package controller import ( "crypto/x509" "fmt" "net/http" "net/url" "strconv" "strings" "time" crdv1alpha1 "github.com/kubernetes-csi/external-snapshotter/pkg/apis/volumesnapshot/v1alpha1" routev1 "github.com/openshift/api/route/v1" "github.com/pkg/errors" v1 "k8s.io/api/core/v1" extensionsv1beta1 "k8s.io/api/extensions/v1beta1" extclientset "k8s.io/apiextensions-apiserver/pkg/client/clientset/clientset" k8serrors "k8s.io/apimachinery/pkg/api/errors" metav1 "k8s.io/apimachinery/pkg/apis/meta/v1" "k8s.io/apimachinery/pkg/types" "k8s.io/apimachinery/pkg/util/intstr" "k8s.io/apimachinery/pkg/util/wait" "k8s.io/client-go/kubernetes" corelisters "k8s.io/client-go/listers/core/v1" "k8s.io/klog" cdiv1 "kubevirt.io/containerized-data-importer/pkg/apis/core/v1alpha1" clientset "kubevirt.io/containerized-data-importer/pkg/client/clientset/versioned" "kubevirt.io/containerized-data-importer/pkg/common" "kubevirt.io/containerized-data-importer/pkg/keys" "kubevirt.io/containerized-data-importer/pkg/operator" "kubevirt.io/containerized-data-importer/pkg/util" "kubevirt.io/containerized-data-importer/pkg/util/cert/triple" ) const ( // DataVolName provides a const to use for creating volumes in pod specs DataVolName = "cdi-data-vol" // CertVolName is the name of the volumecontaining certs CertVolName = "cdi-cert-vol" // ScratchVolName provides a const to use for creating scratch pvc volumes in pod specs ScratchVolName = "cdi-scratch-vol" // ImagePathName provides a const to use for creating volumes in pod specs ImagePathName = "image-path" socketPathName = "socket-path" // SourceHTTP is the source type HTTP, if unspecified or invalid, it defaults to SourceHTTP SourceHTTP = "http" // SourceS3 is the source type S3 SourceS3 = "s3" // SourceGlance is the source type of glance SourceGlance = "glance" // SourceNone means there is no source. SourceNone = "none" // SourceRegistry is the source type of Registry SourceRegistry = "registry" ) type podDeleteRequest struct { namespace string podName string podLister corelisters.PodLister k8sClient kubernetes.Interface } type pvcDeleteRequest struct { namespace string pvcName string pvcLister corelisters.PersistentVolumeClaimLister k8sClient kubernetes.Interface } func checkPVC(pvc *v1.PersistentVolumeClaim, annotation string) bool { if pvc.DeletionTimestamp != nil { return false } // check if we have proper annotation if !metav1.HasAnnotation(pvc.ObjectMeta, annotation) { klog.V(2).Infof("pvc annotation %q not found, skipping pvc \"%s/%s\"\n", annotation, pvc.Namespace, pvc.Name) return false } return true } // returns the endpoint string which contains the full path URI of the target object to be copied. func getEndpoint(pvc *v1.PersistentVolumeClaim) (string, error) { ep, found := pvc.Annotations[AnnEndpoint] if !found || ep == "" { verb := "empty" if !found { verb = "missing" } return ep, errors.Errorf("annotation %q in pvc \"%s/%s\" is %s\n", AnnEndpoint, pvc.Namespace, pvc.Name, verb) } return ep, nil } func getRequestedImageSize(pvc *v1.PersistentVolumeClaim) (string, error) { pvcSize, found := pvc.Spec.Resources.Requests[v1.ResourceStorage] if !found { return "", errors.Errorf("storage request is missing in pvc \"%s/%s\"", pvc.Namespace, pvc.Name) } return pvcSize.String(), nil } // returns the source string which determines the type of source. If no source or invalid source found, default to http func getSource(pvc *v1.PersistentVolumeClaim) string { source, found := pvc.Annotations[AnnSource] if !found { source = "" } switch source { case SourceHTTP, SourceS3, SourceGlance, SourceNone, SourceRegistry: klog.V(2).Infof("pvc source annotation found for pvc \"%s/%s\", value %s\n", pvc.Namespace, pvc.Name, source) default: klog.V(2).Infof("No valid source annotation found for pvc \"%s/%s\", default to http\n", pvc.Namespace, pvc.Name) source = SourceHTTP } return source } // returns the source string which determines the type of source. If no source or invalid source found, default to http func getContentType(pvc *v1.PersistentVolumeClaim) string { contentType, found := pvc.Annotations[AnnContentType] if !found { contentType = "" } switch contentType { case string(cdiv1.DataVolumeKubeVirt), string(cdiv1.DataVolumeArchive): klog.V(2).Infof("pvc content type annotation found for pvc \"%s/%s\", value %s\n", pvc.Namespace, pvc.Name, contentType) default: klog.V(2).Infof("No content type annotation found for pvc \"%s/%s\", default to kubevirt\n", pvc.Namespace, pvc.Name) contentType = string(cdiv1.DataVolumeKubeVirt) } return contentType } // returns the volumeMode which determines if the PVC is block PVC or not. func getVolumeMode(pvc *v1.PersistentVolumeClaim) v1.PersistentVolumeMode { if pvc.Spec.VolumeMode != nil { return *pvc.Spec.VolumeMode } return v1.PersistentVolumeFilesystem } // returns the name of the secret containing endpoint credentials consumed by the importer pod. // A value of "" implies there are no credentials for the endpoint being used. A returned error // causes processNextItem() to stop. func getSecretName(client kubernetes.Interface, pvc *v1.PersistentVolumeClaim) (string, error) { ns := pvc.Namespace name, found := pvc.Annotations[AnnSecret] if !found || name == "" { msg := "getEndpointSecret: " if !found { msg += "annotation %q is missing in pvc \"%s/%s\"" } else { msg += "secret name is missing from annotation %q in pvc \"%s/%s\"" } klog.V(2).Infof(msg+"\n", AnnSecret, ns, pvc.Name) return "", nil // importer pod will not contain secret credentials } klog.V(3).Infof("getEndpointSecret: retrieving Secret \"%s/%s\"\n", ns, name) _, err := client.CoreV1().Secrets(ns).Get(name, metav1.GetOptions{}) if k8serrors.IsNotFound(err) { klog.V(1).Infof("secret %q defined in pvc \"%s/%s\" is missing. Importer pod will run once this secret is created\n", name, ns, pvc.Name) return name, nil } if err != nil { return "", errors.Wrapf(err, "error getting secret %q defined in pvc \"%s/%s\"", name, ns, pvc.Name) } klog.V(1).Infof("retrieved secret %q defined in pvc \"%s/%s\"\n", name, ns, pvc.Name) return name, nil } // Update and return a copy of the passed-in pvc. Only one of the annotation or label maps is required though // both can be passed. // Note: the only pvc changes supported are annotations and labels. func updatePVC(client kubernetes.Interface, pvc *v1.PersistentVolumeClaim, anno, label map[string]string) (*v1.PersistentVolumeClaim, error) { klog.V(3).Infof("updatePVC: updating pvc \"%s/%s\" with anno: %+v and label: %+v", pvc.Namespace, pvc.Name, anno, label) applyUpdt := func(claim *v1.PersistentVolumeClaim, a, l map[string]string) { if a != nil { claim.ObjectMeta.Annotations = addToMap(claim.ObjectMeta.Annotations, a) } if l != nil { claim.ObjectMeta.Labels = addToMap(claim.ObjectMeta.Labels, l) } } var updtPvc *v1.PersistentVolumeClaim nsName := fmt.Sprintf("%s/%s", pvc.Namespace, pvc.Name) // don't mutate the passed-in pvc since it's likely from the shared informer pvcCopy := pvc.DeepCopy() // loop a few times in case the pvc is stale err := wait.PollImmediate(time.Second*1, time.Second*10, func() (bool, error) { var e error applyUpdt(pvcCopy, anno, label) updtPvc, e = client.CoreV1().PersistentVolumeClaims(pvc.Namespace).Update(pvcCopy) if e == nil { return true, nil // successful update } if k8serrors.IsConflict(e) { // pvc is likely stale klog.V(3).Infof("pvc %q is stale, re-trying\n", nsName) pvcCopy, e = client.CoreV1().PersistentVolumeClaims(pvc.Namespace).Get(pvc.Name, metav1.GetOptions{}) if e == nil { return false, nil // retry update } // Get failed, start over pvcCopy = pvc.DeepCopy() } klog.Errorf("%q update/get error: %v\n", nsName, e) return false, nil // retry }) if err == nil { klog.V(3).Infof("updatePVC: pvc %q updated", nsName) return updtPvc, nil } return pvc, errors.Wrapf(err, "error updating pvc %q\n", nsName) } // Sets an annotation `key: val` in the given pvc. Returns the updated pvc. func setPVCAnnotation(client kubernetes.Interface, pvc *v1.PersistentVolumeClaim, key, val string) (*v1.PersistentVolumeClaim, error) { klog.V(3).Infof("setPVCAnnotation: adding annotation \"%s: %s\" to pvc \"%s/%s\"\n", key, val, pvc.Namespace, pvc.Name) return updatePVC(client, pvc, map[string]string{key: val}, nil) } // checks if annotation `key` has a value of `val`. func checkIfAnnoExists(pvc *v1.PersistentVolumeClaim, key string, val string) bool { value, exists := pvc.ObjectMeta.Annotations[key] if exists && value == val { return true } return false } // checks if particular label exists in pvc func checkIfLabelExists(pvc *v1.PersistentVolumeClaim, lbl string, val string) bool { value, exists := pvc.ObjectMeta.Labels[lbl] if exists && value == val { return true } return false } // newScratchPersistentVolumeClaimSpec creates a new PVC based on the size of the passed in PVC. // It also sets the appropriate OwnerReferences on the resource // which allows handleObject to discover the pod resource that 'owns' it, and clean up when needed. func newScratchPersistentVolumeClaimSpec(pvc *v1.PersistentVolumeClaim, pod *v1.Pod, storageClassName string) *v1.PersistentVolumeClaim { labels := map[string]string{ "cdi-controller": pod.Name, "app": "containerized-data-importer", LabelImportPvc: pvc.Name, } pvcDef := &v1.PersistentVolumeClaim{ ObjectMeta: metav1.ObjectMeta{ Name: pvc.Name + "-scratch", Namespace: pvc.Namespace, Labels: labels, OwnerReferences: []metav1.OwnerReference{ { APIVersion: "v1", Kind: "Pod", Name: pod.Name, UID: pod.GetUID(), }, }, }, Spec: v1.PersistentVolumeClaimSpec{ AccessModes: []v1.PersistentVolumeAccessMode{"ReadWriteOnce"}, Resources: pvc.Spec.Resources, }, } if storageClassName != "" { pvcDef.Spec.StorageClassName = &storageClassName } return pvcDef } // CreateScratchPersistentVolumeClaim creates and returns a pointer to a scratch PVC which is created based on the passed-in pvc and storage class name. func CreateScratchPersistentVolumeClaim(client kubernetes.Interface, pvc *v1.PersistentVolumeClaim, pod *v1.Pod, storageClassName string) (*v1.PersistentVolumeClaim, error) { ns := pvc.Namespace scratchPvcSpec := newScratchPersistentVolumeClaimSpec(pvc, pod, storageClassName) scratchPvc, err := client.CoreV1().PersistentVolumeClaims(ns).Create(scratchPvcSpec) if