Mithilfe von Einschränkungsvorlagen können Sie die Funktionsweise einer Einschränkung definieren, aber die Details dieser Einschränkung an eine Einzelperson oder Gruppe mit Fachkenntnissen delegieren. Neben der Trennung von Fragestellungen wird auch die Logik der Einschränkung von der Definition getrennt.
Alle Einschränkungen enthalten den Abschnitt match, in dem die Objekte definiert werden, auf die eine Einschränkung angewendet wird. Weitere Informationen zum Konfigurieren dieses Abschnitts finden Sie unter Abschnitt zum Einschränkungsabgleich.
Nicht alle Einschränkungsvorlagen sind für alle Versionen von Policy Controller verfügbar und Vorlagen können zwischen den Versionen geändert werden. Über folgende Links können Sie Einschränkungen von unterstützten Versionen vergleichen:
Links zu früheren Versionen dieser Seite
Damit Sie vollständigen Support erhalten, empfehlen wir die Verwendung von Einschränkungsvorlagen aus einer unterstützten Version von Policy Controller.
Damit Sie sehen können, wie die Einschränkungsvorlagen funktionieren, enthält jede Vorlage eine Beispieleinschränkung und eine Ressource, die gegen die Einschränkung verstößt.
Verfügbare Einschränkungsvorlagen
| Einschränkungsvorlage | Beschreibung | Referenziell |
|---|---|---|
| AllowedServicePortName | Erfordert, dass Dienstportnamen ein Präfix aus einer angegebenen Liste haben. | Nein |
| AsmAuthzPolicyDefaultDeny | Erzwingen Sie die AuthorizationPolicy für standardmäßige Ablehnung auf Mesh-Ebene. Lesen Sie dazu https://istio.io/latest/docs/ops/best-practices/security/#use-default-deny-patterns. | Ja |
| AsmAuthzPolicyDisallowedPrefix | Erfordert, dass Hauptkonten und Namespaces in "AuthorizationPolicy"-Regeln von Istio kein Präfix aus einer angegebenen Liste haben. https://istio.io/latest/docs/reference/config/security/authorization-policy/ | Nein |
| AsmAuthzPolicyEnforceSourcePrincipals | Erfordert, dass das Istio AuthorizationPolicy-Feld "from", wenn definiert, Quellprinzipien hat, die auf etwas anderes als "*" gesetzt sein müssen. https://istio.io/latest/docs/reference/config/security/authorization-policy/ | Nein |
| AsmAuthzPolicyNormalization | Erzwingen Sie die Normalisierung der AuthorizationPolicy. Lesen Sie dazu https://istio.io/latest/docs/reference/config/security/normalization/. | Nein |
| AsmAuthzPolicySafePattern | Erzwingen Sie die sicheren Muster der AuthorizationPolicy. Lesen Sie dazu https://istio.io/latest/docs/ops/best-practices/security/#safer-authorization-policy-patterns. | Nein |
| AsmIngressgatewayLabel | Erzwingen Sie die Labelverwendung von Istio-Ingressgateways nur für Ingressgateway-Pods. | Nein |
| AsmPeerAuthnMeshStrictMtls | Erzwingen Sie die strikte mTLS PeerAuthentication auf Mesh-Ebene. Lesen Sie dazu https://istio.io/latest/docs/ops/best-practices/security/#mutual-tls. | Ja |
| AsmPeerAuthnStrictMtls | Erzwingen Sie, dass alle PeerAuthentications keine strikten mTLS überschreiben können. Lesen Sie dazu https://istio.io/latest/docs/ops/best-practices/security/#mutual-tls. | Nein |
| AsmRequestAuthnProhibitedOutputHeaders | Erzwingen Sie bei RequestAuthentication, dass das Feld „jwtRules.outPayloadToHeader“ keine bekannten HTTP-Anfrageheader oder benutzerdefinierten verbotenen Header enthält. Referenz zu https://istio.io/latest/docs/reference/config/security/jwt/#JWTRule. | Nein |
| AsmSidecarInjection | Erzwingen Sie, dass die istio-Proxy-Sidecar-Datei immer in die Workload-Pods eingefügt wird. | Nein |
| DestinationRuleTLSEnabled | Verbietet die Deaktivierung von TLS für alle Hosts und Hostteilmengen in Istio DestinationRules. | Nein |
| DisallowedAuthzPrefix | Erfordert, dass Hauptkonten und Namespaces in "AuthorizationPolicy"-Regeln von Istio kein Präfix aus einer angegebenen Liste haben. https://istio.io/latest/docs/reference/config/security/authorization-policy/ | Nein |
| GCPStorageLocationConstraintV1 | Beschränkt die zulässigen "Standorte" für StorageBucket Config Connector-Ressourcen auf die Liste der in der Einschränkung angegebenen Standorte. Bucket-Namen in der -Liste sind ausgenommen. | Nein |
| GkeSpotVMTerminationGrace | Erfordert Pods und Pod-Vorlagen mit "nodeSelector" oder "nodeAfffinty" von "gke-spot" mit einer "terminationGracePeriodSeconds" von maximal 15 Sekunden. | Ja |
| K8sAllowedRepos | Erfordert Container-Images, die mit einem String aus der angegebenen Liste beginnen sollen. | Nein |
| K8sAvoidUseOfSystemMastersGroup | Verhindert die Verwendung der Gruppe "system:masters". Hat keine Auswirkungen während der Prüfung. | Nein |
| K8sBlockAllIngress | Verhindert das Erstellen von Ingress-Objekten (Typen „Ingress“, „Gateway“ und „Service“ von „NodePort“ und „LoadBalancer“). | Nein |
| K8sBlockCreationWithDefaultServiceAccount | Verhindert das Erstellen von Ressourcen mit einem Standarddienstkonto. Hat keine Auswirkungen während der Prüfung. | Nein |
| K8sBlockEndpointEditDefaultRole | Viele Kubernetes-Installationen haben standardmäßig eine system:aggregate-to-edit ClusterRole, die den Zugriff auf die Bearbeitung von Endpunkten nicht richtig einschränkt. Dieses ConstraintTemplate verbietet der system:aggregate-to-edit ClusterRole, die Erlaubnis zur Erstellung/Patch/Aktualisierung von Endpunkten zu erteilen. ClusterRole/system:aggregation-to-edit sollte aufgrund von CVE-2021-25740-Berechtigungen, den Endpoint- und EndpointSlice-Berechtigungen, die Namespace-übergreifende Weiterleitung nicht zulassen, https://github.com/kubernetes/kubernetes/issues/103675. | Nein |
| K8sBlockLoadBalancer | Verhindert alle Dienste mit dem Typ LoadBalancer. https://kubernetes.io/docs/concepts/services-networking/service/#loadbalancer | Nein |
| K8sBlockNodePort | Verhindert alle Dienste mit dem Typ NodePort. https://kubernetes.io/docs/concepts/services-networking/service/#nodeport | Nein |
| K8sBlockObjectsOfType | Verhindert das Objekt unzulässiger Typen. | Nein |
| K8sBlockProcessNamespaceSharing | Unterbindet Pod-Spezifikationen, bei denen „shareProcessNamespace“ auf „true“ gesetzt ist. Dadurch werden Szenarien vermieden, in denen alle Container in einem Pod den gleichen PID-Namespace verwenden und auf das Dateisystem und den Arbeitsspeicher des jeweils anderen Systems zugreifen können. | Nein |
| K8sBlockWildcardIngress | Nutzer sollten keine Ingresses mit einem leeren oder Platzhalternamen (*) erstellen können, da sie dadurch Traffic für andere Dienste im Cluster abfangen könnten, selbst wenn sie keinen Zugriff auf diese Dienste haben. | Nein |
| K8sContainerEphemeralStorageLimit | Erfordert, dass für Container ein sitzungsspezifisches Speicherlimit festgelegt ist und dass das Limit innerhalb der angegebenen Höchstwerte liegt. https://kubernetes.io/docs/concepts/configuration/manage-resources-containers/ | Nein |
| K8sContainerLimits | Erfordert, dass in Containern Speicher- und CPU-Beschränkungen festgelegt sind, die innerhalb der angegebenen Höchstwerte liegen müssen. https://kubernetes.io/docs/concepts/configuration/manage-resources-containers/ | Nein |
| K8sContainerRatios | Legt ein maximales Verhältnis für Containerressourcenlimits zu Anfragen fest. https://kubernetes.io/docs/concepts/configuration/manage-resources-containers/ | Nein |
| K8sContainerRequests | Erfordert, dass in Containern Arbeitsspeicher- und CPU-Anfragen festgelegt sind und dass Anfragen innerhalb der angegebenen Höchstwerte liegen müssen. https://kubernetes.io/docs/concepts/configuration/manage-resources-containers/ | Nein |
| K8sCronJobAllowedRepos | Erfordert Container-Images von CronJobs, die mit einem String aus der angegebenen Liste beginnen sollen. | Nein |
| K8sDisallowAnonymous | Verbietet die Zuordnung von ClusterRole- und Rollen-Ressourcen zum Nutzer "system:anonymous" und zur Gruppe "system:unauthenticated". | Nein |
| K8sDisallowInteractiveTTY | Erfordert, dass für die Objekte die Felder "spec.tty" und "spec.stbin" auf "false" oder nicht festgelegt sind. | Nein |
| K8sDisallowedRepos | Unzulässige Container-Repositories, die mit einem String aus der angegebenen Liste beginnen. | Nein |
| K8sDisallowedRoleBindingSubjects | Unterbindet RoleBinding oder ClusterRoleBinding mit Subjekten, die mit einem als Parameter übergebenen "disallowedSubjects" übereinstimmen. | Nein |
| K8sDisallowedTags | Erfordert, dass Container-Images ein Image-Tag haben, das sich von denen in der angegebenen Liste unterscheidet. https://kubernetes.io/docs/concepts/containers/images/#image-names | Nein |
| K8sEmptyDirHasSizeLimit | Erfordert, dass alle "emptyDir"-Volumes ein "sizeLimit" angeben. Optional kann der Parameter „maxSizeLimit“ in der Einschränkung bereitgestellt werden, um eine maximal zulässige Größenbeschränkung anzugeben. | Nein |
| K8sEnforceCloudArmorBackendConfig | Erzwingt die Cloud Armor-Konfiguration für BackendConfig-Ressourcen | Nein |
| K8sEnforceConfigManagement | Erfordert das Vorhandensein und den Betrieb von Config Management. Einschränkungen, die dieses "ConstraintTemplate" verwenden, werden unabhängig vom "enforcementAction"-Wert nur geprüft. | Ja |
| K8sExternalIPs | Beschränkt externe IP-Adressen auf eine zulässige Liste von IP-Adressen. https://kubernetes.io/docs/concepts/services-networking/service/#external-ips | Nein |
| K8sHorizontalPodAutoscaler | Die folgenden Szenarien sollen bei der Bereitstellung von "HorizontalPodAutoscalers" 1 nicht zugelassen werden. Bereitstellung von HorizontalPodAutoscalern mit ".spec.minReplicas" oder ".spec.maxReplicas" außerhalb der in der Einschränkung 2 definierten Bereiche. Bereitstellung von HorizontalPodAutoscalern, bei denen der Unterschied zwischen ".spec.minReplicas" und ".spec.maxReplicas" kleiner als der konfigurierte "minimumReplicaSpread" 3 ist. Bereitstellung von HorizontalPodAutoscalern, die nicht auf ein gültiges "scaleTargetRef" verweisen (z. B. Deployment, ReplicationController, ReplicaSet, StatefulSet). | Ja |
| K8sHttpsOnly | Erfordert Ingress-Ressourcen, um nur HTTPS zu verwenden. Ingress-Ressourcen müssen die Annotation „kubernetes.io/ingress.allow-http“ enthalten, die auf „false“ gesetzt ist. Standardmäßig ist eine gültige TLS-{}-Konfiguration erforderlich. Dies kann optional durch Festlegen des Parameters „tlsOptional“ auf „true“ erfolgen. https://kubernetes.io/docs/concepts/services-networking/ingress/#TLS | Nein |
| K8sImageDigests | Erfordert, dass Container-Images einen Digest enthalten. https://kubernetes.io/docs/concepts/containers/images/ | Nein |
| K8sLocalStorageRequireSafeToEvict | Erfordert Pods, die lokalen Speicher („emptyDir“ oder „hostPath“) mit der Annotation „cluster-autoscaler.kubernetes.io/safe-to-evict“: „true“ verwenden. Cluster Autoscaler löscht Pods ohne diese Annotation nicht. | Nein |
| K8sMemoryRequestEqualsLimit | Erhöht die Pod-Stabilität. Dafür muss der angeforderte Speicher aller Container genau dem Speicherlimit entsprechen, sodass sich Pods nie in einem Zustand befinden, in dem die Speichernutzung den angeforderten Umfang überschreitet. Andernfalls kann Kubernetes Pods beenden, die zusätzlichen Speicher anfordern, wenn der Speicher auf dem Knoten benötigt wird. | Nein |
| K8sNoEnvVarSecrets | Verbindet Secrets als Umgebungsvariablen in Pod-Containerdefinitionen. Verwenden Sie stattdessen bereitgestellte Secret-Dateien in Datenvolumen: https://kubernetes.io/docs/concepts/configuration/secret/#using-secrets-as-files-from-a-pod | Nein |
| K8sNoExternalServices | Verhindert das Erstellen bekannter Ressourcen, die Arbeitslasten für externe IP-Adressen verfügbar machen. Dazu gehören Istio-Gateway-Ressourcen und Kubernetes Ingress-Ressourcen. Kubernetes-Dienste sind ebenfalls nur zulässig, wenn sie die folgenden Kriterien erfüllen: Jeder Dienst vom Typ "LoadBalancer" in Google Cloud muss die Annotation "networking.gke.io/load-balancer-type": "Intern" haben. Jeder Dienst vom Typ „LoadBalancer“ in AWS muss die Annotation „service.beta.kubernetes.io/aws-load-balancer-internal“ haben. Alle "externen IP-Adressen" (außerhalb des Clusters), die an den Dienst gebunden sind, müssen Mitglied eines Bereichs internen CIDRs sein, wie für die Einschränkung angegeben. | Nein |
| K8sPSPAllowPrivilegeEscalationContainer | Steuert die Einschränkung der Eskalation auf Root-Berechtigungen. Entspricht dem Feld in einer PodSecurityPolicy. Weitere Informationen finden Sie unter https://kubernetes.io/docs/concepts/policy/pod-security-policy/#privilege-escalation | Nein |
| K8sPSPAllowedUsers | Steuert die Nutzer- und Gruppen-IDs des Containers und einige Volumes. Entspricht den Feldern "runAsUser", "runAsGroup"," sichernalGroups" und "fsGroup" in einer PodSecurityPolicy. Weitere Informationen finden Sie unter https://kubernetes.io/docs/concepts/policy/pod-security-policy/#users-and-groups | Nein |
| K8sPSPAppArmor | Konfiguriert eine Zulassungsliste von AppArmor-Profilen für die Verwendung durch Container. Das entspricht bestimmten Annotationen, die auf eine PodSecurityPolicy angewendet werden. Informationen zu AppArmor finden Sie unter https://kubernetes.io/docs/tutorials/clusters/apparmor/ | Nein |
| K8sPSPAutomountServiceAccountTokenPod | Steuert die Fähigkeit jedes Pods, das automountServiceAccountToken zu aktivieren. | Nein |
| K8sPSPCapabilities | Steuert Linux-Funktionen für Container. Entspricht den Feldern "allowedCapabilities" und "requiredDropCapabilities" in einer PodSecurityPolicy. Weitere Informationen finden Sie unter https://kubernetes.io/docs/concepts/policy/pod-security-policy/#functions | Nein |
| K8sPSPFSGroup | Steuert die Zuweisung einer FSGroup, die die Volumes des Pods besitzt. Entspricht dem Feld "fsGroup" in einer PodSecurityPolicy. Weitere Informationen finden Sie unter https://kubernetes.io/docs/concepts/policy/pod-security-policy/#volumes-and-file-systems | Nein |
