(container *api.Container, podVolumes kubecontainer.VolumeMap) (mounts []kubecontainer.Mount) {
+func makeMounts(pod *api.Pod, podDir string, container *api.Container, podVolumes kubecontainer.VolumeMap) ([]kubecontainer.Mount, error) {
+	// Kubernetes only mounts on /etc/hosts if :
+	// - container does not use hostNetwork and
+	// - container is not a infrastructure(pause) container
+	// - container is not already mounting on /etc/hosts
+	// When the pause container is being created, its IP is still unknown. Hence, PodIP will not have been set.
+	mountEtcHostsFile := !pod.Spec.SecurityContext.HostNetwork && len(pod.Status.PodIP) > 0
+	glog.V(4).Infof("Will create hosts mount for container:%q, podIP:%s: %v", container.Name, pod.Status.PodIP, mountEtcHostsFile)
+	mounts := []kubecontainer.Mount{}
 	for _, mount := range container.VolumeMounts {
+		mountEtcHostsFile = mountEtcHostsFile && (mount.MountPath != etcHostsPath)
 		vol, ok := podVolumes[mount.Name]
 		if !ok {
 			glog.Warningf("Mount cannot be satisified for container %q, because the volume is missing: %q", container.Name, mount)
@@ -969,7 +981,44 @@ func makeMounts(container *api.Container, podVolumes kubecontainer.VolumeMap) (m
 			ReadOnly:      mount.ReadOnly,
 		})
 	}
-	return
+	if mountEtcHostsFile {
+		hostsMount, err := makeHostsMount(podDir, pod.Status.PodIP, pod.Name)
+		if err != nil {
+			return nil, err
+		}
+		mounts = append(mounts, *hostsMount)
+	}
+	return mounts, nil
+}
+
+func makeHostsMount(podDir, podIP, podName string) (*kubecontainer.Mount, error) {
+	hostsFilePath := path.Join(podDir, "etc-hosts")
+	if err := ensureHostsFile(hostsFilePath, podIP, podName); err != nil {
+		return nil, err
+	}
+	return &kubecontainer.Mount{
+		Name:          "k8s-managed-etc-hosts",
+		ContainerPath: etcHostsPath,
+		HostPath:      hostsFilePath,
+		ReadOnly:      false,
+	}, nil
+}
+
+func ensureHostsFile(fileName string, hostIP, hostName string) error {
+	if _, err := os.Stat(fileName); os.IsExist(err) {
+		glog.V(4).Infof("kubernetes-managed etc-hosts file exits. Will not be recreated: %q", fileName)
+		return nil
+	}
+	var buffer bytes.Buffer
+	buffer.WriteString("# Kubernetes-managed hosts file.\n")
+	buffer.WriteString("127.0.0.1\tlocalhost\n")                      // ipv4 localhost
+	buffer.WriteString("::1\tlocalhost ip6-localhost ip6-loopback\n") // ipv6 localhost
+	buffer.WriteString("fe00::0\tip6-localnet\n")
+	buffer.WriteString("fe00::0\tip6-mcastprefix\n")
+	buffer.WriteString("fe00::1\tip6-allnodes\n")
+	buffer.WriteString("fe00::2\tip6-allrouters\n")
+	buffer.WriteString(fmt.Sprintf("%s\t%s\n", hostIP, hostName))
+	return ioutil.WriteFile(fileName, buffer.Bytes(), 0644)
 }
 
 func makePortMappings(container *api.Container) (ports []kubecontainer.PortMapping) {
@@ -1014,7 +1063,10 @@ func (kl *Kubelet) GenerateRunContainerOptions(pod *api.Pod, container *api.Cont
 	}
 
 	opts.PortMappings = makePortMappings(container)
-	opts.Mounts = makeMounts(container, vol)
+	opts.Mounts, err = makeMounts(pod, kl.getPodDir(pod.UID), container, vol)
+	if err != nil {
+		return nil, err
+	}
 	opts.Envs, err = kl.makeEnvironmentVariables(pod, container)
 	if err != nil {
 		return nil, err

pkg/kubelet/kubelet_test.go

@@ -152,6 +152,11 @@ func newTestPods(count int) []*api.Pod {
 	pods := make([]*api.Pod, count)
 	for i := 0; i < count; i++ {
 		pods[i] = &api.Pod{
+			Spec: api.PodSpec{
+				SecurityContext: &api.PodSecurityContext{
+					HostNetwork: true,
+				},
+			},
 			ObjectMeta: api.ObjectMeta{
 				Name: fmt.Sprintf("pod%d", i),
 			},
@@ -506,7 +511,7 @@ func TestMakeVolumeMounts(t *testing.T) {
 	container := api.Container{
 		VolumeMounts: []api.VolumeMount{
 			{
-				MountPath: "/mnt/path",
+				MountPath: "/etc/hosts",
 				Name:      "disk",
 				ReadOnly:  false,
 			},
@@ -534,12 +539,20 @@ func TestMakeVolumeMounts(t *testing.T) {
 		"disk5": &stubVolume{"/var/lib/kubelet/podID/volumes/empty/disk5"},
 	}
 