err != nil { return nil, errors.Wrap(err, "scratch PVC API create errored") } klog.V(3).Infof("scratch PVC \"%s/%s\" created\n", scratchPvc.Namespace, scratchPvc.Name) return scratchPvc, nil } // GetScratchPvcStorageClass tries to determine which storage class to use for use with a scratch persistent // volume claim. The order of preference is the following: // 1. Defined value in CDI config map. // 2. If 1 is not available use the 'default' storage class. // 3. If 2 is not available use the storage class name of the original pvc that will own the scratch pvc. // 4. If none of those are available, return blank. func GetScratchPvcStorageClass(client kubernetes.Interface, cdiclient clientset.Interface, pvc *v1.PersistentVolumeClaim) string { config, err := cdiclient.CdiV1alpha1().CDIConfigs().Get(common.ConfigName, metav1.GetOptions{}) if err != nil { klog.Errorf("Unable to find CDI configuration, %v\n", err) } storageClassName := config.Status.ScratchSpaceStorageClass if storageClassName == "" { // Unable to determine default storage class, attempt to read the storage class from the pvc. if pvc.Spec.StorageClassName != nil { storageClassName = *pvc.Spec.StorageClassName if storageClassName != "" { return storageClassName } } } else { return storageClassName } return "" } // CreateImporterPod creates and returns a pointer to a pod which is created based on the passed-in endpoint, secret // name, and pvc. A nil secret means the endpoint credentials are not passed to the // importer pod. func CreateImporterPod(client kubernetes.Interface, image, verbose, pullPolicy string, podEnvVar *importPodEnvVar, pvc *v1.PersistentVolumeClaim, scratchPvcName *string) (*v1.Pod, error) { ns := pvc.Namespace pod := MakeImporterPodSpec(image, verbose, pullPolicy, podEnvVar, pvc, scratchPvcName) pod, err := client.CoreV1().Pods(ns).Create(pod) if err != nil { return nil, errors.Wrap(err, "importer pod API create errored") } klog.V(3).Infof("importer pod \"%s/%s\" (image: %q) created\n", pod.Namespace, pod.Name, image) return pod, nil } // MakeImporterPodSpec creates and return the importer pod spec based on the passed-in endpoint, secret and pvc. func MakeImporterPodSpec(image, verbose, pullPolicy string, podEnvVar *importPodEnvVar, pvc *v1.PersistentVolumeClaim, scratchPvcName *string) *v1.Pod { // importer pod name contains the pvc name podName := fmt.Sprintf("%s-%s-", common.ImporterPodName, pvc.Name) blockOwnerDeletion := true isController := true volumes := []v1.Volume{ { Name: DataVolName, VolumeSource: v1.VolumeSource{ PersistentVolumeClaim: &v1.PersistentVolumeClaimVolumeSource{ ClaimName: pvc.Name, ReadOnly: false, }, }, }, } if scratchPvcName != nil { volumes = append(volumes, v1.Volume{ Name: ScratchVolName, VolumeSource: v1.VolumeSource{ PersistentVolumeClaim: &v1.PersistentVolumeClaimVolumeSource{ ClaimName: *scratchPvcName, ReadOnly: false, }, }, }) } pod := &v1.Pod{ TypeMeta: metav1.TypeMeta{ Kind: "Pod", APIVersion: "v1", }, ObjectMeta: metav1.ObjectMeta{ GenerateName: podName, Annotations: map[string]string{ AnnCreatedBy: "yes", }, Labels: map[string]string{ common.CDILabelKey: common.CDILabelValue, common.CDIComponentLabel: common.ImporterPodName, // this label is used when searching for a pvc's import pod. LabelImportPvc: pvc.Name, common.PrometheusLabel: "", }, OwnerReferences: []metav1.OwnerReference{ { APIVersion: "v1", Kind: "PersistentVolumeClaim", Name: pvc.Name, UID: pvc.GetUID(), BlockOwnerDeletion: &blockOwnerDeletion, Controller: &isController, }, }, }, Spec: v1.PodSpec{ Containers: []v1.Container{ { Name: common.ImporterPodName, Image: image, ImagePullPolicy: v1.PullPolicy(pullPolicy), Args: []string{"-v=" + verbose}, Ports: []v1.ContainerPort{ { Name: "metrics", ContainerPort: 8443, Protocol: v1.ProtocolTCP, }, }, }, }, RestartPolicy: v1.RestartPolicyOnFailure, Volumes: volumes, }, } ownerUID := pvc.UID if len(pvc.OwnerReferences) == 1 { ownerUID = pvc.OwnerReferences[0].UID } if getVolumeMode(pvc) == v1.PersistentVolumeBlock { pod.Spec.Containers[0].VolumeDevices = addVolumeDevices() } else { pod.Spec.Containers[0].VolumeMounts = addVolumeMounts() } if scratchPvcName != nil { pod.Spec.Containers[0].VolumeMounts = append(pod.Spec.Containers[0].VolumeMounts, v1.VolumeMount{ Name: ScratchVolName, MountPath: common.ScratchDataDir, }) } pod.Spec.Containers[0].Env = makeEnv(podEnvVar, ownerUID) if podEnvVar.certConfigMap != "" { vm := v1.VolumeMount{ Name: CertVolName, MountPath: common.ImporterCertDir, } vol := v1.Volume{ Name: CertVolName, VolumeSource: v1.VolumeSource{ ConfigMap: &v1.ConfigMapVolumeSource{ LocalObjectReference: v1.LocalObjectReference{ Name: podEnvVar.certConfigMap, }, }, }, } pod.Spec.Containers[0].VolumeMounts = append(pod.Spec.Containers[0].VolumeMounts, vm) pod.Spec.Volumes = append(pod.Spec.Volumes, vol) } return pod } // this is being called for pods using PV with block volume mode func addVolumeDevices() []v1.VolumeDevice { volumeDevices := []v1.VolumeDevice{ { Name: DataVolName, DevicePath: common.ImporterWriteBlockPath, }, } return volumeDevices } // this