| K8sPSPFlexVolumes | Steuert die Zulassungsliste der FlexVolume-Treiber. Entspricht dem Feld "allowedFlexVolumes" in PodSecurityPolicy. Weitere Informationen finden Sie unter https://kubernetes.io/docs/concepts/policy/pod-security-policy/#flexvolume-drivers. | Nein |
| K8sPSPForbiddenSysctls | Steuert das von Containern verwendete "sysctl"-Profil. Entspricht den Feldern "allowedUnsafeSysctls" und "ForbiddenSysctls" in einer PodSecurityPolicy. Wenn angegeben, wird jeder sysctl, der nicht im Parameter „allowedSysctls“ enthalten ist, als unzulässig betrachtet. Der Parameter "forbiddenSysctls" hat Vorrang vor dem Parameter "allowedSysctls". Weitere Informationen finden Sie unter https://kubernetes.io/docs/tasks/administer-cluster/sysctl-cluster/ | Nein |
| K8sPSPHostFilesystem | Steuert die Nutzung des Hostdateisystems. Entspricht dem Feld in einer "allowedHostPaths". Weitere Informationen finden Sie unter https://kubernetes.io/docs/concepts/policy/pod-security-policy/#volumes-and-file-systems | Nein |
| K8sPSPHostNamespace | Verhindert die Freigabe von Host-PID- und IPC-Namespaces nach Pod-Containern. Entspricht den Feldern "hostPID" und "hostIPC" in einer PodSecurityPolicy. Weitere Informationen finden Sie unter https://kubernetes.io/docs/concepts/policy/pod-security-policy/#host-namespaces | Nein |
| K8sPSPHostNetworkingPorts | Steuert die Nutzung des Hostnetzwerk-Namespace nach Pod-Containern. Bestimmte Ports müssen angegeben werden. Entspricht den Feldern "hostNetwork" und "hostPorts" in einer PodSecurityPolicy. Weitere Informationen finden Sie unter https://kubernetes.io/docs/concepts/policy/pod-security-policy/#host-namespaces | Nein |
| K8sPSPPrivilegedContainer | Steuert die Fähigkeit eines Containers, den privilegierten Modus zu aktivieren. Entspricht dem Feld "privileged" in einer PodSecurityPolicy. Weitere Informationen finden Sie unter https://kubernetes.io/docs/concepts/policy/pod-security-policy/#privileged. | Nein |
| K8sPSPProcMount | Steuert die zulässigen "procMount"-Typen für den Container. Entspricht dem Feld allowedProcMountTypes" in einer PodSecurityPolicy. Weitere Informationen finden Sie unter https://kubernetes.io/docs/concepts/policy/pod-security-policy/#allowprocmounttypes. | Nein |
| K8sPSPReadOnlyRootFilesystem | Erfordert die Verwendung eines schreibgeschützten Root-Dateisystems durch Pod-Container. Entspricht dem Feld `readOnlyRootFilesystem` in einer PodSecurityPolicy. Weitere Informationen finden Sie unter https://kubernetes.io/docs/concepts/policy/pod-security-policy/#volumes-and-file-systems | Nein |
| K8sPSPSELinuxV2 | Definiert eine Zulassungsliste von seLinuxOptions-Konfigurationen für Pod-Container. Entspricht einer PodSecurityPolicy, die SELinux-Konfigurationen erfordert. Weitere Informationen finden Sie unter https://kubernetes.io/docs/concepts/policy/pod-security-policy/#selinux | Nein |
| K8sPSPSeccomp | Steuert das von Containern verwendete seccomp-Profil. Entspricht der Annotation „seccomp.security.alpha.kubernetes.io/allowedProfileNames“ in einer PodSecurityPolicy. Weitere Informationen finden Sie unter https://kubernetes.io/docs/concepts/policy/pod-security-policy/#seccomp | Nein |
| K8sPSPVolumeTypes | Beschränkt die bereitstellbaren Volume-Typen auf die vom Nutzer angegebenen Volumes. Entspricht dem Feld `volumes` in einer PodSecurityPolicy. Weitere Informationen finden Sie unter https://kubernetes.io/docs/concepts/policy/pod-security-policy/#volumes-and-file-systems | Nein |
| K8sPSPWindowsHostProcess | Schränkt die Ausführung von Windows HostProcess-Containern/-Pods ein. Weitere Informationen finden Sie unter https://kubernetes.io/docs/tasks/configure-pod-container/create-hostprocess-pod/. | Nein |
| K8sPSSRunAsNonRoot | Erfordert, dass Container als Nutzer ohne Root-Berechtigung ausgeführt werden. Weitere Informationen finden Sie unter https://kubernetes.io/docs/concepts/security/pod-security-standards/ | Nein |
| K8sPodDisruptionBudget | Die folgenden Szenarien sollen bei der Bereitstellung von PodDisruptionBudgets oder Ressourcen, die die Replikatunterressource implementieren (z. B. Deployment, ReplicationController, ReplicaSet, StatefulSet), nicht zugelassen werden: 1. Bereitstellung von PodDisruptionBudgets mit .spec.maxUnavailable == 0. 2. Bereitstellung von PodDisruptionBudgets mit .spec.minAvailable == .spec.replicas der Ressource mit Replikatunterressource. Dies verhindert, dass PodDisruptionBudgets freiwillige Störungen wie das Leeren von Knoten blockieren. https://kubernetes.io/docs/concepts/workloads/Pods/Unterbrechungen/ | Ja |
| K8sPodResourcesBestPractices | Erfordert, dass Container nicht auf Best-Effort-Basis (durch Festlegen von CPU- und Arbeitsspeicheranforderungen) und gemäß den Best Practices für Bursts funktionieren (die Arbeitsspeicherspeicheranfrage muss genau dem Limit entsprechen). Optional können Annotationsschlüssel konfiguriert werden, um das Überspringen der verschiedenen Validierungen zu ermöglichen. | Nein |
| K8sPodsRequireSecurityContext | Erfordert, dass alle Pods securityContext definieren. Erfordert, dass alle in Pods definierten Container einen SecurityContext auf Pod- oder Containerebene definieren. | Nein |
| K8sProhibitRoleWildcardAccess | Erfordert, dass Roles und ClusterRoles den Ressourcenzugriff nicht auf einen Platzhalterwert ("*") festlegen. Dies gilt nicht für die ausgenommenen Roles und ClusterRoles, die als Ausnahmen angegeben werden. Beschränkt den Platzhalterzugriff auf untergeordnete Ressourcen wie „*/status“. | Nein |
| K8sReplicaLimits | Erfordert, dass Objekte mit dem Feld "spec.replicas" (Deployments, ReplicaSets usw.) eine Anzahl von Replikaten innerhalb definierter Bereiche angeben. | Nein |
| K8sRequireAdmissionController | Erfordert entweder "Pod Security Admission" oder ein externes Richtlinienkontrollsystem | Ja |
| K8sRequireBinAuthZ | Erfordert validierenden Zulassungs-Webhook der Binärautorisierung. Einschränkungen, die dieses "ConstraintTemplate" verwenden, werden unabhängig vom "enforcementAction"-Wert nur geprüft. | Ja |
| K8sRequireCosNodeImage | Erzwingt die Verwendung von Container-Optimized OS von Google auf Knoten. | Nein |
| K8sRequireDaemonsets | Erfordert die vorhandene Liste der Daemonsets. | Ja |
| K8sRequireDefaultDenyEgressPolicy | Erfordert, dass jeder im Cluster definierte Namespace eine standardmäßige NetworkPolicy für ausgehenden Traffic hat. | Ja |
| K8sRequireNamespaceNetworkPolicies | Erfordert, dass jeder im Cluster definierte Namespace eine NetworkPolicy hat. | Ja |
| K8sRequireValidRangesForNetworks | Erzwingt, welche CIDR-Blöcke für eingehenden und ausgehenden Netzwerktraffic zulässig sind. | Nein |
| K8sRequiredAnnotations | Erfordert Ressourcen, die bestimmte Annotationen enthalten, wobei Werte mit den regulären Ausdrücken übereinstimmen. | Nein |
| K8sRequiredLabels | Erfordert, dass Ressourcen bestimmte Labels enthalten, wobei Werte mit den regulären Ausdrücken übereinstimmen. | Nein |
| K8sRequiredProbes | Erfordert, dass Pods über Bereitschafts- und/oder Aktivitätsprüfungen verfügen. | Nein |
| K8sRequiredResources | Erfordert, dass Container definierte Ressourcen festlegen. https://kubernetes.io/docs/concepts/configuration/manage-resources-containers/ | Nein |
| K8sRestrictAdmissionController | Dynamische Zulassungs-Controller auf zulässige Zulassungs-Controller beschränken | Nein |
| K8sRestrictAutomountServiceAccountTokens | Beschränkt die Verwendung von Dienstkonto-Tokens. | Nein |
| K8sRestrictLabels | Verhindert, dass Ressourcen bestimmte Labels enthalten, es sei denn, es gibt eine Ausnahme für die jeweilige Ressource. | Nein |
| K8sRestrictNamespaces | Schränkt die Verwendung von Namespaces durch Ressourcen ein, die unter dem Parameter restrictedNamespaces aufgeführt sind. | Nein |
| K8sRestrictNfsUrls | Verhindert, dass Ressourcen NFS-URLs enthalten, sofern nicht anders angegeben. | Nein |
| K8sRestrictRbacSubjects | Schränkt die Verwendung von Namen in RBAC auf zulässige Werte ein. | Nein |
| K8sRestrictRoleBindings | Beschränkt die in ClusterRoleBindings und RoleBindings angegebenen Subjekte auf eine Liste zulässiger Subjekte. | Nein |
| K8sRestrictRoleRules | Schränkt Regeln ein, die für Role- und ClusterRole-Objekte festgelegt werden können. | Nein |
| K8sStorageClass | Erfordert die Angabe von Speicherklassen, wenn verwendet. Nur Gatekeeper 3.9+ und nicht sitzungsspezifische Container werden unterstützt. | Ja |
| K8sUniqueIngressHost | Alle Ingress-Regelhosts müssen eindeutig sein. Unterstützt keine Hostnamen-Platzhalter: https://kubernetes.io/docs/concepts/services-networking/ingress/ | Ja |
| K8sUniqueServiceSelector | Dienste müssen einen eindeutigen Selektor innerhalb eines Namespace haben. Selektoren werden als identisch betrachtet, wenn sie identische Schlüssel und Werte haben. Selektoren können ein Schlüssel/Wert-Paar verwenden, solange sich mindestens ein separates Schlüssel/Wert-Paar zwischen ihnen befindet. https://kubernetes.io/docs/concepts/services-networking/service/#defining-a-service | Ja |
| NoUpdateServiceAccount | Blockiert das Aktualisieren des Dienstkontos für Ressourcen, die über Pods abstrahiert werden. Diese Richtlinie wird im Prüfmodus ignoriert. | Nein |
| PolicyStrictOnly | Erfordert, dass die gegenseitige TLS-Authentifizierung von IstioT immer angegeben wird, wenn [PeerAuthentication](https://istio.io/latest/docs/reference/config/security/peer_authentication/) verwendet wird. Mit dieser Einschränkung wird auch sichergestellt, dass die verworfene [Richtlinie](https://istio.io/v1.4/docs/reference/config/security/istio.authentication.v1alpha1/#Policy) und die MeshPolicy-Ressourcen gegenseitiges TLS des Typs „STRICT“ erzwingen. Weitere Informationen finden Sie unter https://istio.io/latest/docs/tasks/security/authentication/mtls-migration/#lock-down-mutual-tls-for-the-entire-mesh | Nein |
| RestrictNetworkExclusions | Steuert, welche Eingangs-Ports, Ausgangs-Ports und Ausgangs-IP-Bereiche von der Istio-Netzwerkerfassung ausgeschlossen werden können. Ports und IP-Bereiche, die die Istio-Netzwerkerfassung umgehen, werden vom Istio-Proxy nicht verarbeitet und unterliegen nicht der Authentifizierung, der Autorisierungsrichtlinie und anderen Istio-Features von Istio mTLS. Mithilfe dieser Einschränkung können Einschränkungen auf die Verwendung der folgenden Annotationen angewendet werden:
Weitere Informationen finden Sie unter https://istio.io/latest/docs/reference/config/annotations/. Beim Einschränken von Ausgangs-IP-Bereichen berechnet die Einschränkung, ob ausgeschlossene IP-Bereiche übereinstimmen oder ein Teil der zulässigen IP-Bereichsausschlüsse sind. Wenn Sie diese Einschränkung verwenden, müssen alle Eingangs- und Ausgangs-IP-Bereiche immer einbezogen werden. Setzen Sie dazu entweder die entsprechenden "include"-Annotationen auf "*" oder legen Sie sie nicht fest. Das Festlegen einer der folgenden Anmerkungen auf etwas anderes als"*" ist nicht zulässig:
Mit dieser Einschränkung kann Port 15020 immer ausgeschlossen werden, da der Istio-Sidecar-Injektor immer die Annotation |
Nein |
| SourceNotAllAuthz | Erfordert, dass die AuthorizationPolicy-Regeln von Istio Quellhauptkonten auf etwas anderes als "*" gesetzt haben. https://istio.io/latest/docs/reference/config/security/authorization-policy/ | Nein |
| VerifyDeprecatedAPI | Verifiziert verworfene Kubernetes APIs, um sicherzustellen, dass alle API-Versionen aktuell sind. Diese Vorlage gilt nicht für Audits, da Audits die Ressourcen betrachten, die bereits im Cluster mit nicht verworfenen API-Versionen vorhanden sind. | Nein |
AllowedServicePortName
Allowed Service Port Names v1.0.1
Erfordert, dass Dienstportnamen ein Präfix aus einer angegebenen Liste haben.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: AllowedServicePortName
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# prefixes <array>: Prefixes of allowed service port names.
prefixes:
- <string>
Beispiele
port-name-constraint
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: AllowedServicePortName metadata: name: port-name-constraint spec: enforcementAction: deny match: kinds: - apiGroups: - "" kinds: - Service parameters: prefixes: - http- - http2- - grpc- - mongo- - redis- - tcp-
Zulässig
apiVersion: v1 kind: Service metadata: labels: app: helloworld name: port-name-http spec: ports: - name: http-helloport port: 5000 selector: app: helloworld
Nicht zugelassen
apiVersion: v1 kind: Service metadata: labels: app: helloworld name: port-name-tcp spec: ports: - name: foo-helloport port: 5000 selector: app: helloworld
apiVersion: v1 kind: Service metadata: labels: app: helloworld name: port-name-bad spec: ports: - name: helloport port: 5000 selector: app: helloworld
AsmAuthzPolicyDefaultDeny
ASM AuthorizationPolicy Default Deny v1.0.4
Erzwingen Sie die AuthorizationPolicy für standardmäßige Ablehnung auf Mesh-Ebene. Lesen Sie dazu https://istio.io/latest/docs/ops/best-practices/security/#use-default-deny-patterns.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: AsmAuthzPolicyDefaultDeny
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# rootNamespace <string>: Anthos Service Mesh root namespace, default value
# is "istio-system" if not specified.
rootNamespace: <string>
# strictnessLevel <string>: Level of AuthorizationPolicy strictness.
# Allowed Values: Low, High
strictnessLevel: <string>
Referenzielle Einschränkung
Diese Einschränkung ist referenziell. Vor der Verwendung müssen Sie referenzielle Einschränkungen aktivieren und eine Konfiguration erstellen, die Policy Controller mitteilt, welche Arten von Objekten beobachtet werden sollen.