-	mounts := makeMounts(&container, podVolumes)
+	pod := api.Pod{
+		Spec: api.PodSpec{
+			SecurityContext: &api.PodSecurityContext{
+				HostNetwork: true,
+			},
+		},
+	}
+
+	mounts, _ := makeMounts(&pod, "/pod", &container, podVolumes)
 
 	expectedMounts := []kubecontainer.Mount{
 		{
 			"disk",
-			"/mnt/path",
+			"/etc/hosts",
 			"/mnt/disk",
 			false,
 		},

test/e2e/kubelet_etc_hosts.go

+/*
+Copyright 2014 The Kubernetes Authors All rights reserved.
+
+Licensed under the Apache License, Version 2.0 (the "License");
+you may not use this file except in compliance with the License.
+You may obtain a copy of the License at
+
+    http://www.apache.org/licenses/LICENSE-2.0
+
+Unless required by applicable law or agreed to in writing, software
+distributed under the License is distributed on an "AS IS" BASIS,
+WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+See the License for the specific language governing permissions and
+limitations under the License.
+*/
+
+package e2e
+
+import (
+	"fmt"
+	. "github.com/onsi/ginkgo"
+	api "k8s.io/kubernetes/pkg/api"
+	"k8s.io/kubernetes/pkg/api/latest"
+	"k8s.io/kubernetes/pkg/api/unversioned"
+	client "k8s.io/kubernetes/pkg/client/unversioned"
+	"strings"
+)
+
+const (
+	kubeletEtcHostsImageName          = "gcr.io/google_containers/netexec:1.0"
+	kubeletEtcHostsPodName            = "test-pod"
+	kubeletEtcHostsHostNetworkPodName = "test-host-network-pod"
+	etcHostsPartialContent            = "# Kubernetes-managed hosts file."
+)
+
+type KubeletManagedHostConfig struct {
+	hostNetworkPod *api.Pod
+	pod            *api.Pod
+	f              *Framework
+}
+
+var _ = Describe("KubeletManagedEtcHosts", func() {
+	f := NewFramework("e2e-kubelet-etc-hosts")
+	config := &KubeletManagedHostConfig{
+		f: f,
+	}
+
+	It("should test kubelet managed /etc/hosts file", func() {
+		By("Setting up the test")
+		config.setup()
+
+		By("Running the test")
+		config.verifyEtcHosts()
+	})
+})
+
+func (config *KubeletManagedHostConfig) verifyEtcHosts() {
+	By("Verifying /etc/hosts of container is kubelet-managed for pod with hostNetwork=false")
+	stdout := config.getEtcHostsContent(kubeletEtcHostsPodName, "busybox-1")
+	assertEtcHostsIsKubeletManaged(stdout)
+	stdout = config.getEtcHostsContent(kubeletEtcHostsPodName, "busybox-2")
+	assertEtcHostsIsKubeletManaged(stdout)
+
+	By("Verifying /etc/hosts of container is not kubelet-managed since container specifies /etc/hosts mount")
+	stdout = config.getEtcHostsContent(kubeletEtcHostsPodName, "busybox-3")
+	assertEtcHostsIsNotKubeletManaged(stdout)
+
+	By("Verifying /etc/hosts content of container is not kubelet-managed for pod with hostNetwork=true")
+	stdout = config.getEtcHostsContent(kubeletEtcHostsHostNetworkPodName, "busybox-1")
+	assertEtcHostsIsNotKubeletManaged(stdout)
+	stdout = config.getEtcHostsContent(kubeletEtcHostsHostNetworkPodName, "busybox-2")
+	assertEtcHostsIsNotKubeletManaged(stdout)
+}
+
+func (config *KubeletManagedHostConfig) setup() {
+	By("Creating hostNetwork=false pod")
+	config.createPodWithoutHostNetwork()
+
+	By("Creating hostNetwork=true pod")
+	config.createPodWithHostNetwork()
+}
+
+func (config *KubeletManagedHostConfig) createPodWithoutHostNetwork() {
+	podSpec := config.createPodSpec(kubeletEtcHostsPodName)
+	config.pod = config.createPod(podSpec)
+}
+
+func (config *KubeletManagedHostConfig) createPodWithHostNetwork() {
+	podSpec := config.createPodSpecWithHostNetwork(kubeletEtcHostsHostNetworkPodName)
+	config.hostNetworkPod = config.createPod(podSpec)
+}
+
+func (config *KubeletManagedHostConfig) createPod(podSpec *api.Pod) *api.Pod {
+	createdPod, err := config.getPodClient().Create(podSpec)
+	if err != nil {
+		Failf("Failed to create %s pod: %v", podSpec.Name, err)
+	}
+	expectNoError(config.f.WaitForPodRunning(podSpec.Name))