is being called for pods using PV with filesystem volume mode func addVolumeMounts() []v1.VolumeMount { volumeMounts := []v1.VolumeMount{ { Name: DataVolName, MountPath: common.ImporterDataDir, }, } return volumeMounts } // return the Env portion for the importer container. func makeEnv(podEnvVar *importPodEnvVar, uid types.UID) []v1.EnvVar { env := []v1.EnvVar{ { Name: common.ImporterSource, Value: podEnvVar.source, }, { Name: common.ImporterEndpoint, Value: podEnvVar.ep, }, { Name: common.ImporterContentType, Value: podEnvVar.contentType, }, { Name: common.ImporterImageSize, Value: podEnvVar.imageSize, }, { Name: common.OwnerUID, Value: string(uid), }, { Name: common.InsecureTLSVar, Value: strconv.FormatBool(podEnvVar.insecureTLS), }, } if podEnvVar.secretName != "" { env = append(env, v1.EnvVar{ Name: common.ImporterAccessKeyID, ValueFrom: &v1.EnvVarSource{ SecretKeyRef: &v1.SecretKeySelector{ LocalObjectReference: v1.LocalObjectReference{ Name: podEnvVar.secretName, }, Key: common.KeyAccess, }, }, }, v1.EnvVar{ Name: common.ImporterSecretKey, ValueFrom: &v1.EnvVarSource{ SecretKeyRef: &v1.SecretKeySelector{ LocalObjectReference: v1.LocalObjectReference{ Name: podEnvVar.secretName, }, Key: common.KeySecret, }, }, }) } if podEnvVar.certConfigMap != "" { env = append(env, v1.EnvVar{ Name: common.ImporterCertDirVar, Value: common.ImporterCertDir, }) } return env } // Return a new map consisting of map1 with map2 added. In general, map2 is expected to have a single key. eg // a single annotation or label. If map1 has the same key as map2 then map2's value is used. func addToMap(m1, m2 map[string]string) map[string]string { if m1 == nil { m1 = make(map[string]string) } for k, v := range m2 { m1[k] = v } return m1 } // returns the CloneRequest string which contains the pvc name (and namespace) from which we want to clone the image. func getCloneRequestPVCAnnotation(pvc *v1.PersistentVolumeClaim) (string, error) { cr, found := pvc.Annotations[AnnCloneRequest] if !found || cr == "" { verb := "empty" if !found { verb = "missing" } return cr, errors.Errorf("annotation %q in pvc \"%s/%s\" is %s\n", AnnCloneRequest, pvc.Namespace, pvc.Name, verb) } return cr, nil } // returns the CloneRequest string which contains the pvc name (and namespace) from which we want to clone the image. func getCloneRequestSourcePVC(pvc *v1.PersistentVolumeClaim, pvcLister corelisters.PersistentVolumeClaimLister) (*v1.PersistentVolumeClaim, error) { ann, err := getCloneRequestPVCAnnotation(pvc) if err != nil { return nil, err } namespace, name := ParseSourcePvcAnnotation(ann, "/") if namespace == "" || name == "" { return nil, errors.New("Unable to parse to source PVC annotation") } return pvcLister.PersistentVolumeClaims(namespace).Get(name) } // ParseSourcePvcAnnotation parses out the annotations for a CDI PVC, splitting the string based on the delimiter argument func ParseSourcePvcAnnotation(sourcePvcAnno, del string) (namespace, name string) { strArr := strings.Split(sourcePvcAnno, del) if strArr == nil || len(strArr) < 2 { klog.V(3).Infof("Bad CloneRequest Annotation") return "", "" } return strArr[0], strArr[1] } // ValidateCanCloneSourceAndTargetSpec validates the specs passed in are compatible for cloning. func ValidateCanCloneSourceAndTargetSpec(sourceSpec, targetSpec *v1.PersistentVolumeClaimSpec) error { sourceRequest := sourceSpec.Resources.Requests[v1.ResourceStorage] targetRequest := targetSpec.Resources.Requests[v1.ResourceStorage] // Verify that the target PVC size is equal or larger than the source. if sourceRequest.Value() > targetRequest.Value() { return errors.New("Target resources requests storage size is smaller than the source") } // Verify that the source and target volume modes are the same. sourceVolumeMode := v1.PersistentVolumeFilesystem if sourceSpec.VolumeMode != nil && *sourceSpec.VolumeMode == v1.PersistentVolumeBlock { sourceVolumeMode = v1.PersistentVolumeBlock } targetVolumeMode := v1.PersistentVolumeFilesystem if targetSpec.VolumeMode != nil && *targetSpec.VolumeMode == v1.PersistentVolumeBlock { targetVolumeMode = v1.PersistentVolumeBlock } if sourceVolumeMode != targetVolumeMode { return errors.New("Source and target volume modes do not match") } // Can clone. return nil } // CreateCloneSourcePod creates our cloning src pod which will be used for out of band cloning to read the contents of the src PVC func CreateCloneSourcePod(client kubernetes.Interface, image string, pullPolicy string, cr string, pvc *v1.PersistentVolumeClaim) (*v1.Pod, error) { sourcePvcNamespace, sourcePvcName := ParseSourcePvcAnnotation(cr, "/") if sourcePvcNamespace == "" || sourcePvcName == "" { return nil, errors.Errorf("Bad CloneRequest Annotation") } pod := MakeCloneSourcePodSpec(image, pullPolicy, sourcePvcName, pvc) pod, err := client.CoreV1().Pods(sourcePvcNamespace).Create(pod) if err != nil { return nil, errors.Wrap(err, "source pod API create errored") } klog.V(1).Infof("cloning source pod \"%s/%s\" (image: %q) created\n", pod.Namespace, pod.Name, image) return pod, nil } // MakeCloneSourcePodSpec creates and returns the