Ihr Policy Controller Config erfordert einen syncOnly-Eintrag ähnlich dem folgenden:
spec:
sync:
syncOnly:
- group: "security.istio.io"
version: "v1beta1"
kind: "AuthorizationPolicy"
Beispiele
asm-authz-policy-default-deny-with-input-constraint
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: AsmAuthzPolicyDefaultDeny metadata: name: asm-authz-policy-default-deny-with-input-constraint spec: enforcementAction: dryrun parameters: rootNamespace: istio-system strictnessLevel: High
Zulässig
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: AsmAuthzPolicyDefaultDeny metadata: name: asm-authz-policy-default-deny-with-input-constraint spec: enforcementAction: dryrun parameters: rootNamespace: istio-system strictnessLevel: High --- # Referential Data apiVersion: security.istio.io/v1beta1 kind: AuthorizationPolicy metadata: name: default-deny-no-action namespace: istio-system spec: null
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: AsmAuthzPolicyDefaultDeny metadata: name: asm-authz-policy-default-deny-with-input-constraint spec: enforcementAction: dryrun parameters: rootNamespace: istio-system strictnessLevel: High --- # Referential Data apiVersion: security.istio.io/v1beta1 kind: AuthorizationPolicy metadata: name: default-deny-with-action namespace: istio-system spec: action: ALLOW
Nicht zugelassen
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: AsmAuthzPolicyDefaultDeny metadata: name: asm-authz-policy-default-deny-with-input-constraint spec: enforcementAction: dryrun parameters: rootNamespace: istio-system strictnessLevel: High --- # Referential Data apiVersion: security.istio.io/v1beta1 kind: AuthorizationPolicy metadata: name: not-default-deny namespace: istio-system spec: action: DENY rules: - to: - operation: notMethods: - GET - POST
asm-authz-policy-default-deny-no-input-constraint
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: AsmAuthzPolicyDefaultDeny metadata: name: asm-authz-policy-default-deny-no-input-constraint spec: enforcementAction: dryrun parameters: strictnessLevel: High
Zulässig
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: AsmAuthzPolicyDefaultDeny metadata: name: asm-authz-policy-default-deny-no-input-constraint spec: enforcementAction: dryrun parameters: strictnessLevel: High --- # Referential Data apiVersion: security.istio.io/v1beta1 kind: AuthorizationPolicy metadata: name: default-deny-no-action namespace: istio-system spec: null
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: AsmAuthzPolicyDefaultDeny metadata: name: asm-authz-policy-default-deny-no-input-constraint spec: enforcementAction: dryrun parameters: strictnessLevel: High --- # Referential Data apiVersion: security.istio.io/v1beta1 kind: AuthorizationPolicy metadata: name: default-deny-with-action namespace: istio-system spec: action: ALLOW
Nicht zugelassen
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: AsmAuthzPolicyDefaultDeny metadata: name: asm-authz-policy-default-deny-no-input-constraint spec: enforcementAction: dryrun parameters: strictnessLevel: High --- # Referential Data apiVersion: security.istio.io/v1beta1 kind: AuthorizationPolicy metadata: name: not-default-deny namespace: istio-system spec: action: DENY rules: - to: - operation: notMethods: - GET - POST
AsmAuthzPolicyDisallowedPrefix
ASM AuthorizationPolicy Disallowed Prefixes v1.0.2
Erfordert, dass Hauptkonten und Namespaces in AuthorizationPolicy-Regeln von Istio kein Präfix aus einer angegebenen Liste haben.
https://istio.io/latest/docs/reference/config/security/authorization-policy/
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: AsmAuthzPolicyDisallowedPrefix
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# disallowedNamespacePrefixes <array>: Disallowed prefixes for namespaces.
disallowedNamespacePrefixes:
- <string>
# disallowedPrincipalPrefixes <array>: Disallowed prefixes for principals.
disallowedPrincipalPrefixes:
- <string>
Beispiele
asm-authz-policy-disallowed-prefix-constraint
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: AsmAuthzPolicyDisallowedPrefix metadata: name: asm-authz-policy-disallowed-prefix-constraint spec: enforcementAction: dryrun match: kinds: - apiGroups: - security.istio.io kinds: - AuthorizationPolicy parameters: disallowedNamespacePrefixes: - bad-ns-prefix - worse-ns-prefix disallowedPrincipalPrefixes: - bad-principal-prefix - worse-principal-prefix
Zulässig
apiVersion: security.istio.io/v1beta1 kind: AuthorizationPolicy metadata: name: valid-authz-policy spec: rules: - from: - source: principals: - cluster.local/ns/default/sa/sleep - source: namespaces: - test selector: matchLabels: app: httpbin
Nicht zugelassen
apiVersion: security.istio.io/v1beta1 kind: AuthorizationPolicy metadata: name: bad-source-principal spec: rules: - from: - source: principals: - cluster.local/ns/default/sa/worse-principal-prefix-sleep - source: namespaces: - test selector: matchLabels: app: httpbin
apiVersion: security.istio.io/v1beta1 kind: AuthorizationPolicy metadata: name: bad-source-namespace spec: rules: - from: - source: principals: - cluster.local/ns/default/sa/sleep - source: namespaces: - bad-ns-prefix-test selector: matchLabels: app: httpbin
AsmAuthzPolicyEnforceSourcePrincipals
ASM AuthorizationPolicy Enforcement Principals v1.0.2
Erfordert, dass das Istio AuthorizationPolicy-Feld "from", wenn definiert, Quellprinzipien hat, die auf etwas anderes als "*" gesetzt sein müssen. https://istio.io/latest/docs/reference/config/security/authorization-policy/
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: AsmAuthzPolicyEnforceSourcePrincipals
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
Beispiele
asm-authz-policy-enforce-source-principals-constraint
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: AsmAuthzPolicyEnforceSourcePrincipals metadata: name: asm-authz-policy-enforce-source-principals-constraint spec: enforcementAction: dryrun match: kinds: - apiGroups: - security.istio.io kinds: - AuthorizationPolicy
Zulässig
apiVersion: security.istio.io/v1beta1 kind: AuthorizationPolicy metadata: name: valid-authz-policy spec: rules: - from: - source: principals: - cluster.local/ns/default/sa/sleep - source: namespaces: - test to: - operation: methods: - GET paths: - /info* - operation: methods: - POST paths: - /data when: - key: request.auth.claims[iss] values: - https://accounts.google.com selector: matchLabels: app: httpbin
Nicht zugelassen
apiVersion: security.istio.io/v1beta1 kind: AuthorizationPolicy metadata: name: no-source-principals spec: rules: - from: - source: namespaces: - test to: - operation: methods: - GET paths: - /info* - operation: methods: - POST paths: - /data when: - key: request.auth.claims[iss] values: - https://accounts.google.com selector: matchLabels: app: httpbin
apiVersion: security.istio.io/v1beta1 kind: AuthorizationPolicy metadata: name: source-principals-wildcard spec: rules: - from: - source: principals: - '*' - source: namespaces: - test to: - operation: methods: - GET paths: - /info* - operation: methods: - POST paths: - /data when: - key: request.auth.claims[iss] values: - https://accounts.google.com selector: matchLabels: app: httpbin
apiVersion: security.istio.io/v1beta1 kind: AuthorizationPolicy metadata: name: source-principals-contains-wildcard spec: rules: - from: - source: principals: - cluster.local/ns/default/sa/sleep - '*' - source: namespaces: - test to: - operation: methods: - GET paths: - /info* - operation: methods: - POST paths: - /data when: - key: request.auth.claims[iss] values: - https://accounts.google.com selector: matchLabels: app: httpbin
AsmAuthzPolicyNormalization
ASM AuthorizationPolicy Normalization v1.0.2
Erzwingen Sie die Normalisierung der AuthorizationPolicy. Lesen Sie dazu https://istio.io/latest/docs/reference/config/security/normalization/.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: AsmAuthzPolicyNormalization
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
Beispiele
asm-authz-policy-normalization-sample
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: AsmAuthzPolicyNormalization metadata: name: asm-authz-policy-normalization-sample spec: enforcementAction: dryrun match: kinds: - apiGroups: - security.istio.io kinds: - AuthorizationPolicy
Zulässig
apiVersion: security.istio.io/v1beta1 kind: AuthorizationPolicy metadata: name: good-authz-policy spec: action: ALLOW rules: - to: - operation: methods: - GET paths: - /test/foo - when: - key: source.ip values: - 10.1.2.3 - 10.2.0.0/16 - key: request.headers[User-Agent] values: - Mozilla/* selector: matchLabels: app: httpbin
Nicht zugelassen
apiVersion: security.istio.io/v1beta1 kind: AuthorizationPolicy metadata: name: bad-method-lowercase spec: action: ALLOW rules: - to: - operation: methods: - get selector: matchLabels: app: httpbin
apiVersion: security.istio.io/v1beta1 kind: AuthorizationPolicy metadata: name: bad-request-header-whitespace spec: action: ALLOW rules: - to: - operation: methods: - GET - when: - key: source.ip values: - 10.1.2.3 - 10.2.0.0/16 - key: request.headers[User-Ag ent] values: - Mozilla/* selector: matchLabels: app: httpbin
apiVersion: security.istio.io/v1beta1 kind: AuthorizationPolicy metadata: name: path-unnormalized spec: action: ALLOW rules: - to: - operation: methods: - GET paths: - /test\/foo - when: - key: source.ip values: - 10.1.2.3 - 10.2.0.0/16 - key: request.headers[User-Agent] values: - Mozilla/* selector: matchLabels: app: httpbin
AsmAuthzPolicySafePattern
ASM AuthorizationPolicy Safe Patterns v1.0.4
Erzwingen Sie die sicheren Muster der AuthorizationPolicy. Lesen Sie dazu https://istio.io/latest/docs/ops/best-practices/security/#safer-authorization-policy-patterns.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: AsmAuthzPolicySafePattern
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# strictnessLevel <string>: Level of AuthorizationPolicy strictness.
# Allowed Values: Low, High
strictnessLevel: <string>
Beispiele
asm-authz-policy-safe-pattern-sample
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: AsmAuthzPolicySafePattern metadata: name: asm-authz-policy-safe-pattern-sample spec: enforcementAction: dryrun match: kinds: - apiGroups: - security.istio.io kinds: - AuthorizationPolicy parameters: strictnessLevel: High
Zulässig
apiVersion: security.istio.io/v1beta1 kind: AuthorizationPolicy metadata: name: good-authz-policy-istio-ingress spec: action: ALLOW rules: - to: - operation: hosts: - test.com - test.com:* methods: - GET selector: matchLabels: istio: ingressgateway
apiVersion: security.istio.io/v1beta1 kind: AuthorizationPolicy metadata: name: good-authz-policy-asm-ingress spec: action: ALLOW rules: - to: - operation: hosts: - test.com - test.com:* methods: - GET selector: matchLabels: asm: ingressgateway
Nicht zugelassen
apiVersion: security.istio.io/v1beta1 kind: AuthorizationPolicy metadata: name: hosts-on-noningress spec: action: ALLOW rules: - to: - operation: hosts: - test.com - test.com:* methods: - GET
apiVersion: security.istio.io/v1beta1 kind: AuthorizationPolicy metadata: name: invalid-hosts spec: action: ALLOW rules: - to: - operation: hosts: - test.com methods: - GET selector: matchLabels: istio: ingressgateway
apiVersion: security.istio.io/v1beta1 kind: AuthorizationPolicy metadata: name: allow-negative-match spec: action: ALLOW rules: - to: - operation: hosts: - test.com - test.com:* notMethods: - GET selector: matchLabels: istio: ingressgateway
apiVersion: security.istio.io/v1beta1 kind: AuthorizationPolicy metadata: name: deny-positive-match spec: action: DENY rules: - to: - operation: hosts: - test.com - test.com:* methods: - GET selector: matchLabels: istio: ingressgateway
AsmIngressgatewayLabel
ASM Ingress Gateway Label v1.0.3
Erzwingen Sie die Labelverwendung von Istio-Ingressgateways nur für Ingressgateway-Pods.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: AsmIngressgatewayLabel
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
Beispiele
asm-ingressgateway-label-sample
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: AsmIngressgatewayLabel metadata: name: asm-ingressgateway-label-sample spec: enforcementAction: dryrun match: kinds: - apiGroups: - "" kinds: - Pod
Zulässig
apiVersion: v1 kind: Pod metadata: labels: app: sleep istio: istio name: sleep spec: containers: - image: curlimages/curl name: sleep - image: gcr.io/gke-release/asm/proxyv2:release name: istio-proxy ports: - containerPort: 15090 name: http-envoy-prom protocol: TCP
apiVersion: v1 kind: Pod metadata: labels: app: istio-ingressgateway istio: ingressgateway name: istio-ingressgateway spec: containers: - image: gcr.io/gke-release/asm/proxyv2:release name: istio-proxy ports: - containerPort: 15090 name: http-envoy-prom protocol: TCP
apiVersion: v1 kind: Pod metadata: labels: app: asm-ingressgateway asm: ingressgateway name: asm-ingressgateway spec: containers: - image: gcr.io/gke-release/asm/proxyv2:release name: istio-proxy ports: - containerPort: 15090 name: http-envoy-prom protocol: TCP
Nicht zugelassen
apiVersion: v1 kind: Pod metadata: labels: app: sleep istio: ingressgateway name: sleep spec: containers: - image: curlimages/curl name: sleep
apiVersion: v1 kind: Pod metadata: labels: app: sleep asm: ingressgateway name: sleep spec: containers: - image: curlimages/curl name: sleep
apiVersion: v1 kind: Pod metadata: labels: app: sleep istio: ingressgateway name: sleep spec: containers: - image: curlimages/curl name: sleep - image: gcr.io/gke-release/asm/proxyv2:release name: istio-proxy ports: - containerPort: 15090 name: http-envoy-prom protocol: TCP
AsmPeerAuthnMeshStrictMtls
ASM Peer Authentication Mesh Strict mTLS v1.0.4
Erzwingen Sie die strikte mTLS PeerAuthentication auf Mesh-Ebene. Lesen Sie dazu https://istio.io/latest/docs/ops/best-practices/security/#mutual-tls.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: AsmPeerAuthnMeshStrictMtls
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# rootNamespace <string>: Anthos Service Mesh root namespace, default value
# is "istio-system" if not specified.
rootNamespace: <string>
# strictnessLevel <string>: Level of PeerAuthentication strictness.
# Allowed Values: Low, High
strictnessLevel: <string>
Referenzielle Einschränkung
Diese Einschränkung ist referenziell. Vor der Verwendung müssen Sie referenzielle Einschränkungen aktivieren und eine Konfiguration erstellen, die Policy Controller mitteilt, welche Arten von Objekten beobachtet werden sollen.
Ihr Policy Controller Config erfordert einen syncOnly-Eintrag ähnlich dem folgenden:
spec:
sync:
syncOnly:
- group: "security.istio.io"
version: "v1beta1"
kind: "PeerAuthentication"
Beispiele
asm-peer-authn-mesh-strict-mtls-with-input-constraint
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: AsmPeerAuthnMeshStrictMtls metadata: name: asm-peer-authn-mesh-strict-mtls-with-input-constraint spec: enforcementAction: dryrun parameters: rootNamespace: asm-root strictnessLevel: High
Zulässig
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: AsmPeerAuthnMeshStrictMtls metadata: name: asm-peer-authn-mesh-strict-mtls-with-input-constraint spec: enforcementAction: dryrun parameters: rootNamespace: asm-root strictnessLevel: High --- # Referential Data apiVersion: security.istio.io/v1beta1 kind: PeerAuthentication metadata: name: mesh-strict-mtls namespace: asm-root spec: mtls: mode: STRICT
Nicht zugelassen
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: AsmPeerAuthnMeshStrictMtls metadata: name: asm-peer-authn-mesh-strict-mtls-with-input-constraint spec: enforcementAction: dryrun parameters: rootNamespace: asm-root strictnessLevel: High --- # Referential Data apiVersion: security.istio.io/v1beta1 kind: PeerAuthentication metadata: name: mesh-permissive-mtls namespace: asm-root spec: mtls: mode: PERMISSIVE
asm-peer-authn-mesh-strict-mtls-no-input-constraint
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: AsmPeerAuthnMeshStrictMtls metadata: name: asm-peer-authn-mesh-strict-mtls-no-input-constraint spec: enforcementAction: dryrun parameters: strictnessLevel: High
Zulässig
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: AsmPeerAuthnMeshStrictMtls metadata: name: asm-peer-authn-mesh-strict-mtls-no-input-constraint spec: enforcementAction: dryrun parameters: strictnessLevel: High --- # Referential Data apiVersion: security.istio.io/v1beta1 kind: PeerAuthentication metadata: name: mesh-strict-mtls namespace: istio-system spec: mtls: mode: STRICT
Nicht zugelassen
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: AsmPeerAuthnMeshStrictMtls metadata: name: asm-peer-authn-mesh-strict-mtls-no-input-constraint spec: enforcementAction: dryrun parameters: strictnessLevel: High --- # Referential Data apiVersion: security.istio.io/v1beta1 kind: PeerAuthentication metadata: name: mesh-permissive-mtls namespace: istio-system spec: mtls: mode: PERMISSIVE
AsmPeerAuthnStrictMtls
ASM Peer Authentication Strict mTLS v1.0.3
Erzwingen Sie, dass alle PeerAuthentications keine strikten mTLS überschreiben können. Lesen Sie dazu https://istio.io/latest/docs/ops/best-practices/security/#mutual-tls.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: AsmPeerAuthnStrictMtls
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# strictnessLevel <string>: Level of PeerAuthentication strictness.