+	createdPod, err = config.getPodClient().Get(podSpec.Name)
+	if err != nil {
+		Failf("Failed to retrieve %s pod: %v", podSpec.Name, err)
+	}
+	return createdPod
+}
+
+func (config *KubeletManagedHostConfig) getPodClient() client.PodInterface {
+	return config.f.Client.Pods(config.f.Namespace.Name)
+}
+
+func assertEtcHostsIsKubeletManaged(etcHostsContent string) {
+	isKubeletManaged := strings.Contains(etcHostsContent, etcHostsPartialContent)
+	if !isKubeletManaged {
+		Failf("/etc/hosts file should be kubelet managed, but is not: %q", etcHostsContent)
+	}
+}
+
+func assertEtcHostsIsNotKubeletManaged(etcHostsContent string) {
+	isKubeletManaged := strings.Contains(etcHostsContent, etcHostsPartialContent)
+	if isKubeletManaged {
+		Failf("/etc/hosts file should not be kubelet managed, but is: %q", etcHostsContent)
+	}
+}
+
+func (config *KubeletManagedHostConfig) getEtcHostsContent(podName, containerName string) string {
+	cmd := kubectlCmd("exec", fmt.Sprintf("--namespace=%v", config.f.Namespace.Name), podName, "-c", containerName, "cat", "/etc/hosts")
+	stdout, stderr, err := startCmdAndStreamOutput(cmd)
+	if err != nil {
+		Failf("Failed to retrieve /etc/hosts, err: %q", err)
+	}
+	defer stdout.Close()
+	defer stderr.Close()
+
+	buf := make([]byte, 1000)
+	var n int
+	Logf("reading from `kubectl exec` command's stdout")
+	if n, err = stdout.Read(buf); err != nil {
+		Failf("Failed to read from kubectl exec stdout: %v", err)
+	}
+	return string(buf[:n])
+}
+
+func (config *KubeletManagedHostConfig) createPodSpec(podName string) *api.Pod {
+	pod := &api.Pod{
+		TypeMeta: unversioned.TypeMeta{
+			Kind:       "Pod",
+			APIVersion: latest.GroupOrDie("").Version,
+		},
+		ObjectMeta: api.ObjectMeta{
+			Name:      podName,
+			Namespace: config.f.Namespace.Name,
+		},
+		Spec: api.PodSpec{
+			Containers: []api.Container{
+				{
+					Name:            "busybox-1",
+					Image:           kubeletEtcHostsImageName,
+					ImagePullPolicy: api.PullIfNotPresent,
+					Command: []string{
+						"sleep",
+						"900",
+					},
+				},
+				{
+					Name:            "busybox-2",
+					Image:           kubeletEtcHostsImageName,
+					ImagePullPolicy: api.PullIfNotPresent,
+					Command: []string{
+						"sleep",
+						"900",
+					},
+				},
+				{
+					Name:            "busybox-3",
+					Image:           kubeletEtcHostsImageName,
+					ImagePullPolicy: api.PullIfNotPresent,
+					Command: []string{
+						"sleep",
+						"900",
+					},
+					VolumeMounts: []api.VolumeMount{
+						{
+							Name:      "host-etc-hosts",
+							MountPath: "/etc/hosts",
+						},
+					},
+				},
+			},
+			Volumes: []api.Volume{
+				{
+					Name: "host-etc-hosts",
+					VolumeSource: api.VolumeSource{
+						HostPath: &api.HostPathVolumeSource{
+							Path: "/etc/hosts",
+						},
+					},
+				},
+			},
+		},
+	}
+	return pod
+}
+
+func (config *KubeletManagedHostConfig) createPodSpecWithHostNetwork(podName string) *api.Pod {
+	pod := &api.Pod{
+		TypeMeta: unversioned.TypeMeta{
+			Kind:       "Pod",
+			APIVersion: latest.GroupOrDie("").Version,
+		},
+		ObjectMeta: api.ObjectMeta{
+			Name:      podName,
+			Namespace: config.f.Namespace.Name,
+		},
+		Spec: api.PodSpec{
+			SecurityContext: &api.PodSecurityContext{
+				HostNetwork: true,
+			},
+			Containers: []api.Container{
+				{
+					Name:            "busybox-1",
+					Image:           kubeletEtcHostsImageName,
+					ImagePullPolicy: api.PullIfNotPresent,
+					Command: []string{
+						"sleep",
+						"900",
+					},
+				},
+				{
+					Name:            "busybox-2",
+					Image:           kubeletEtcHostsImageName,
+					ImagePullPolicy: api.PullIfNotPresent,
+					Command: []string{
+						"sleep",
+						"900",
+					},
+				},
+			},
+		},
+	}
+	return pod
+}