clone source pod spec based on the target pvc. func MakeCloneSourcePodSpec(image, pullPolicy, sourcePvcName string, pvc *v1.PersistentVolumeClaim) *v1.Pod { // source pod name contains the pvc name podName := fmt.Sprintf("%s-", common.ClonerSourcePodName) id := string(pvc.GetUID()) blockOwnerDeletion := true isController := true pod := &v1.Pod{ TypeMeta: metav1.TypeMeta{ Kind: "Pod", APIVersion: "v1", }, ObjectMeta: metav1.ObjectMeta{ GenerateName: podName, Annotations: map[string]string{ AnnCreatedBy: "yes", AnnTargetPodNamespace: pvc.Namespace, }, Labels: map[string]string{ common.CDILabelKey: common.CDILabelValue, //filtered by the podInformer common.CDIComponentLabel: common.ClonerSourcePodName, common.CloningLabelKey: common.CloningLabelValue + "-" + id, //used by podAffity // this label is used when searching for a pvc's cloner source pod. CloneUniqueID: id + "-source-pod", }, OwnerReferences: []metav1.OwnerReference{ { APIVersion: "v1", Kind: "PersistentVolumeClaim", Name: pvc.Name, UID: pvc.GetUID(), BlockOwnerDeletion: &blockOwnerDeletion, Controller: &isController, }, }, }, Spec: v1.PodSpec{ // We create initContainer just to set the pod as privileged. // The pod has to be privileged as it has to have access to the hostPath in the node. // However, currently there is a bug that we cannot attach block device to the pod if the pod (container) // is privileged: //https://github.com/kubernetes/kubernetes/issues/58251 //https://github.com/kubernetes/kubernetes/issues/62560 // As a result of that, instead setting the SecurityContext field as Privileged, in the main container // where we need to have access to the hostPath, we set the SecurityContext field in the initContainer, // and that will do it. InitContainers: []v1.Container{ { Name: "init", Image: image, VolumeMounts: []v1.VolumeMount{ { Name: socketPathName, MountPath: common.ClonerSocketPath + "/" + id, }, }, SecurityContext: &v1.SecurityContext{ Privileged: &[]bool{true}[0], RunAsUser: &[]int64{0}[0], }, Command: []string{"sh", "-c", "echo setting the pod as privileged"}, }, }, Containers: []v1.Container{ { Name: common.ClonerSourcePodName, Image: image, ImagePullPolicy: v1.PullPolicy(pullPolicy), }, }, RestartPolicy: v1.RestartPolicyNever, Volumes: []v1.Volume{ { Name: DataVolName, VolumeSource: v1.VolumeSource{ PersistentVolumeClaim: &v1.PersistentVolumeClaimVolumeSource{ ClaimName: sourcePvcName, ReadOnly: false, }, }, }, { Name: socketPathName, VolumeSource: v1.VolumeSource{ HostPath: &v1.HostPathVolumeSource{ Path: common.ClonerSocketPath + "/" + id, }, }, }, }, }, } var volumeMode v1.PersistentVolumeMode if pvc.Spec.VolumeMode != nil { volumeMode = *pvc.Spec.VolumeMode } else { volumeMode = v1.PersistentVolumeFilesystem } if volumeMode == v1.PersistentVolumeBlock { pod.Spec.Containers[0].VolumeDevices = addVolumeDevices() pod.Spec.Containers[0].VolumeMounts = addCloneVolumeMounts("Block", id) pod.Spec.Containers[0].Args = addArgs("source", id, "block") } else { pod.Spec.Containers[0].VolumeMounts = addCloneVolumeMounts("FS", id) pod.Spec.Containers[0].Args = addArgs("source", id, "FS") } return pod } func addArgs(source, id, volumeMode string) []string { args := []string{ source, id, volumeMode, } return args } func addCloneVolumeMounts(volumeMode, id string) []v1.VolumeMount { if volumeMode == "Block" { volumeMounts := []v1.VolumeMount{ { Name: socketPathName, MountPath: common.ClonerSocketPath + "/" + id, }, } return volumeMounts } volumeMounts := []v1.VolumeMount{ { Name: DataVolName, MountPath: common.ClonerImagePath, }, { Name: socketPathName, MountPath: common.ClonerSocketPath + "/" + id, }, } return volumeMounts } // CreateCloneTargetPod creates our cloning tgt pod which will be used for out of band cloning to write the contents of the tgt PVC func CreateCloneTargetPod(client kubernetes.Interface, image string, pullPolicy string, pvc *v1.PersistentVolumeClaim, podAffinityNamespace string) (*v1.Pod, error) { ns := pvc.Namespace pod := MakeCloneTargetPodSpec(image, pullPolicy, podAffinityNamespace, pvc) pod, err := client.CoreV1().Pods(ns).Create(pod) if err != nil { return nil, errors.Wrap(err, "clone target pod API create errored") } klog.V(1).Infof("cloning target pod \"%s/%s\" (image: %q) created\n", pod.Namespace, pod.Name, image) return pod, nil } // MakeCloneTargetPodSpec creates and returns the clone target pod spec based on the target pvc. func MakeCloneTargetPodSpec(image, pullPolicy, podAffinityNamespace string, pvc *v1.PersistentVolumeClaim) *v1.Pod { // target pod name contains the pvc name podName := fmt.Sprintf("%s-", common.ClonerTargetPodName) id := string(pvc.GetUID()) blockOwnerDeletion := true isController := true ownerUID := pvc.UID if len(pvc.OwnerReferences) == 1 { ownerUID = pvc.OwnerReferences[0].UID } pod := &v1.Pod{ TypeMeta: metav1.TypeMeta{ Kind: "Pod", APIVersion: "v1", }, ObjectMeta: metav1.ObjectMeta{ GenerateName: podName, Annotations: map[string]string{ AnnCreatedBy: "yes", AnnTargetPodNamespace: pvc.Namespace, }, Labels: map[string]string{ common.CDILabelKey: common.CDILabelValue, //filtered by the