# Allowed Values: Low, High
strictnessLevel: <string>
Beispiele
asm-peer-authn-strict-mtls-constraint
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: AsmPeerAuthnStrictMtls metadata: name: asm-peer-authn-strict-mtls-constraint spec: enforcementAction: dryrun match: kinds: - apiGroups: - security.istio.io kinds: - PeerAuthentication parameters: strictnessLevel: High
Zulässig
apiVersion: security.istio.io/v1beta1 kind: PeerAuthentication metadata: name: valid-strict-mtls-pa namespace: foo spec: mtls: mode: UNSET portLevelMtls: "80": mode: UNSET "443": mode: STRICT selector: matchLabels: app: bar
Nicht zugelassen
apiVersion: security.istio.io/v1beta1 kind: PeerAuthentication metadata: name: invalid-permissive-mtls-pa namespace: foo spec: mtls: mode: PERMISSIVE portLevelMtls: "80": mode: UNSET "443": mode: STRICT selector: matchLabels: app: bar
apiVersion: security.istio.io/v1beta1 kind: PeerAuthentication metadata: name: invalid-port-disable-mtls-pa namespace: foo spec: mtls: mode: UNSET portLevelMtls: "80": mode: DISABLE "443": mode: STRICT selector: matchLabels: app: bar
AsmRequestAuthnProhibitedOutputHeaders
ASM RequestAuthentication Prohibited Output Headers v1.0.2
Erzwingen Sie bei RequestAuthentication, dass das Feld jwtRules.outPayloadToHeader keine bekannten HTTP-Anfrageheader oder benutzerdefinierten verbotenen Header enthält. Referenz zu https://istio.io/latest/docs/reference/config/security/jwt/#JWTRule.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: AsmRequestAuthnProhibitedOutputHeaders
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# prohibitedHeaders <array>: User predefined prohibited headers.
prohibitedHeaders:
- <string>
Beispiele
asm-request-authn-prohibited-output-headers-constraint
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: AsmRequestAuthnProhibitedOutputHeaders metadata: name: asm-request-authn-prohibited-output-headers-constraint spec: enforcementAction: dryrun match: kinds: - apiGroups: - security.istio.io kinds: - RequestAuthentication parameters: prohibitedHeaders: - Bad-Header - X-Bad-Header
Zulässig
apiVersion: security.istio.io/v1beta1 kind: RequestAuthentication metadata: name: valid-request-authn namespace: istio-system spec: jwtRules: - issuer: example.com outputPayloadToHeader: Good-Header selector: matchLabels: app: istio-ingressgateway
Nicht zugelassen
apiVersion: security.istio.io/v1beta1 kind: RequestAuthentication metadata: name: deny-predefined-output-header namespace: istio-system spec: jwtRules: - issuer: example.com outputPayloadToHeader: Host selector: matchLabels: app: istio-ingressgateway
apiVersion: security.istio.io/v1beta1 kind: RequestAuthentication metadata: name: deny-predefined-output-header namespace: istio-system spec: jwtRules: - issuer: example.com outputPayloadToHeader: X-Bad-Header selector: matchLabels: app: istio-ingressgateway
AsmSidecarInjection
ASM Sidecar Injection v1.0.2
Erzwingen Sie, dass die istio-Proxy-Sidecar-Datei immer in die Workload-Pods eingefügt wird.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: AsmSidecarInjection
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# strictnessLevel <string>: Level of sidecar injection strictness.
# Allowed Values: Low, High
strictnessLevel: <string>
Beispiele
asm-sidecar-injection-sample
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: AsmSidecarInjection metadata: name: asm-sidecar-injection-sample spec: enforcementAction: dryrun match: kinds: - apiGroups: - "" kinds: - Pod parameters: strictnessLevel: High
Zulässig
apiVersion: v1 kind: Pod metadata: annotations: sidecar.istio.io/inject: "true" name: sleep spec: containers: - image: curlimages/curl name: sleep - image: gcr.io/gke-release/asm/proxyv2:release name: istio-proxy ports: - containerPort: 15090 name: http-envoy-prom protocol: TCP
apiVersion: v1 kind: Pod metadata: annotations: "false": "false" name: sleep spec: containers: - image: curlimages/curl name: sleep - image: gcr.io/gke-release/asm/proxyv2:release name: istio-proxy ports: - containerPort: 15090 name: http-envoy-prom protocol: TCP
Nicht zugelassen
apiVersion: v1 kind: Pod metadata: annotations: sidecar.istio.io/inject: "false" name: sleep spec: containers: - image: curlimages/curl name: sleep
DestinationRuleTLSEnabled
Destination Rule TLS Enabled v1.0.1
Verbietet die Deaktivierung von TLS für alle Hosts und Hostteilmengen in Istio DestinationRules.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: DestinationRuleTLSEnabled
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
Beispiele
dr-tls-enabled
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: DestinationRuleTLSEnabled metadata: name: dr-tls-enabled spec: enforcementAction: dryrun match: kinds: - apiGroups: - networking.istio.io kinds: - DestinationRule
Unzulässig
apiVersion: networking.istio.io/v1alpha3 kind: DestinationRule metadata: name: dr-subset-tls-disable namespace: default spec: host: myservice subsets: - name: v1 trafficPolicy: tls: mode: DISABLE - name: v2 trafficPolicy: tls: mode: SIMPLE
apiVersion: networking.istio.io/v1alpha3 kind: DestinationRule metadata: name: dr-traffic-tls-disable namespace: default spec: host: myservice trafficPolicy: tls: mode: DISABLE
DisallowedAuthzPrefix
Disallow Istio AuthorizationPolicy Prefixes v1.0.2
Erfordert, dass Hauptkonten und Namespaces in AuthorizationPolicy-Regeln von Istio kein Präfix aus einer angegebenen Liste haben.
https://istio.io/latest/docs/reference/config/security/authorization-policy/
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: DisallowedAuthzPrefix
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# disallowedprefixes <array>: Disallowed prefixes of principals and
# namespaces.
disallowedprefixes:
- <string>
Beispiele
disallowed-authz-prefix-constraint
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: DisallowedAuthzPrefix metadata: name: disallowed-authz-prefix-constraint spec: enforcementAction: dryrun match: kinds: - apiGroups: - security.istio.io kinds: - AuthorizationPolicy parameters: disallowedprefixes: - badprefix - reallybadprefix
Zulässig
apiVersion: security.istio.io/v1beta1 kind: AuthorizationPolicy metadata: name: good namespace: foo spec: rules: - from: - source: principals: - cluster.local/ns/default/sa/sleep - source: namespaces: - test to: - operation: methods: - GET paths: - /info* - operation: methods: - POST paths: - /data when: - key: request.auth.claims[iss] values: - https://accounts.google.com selector: matchLabels: app: httpbin version: v1
Nicht zugelassen
apiVersion: security.istio.io/v1beta1 kind: AuthorizationPolicy metadata: name: bad-source-principal namespace: foo spec: rules: - from: - source: principals: - cluster.local/ns/default/sa/badprefix-sleep - source: namespaces: - test to: - operation: methods: - GET paths: - /info* - operation: methods: - POST paths: - /data when: - key: request.auth.claims[iss] values: - https://accounts.google.com selector: matchLabels: app: httpbin version: v1
apiVersion: security.istio.io/v1beta1 kind: AuthorizationPolicy metadata: name: bad-source-namespace namespace: foo spec: rules: - from: - source: principals: - cluster.local/ns/default/sa/sleep - source: namespaces: - badprefix-test to: - operation: methods: - GET paths: - /info* - operation: methods: - POST paths: - /data when: - key: request.auth.claims[iss] values: - https://accounts.google.com selector: matchLabels: app: httpbin version: v1
GCPStorageLocationConstraintV1
GCP Storage Location Constraint v1.0.3
Beschränkt den zulässigen locations für StorageBucket Config Connector-Ressourcen auf die Liste der in der Einschränkung angegebenen Standorte. Bucket-Namen in der exemptions-Liste sind ausgenommen.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: GCPStorageLocationConstraintV1
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# exemptions <array>: A list of bucket names that are exempt from this
# constraint.
exemptions:
- <string>
# locations <array>: A list of locations that a bucket is permitted to
# have.
locations:
- <string>
Beispiele
singapore-and-jakarta-only
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: GCPStorageLocationConstraintV1 metadata: name: singapore-and-jakarta-only spec: enforcementAction: deny match: kinds: - apiGroups: - storage.cnrm.cloud.google.com kinds: - StorageBucket parameters: exemptions: - my_project_id_cloudbuild locations: - asia-southeast1 - asia-southeast2
Zulässig
apiVersion: storage.cnrm.cloud.google.com/v1beta1 kind: StorageBucket metadata: name: bucket-in-permitted-location spec: location: asia-southeast1
Nicht zugelassen
apiVersion: storage.cnrm.cloud.google.com/v1beta1 kind: StorageBucket metadata: name: bucket-in-disallowed-location spec: location: us-central1
apiVersion: storage.cnrm.cloud.google.com/v1beta1 kind: StorageBucket metadata: name: bucket-without-specific-location spec: null
GkeSpotVMTerminationGrace
Restricts terminationGracePeriodSeconds for GKE Spot VMs v1.1.3
Erfordert Pods und Pod-Vorlagen mit nodeSelector oder nodeAfffinty von gke-spot, um einen terminationGracePeriodSeconds von maximal 15 Sekunden zu haben.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: GkeSpotVMTerminationGrace
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# includePodOnSpotNodes <boolean>: Require `terminationGracePeriodSeconds`
# of 15s or less for all `Pod` on a `gke-spot` Node.
includePodOnSpotNodes: <boolean>
Referenzielle Einschränkung
Diese Einschränkung ist referenziell. Vor der Verwendung müssen Sie referenzielle Einschränkungen aktivieren und eine Konfiguration erstellen, die Policy Controller mitteilt, welche Arten von Objekten beobachtet werden sollen.
Ihr Policy Controller Config erfordert einen syncOnly-Eintrag ähnlich dem folgenden:
spec:
sync:
syncOnly:
- group: ""
version: "v1"
kind: "Node"
Beispiele
spotvm-termination-grace
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: GkeSpotVMTerminationGrace metadata: name: spotvm-termination-grace spec: enforcementAction: dryrun match: kinds: - apiGroups: - "" kinds: - Pod parameters: includePodOnSpotNodes: true
Zulässig
apiVersion: v1 kind: Pod metadata: name: example-allowed spec: containers: - image: nginx name: nginx nodeSelector: cloud.google.com/gke-spot: "true" terminationGracePeriodSeconds: 15
apiVersion: v1 kind: Pod metadata: name: example-allowed spec: containers: - image: nginx name: nginx nodeSelector: cloud.google.com/gke-spot: "true" terminationGracePeriodSeconds: 15
apiVersion: v1 kind: Pod metadata: name: example-with-termGrace spec: Nodename: default containers: - image: nginx name: nginx terminationGracePeriodSeconds: 15 --- # Referential Data apiVersion: v1 kind: Node metadata: labels: cloud.google.com/gke-spot: "true" name: default
apiVersion: v1 kind: Pod metadata: name: example-with-termGrace spec: Nodename: default containers: - image: nginx name: nginx terminationGracePeriodSeconds: 15 --- # Referential Data apiVersion: v1 kind: Node metadata: name: default
apiVersion: v1 kind: Pod metadata: name: example-without-termGrace spec: Nodename: default containers: - image: nginx name: nginx --- # Referential Data apiVersion: v1 kind: Node metadata: name: default
Unzulässig
apiVersion: v1 kind: Pod metadata: name: example-disallowed spec: affinity: nodeAffinity: requiredDuringSchedulingIgnoredDuringExecution: nodeSelectorTerms: - matchExpressions: - key: cloud.google.com/gke-spot operator: In values: - "true" containers: - image: nginx name: nginx terminationGracePeriodSeconds: 30
apiVersion: v1 kind: Pod metadata: name: example-disallowed spec: affinity: nodeAffinity: requiredDuringSchedulingIgnoredDuringExecution: nodeSelectorTerms: - matchExpressions: - key: cloud.google.com/gke-spot operator: In values: - "true" containers: - image: nginx name: nginx
apiVersion: v1 kind: Pod metadata: name: example-disallowed spec: containers: - image: nginx name: nginx nodeSelector: cloud.google.com/gke-spot: "true" terminationGracePeriodSeconds: 30
apiVersion: v1 kind: Pod metadata: name: example-disallowed spec: affinity: nodeAffinity: requiredDuringSchedulingIgnoredDuringExecution: nodeSelectorTerms: - matchExpressions: - key: cloud.google.com/gke-spot operator: In values: - "true" containers: - image: nginx name: nginx
apiVersion: v1 kind: Pod metadata: name: example-without-termGrace spec: Nodename: default containers: - image: nginx name: nginx --- # Referential Data apiVersion: v1 kind: Node metadata: labels: cloud.google.com/gke-spot: "true" name: default
K8sAllowedRepos
Erlaubte Repositories v1.0.1
Erfordert Container-Images, die mit einem String aus der angegebenen Liste beginnen sollen.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sAllowedRepos
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# repos <array>: The list of prefixes a container image is allowed to have.
repos:
- <string>
Beispiele
repo-is-openpolicyagent
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sAllowedRepos metadata: name: repo-is-openpolicyagent spec: match: kinds: - apiGroups: - "" kinds: - Pod namespaces: - default parameters: repos: - openpolicyagent/
Zulässig
apiVersion: v1 kind: Pod metadata: name: opa-allowed spec: containers: - args: - run - --server - --addr=localhost:8080 image: openpolicyagent/opa:0.9.2 name: opa resources: limits: cpu: 100m memory: 30Mi
Nicht zugelassen
apiVersion: v1 kind: Pod metadata: name: nginx-disallowed spec: containers: - image: nginx name: nginx resources: limits: cpu: 100m memory: 30Mi
apiVersion: v1 kind: Pod metadata: name: nginx-disallowed spec: containers: - args: - run - --server - --addr=localhost:8080 image: openpolicyagent/opa:0.9.2 name: opa resources: limits: cpu: 100m memory: 30Mi initContainers: - image: nginx name: nginxinit resources: limits: cpu: 100m memory: 30Mi
apiVersion: v1 kind: Pod metadata: name: nginx-disallowed spec: containers: - image: nginx name: nginx resources: limits: cpu: 100m memory: 30Mi initContainers: - image: nginx name: nginxinit resources: limits: cpu: 100m memory: 30Mi
apiVersion: v1 kind: Pod metadata: name: nginx-disallowed spec: containers: - image: nginx name: nginx resources: limits: cpu: 100m memory: 30Mi ephemeralContainers: - image: nginx name: nginx resources: limits: cpu: 100m memory: 30Mi initContainers: - image: nginx name: nginx resources: limits: cpu: 100m memory: 30Mi
K8sAvoidUseOfSystemMastersGroup
Disallow the use of 'system:masters' group v1.0.0
Verhindert die Verwendung der Gruppe "system:masters". Hat keine Auswirkungen während der Prüfung.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sAvoidUseOfSystemMastersGroup
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# allowlistedUsernames <array>: allowlistedUsernames is the list of
# usernames that are allowed to use system:masters group.
allowlistedUsernames:
- <string>
Beispiele
avoid-use-of-system-masters-group
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sAvoidUseOfSystemMastersGroup metadata: name: avoid-use-of-system-masters-group
Zugelassen
apiVersion: v1 kind: Namespace metadata: name: example-namespace
K8sBlockAllIngress
Ingress v1.0.4 vollständig blockieren
Verhindert das Erstellen von Ingress-Objekten (die Typen Ingress, Gateway und Service von NodePort und LoadBalancer).
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sBlockAllIngress
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# allowList <array>: A list of regular expressions for the Ingress object
# names that are exempt from the constraint.
allowList:
- <string>
Beispiele
block-all-ingress
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sBlockAllIngress metadata: name: block-all-ingress spec: enforcementAction: dryrun parameters: allowList: - name1 - name2 - name3 - my-*
Zulässig
apiVersion: v1 kind: Service metadata: name: my-service spec: ports: - port: 80 protocol: TCP targetPort: 9376 selector: app.kubernetes.io/name: MyApp type: LoadBalancer
apiVersion: v1 kind: Service metadata: name: allowed-clusterip-service-example spec: ports: - port: 80 protocol: TCP targetPort: 9376 selector: app.kubernetes.io/name: MyApp type: ClusterIP
Nicht zugelassen
apiVersion: v1 kind: Service metadata: name: disallowed-service-example spec: ports: - port: 80 protocol: TCP targetPort: 9376 selector: app.kubernetes.io/name: MyApp type: LoadBalancer
apiVersion: v1 kind: Service metadata: name: disallowed-service-example spec: ports: - port: 80 protocol: TCP targetPort: 9376 selector: app.kubernetes.io/name: MyApp type: LoadBalancer
apiVersion: gateway.networking.k8s.io/v1 kind: Gateway metadata: name: disallowed-gateway-example spec: gatewayClassName: istio listeners: - allowedRoutes: namespaces: from: All hostname: '*.example.com' name: default port: 80 protocol: HTTP
K8sBlockCreationWithDefaultServiceAccount
Block Creation with Default Service Account v1.0.2
Verhindert das Erstellen von Ressourcen mit einem Standarddienstkonto. Hat keine Auswirkungen während der Prüfung.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sBlockCreationWithDefaultServiceAccount
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
Beispiele
block-creation-with-default-serviceaccount
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sBlockCreationWithDefaultServiceAccount metadata: name: block-creation-with-default-serviceaccount spec: enforcementAction: dryrun
Zugelassen
apiVersion: v1 kind: Namespace metadata: name: example-namespace
K8sBlockEndpointEditDefaultRole
Block Endpoint Edit Default Role v1.0.0
Viele Kubernetes-Installationen haben standardmäßig eine system:aggregate-to-edit ClusterRole, die den Zugriff auf die Bearbeitung von Endpunkten nicht richtig einschränkt. Dieses ConstraintTemplate verbietet der system:aggregate-to-edit ClusterRole, die Erlaubnis zur Erstellung/Patch/Aktualisierung von Endpunkten zu erteilen. ClusterRole/system:aggregation-to-edit sollte aufgrund von CVE-2021-25740-Berechtigungen, den Endpoint- und EndpointSlice-Berechtigungen, die Namespace-übergreifende Weiterleitung nicht zulassen, https://github.com/kubernetes/kubernetes/issues/103675.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sBlockEndpointEditDefaultRole
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
Beispiele
block-endpoint-edit-default-role
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sBlockEndpointEditDefaultRole metadata: name: block-endpoint-edit-default-role spec: match: kinds: - apiGroups: - rbac.authorization.k8s.io kinds: - ClusterRole
Zulässig
apiVersion: rbac.authorization.k8s.io/v1 kind: ClusterRole metadata: annotations: rbac.authorization.kubernetes.io/autoupdate: "true" labels: kubernetes.io/bootstrapping: rbac-defaults rbac.authorization.k8s.io/aggregate-to-edit: "true" name: system:aggregate-to-edit rules: - apiGroups: - "" resources: - pods/attach - pods/exec - pods/portforward - pods/proxy - secrets - services/proxy verbs: - get - list - watch - apiGroups: - "" resources: - serviceaccounts verbs: - impersonate - apiGroups: - "" resources: - pods - pods/attach - pods/exec - pods/portforward - pods/proxy verbs: - create - delete - deletecollection - patch - update - apiGroups: - "" resources: - configmaps - persistentvolumeclaims - replicationcontrollers - replicationcontrollers/scale - secrets - serviceaccounts - services - services/proxy verbs: - create - delete - deletecollection - patch - update - apiGroups: - apps resources: - daemonsets - deployments - deployments/rollback - deployments/scale - replicasets - replicasets/scale - statefulsets - statefulsets/scale verbs: - create - delete - deletecollection - patch - update - apiGroups: - autoscaling resources: - horizontalpodautoscalers verbs: - create - delete - deletecollection - patch - update - apiGroups: - batch resources: - cronjobs - jobs verbs: - create - delete - deletecollection - patch - update - apiGroups: - extensions resources: - daemonsets - deployments - deployments/rollback - deployments/scale - ingresses - networkpolicies - replicasets - replicasets/scale - replicationcontrollers/scale verbs: - create - delete - deletecollection - patch - update - apiGroups: - policy resources: - poddisruptionbudgets verbs: - create - delete - deletecollection - patch - update - apiGroups: - networking.k8s.io resources: - ingresses - networkpolicies verbs: - create - delete - deletecollection - patch - update
Nicht zugelassen
apiVersion: rbac.authorization.k8s.io/v1 kind: ClusterRole metadata: annotations: rbac.authorization.kubernetes.io/autoupdate: "true" labels: kubernetes.io/bootstrapping: rbac-defaults rbac.authorization.k8s.io/aggregate-to-edit: "true" name: system:aggregate-to-edit rules: - apiGroups: - "" resources: - pods/attach - pods/exec - pods/portforward - pods/proxy - secrets - services/proxy verbs: - get - list - watch - apiGroups: - "" resources: - serviceaccounts verbs: - impersonate - apiGroups: - "" resources: - pods - pods/attach - pods/exec - pods/portforward - pods/proxy verbs: - create - delete - deletecollection - patch - update - apiGroups: - "" resources: - configmaps - persistentvolumeclaims - replicationcontrollers - replicationcontrollers/scale - secrets - serviceaccounts - services - services/proxy verbs: - create - delete - deletecollection - patch - update - apiGroups: - apps resources: - daemonsets - deployments - deployments/rollback - deployments/scale - endpoints - replicasets - replicasets/scale - statefulsets - statefulsets/scale verbs: - create - delete - deletecollection - patch - update
K8sBlockLoadBalancer
Dienste mit dem Typ LoadBalancer v1.0.0 blockieren
Verhindert alle Dienste mit dem Typ LoadBalancer. https://kubernetes.io/docs/concepts/services-networking/service/#loadbalancer
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sBlockLoadBalancer
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
Beispiele
block-load-balancer
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sBlockLoadBalancer metadata: name: block-load-balancer spec: match: kinds: - apiGroups: - "" kinds: - Service
Zulässig
apiVersion: v1 kind: Service metadata: name: my-service-allowed spec: ports: - port: 80 targetPort: 80 type: ClusterIP
Unzulässig
apiVersion: v1 kind: Service metadata: name: my-service-disallowed spec: ports: - nodePort: 30007 port: 80 targetPort: 80 type: LoadBalancer
K8sBlockNodePort
NodePort v1.0.0 blockieren
Verhindert alle Dienste mit dem Typ NodePort. https://kubernetes.io/docs/concepts/services-networking/service/#nodeport
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sBlockNodePort
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
Beispiele
block-node-port
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sBlockNodePort metadata: name: block-node-port spec: match: kinds: - apiGroups: - "" kinds: - Service
Nicht zugelassen
apiVersion: v1 kind: Service metadata: name: my-service-disallowed spec: ports: - nodePort: 30007 port: 80 targetPort: 80 type: NodePort
K8sBlockObjectsOfType
Objekte des Typs v1.0.1 blockieren
Verhindert das Objekt unzulässiger Typen.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sBlockObjectsOfType
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
forbiddenTypes:
- <string>
Beispiele
block-secrets-of-type-basic-auth
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sBlockObjectsOfType metadata: name: block-secrets-of-type-basic-auth spec: match: kinds: - apiGroups: - "" kinds: - Secret parameters: forbiddenTypes: - kubernetes.io/basic-auth
Zulässig
apiVersion: v1 data: password: ZHVtbXlwYXNz username: ZHVtbXl1c2Vy kind: Secret metadata: name: credentials namespace: default type: Opaque
Nicht zugelassen
apiVersion: v1 data: password: YmFzaWMtcGFzc3dvcmQ= username: YmFzaWMtdXNlcm5hbWU= kind: Secret metadata: name: secret-basic-auth namespace: default type: kubernetes.io/basic-auth
K8sBlockProcessNamespaceSharing
Prozess-Namespace-Freigabe blockieren v1.0.1
Unterbindet Pod-Spezifikationen, bei denen shareProcessNamespace auf true festgelegt ist. Dadurch werden Szenarien vermieden, in denen alle Container in einem Pod den gleichen PID-Namespace verwenden und auf das Dateisystem und den Arbeitsspeicher des jeweils anderen Systems zugreifen können.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sBlockProcessNamespaceSharing
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
Beispiele
block-process-namespace-sharing
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sBlockProcessNamespaceSharing metadata: name: block-process-namespace-sharing
Zulässig
apiVersion: v1 kind: Pod metadata: name: good-pod namespace: default spec: containers: - image: nginx name: nginx
Nicht zugelassen
apiVersion: v1 kind: Pod metadata: name: bad-pod namespace: default spec: containers: - image: nginx name: nginx shareProcessNamespace: true
K8sBlockWildcardIngress
Platzhalter Ingress Version 1.0.1 blockieren
Nutzer sollten keine Ingresses mit einem leeren oder Platzhalternamen (*) erstellen können, da sie dadurch Traffic für andere Dienste im Cluster abfangen könnten, selbst wenn sie keinen Zugriff auf diese Dienste haben.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sBlockWildcardIngress
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
Beispiele
block-wildcard-ingress
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sBlockWildcardIngress metadata: name: block-wildcard-ingress spec: match: kinds: - apiGroups: - extensions - networking.k8s.io kinds: - Ingress
Zulässig
apiVersion: networking.k8s.io/v1 kind: Ingress metadata: name: non-wildcard-ingress spec: rules: - host: myservice.example.com http: paths: - backend: service: name: example port: number: 80 path: / pathType: Prefix
Nicht zugelassen
apiVersion: networking.k8s.io/v1 kind: Ingress metadata: name: wildcard-ingress spec: rules: - host: "" http: paths: - backend: service: name: example port: number: 80 path: / pathType: Prefix
apiVersion: networking.k8s.io/v1 kind: Ingress metadata: name: wildcard-ingress spec: rules: - http: paths: - backend: service: name: example port: number: 80 path: / pathType: Prefix
apiVersion: networking.k8s.io/v1 kind: Ingress metadata: name: wildcard-ingress spec: rules: - host: '*.example.com' http: paths: - backend: service: name: example port: number: 80 path: / pathType: Prefix - host: valid.example.com http: paths: - backend: service: name: example port: number: 80 path: / pathType: Prefix
K8sContainerEphemeralStorageLimit
Flüchtiger Container-Speicherlimit (Version 1.0.2)
Erfordert, dass für Container ein sitzungsspezifisches Speicherlimit festgelegt ist und dass das Limit innerhalb der angegebenen Höchstwerte liegt. https://kubernetes.io/docs/concepts/configuration/manage-resources-containers/
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sContainerEphemeralStorageLimit
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# ephemeral-storage <string>: The maximum allowed ephemeral storage limit
# on a Pod, exclusive.
ephemeral-storage: <string>
# exemptImages <array>: Any container that uses an image that matches an
# entry in this list will be excluded from enforcement. Prefix-matching can
# be signified with `*`. For example: `my-image-*`. It is recommended that
# users use the fully-qualified Docker image name (e.g. start with a domain
# name) in order to avoid unexpectedly exempting images from an untrusted
# repository.
exemptImages:
- <string>
Beispiele
container-ephemeral-storage-limit
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sContainerEphemeralStorageLimit metadata: name: container-ephemeral-storage-limit spec: match: kinds: - apiGroups: - "" kinds: - Pod parameters: ephemeral-storage: 500Mi
Zulässig
apiVersion: v1 kind: Pod metadata: labels: owner: me.agilebank.demo name: opa-allowed spec: containers: - args: - run - --server - --addr=localhost:8080 image: openpolicyagent/opa:0.9.2 name: opa resources: limits: cpu: 100m ephemeral-storage: 100Mi memory: 1Gi
apiVersion: v1 kind: Pod metadata: labels: owner: me.agilebank.demo name: opa-allowed spec: containers: - args: - run - --server - --addr=localhost:8080 image: openpolicyagent/opa:0.9.2 name: opa resources: limits: cpu: 100m ephemeral-storage: 100Mi memory: 1Gi initContainers: - args: - run - --server - --addr=localhost:8080 image: openpolicyagent/opa:0.9.2 name: init-opa resources: limits: cpu: 100m ephemeral-storage: 100Mi memory: 1Gi
Unzulässig
apiVersion: v1 kind: Pod metadata: labels: owner: me.agilebank.demo name: opa-disallowed spec: containers: - args: - run - --server - --addr=localhost:8080 image: openpolicyagent/opa:0.9.2 name: opa resources: limits: cpu: 100m memory: 2Gi
apiVersion: v1 kind: Pod metadata: labels: owner: me.agilebank.demo name: opa-disallowed spec: containers: - args: - run - --server - --addr=localhost:8080 image: openpolicyagent/opa:0.9.2 name: opa resources: limits: cpu: 100m ephemeral-storage: 1Pi memory: 1Gi
apiVersion: v1 kind: Pod metadata: labels: owner: me.agilebank.demo name: opa-disallowed spec: containers: - args: - run - --server - --addr=localhost:8080 image: openpolicyagent/opa:0.9.2 name: opa resources: limits: cpu: 100m ephemeral-storage: 100Mi memory: 1Gi initContainers: - args: - run - --server - --addr=localhost:8080 image: openpolicyagent/opa:0.9.2 name: init-opa resources: limits: cpu: 100m ephemeral-storage: 1Pi memory: 1Gi
K8sContainerLimits
Containerlimits v1.0.1
Erfordert, dass in Containern Speicher- und CPU-Beschränkungen festgelegt sind, die innerhalb der angegebenen Höchstwerte liegen müssen. https://kubernetes.io/docs/concepts/configuration/manage-resources-containers/
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sContainerLimits
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# cpu <string>: The maximum allowed cpu limit on a Pod, exclusive.
cpu: <string>
# exemptImages <array>: Any container that uses an image that matches an
# entry in this list will be excluded from enforcement. Prefix-matching can
# be signified with `*`. For example: `my-image-*`. It is recommended that
# users use the fully-qualified Docker image name (e.g. start with a domain
# name) in order to avoid unexpectedly exempting images from an untrusted
# repository.
exemptImages:
- <string>
# memory <string>: The maximum allowed memory limit on a Pod, exclusive.
memory: <string>
Beispiele
container-must-have-limits
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sContainerLimits metadata: name: container-must-have-limits spec: match: kinds: - apiGroups: - "" kinds: - Pod parameters: cpu: 200m memory: 1Gi
Zulässig
apiVersion: v1 kind: Pod metadata: labels: owner: me.agilebank.demo name: opa-allowed spec: containers: - args: - run - --server - --addr=localhost:8080 image: openpolicyagent/opa:0.9.2 name: opa resources: limits: cpu: 100m memory: 1Gi
Unzulässig
apiVersion: v1 kind: Pod metadata: labels: owner: me.agilebank.demo name: opa-disallowed spec: containers: - args: - run - --server - --addr=localhost:8080 image: openpolicyagent/opa:0.9.2 name: opa resources: limits: cpu: 100m memory: 2Gi
K8sContainerRatios
Containerverhältnisse v1.0.1
Legt ein maximales Verhältnis für Containerressourcenlimits zu Anfragen fest. https://kubernetes.io/docs/concepts/configuration/manage-resources-containers/
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sContainerRatios
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# cpuRatio <string>: The maximum allowed ratio of `resources.limits.cpu` to
# `resources.requests.cpu` on a container. If not specified, equal to
# `ratio`.
cpuRatio: <string>
# exemptImages <array>: Any container that uses an image that matches an
# entry in this list will be excluded from enforcement. Prefix-matching can
# be signified with `*`. For example: `my-image-*`. It is recommended that
# users use the fully-qualified Docker image name (e.g. start with a domain
# name) in order to avoid unexpectedly exempting images from an untrusted
# repository.
exemptImages:
- <string>
# ratio <string>: The maximum allowed ratio of `resources.limits` to
# `resources.requests` on a container.
ratio: <string>
Beispiele
container-must-meet-ratio
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sContainerRatios metadata: name: container-must-meet-ratio spec: match: kinds: - apiGroups: - "" kinds: - Pod parameters: ratio: "2"
Zulässig
apiVersion: v1 kind: Pod metadata: labels: owner: me.agilebank.demo name: opa-disallowed spec: containers: - args: - run - --server - --addr=localhost:8080 image: openpolicyagent/opa:0.9.2 name: opa resources: limits: cpu: 200m memory: 200Mi requests: cpu: 100m memory: 100Mi
Nicht zugelassen
apiVersion: v1 kind: Pod metadata: labels: owner: me.agilebank.demo name: opa-disallowed spec: containers: - args: - run - --server - --addr=localhost:8080 image: openpolicyagent/opa:0.9.2 name: opa resources: limits: cpu: 800m memory: 2Gi requests: cpu: 100m memory: 100Mi
container-must-meet-memory-and-cpu-ratio
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sContainerRatios metadata: name: container-must-meet-memory-and-cpu-ratio spec: match: kinds: - apiGroups: - "" kinds: - Pod parameters: cpuRatio: "10" ratio: "1"
Zulässig
apiVersion: v1 kind: Pod metadata: labels: owner: me.agilebank.demo name: opa-allowed spec: containers: - args: - run - --server - --addr=localhost:8080 image: openpolicyagent/opa:0.9.2 name: opa resources: limits: cpu: "4" memory: 2Gi requests: cpu: "1" memory: 2Gi
Nicht zugelassen
apiVersion: v1 kind: Pod metadata: labels: owner: me.agilebank.demo name: opa-disallowed spec: containers: - args: - run - --server - --addr=localhost:8080 image: openpolicyagent/opa:0.9.2 name: opa resources: limits: cpu: "4" memory: 2Gi requests: cpu: 100m memory: 2Gi
K8sContainerRequests
Containeranfragen v1.0.1
Erfordert, dass in Containern Arbeitsspeicher- und CPU-Anfragen festgelegt sind und dass Anfragen innerhalb der angegebenen Höchstwerte liegen müssen. https://kubernetes.io/docs/concepts/configuration/manage-resources-containers/
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sContainerRequests
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# cpu <string>: The maximum allowed cpu request on a Pod, exclusive.