podInformer common.CDIComponentLabel: common.ClonerTargetPodName, // this label is used when searching for a pvc's cloner target pod. CloneUniqueID: id + "-target-pod", common.PrometheusLabel: "", }, OwnerReferences: []metav1.OwnerReference{ { APIVersion: "v1", Kind: "PersistentVolumeClaim", Name: pvc.Name, UID: pvc.GetUID(), BlockOwnerDeletion: &blockOwnerDeletion, Controller: &isController, }, }, }, Spec: v1.PodSpec{ Affinity: &v1.Affinity{ PodAffinity: &v1.PodAffinity{ RequiredDuringSchedulingIgnoredDuringExecution: []v1.PodAffinityTerm{ { LabelSelector: &metav1.LabelSelector{ MatchExpressions: []metav1.LabelSelectorRequirement{ { Key: common.CloningLabelKey, Operator: metav1.LabelSelectorOpIn, Values: []string{common.CloningLabelValue + "-" + id}, }, }, }, Namespaces: []string{podAffinityNamespace}, //the scheduler looks for the namespace of the source pod TopologyKey: common.CloningTopologyKey, }, }, }, }, // We create initContainer just to set the pod as privileged. // The pod has to be privileged as it has to have access to the hostPath in the node. // However, currently there is a bug that we cannot attach block device to the pod if the pod (container) // is privileged: //https://github.com/kubernetes/kubernetes/issues/58251 //https://github.com/kubernetes/kubernetes/issues/62560 // As a result of that, instead setting the SecurityContext field as Privileged, in the main container // where we need to have access to the hostPath, we set the SecurityContext field in the initContainer, // and that will do it. InitContainers: []v1.Container{ { Name: "init", Image: image, VolumeMounts: []v1.VolumeMount{ { Name: socketPathName, MountPath: common.ClonerSocketPath + "/" + id, }, }, SecurityContext: &v1.SecurityContext{ Privileged: &[]bool{true}[0], RunAsUser: &[]int64{0}[0], }, Command: []string{"sh", "-c", "echo setting the pod as privileged"}, }, }, Containers: []v1.Container{ { Name: common.ClonerTargetPodName, Image: image, ImagePullPolicy: v1.PullPolicy(pullPolicy), Ports: []v1.ContainerPort{ { Name: "metrics", ContainerPort: 8443, Protocol: v1.ProtocolTCP, }, }, Env: []v1.EnvVar{ { Name: common.OwnerUID, Value: string(ownerUID), }, }, }, }, RestartPolicy: v1.RestartPolicyNever, Volumes: []v1.Volume{ { Name: DataVolName, VolumeSource: v1.VolumeSource{ PersistentVolumeClaim: &v1.PersistentVolumeClaimVolumeSource{ ClaimName: pvc.Name, ReadOnly: false, }, }, }, { Name: socketPathName, VolumeSource: v1.VolumeSource{ HostPath: &v1.HostPathVolumeSource{ Path: common.ClonerSocketPath + "/" + id, }, }, }, }, }, } var volumeMode v1.PersistentVolumeMode if pvc.Spec.VolumeMode != nil { volumeMode = *pvc.Spec.VolumeMode } else { volumeMode = v1.PersistentVolumeFilesystem } if volumeMode == v1.PersistentVolumeBlock { pod.Spec.Containers[0].VolumeDevices = addVolumeDevices() pod.Spec.Containers[0].VolumeMounts = addCloneVolumeMounts("Block", id) pod.Spec.Containers[0].Args = addArgs("target", id, "block") } else { pod.Spec.Containers[0].VolumeMounts = addCloneVolumeMounts("FS", id) pod.Spec.Containers[0].Args = addArgs("target", id, "FS") } return pod } // CreateUploadPod creates upload service pod manifest and sends to server func CreateUploadPod(client kubernetes.Interface, caKeyPair *triple.KeyPair, clientCACert *x509.Certificate, image string, verbose string, pullPolicy string, name string, pvc *v1.PersistentVolumeClaim, scratchPvcName string) (*v1.Pod, error) { ns := pvc.Namespace commonName := name + "." + ns secretName := name + "-server-tls" pod := MakeUploadPodSpec(image, verbose, pullPolicy, name, pvc, scratchPvcName, secretName) pod, err := client.CoreV1().Pods(ns).Create(pod) if err != nil { if k8serrors.IsAlreadyExists(err) { pod, err = client.CoreV1().Pods(ns).Get(name, metav1.GetOptions{}) if err != nil { return nil, errors.Wrap(err, "upload pod should exist but couldn't retrieve it") } } else { return nil, errors.Wrap(err, "upload pod API create errored") } } podOwner := MakePodOwnerReference(pod) _, err = keys.GetOrCreateServerKeyPairAndCert(client, ns, secretName, caKeyPair, clientCACert, commonName, name, &podOwner) if err != nil { // try to clean up client.CoreV1().Pods(ns).Delete(pod.Name, &metav1.DeleteOptions{}) return nil, errors.Wrap(err, "Error creating server key pair") } klog.V(1).Infof("upload pod \"%s/%s\" (image: %q) created\n", pod.Namespace, pod.Name, image) return pod, nil } // MakePVCOwnerReference makes owner reference from a PVC func MakePVCOwnerReference(pvc *v1.PersistentVolumeClaim) metav1.OwnerReference { blockOwnerDeletion := true isController := true return metav1.OwnerReference{ APIVersion: "v1", Kind: "PersistentVolumeClaim", Name: pvc.Name, UID: pvc.GetUID(), BlockOwnerDeletion: &blockOwnerDeletion, Controller: &isController, } } // MakePodOwnerReference makes owner reference from a Pod func MakePodOwnerReference(pod *v1.Pod) metav1.OwnerReference { blockOwnerDeletion := true isController := true return metav1.OwnerReference{ APIVersion: "v1", Kind: "Pod", Name: pod.Name, UID: pod.GetUID(), BlockOwnerDeletion: &blockOwnerDeletion, Controller: &isController, } } // MakeUploadPodSpec creates upload service pod manifest func MakeUploadPodSpec(image, verbose, pullPolicy, name string, pvc *v1.PersistentVolumeClaim, scratchName, secretName string) *v1.Pod { requestImageSize, _ := getRequestedImageSize(pvc) pod := &v1.Pod{ TypeMeta: metav1.TypeMeta{ Kind: "Pod", APIVersion: "v1", }, ObjectMeta: metav1.ObjectMeta{ Name: name, Annotations: map[string]string{ annCreatedByUpload: "yes", }, Labels: map[string]string{ common.CDILabelKey: common.CDILabelValue, common.CDIComponentLabel: common.UploadServerCDILabel, common.UploadServerServiceLabel: name, }, OwnerReferences: []metav1.OwnerReference{ MakePVCOwnerReference(pvc), }, }, Spec: v1.PodSpec{ Containers: []v1.Container{ { Name: common.UploadServerPodname, Image: image, ImagePullPolicy: v1.PullPolicy(pullPolicy), Env: []v1.EnvVar{ { Name: "TLS_KEY", ValueFrom: &v1.EnvVarSource{ SecretKeyRef: &v1.SecretKeySelector{ LocalObjectReference: v1.LocalObjectReference{ Name: secretName, }, Key: keys.KeyStoreTLSKeyFile, }, }, }, { Name: "TLS_CERT", ValueFrom: &v1.EnvVarSource{ SecretKeyRef: &v1.SecretKeySelector{ LocalObjectReference: v1.LocalObjectReference{ Name: secretName, }, Key: keys.KeyStoreTLSCertFile, }, }, }, { Name: "CLIENT_CERT", ValueFrom: &v1.EnvVarSource{ SecretKeyRef: &v1.SecretKeySelector{ LocalObjectReference: v1.LocalObjectReference{ Name: secretName, }, Key: keys.KeyStoreTLSCAFile, }, }, }, { Name: common.UploadImageSize, Value: requestImageSize, }, }, Args: []string{"-v=" + verbose}, ReadinessProbe: &v1.Probe{ Handler: v1.Handler{ HTTPGet: &v1.HTTPGetAction{ Path: "/healthz", Port: intstr.IntOrString{ Type: intstr.Int, IntVal: 8080, }, }, }, InitialDelaySeconds: 2, PeriodSeconds: 5, }, }, }, RestartPolicy: v1.RestartPolicyOnFailure, Volumes: []v1.Volume{ { Name: DataVolName, VolumeSource: v1.VolumeSource{ PersistentVolumeClaim: &v1.PersistentVolumeClaimVolumeSource{ ClaimName: pvc.Name, ReadOnly: false, }, }, }, { Name: ScratchVolName, VolumeSource: v1.VolumeSource{ PersistentVolumeClaim: &v1.PersistentVolumeClaimVolumeSource{ ClaimName: scratchName, ReadOnly: false, }, }, }, }, }, } if getVolumeMode(pvc) == v1.PersistentVolumeBlock { pod.Spec.Containers[0].VolumeDevices = addVolumeDevicesForUpload() pod.Spec.Containers[0].Env = append(pod.Spec.Containers[0].Env, v1.EnvVar{ Name: "DESTINATION", Value: "/dev/blockDevice", }) } else { pod.Spec.Containers[0].VolumeMounts = addVolumeMountsForUpload() } pod.Spec.Containers[0].VolumeMounts = append(pod.Spec.Containers[0].VolumeMounts, v1.VolumeMount{ Name: ScratchVolName, MountPath: common.ScratchDataDir, }) return pod } func addVolumeDevicesForUpload() []v1.VolumeDevice { volumeDevices := []v1.VolumeDevice{ { Name: DataVolName, DevicePath: common.ImporterWriteBlockPath, }, } return volumeDevices } func addVolumeMountsForUpload() []v1.VolumeMount { volumeMounts := []v1.VolumeMount{ { Name: DataVolName, MountPath: common.UploadServerDataDir, }, } return volumeMounts } // CreateUploadService creates upload service service manifest and sends to server func CreateUploadService(client kubernetes.Interface, name string, pvc *v1.PersistentVolumeClaim) (*v1.Service, error) { ns := pvc.Namespace service := MakeUploadServiceSpec(name, pvc) service, err := client.CoreV1().Services(ns).Create(service) if err != nil { if k8serrors.IsAlreadyExists(err) { service, err = client.CoreV1().Services(ns).Get(name, metav1.GetOptions{}) if err != nil { return nil, errors.Wrap(err, "upload service should exist but couldn't retrieve it") } } else { return nil, errors.Wrap(err, "upload pod API create errored") } } klog.V(1).Infof("upload service \"%s/%s\" created\n", service.Namespace, service.Name) return service, nil } // MakeUploadServiceSpec creates upload service service manifest func MakeUploadServiceSpec(name string, pvc *v1.PersistentVolumeClaim) *v1.Service { blockOwnerDeletion := true isController := true service := &v1.Service{ TypeMeta: metav1.TypeMeta{ Kind: "Service", APIVersion: "v1", }, ObjectMeta: metav1.ObjectMeta{ Name: name, Annotations: map[string]string{ annCreatedByUpload: "yes", }, Labels: map[string]string{ common.CDILabelKey: common.CDILabelValue, common.CDIComponentLabel: common.UploadServerCDILabel, }, OwnerReferences: []metav1.OwnerReference{ { APIVersion: "v1", Kind: "PersistentVolumeClaim", Name: pvc.Name, UID: pvc.GetUID(), BlockOwnerDeletion: &blockOwnerDeletion, Controller: &isController, }, }, }, Spec: v1.ServiceSpec{ Ports: []v1.ServicePort{ { Protocol: "TCP", Port: 443, TargetPort: intstr.IntOrString{ Type: intstr.Int, IntVal: 8443, }, }, }, Selector: map[string]string{ common.UploadServerServiceLabel: name, }, }, } return service } // EnsureCDIConfigExists creates an empty CDIConfig if necessary func EnsureCDIConfigExists(client kubernetes.Interface, cdiClient clientset.Interface, name string) error { cfg := MakeEmptyCDIConfigSpec(name) err := operator.SetOwner(client, cfg) if err != nil { return errors.Wrap(err, "Error setting CDI config owner ref") } config, err := cdiClient.CdiV1alpha1().CDIConfigs().Create(cfg) if err != nil { if k8serrors.IsAlreadyExists(err) { return