cpu: <string>
# exemptImages <array>: Any container that uses an image that matches an
# entry in this list will be excluded from enforcement. Prefix-matching can
# be signified with `*`. For example: `my-image-*`. It is recommended that
# users use the fully-qualified Docker image name (e.g. start with a domain
# name) in order to avoid unexpectedly exempting images from an untrusted
# repository.
exemptImages:
- <string>
# memory <string>: The maximum allowed memory request on a Pod, exclusive.
memory: <string>
Beispiele
container-must-have-requests
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sContainerRequests metadata: name: container-must-have-requests spec: match: kinds: - apiGroups: - "" kinds: - Pod parameters: cpu: 200m memory: 1Gi
Zulässig
apiVersion: v1 kind: Pod metadata: labels: owner: me.agilebank.demo name: opa-allowed spec: containers: - args: - run - --server - --addr=localhost:8080 image: openpolicyagent/opa:0.9.2 name: opa resources: requests: cpu: 100m memory: 1Gi
Nicht zugelassen
apiVersion: v1 kind: Pod metadata: labels: owner: me.agilebank.demo name: opa-disallowed spec: containers: - args: - run - --server - --addr=localhost:8080 image: openpolicyagent/opa:0.9.2 name: opa resources: requests: cpu: 100m memory: 2Gi
K8sCronJobAllowedRepos
CronJob-Zulässige Repositories v1.0.1
Erfordert Container-Images von CronJobs, die mit einem String aus der angegebenen Liste beginnen sollen.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sCronJobAllowedRepos
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# repos <array>: The list of prefixes a container image is allowed to have.
repos:
- <string>
Beispiele
cronjob-restrict-repos
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sCronJobAllowedRepos metadata: name: cronjob-restrict-repos spec: match: kinds: - apiGroups: - batch kinds: - CronJob parameters: repos: - gke.gcr.io/
Zulässig
apiVersion: batch/v1 kind: CronJob metadata: name: hello spec: jobTemplate: spec: template: spec: containers: - image: gke.gcr.io/busybox:1.28 name: hello schedule: '* * * * *'
Nicht zugelassen
apiVersion: batch/v1 kind: CronJob metadata: name: hello spec: jobTemplate: spec: template: spec: containers: - image: busybox:1.28 name: hello schedule: '* * * * *'
K8sDisallowAnonymous
Anonymen Zugriff nicht zulassen v1.0.0
Verbietet die Zuordnung von ClusterRole- und Rollen-Ressourcen zum Nutzer "system:anonymous" und zur Gruppe "system:unauthenticated".
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sDisallowAnonymous
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# allowedRoles <array>: The list of ClusterRoles and Roles that may be
# associated with the `system:unauthenticated` group and `system:anonymous`
# user.
allowedRoles:
- <string>
Beispiele
no-anonymous
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sDisallowAnonymous metadata: name: no-anonymous spec: match: kinds: - apiGroups: - rbac.authorization.k8s.io kinds: - ClusterRoleBinding - apiGroups: - rbac.authorization.k8s.io kinds: - RoleBinding parameters: allowedRoles: - cluster-role-1
Zulässig
apiVersion: rbac.authorization.k8s.io/v1 kind: ClusterRoleBinding metadata: name: cluster-role-binding-1 roleRef: apiGroup: rbac.authorization.k8s.io kind: ClusterRole name: cluster-role-1 subjects: - apiGroup: rbac.authorization.k8s.io kind: Group name: system:authenticated - apiGroup: rbac.authorization.k8s.io kind: Group name: system:unauthenticated
Nicht zugelassen
apiVersion: rbac.authorization.k8s.io/v1 kind: ClusterRoleBinding metadata: name: cluster-role-binding-2 roleRef: apiGroup: rbac.authorization.k8s.io kind: ClusterRole name: cluster-role-2 subjects: - apiGroup: rbac.authorization.k8s.io kind: Group name: system:authenticated - apiGroup: rbac.authorization.k8s.io kind: Group name: system:unauthenticated
K8sDisallowInteractiveTTY
Interaktive TTY-Container v1.0.0 nicht zulassen
Erfordert, dass die Felder spec.tty und spec.stdin der Objekte auf "false" gesetzt oder nicht festgelegt sind.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sDisallowInteractiveTTY
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# exemptImages <array>: Any container that uses an image that matches an
# entry in this list will be excluded from enforcement. Prefix-matching can
# be signified with `*`. For example: `my-image-*`. It is recommended that
# users use the fully-qualified Docker image name (e.g. start with a domain
# name) in order to avoid unexpectedly exempting images from an untrusted
# repository.
exemptImages:
- <string>
Beispiele
no-interactive-tty-containers
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sDisallowInteractiveTTY metadata: name: no-interactive-tty-containers spec: match: kinds: - apiGroups: - "" kinds: - Pod
Zulässig
apiVersion: v1 kind: Pod metadata: labels: app: nginx-interactive-tty name: nginx-interactive-tty-allowed spec: containers: - image: nginx name: nginx stdin: false tty: false
Nicht zugelassen
apiVersion: v1 kind: Pod metadata: labels: app: nginx-privilege-escalation name: nginx-privilege-escalation-disallowed spec: containers: - image: nginx name: nginx stdin: true tty: true
K8sDisallowedRepos
Unzulässige Repositories v1.0.0
Unzulässige Container-Repositories, die mit einem String aus der angegebenen Liste beginnen.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sDisallowedRepos
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# repos <array>: The list of prefixes a container image is not allowed to
# have.
repos:
- <string>
Beispiele
repo-must-not-be-k8s-gcr-io
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sDisallowedRepos metadata: name: repo-must-not-be-k8s-gcr-io spec: match: kinds: - apiGroups: - "" kinds: - Pod parameters: repos: - k8s.gcr.io/
Zulässig
apiVersion: v1 kind: Pod metadata: name: kustomize-allowed spec: containers: - image: registry.k8s.io/kustomize/kustomize:v3.8.9 name: kustomize
Nicht zugelassen
apiVersion: v1 kind: Pod metadata: name: kustomize-disallowed spec: containers: - image: k8s.gcr.io/kustomize/kustomize:v3.8.9 name: kustomize
apiVersion: v1 kind: Pod metadata: name: kustomize-disallowed spec: containers: - image: registry.k8s.io/kustomize/kustomize:v3.8.9 name: kustomize initContainers: - image: k8s.gcr.io/kustomize/kustomize:v3.8.9 name: kustomizeinit
apiVersion: v1 kind: Pod metadata: name: kustomize-disallowed spec: containers: - image: k8s.gcr.io/kustomize/kustomize:v3.8.9 name: kustomize initContainers: - image: k8s.gcr.io/kustomize/kustomize:v3.8.9 name: kustomizeinit
apiVersion: v1 kind: Pod metadata: name: kustomize-disallowed spec: containers: - image: k8s.gcr.io/kustomize/kustomize:v3.8.9 name: kustomize ephemeralContainers: - image: k8s.gcr.io/kustomize/kustomize:v3.8.9 name: kustomize initContainers: - image: k8s.gcr.io/kustomize/kustomize:v3.8.9 name: kustomize
K8sDisallowedRoleBindingSubjects
Unzulässige Rollenbindungsthemen v1.0.1
Unterbindet RoleBinding oder ClusterRoleBinding mit Subjekten, die mit einem als Parameter übergebenen disallowedSubjects übereinstimmen.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sDisallowedRoleBindingSubjects
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# disallowedSubjects <array>: A list of subjects that cannot appear in a
# RoleBinding.
disallowedSubjects:
- # apiGroup <string>: The Kubernetes API group of the disallowed role
# binding subject. Currently ignored.
apiGroup: <string>
# kind <string>: The kind of the disallowed role binding subject.
kind: <string>
# name <string>: The name of the disallowed role binding subject.
name: <string>
Beispiele
disallowed-rolebinding-subjects
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sDisallowedRoleBindingSubjects metadata: name: disallowed-rolebinding-subjects spec: parameters: disallowedSubjects: - apiGroup: rbac.authorization.k8s.io kind: Group name: system:unauthenticated
Zulässig
apiVersion: rbac.authorization.k8s.io/v1 kind: ClusterRoleBinding metadata: name: good-clusterrolebinding roleRef: apiGroup: rbac.authorization.k8s.io kind: ClusterRole name: my-role subjects: - apiGroup: rbac.authorization.k8s.io kind: Group name: system:authenticated
Unzulässig
apiVersion: rbac.authorization.k8s.io/v1 kind: ClusterRoleBinding metadata: name: bad-clusterrolebinding roleRef: apiGroup: rbac.authorization.k8s.io kind: ClusterRole name: my-role subjects: - apiGroup: rbac.authorization.k8s.io kind: Group name: system:unauthenticated
K8sDisallowedTags
Tags v1.0.1 nicht zulassen
Erfordert, dass Container-Images ein Image-Tag haben, das sich von denen in der angegebenen Liste unterscheidet. https://kubernetes.io/docs/concepts/containers/images/#image-names
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sDisallowedTags
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# exemptImages <array>: Any container that uses an image that matches an
# entry in this list will be excluded from enforcement. Prefix-matching can
# be signified with `*`. For example: `my-image-*`. It is recommended that
# users use the fully-qualified Docker image name (e.g. start with a domain
# name) in order to avoid unexpectedly exempting images from an untrusted
# repository.
exemptImages:
- <string>
# tags <array>: Disallowed container image tags.
tags:
- <string>
Beispiele
container-image-must-not-have-latest-tag
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sDisallowedTags metadata: name: container-image-must-not-have-latest-tag spec: match: kinds: - apiGroups: - "" kinds: - Pod namespaces: - default parameters: exemptImages: - openpolicyagent/opa-exp:latest - openpolicyagent/opa-exp2:latest tags: - latest
Zulässig
apiVersion: v1 kind: Pod metadata: name: opa-allowed spec: containers: - args: - run - --server - --addr=localhost:8080 image: openpolicyagent/opa:0.9.2 name: opa
apiVersion: v1 kind: Pod metadata: name: opa-exempt-allowed spec: containers: - args: - run - --server - --addr=localhost:8080 image: openpolicyagent/opa-exp:latest name: opa-exp - args: - run - --server - --addr=localhost:8080 image: openpolicyagent/init:v1 name: opa-init - args: - run - --server - --addr=localhost:8080 image: openpolicyagent/opa-exp2:latest name: opa-exp2
Unzulässig
apiVersion: v1 kind: Pod metadata: name: opa-disallowed spec: containers: - args: - run - --server - --addr=localhost:8080 image: openpolicyagent/opa name: opa
apiVersion: v1 kind: Pod metadata: name: opa-disallowed-2 spec: containers: - args: - run - --server - --addr=localhost:8080 image: openpolicyagent/opa:latest name: opa
apiVersion: v1 kind: Pod metadata: name: opa-disallowed-ephemeral spec: containers: - args: - run - --server - --addr=localhost:8080 image: openpolicyagent/opa:0.9.2 name: opa ephemeralContainers: - args: - run - --server - --addr=localhost:8080 image: openpolicyagent/opa:latest name: opa
apiVersion: v1 kind: Pod metadata: name: opa-disallowed-3 spec: containers: - args: - run - --server - --addr=localhost:8080 image: openpolicyagent/opa-exp:latest name: opa - args: - run - --server - --addr=localhost:8080 image: openpolicyagent/init:latest name: opa-init - args: - run - --server - --addr=localhost:8080 image: openpolicyagent/opa-exp2:latest name: opa-exp2 - args: - run - --server - --addr=localhost:8080 image: openpolicyagent/monitor:latest name: opa-monitor
K8sEmptyDirHasSizeLimit
Leeres Verzeichnis hat eine Größenbeschränkung von v1.0.5
Erfordert, dass alle emptyDir-Volumes einen sizeLimit angeben. Optional kann in der Einschränkung der Parameter maxSizeLimit angegeben werden, um ein maximal zulässiges Größenlimit anzugeben.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sEmptyDirHasSizeLimit
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# exemptVolumesRegex <array>: Exempt Volume names as regex match.
exemptVolumesRegex:
- <string>
# maxSizeLimit <string>: When set, the declared size limit for each volume
# must be less than `maxSizeLimit`.
maxSizeLimit: <string>
Beispiele
empty-dir-has-size-limit
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sEmptyDirHasSizeLimit metadata: name: empty-dir-has-size-limit spec: match: excludedNamespaces: - istio-system - kube-system - gatekeeper-system parameters: exemptVolumesRegex: - ^istio-[a-z]+$ maxSizeLimit: 4Gi
Zulässig
apiVersion: v1 kind: Pod metadata: name: good-pod namespace: default spec: containers: - image: nginx name: nginx volumes: - emptyDir: sizeLimit: 2Gi name: good-pod-volume
apiVersion: v1 kind: Pod metadata: name: exempt-pod namespace: default spec: containers: - image: nginx name: nginx volumes: - emptyDir: {} name: istio-envoy
Nicht zugelassen
apiVersion: v1 kind: Pod metadata: name: bad-pod namespace: default spec: containers: - image: nginx name: nginx volumes: - emptyDir: {} name: bad-pod-volume
K8sEnforceCloudArmorBackendConfig
Cloud Armor für BackendConfig-Ressourcen v1.0.2 erzwingen
Erzwingt die Cloud Armor-Konfiguration für BackendConfig-Ressourcen
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sEnforceCloudArmorBackendConfig
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
Beispiele
enforce-cloudarmor-backendconfig
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sEnforceCloudArmorBackendConfig metadata: name: enforce-cloudarmor-backendconfig spec: enforcementAction: dryrun
Zulässig
apiVersion: cloud.google.com/v1 kind: BackendConfig metadata: name: my-backendconfig namespace: examplenamespace spec: securityPolicy: name: example-security-policy
apiVersion: cloud.google.com/v1 kind: BackendConfig metadata: name: second-backendconfig spec: securityPolicy: name: my-security-policy
Nicht zugelassen
apiVersion: cloud.google.com/v1 kind: BackendConfig metadata: name: my-backendconfig namespace: examplenamespace spec: securityPolicy: name: null
apiVersion: cloud.google.com/v1 kind: BackendConfig metadata: name: my-backendconfig namespace: examplenamespace spec: securityPolicy: name: ""
apiVersion: cloud.google.com/v1 kind: BackendConfig metadata: name: my-backendconfig spec: logging: enable: true sampleRate: 0.5
K8sEnforceConfigManagement
Config Management Version 1.1.6 erzwingen
Erfordert das Vorhandensein und den Betrieb von Config Management. Einschränkungen mit diesem ConstraintTemplate werden unabhängig vom enforcementAction-Wert nur geprüft.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sEnforceConfigManagement
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# requireDriftPrevention <boolean>: Require Config Sync drift prevention to
# prevent config drift.
requireDriftPrevention: <boolean>
# requireRootSync <boolean>: Require a Config Sync `RootSync` object for
# cluster config management.
requireRootSync: <boolean>
Referenzielle Einschränkung
Diese Einschränkung ist referenziell. Vor der Verwendung müssen Sie referenzielle Einschränkungen aktivieren und eine Konfiguration erstellen, die Policy Controller mitteilt, welche Arten von Objekten beobachtet werden sollen.
Ihr Policy Controller Config erfordert einen syncOnly-Eintrag ähnlich dem folgenden:
spec:
sync:
syncOnly:
- group: "configsync.gke.io"
version: "v1beta1"
kind: "RootSync"
Beispiele
enforce-config-management
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sEnforceConfigManagement metadata: name: enforce-config-management spec: enforcementAction: dryrun match: kinds: - apiGroups: - configmanagement.gke.io kinds: - ConfigManagement
Zulässig
apiVersion: configmanagement.gke.io/v1 kind: ConfigManagement metadata: annotations: configmanagement.gke.io/managed-by-hub: "true" configmanagement.gke.io/update-time: "1663586155" name: config-management spec: binauthz: enabled: true clusterName: tec6ea817b5b4bb2-cluster enableMultiRepo: true git: proxy: {} syncRepo: git@test-git-server.config-management-system-test:/git-server/repos/sot.git hierarchyController: {} policyController: auditIntervalSeconds: 60 enabled: true monitoring: backends: - prometheus - cloudmonitoring mutation: {} referentialRulesEnabled: true templateLibraryInstalled: true status: configManagementVersion: v1.12.2-rc.2 healthy: true
Unzulässig
apiVersion: configmanagement.gke.io/v1 kind: ConfigManagement metadata: annotations: configmanagement.gke.io/managed-by-hub: "true" configmanagement.gke.io/update-time: "1663586155" name: config-management spec: binauthz: enabled: true clusterName: tec6ea817b5b4bb2-cluster enableMultiRepo: true git: syncRepo: git@test-git-server.config-management-system-test:/git-server/repos/sot.git hierarchyController: {} policyController: auditIntervalSeconds: 60 enabled: true monitoring: backends: - prometheus - cloudmonitoring mutation: {} referentialRulesEnabled: true templateLibraryInstalled: true status: configManagementVersion: v1.12.2-rc.2
K8sExternalIPs
Externe IP-Adressen v1.0.0
Beschränkt externe IP-Adressen auf eine zulässige Liste von IP-Adressen. https://kubernetes.io/docs/concepts/services-networking/service/#external-ips
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sExternalIPs
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# allowedIPs <array>: An allow-list of external IP addresses.
allowedIPs:
- <string>
Beispiele
external-ips
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sExternalIPs metadata: name: external-ips spec: match: kinds: - apiGroups: - "" kinds: - Service parameters: allowedIPs: - 203.0.113.0
Zulässig
apiVersion: v1 kind: Service metadata: name: allowed-external-ip spec: externalIPs: - 203.0.113.0 ports: - name: http port: 80 protocol: TCP targetPort: 8080 selector: app: MyApp
Nicht zugelassen
apiVersion: v1 kind: Service metadata: name: disallowed-external-ip spec: externalIPs: - 1.1.1.1 ports: - name: http port: 80 protocol: TCP targetPort: 8080 selector: app: MyApp
K8sHorizontalPodAutoscaler
Horizontales Pod-Autoscaling v1.0.1
Die folgenden Szenarien sollen bei der Bereitstellung von HorizontalPodAutoscalers 1 nicht zugelassen werden. Bereitstellung von HorizontalPodAutoscalern mit .spec.minReplicas oder .spec.maxReplicas außerhalb der in der Einschränkung 2 definierten Bereiche Bereitstellung von HorizontalPodAutoscalern, bei denen der Unterschied zwischen .spec.minReplicas und .spec.maxReplicas kleiner als der konfigurierte minimumReplicaSpread 3 ist Bereitstellung von HorizontalPodAutoscalern, die nicht auf ein gültiges scaleTargetRef verweisen (z.B. Deployment, ReplicationController, ReplicaSet, StatefulSet).
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sHorizontalPodAutoscaler
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# enforceScaleTargetRef <boolean>: If set to true it validates the HPA
# scaleTargetRef exists
enforceScaleTargetRef: <boolean>
# minimumReplicaSpread <integer>: If configured it enforces the minReplicas
# and maxReplicas in an HPA must have a spread of at least this many
# replicas
minimumReplicaSpread: <integer>
# ranges <array>: Allowed ranges for numbers of replicas. Values are
# inclusive.
ranges:
# <list item: object>: A range of allowed replicas. Values are
# inclusive.
- # max_replicas <integer>: The maximum number of replicas allowed,
# inclusive.
max_replicas: <integer>
# min_replicas <integer>: The minimum number of replicas allowed,
# inclusive.
min_replicas: <integer>
Referenzielle Einschränkung
Diese Einschränkung ist referenziell. Vor der Verwendung müssen Sie referenzielle Einschränkungen aktivieren und eine Konfiguration erstellen, die Policy Controller mitteilt, welche Arten von Objekten beobachtet werden sollen.
Ihr Policy Controller Config erfordert einen syncOnly-Eintrag ähnlich dem folgenden:
spec:
sync:
syncOnly:
- group: "apps"
version: "v1"
kind: "Deployment"
OR
- group: "apps"
version: "v1"
kind: "StatefulSet"
Beispiele
horizontal-pod-autoscaler
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sHorizontalPodAutoscaler metadata: name: horizontal-pod-autoscaler spec: enforcementAction: deny match: kinds: - apiGroups: - autoscaling kinds: - HorizontalPodAutoscaler parameters: enforceScaleTargetRef: true minimumReplicaSpread: 1 ranges: - max_replicas: 6 min_replicas: 3
Zulässig
apiVersion: autoscaling/v2 kind: HorizontalPodAutoscaler metadata: name: nginx-hpa-allowed namespace: default spec: maxReplicas: 6 metrics: - resource: name: cpu target: averageUtilization: 900 type: Utilization type: Resource minReplicas: 3 scaleTargetRef: apiVersion: apps/v1 kind: Deployment name: nginx-deployment --- # Referential Data apiVersion: apps/v1 kind: Deployment metadata: labels: app: nginx name: nginx-deployment namespace: default spec: replicas: 3 selector: matchLabels: app: nginx example: allowed-deployment template: metadata: labels: app: nginx example: allowed-deployment spec: containers: - image: nginx:1.14.2 name: nginx ports: - containerPort: 80
Unzulässig
apiVersion: autoscaling/v2 kind: HorizontalPodAutoscaler metadata: name: nginx-hpa-disallowed-replicas namespace: default spec: maxReplicas: 7 metrics: - resource: name: cpu target: averageUtilization: 900 type: Utilization type: Resource minReplicas: 2 scaleTargetRef: apiVersion: apps/v1 kind: Deployment name: nginx-deployment --- # Referential Data apiVersion: apps/v1 kind: Deployment metadata: labels: app: nginx name: nginx-deployment namespace: default spec: replicas: 3 selector: matchLabels: app: nginx example: allowed-deployment template: metadata: labels: app: nginx example: allowed-deployment spec: containers: - image: nginx:1.14.2 name: nginx ports: - containerPort: 80
apiVersion: autoscaling/v2 kind: HorizontalPodAutoscaler metadata: name: nginx-hpa-disallowed-replicaspread namespace: default spec: maxReplicas: 4 metrics: - resource: name: cpu target: averageUtilization: 900 type: Utilization type: Resource minReplicas: 4 scaleTargetRef: apiVersion: apps/v1 kind: Deployment name: nginx-deployment --- # Referential Data apiVersion: apps/v1 kind: Deployment metadata: labels: app: nginx name: nginx-deployment namespace: default spec: replicas: 3 selector: matchLabels: app: nginx example: allowed-deployment template: metadata: labels: app: nginx example: allowed-deployment spec: containers: - image: nginx:1.14.2 name: nginx ports: - containerPort: 80
apiVersion: autoscaling/v2 kind: HorizontalPodAutoscaler metadata: name: nginx-hpa-disallowed-scaletarget namespace: default spec: maxReplicas: 6 metrics: - resource: name: cpu target: averageUtilization: 900 type: Utilization type: Resource minReplicas: 3 scaleTargetRef: apiVersion: apps/v1 kind: Deployment name: nginx-deployment-missing --- # Referential Data apiVersion: apps/v1 kind: Deployment metadata: labels: app: nginx name: nginx-deployment namespace: default spec: replicas: 3 selector: matchLabels: app: nginx example: allowed-deployment template: metadata: labels: app: nginx example: allowed-deployment spec: containers: - image: nginx:1.14.2 name: nginx ports: - containerPort: 80
K8sHttpsOnly
Nur HTTPS v1.0.2
Erfordert Ingress-Ressourcen, um nur HTTPS zu verwenden. Ingress-Ressourcen müssen die kubernetes.io/ingress.allow-http-Annotation enthalten, die auf false festgelegt sein muss. Standardmäßig ist eine gültige TLS-{}-Konfiguration erforderlich. Dies kann optional durch das Festlegen des Parameters tlsOptional auf true erfolgen.
https://kubernetes.io/docs/concepts/services-networking/ingress/#tls
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sHttpsOnly
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# tlsOptional <boolean>: When set to `true` the TLS {} is optional,
# defaults to false.
tlsOptional: <boolean>
Beispiele
ingress-https-only
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sHttpsOnly metadata: name: ingress-https-only spec: match: kinds: - apiGroups: - extensions - networking.k8s.io kinds: - Ingress
Zulässig
apiVersion: networking.k8s.io/v1 kind: Ingress metadata: annotations: kubernetes.io/ingress.allow-http: "false" name: ingress-demo-allowed spec: rules: - host: example-host.example.com http: paths: - backend: service: name: nginx port: number: 80 path: / pathType: Prefix tls: - {}
Nicht zugelassen
apiVersion: networking.k8s.io/v1 kind: Ingress metadata: name: ingress-demo-disallowed spec: rules: - host: example-host.example.com http: paths: - backend: service: name: nginx port: number: 80 path: / pathType: Prefix
ingress-https-only-tls-optional
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sHttpsOnly metadata: name: ingress-https-only-tls-optional spec: match: kinds: - apiGroups: - extensions - networking.k8s.io kinds: - Ingress parameters: tlsOptional: true
Zulässig
apiVersion: networking.k8s.io/v1 kind: Ingress metadata: annotations: kubernetes.io/ingress.allow-http: "false" name: ingress-demo-allowed-tls-optional spec: rules: - host: example-host.example.com http: paths: - backend: service: name: nginx port: number: 80 path: / pathType: Prefix
Nicht zugelassen
apiVersion: networking.k8s.io/v1 kind: Ingress metadata: name: ingress-demo-disallowed-tls-optional spec: rules: - host: example-host.example.com http: paths: - backend: service: name: nginx port: number: 80 path: / pathType: Prefix
K8sImageDigests
Image Digests v1.0.1
Erfordert, dass Container-Images einen Digest enthalten. https://kubernetes.io/docs/concepts/containers/images/
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sImageDigests
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# exemptImages <array>: Any container that uses an image that matches an
# entry in this list will be excluded from enforcement. Prefix-matching can
# be signified with `*`. For example: `my-image-*`. It is recommended that
# users use the fully-qualified Docker image name (e.g. start with a domain
# name) in order to avoid unexpectedly exempting images from an untrusted
# repository.
exemptImages:
- <string>
Beispiele
container-image-must-have-digest
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sImageDigests metadata: name: container-image-must-have-digest spec: match: kinds: - apiGroups: - "" kinds: - Pod namespaces: - default
Zulässig
apiVersion: v1 kind: Pod metadata: name: opa-allowed spec: containers: - args: - run - --server - --addr=localhost:8080 image: openpolicyagent/opa:0.9.2@sha256:04ff8fce2afd1a3bc26260348e5b290e8d945b1fad4b4c16d22834c2f3a1814a name: opa
Unzulässig
apiVersion: v1 kind: Pod metadata: name: opa-disallowed spec: containers: - args: - run - --server - --addr=localhost:8080 image: openpolicyagent/opa:0.9.2 name: opa initContainers: - args: - run - --server - --addr=localhost:8080 image: openpolicyagent/opa:0.9.2 name: opainit
apiVersion: v1 kind: Pod metadata: name: opa-disallowed spec: containers: - args: - run - --server - --addr=localhost:8080 image: openpolicyagent/opa:0.9.2 name: opa ephemeralContainers: - args: - run - --server - --addr=localhost:8080 image: openpolicyagent/opa:0.9.2 name: opa initContainers: - args: - run - --server - --addr=localhost:8080 image: openpolicyagent/opa:0.9.2 name: opainit
K8sLocalStorageRequireSafeToEvict
Lokaler Speicher erfordert Sicherheit zum Entfernen von v1.0.1
Erfordert Pods, die lokalen Speicher (emptyDir oder hostPath) mit der Annotation "cluster-autoscaler.kubernetes.io/safe-to-evict": "true" verwenden. Cluster Autoscaler löscht Pods ohne diese Annotation nicht.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sLocalStorageRequireSafeToEvict
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
Beispiele
local-storage-require-safe-to-evict
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sLocalStorageRequireSafeToEvict metadata: name: local-storage-require-safe-to-evict spec: match: excludedNamespaces: - kube-system - istio-system - gatekeeper-system
Zulässig
apiVersion: v1 kind: Pod metadata: annotations: cluster-autoscaler.kubernetes.io/safe-to-evict: "true" name: good-pod namespace: default spec: containers: - image: redis name: redis volumeMounts: - mountPath: /data/redis name: redis-storage volumes: - emptyDir: {} name: redis-storage
Unzulässig
apiVersion: v1 kind: Pod metadata: name: bad-pod namespace: default spec: containers: - image: redis name: redis volumeMounts: - mountPath: /data/redis name: redis-storage volumes: - emptyDir: {} name: redis-storage
K8sMemoryRequestEqualsLimit
Speicheranfrage entspricht Limit v1.0.4
Erhöht die Pod-Stabilität. Dafür muss der angeforderte Speicher aller Container genau dem Speicherlimit entsprechen, sodass sich Pods nie in einem Zustand befinden, in dem die Speichernutzung den angeforderten Umfang überschreitet. Andernfalls kann Kubernetes Pods beenden, die zusätzlichen Speicher anfordern, wenn der Speicher auf dem Knoten benötigt wird.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sMemoryRequestEqualsLimit
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# exemptContainersRegex <array>: Exempt Container names as regex match.
exemptContainersRegex:
- <string>
Beispiele
container-must-request-limit
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sMemoryRequestEqualsLimit metadata: name: container-must-request-limit spec: match: excludedNamespaces: - kube-system - resource-group-system - asm-system - istio-system - config-management-system - config-management-monitoring parameters: exemptContainersRegex: - ^istio-[a-z]+$
Zulässig
apiVersion: v1 kind: Pod metadata: name: good-pod namespace: default spec: containers: - image: nginx name: nginx resources: limits: cpu: 100m memory: 4Gi requests: cpu: 50m memory: 4Gi
apiVersion: v1 kind: Pod metadata: name: exempt-pod namespace: default spec: containers: - image: auto name: istio-proxy resources: limits: cpu: 100m memory: 4Gi requests: cpu: 50m memory: 2Gi
Unzulässig
apiVersion: v1 kind: Pod metadata: name: bad-pod namespace: default spec: containers: - image: nginx name: nginx resources: limits: cpu: 100m memory: 4Gi requests: cpu: 50m memory: 2Gi
K8sNoEnvVarSecrets
Keine Secrets für Umgebungsvariablen v1.0.1
Verbindet Secrets als Umgebungsvariablen in Pod-Containerdefinitionen. Verwenden Sie stattdessen bereitgestellte Secret-Dateien in Datenvolumen: https://kubernetes.io/docs/concepts/configuration/secret/#using-secrets-as-files-from-a-pod
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sNoEnvVarSecrets
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
Beispiele
no-secrets-as-env-vars-sample
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sNoEnvVarSecrets metadata: name: no-secrets-as-env-vars-sample spec: enforcementAction: dryrun
Zulässig
apiVersion: v1 kind: Pod metadata: name: allowed-example spec: containers: - image: redis name: test volumeMounts: - mountPath: /etc/test name: test readOnly: true volumes: - name: test secret: secretName: mysecret
Unzulässig
apiVersion: v1 kind: Pod metadata: name: disallowed-example spec: containers: - env: - name: MY_PASSWORD valueFrom: secretKeyRef: key: password name: mysecret image: redis name: test
K8sNoExternalServices
Keine externen Dienste v1.0.3
Verhindert das Erstellen bekannter Ressourcen, die Arbeitslasten für externe IP-Adressen verfügbar machen. Dazu gehören Istio-Gateway-Ressourcen und Kubernetes Ingress-Ressourcen. Kubernetes-Dienste sind ebenfalls nur zulässig, wenn sie die folgenden Kriterien erfüllen: Jeder Dienst vom Typ LoadBalancer in Google Cloud muss eine "networking.gke.io/load-balancer-type": "Internal"-Annotation haben.