nil } return errors.Wrap(err, "CDI config create errored") } klog.V(1).Infof("CDI config \"%s\" created\n", config.Name) return nil } // MakeEmptyCDIConfigSpec creates cdi config manifest func MakeEmptyCDIConfigSpec(name string) *cdiv1.CDIConfig { config := &cdiv1.CDIConfig{ ObjectMeta: metav1.ObjectMeta{ Name: name, Labels: map[string]string{ common.CDILabelKey: common.CDILabelValue, common.CDIComponentLabel: "", }, }, } return config } func updateCDIConfig(cdiClient clientset.Interface, config *cdiv1.CDIConfig) error { _, err := cdiClient.CdiV1alpha1().CDIConfigs().Update(config) if err != nil { return err } return nil } func deletePod(req podDeleteRequest) error { pod, err := req.podLister.Pods(req.namespace).Get(req.podName) if k8serrors.IsNotFound(err) { return nil } if err == nil && pod.DeletionTimestamp == nil { err = req.k8sClient.CoreV1().Pods(req.namespace).Delete(req.podName, &metav1.DeleteOptions{}) if k8serrors.IsNotFound(err) { return nil } } if err != nil { klog.V(1).Infof("error encountered deleting pod (%s): %s", req.podName, err.Error()) } return err } func createImportEnvVar(client kubernetes.Interface, pvc *v1.PersistentVolumeClaim) (*importPodEnvVar, error) { podEnvVar := &importPodEnvVar{} podEnvVar.source = getSource(pvc) podEnvVar.contentType = getContentType(pvc) var err error if podEnvVar.source != SourceNone { podEnvVar.ep, err = getEndpoint(pvc) if err != nil { return nil, err } podEnvVar.secretName, err = getSecretName(client, pvc) if err != nil { return nil, err } if podEnvVar.secretName == "" { klog.V(2).Infof("no secret will be supplied to endpoint %q\n", podEnvVar.ep) } podEnvVar.certConfigMap, err = getCertConfigMap(client, pvc) if err != nil { return nil, err } podEnvVar.insecureTLS, err = isInsecureTLS(client, pvc) if err != nil { return nil, err } } //get the requested image size. podEnvVar.imageSize, err = getRequestedImageSize(pvc) if err != nil { return nil, err } return podEnvVar, nil } func getCertConfigMap(client kubernetes.Interface, pvc *v1.PersistentVolumeClaim) (string, error) { value, ok := pvc.Annotations[AnnCertConfigMap] if !ok || value == "" { return "", nil } _, err := client.CoreV1().ConfigMaps(pvc.Namespace).Get(value, metav1.GetOptions{}) if err != nil { if k8serrors.IsNotFound(err) { klog.Warningf("Configmap %s does not exist, pod will not start until it does", value) return value, nil } return "", err } return value, nil } func getURLFromIngress(ing *extensionsv1beta1.Ingress, uploadProxyServiceName string) string { if ing.Spec.Backend != nil { if ing.Spec.Backend.ServiceName != uploadProxyServiceName { return "" } return ing.Spec.Rules[0].Host } for _, rule := range ing.Spec.Rules { for _, path := range rule.HTTP.Paths { if path.Backend.ServiceName == uploadProxyServiceName { if rule.Host != "" { return rule.Host } } } } return "" } func getURLFromRoute(route *routev1.Route, uploadProxyServiceName string) string { if route.Spec.To.Name == uploadProxyServiceName { if len(route.Status.Ingress) > 0 { return route.Status.Ingress[0].Host } } return "" } //IsOpenshift checks if we are on OpenShift platform func IsOpenshift(client kubernetes.Interface) bool { result := client.Discovery().RESTClient().Get().AbsPath("/oapi/v1").Do() var statusCode int result.StatusCode(&statusCode) if result.Error() == nil { // It is OpenShift if statusCode == http.StatusOK { return true } } else { // Got 404 so this is not Openshift if statusCode == http.StatusNotFound { return false } } return false } func isInsecureTLS(client kubernetes.Interface, pvc *v1.PersistentVolumeClaim) (bool, error) { var configMapName string value, ok := pvc.Annotations[AnnEndpoint] if !ok || value == "" { return false, nil } url, err := url.Parse(value) if err != nil { return false, err } switch url.Scheme { case "docker": configMapName = common.InsecureRegistryConfigMap default: return false, nil } klog.V(3).Infof("Checking configmap %s for host %s", configMapName, url.Host) cm, err := client.CoreV1().ConfigMaps(util.GetNamespace()).Get(configMapName, metav1.GetOptions{}) if err != nil { if k8serrors.IsNotFound(err) { klog.Warningf("Configmap %s does not exist", configMapName) return false, nil } return false, err } for key, value := range cm.Data { klog.V(3).Infof("Checking %q against %q: %q", url.Host, key, value) if value == url.Host { return true, nil } } return false, nil } // IsCsiCrdsDeployed checks whether the CSI snapshotter CRD are deployed func IsCsiCrdsDeployed(c extclientset.Interface) bool { vsClass := crdv1alpha1.VolumeSnapshotClassResourcePlural + "." + crdv1alpha1.GroupName vsContent := crdv1alpha1.VolumeSnapshotContentResourcePlural + "." + crdv1alpha1.GroupName vs := crdv1alpha1.VolumeSnapshotResourcePlural + "." + crdv1alpha1.GroupName return isCrdDeployed(c, vsClass) && isCrdDeployed(c, vsContent) && isCrdDeployed(c, vs) } func isCrdDeployed(c extclientset.Interface, name string) bool { obj, err := c.ApiextensionsV1beta1().CustomResourceDefinitions().Get(name, metav1.GetOptions{}) if err != nil { if k8serrors.IsNotFound(err) { return false } return false } return obj != nil }