Jeder Dienst vom Typ LoadBalancer in AWS muss eine service.beta.kubernetes.io/aws-load-balancer-internal: "true-Annotation haben.
Alle "externen IP-Adressen" (außerhalb des Clusters), die an den Dienst gebunden sind, müssen Mitglied eines Bereichs internen CIDRs sein, wie für die Einschränkung angegeben.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sNoExternalServices
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# cloudPlatform <string>: The hosting cloud platform. Only `GCP` and `AWS`
# are supported currently.
cloudPlatform: <string>
# internalCIDRs <array>: A list of CIDRs that are only accessible
# internally, for example: `10.3.27.0/24`. Which IP ranges are
# internal-only is determined by the underlying network infrastructure.
internalCIDRs:
- <string>
Beispiele
no-external
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sNoExternalServices metadata: name: no-external spec: parameters: internalCIDRs: - 10.0.0.1/32
Zulässig
apiVersion: v1 kind: Service metadata: name: good-service namespace: default spec: externalIPs: - 10.0.0.1 ports: - port: 8888 protocol: TCP targetPort: 8888
apiVersion: v1 kind: Service metadata: annotations: networking.gke.io/load-balancer-type: Internal name: allowed-internal-load-balancer namespace: default spec: type: LoadBalancer
Nicht zugelassen
apiVersion: v1 kind: Service metadata: name: bad-service namespace: default spec: externalIPs: - 10.0.0.2 ports: - port: 8888 protocol: TCP targetPort: 8888
no-external-aws
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sNoExternalServices metadata: name: no-external-aws spec: parameters: cloudPlatform: AWS
Zulässig
apiVersion: v1 kind: Service metadata: annotations: service.beta.kubernetes.io/aws-load-balancer-internal: "true" name: good-aws-service namespace: default spec: type: LoadBalancer
Unzulässig
apiVersion: v1 kind: Service metadata: annotations: cloud.google.com/load-balancer-type: Internal name: bad-aws-service namespace: default spec: type: LoadBalancer
K8sPSPAllowPrivilegeEscalationContainer
Rechteausweitung in Container v1.0.1 zulassen
Steuert die Einschränkung der Eskalation auf Root-Berechtigungen. Entspricht dem Feld allowPrivilegeEscalation in einer PodSecurityPolicy. Weitere Informationen finden Sie unter https://kubernetes.io/docs/concepts/policy/pod-security-policy/#privilege-escalation
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sPSPAllowPrivilegeEscalationContainer
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# exemptImages <array>: Any container that uses an image that matches an
# entry in this list will be excluded from enforcement. Prefix-matching can
# be signified with `*`. For example: `my-image-*`. It is recommended that
# users use the fully-qualified Docker image name (e.g. start with a domain
# name) in order to avoid unexpectedly exempting images from an untrusted
# repository.
exemptImages:
- <string>
Beispiele
psp-allow-privilege-escalation-container-sample
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sPSPAllowPrivilegeEscalationContainer metadata: name: psp-allow-privilege-escalation-container-sample spec: match: kinds: - apiGroups: - "" kinds: - Pod
Zulässig
apiVersion: v1 kind: Pod metadata: labels: app: nginx-privilege-escalation name: nginx-privilege-escalation-allowed spec: containers: - image: nginx name: nginx securityContext: allowPrivilegeEscalation: false
Unzulässig
apiVersion: v1 kind: Pod metadata: labels: app: nginx-privilege-escalation name: nginx-privilege-escalation-disallowed spec: containers: - image: nginx name: nginx securityContext: allowPrivilegeEscalation: true
apiVersion: v1 kind: Pod metadata: labels: app: nginx-privilege-escalation name: nginx-privilege-escalation-disallowed spec: ephemeralContainers: - image: nginx name: nginx securityContext: allowPrivilegeEscalation: true
K8sPSPAllowedUsers
Zulässige Nutzer v1.0.2
Steuert die Nutzer- und Gruppen-IDs des Containers und einige Volumes. Entspricht den Feldern runAsUser, runAsGroup, supplementalGroups und fsGroup in einer PodSecurityPolicy. Weitere Informationen finden Sie unter https://kubernetes.io/docs/concepts/policy/pod-security-policy/#users-and-groups
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sPSPAllowedUsers
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# exemptImages <array>: Any container that uses an image that matches an
# entry in this list will be excluded from enforcement. Prefix-matching can
# be signified with `*`. For example: `my-image-*`. It is recommended that
# users use the fully-qualified Docker image name (e.g. start with a domain
# name) in order to avoid unexpectedly exempting images from an untrusted
# repository.
exemptImages:
- <string>
# fsGroup <object>: Controls the fsGroup values that are allowed in a Pod
# or container-level SecurityContext.
fsGroup:
# ranges <array>: A list of group ID ranges affected by the rule.
ranges:
# <list item: object>: The range of group IDs affected by the rule.
- # max <integer>: The maximum group ID in the range, inclusive.
max: <integer>
# min <integer>: The minimum group ID in the range, inclusive.
min: <integer>
# rule <string>: A strategy for applying the fsGroup restriction.
# Allowed Values: MustRunAs, MayRunAs, RunAsAny
rule: <string>
# runAsGroup <object>: Controls which group ID values are allowed in a Pod
# or container-level SecurityContext.
runAsGroup:
# ranges <array>: A list of group ID ranges affected by the rule.
ranges:
# <list item: object>: The range of group IDs affected by the rule.
- # max <integer>: The maximum group ID in the range, inclusive.
max: <integer>
# min <integer>: The minimum group ID in the range, inclusive.
min: <integer>
# rule <string>: A strategy for applying the runAsGroup restriction.
# Allowed Values: MustRunAs, MayRunAs, RunAsAny
rule: <string>
# runAsUser <object>: Controls which user ID values are allowed in a Pod or
# container-level SecurityContext.
runAsUser:
# ranges <array>: A list of user ID ranges affected by the rule.
ranges:
# <list item: object>: The range of user IDs affected by the rule.
- # max <integer>: The maximum user ID in the range, inclusive.
max: <integer>
# min <integer>: The minimum user ID in the range, inclusive.
min: <integer>
# rule <string>: A strategy for applying the runAsUser restriction.
# Allowed Values: MustRunAs, MustRunAsNonRoot, RunAsAny
rule: <string>
# supplementalGroups <object>: Controls the supplementalGroups values that
# are allowed in a Pod or container-level SecurityContext.
supplementalGroups:
# ranges <array>: A list of group ID ranges affected by the rule.
ranges:
# <list item: object>: The range of group IDs affected by the rule.
- # max <integer>: The maximum group ID in the range, inclusive.
max: <integer>
# min <integer>: The minimum group ID in the range, inclusive.
min: <integer>
# rule <string>: A strategy for applying the supplementalGroups
# restriction.
# Allowed Values: MustRunAs, MayRunAs, RunAsAny
rule: <string>
Beispiele
psp-pods-allowed-user-ranges
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sPSPAllowedUsers metadata: name: psp-pods-allowed-user-ranges spec: match: kinds: - apiGroups: - "" kinds: - Pod parameters: fsGroup: ranges: - max: 200 min: 100 rule: MustRunAs runAsGroup: ranges: - max: 200 min: 100 rule: MustRunAs runAsUser: ranges: - max: 200 min: 100 rule: MustRunAs supplementalGroups: ranges: - max: 200 min: 100 rule: MustRunAs
Zulässig
apiVersion: v1 kind: Pod metadata: labels: app: nginx-users name: nginx-users-allowed spec: containers: - image: nginx name: nginx securityContext: runAsGroup: 199 runAsUser: 199 securityContext: fsGroup: 199 supplementalGroups: - 199
Nicht zugelassen
apiVersion: v1 kind: Pod metadata: labels: app: nginx-users name: nginx-users-disallowed spec: containers: - image: nginx name: nginx securityContext: runAsGroup: 250 runAsUser: 250 securityContext: fsGroup: 250 supplementalGroups: - 250
apiVersion: v1 kind: Pod metadata: labels: app: nginx-users name: nginx-users-disallowed spec: ephemeralContainers: - image: nginx name: nginx securityContext: runAsGroup: 250 runAsUser: 250 securityContext: fsGroup: 250 supplementalGroups: - 250
K8sPSPAppArmor
App Armor v1.0.0
Konfiguriert eine Zulassungsliste von AppArmor-Profilen für die Verwendung durch Container. Das entspricht bestimmten Annotationen, die auf eine PodSecurityPolicy angewendet werden. Informationen zu AppArmor finden Sie unter https://kubernetes.io/docs/tutorials/clusters/apparmor/
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sPSPAppArmor
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# allowedProfiles <array>: An array of AppArmor profiles. Examples:
# `runtime/default`, `unconfined`.
allowedProfiles:
- <string>
# exemptImages <array>: Any container that uses an image that matches an
# entry in this list will be excluded from enforcement. Prefix-matching can
# be signified with `*`. For example: `my-image-*`. It is recommended that
# users use the fully-qualified Docker image name (e.g. start with a domain
# name) in order to avoid unexpectedly exempting images from an untrusted
# repository.
exemptImages:
- <string>
Beispiele
psp-apparmor
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sPSPAppArmor metadata: name: psp-apparmor spec: match: kinds: - apiGroups: - "" kinds: - Pod parameters: allowedProfiles: - runtime/default
Zulässig
apiVersion: v1 kind: Pod metadata: annotations: container.apparmor.security.beta.kubernetes.io/nginx: runtime/default labels: app: nginx-apparmor name: nginx-apparmor-allowed spec: containers: - image: nginx name: nginx
Nicht zugelassen
apiVersion: v1 kind: Pod metadata: annotations: container.apparmor.security.beta.kubernetes.io/nginx: unconfined labels: app: nginx-apparmor name: nginx-apparmor-disallowed spec: containers: - image: nginx name: nginx
apiVersion: v1 kind: Pod metadata: annotations: container.apparmor.security.beta.kubernetes.io/nginx: unconfined labels: app: nginx-apparmor name: nginx-apparmor-disallowed spec: ephemeralContainers: - image: nginx name: nginx
K8sPSPAutomountServiceAccountTokenPod
Automount-Dienstkonto-Token für Pod v1.0.1
Steuert die Fähigkeit jedes Pods, das automountServiceAccountToken zu aktivieren.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sPSPAutomountServiceAccountTokenPod
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
<object>
Beispiele
psp-automount-serviceaccount-token-pod
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sPSPAutomountServiceAccountTokenPod metadata: name: psp-automount-serviceaccount-token-pod spec: match: excludedNamespaces: - kube-system kinds: - apiGroups: - "" kinds: - Pod
Zulässig
apiVersion: v1 kind: Pod metadata: labels: app: nginx-not-automountserviceaccounttoken name: nginx-automountserviceaccounttoken-allowed spec: automountServiceAccountToken: false containers: - image: nginx name: nginx
Unzulässig
apiVersion: v1 kind: Pod metadata: labels: app: nginx-automountserviceaccounttoken name: nginx-automountserviceaccounttoken-disallowed spec: automountServiceAccountToken: true containers: - image: nginx name: nginx
K8sPSPCapabilities
Funktionen Version 1.0.2
Steuert Linux-Funktionen für Container. Entspricht den Feldern allowedCapabilities und requiredDropCapabilities in einer PodSecurityPolicy. Weitere Informationen finden Sie unter https://kubernetes.io/docs/concepts/policy/pod-security-policy/#capabilities
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sPSPCapabilities
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# allowedCapabilities <array>: A list of Linux capabilities that can be
# added to a container.
allowedCapabilities:
- <string>
# exemptImages <array>: Any container that uses an image that matches an
# entry in this list will be excluded from enforcement. Prefix-matching can
# be signified with `*`. For example: `my-image-*`. It is recommended that
# users use the fully-qualified Docker image name (e.g. start with a domain
# name) in order to avoid unexpectedly exempting images from an untrusted
# repository.
exemptImages:
- <string>
# requiredDropCapabilities <array>: A list of Linux capabilities that are
# required to be dropped from a container.
requiredDropCapabilities:
- <string>
Beispiele
capabilities-demo
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sPSPCapabilities metadata: name: capabilities-demo spec: match: kinds: - apiGroups: - "" kinds: - Pod namespaces: - default parameters: allowedCapabilities: - something requiredDropCapabilities: - must_drop
Zulässig
apiVersion: v1 kind: Pod metadata: labels: owner: me.agilebank.demo name: opa-allowed spec: containers: - args: - run - --server - --addr=localhost:8080 image: openpolicyagent/opa:0.9.2 name: opa resources: limits: cpu: 100m memory: 30Mi securityContext: capabilities: add: - something drop: - must_drop - another_one
Nicht zugelassen
apiVersion: v1 kind: Pod metadata: labels: owner: me.agilebank.demo name: opa-disallowed spec: containers: - args: - run - --server - --addr=localhost:8080 image: openpolicyagent/opa:0.9.2 name: opa resources: limits: cpu: 100m memory: 30Mi securityContext: capabilities: add: - disallowedcapability
apiVersion: v1 kind: Pod metadata: labels: owner: me.agilebank.demo name: opa-disallowed spec: ephemeralContainers: - args: - run - --server - --addr=localhost:8080 image: openpolicyagent/opa:0.9.2 name: opa resources: limits: cpu: 100m memory: 30Mi securityContext: capabilities: add: - disallowedcapability
K8sPSPFSGroup
FS Group v1.0.2
Steuert die Zuweisung einer FSGroup, die die Volumes des Pods besitzt. Entspricht dem Feld fsGroup in einer PodSecurityPolicy. Weitere Informationen finden Sie unter https://kubernetes.io/docs/concepts/policy/pod-security-policy/#volumes-and-file-systems
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sPSPFSGroup
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# ranges <array>: GID ranges affected by the rule.
ranges:
- # max <integer>: The maximum GID in the range, inclusive.
max: <integer>
# min <integer>: The minimum GID in the range, inclusive.
min: <integer>
# rule <string>: An FSGroup rule name.
# Allowed Values: MayRunAs, MustRunAs, RunAsAny
rule: <string>
Beispiele
psp-fsgroup
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sPSPFSGroup metadata: name: psp-fsgroup spec: match: kinds: - apiGroups: - "" kinds: - Pod parameters: ranges: - max: 1000 min: 1 rule: MayRunAs
Zulässig
apiVersion: v1 kind: Pod metadata: name: fsgroup-disallowed spec: containers: - command: - sh - -c - sleep 1h image: busybox name: fsgroup-demo volumeMounts: - mountPath: /data/demo name: fsgroup-demo-vol securityContext: fsGroup: 500 volumes: - emptyDir: {} name: fsgroup-demo-vol
Nicht zugelassen
apiVersion: v1 kind: Pod metadata: name: fsgroup-disallowed spec: containers: - command: - sh - -c - sleep 1h image: busybox name: fsgroup-demo volumeMounts: - mountPath: /data/demo name: fsgroup-demo-vol securityContext: fsGroup: 2000 volumes: - emptyDir: {} name: fsgroup-demo-vol
K8sPSPFlexVolumes
FlexVolumes Version 1.0.1
Steuert die Zulassungsliste der FlexVolume-Treiber. Entspricht dem Feld allowedFlexVolumes in PodSecurityPolicy. Weitere Informationen finden Sie unter https://kubernetes.io/docs/concepts/policy/pod-security-policy/#flexvolume-drivers.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sPSPFlexVolumes
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# allowedFlexVolumes <array>: An array of AllowedFlexVolume objects.
allowedFlexVolumes:
- # driver <string>: The name of the FlexVolume driver.
driver: <string>
Beispiele
psp-flexvolume-drivers
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sPSPFlexVolumes metadata: name: psp-flexvolume-drivers spec: match: kinds: - apiGroups: - "" kinds: - Pod parameters: allowedFlexVolumes: - driver: example/lvm - driver: example/cifs
Zulässig
apiVersion: v1 kind: Pod metadata: labels: app: nginx-flexvolume-driver name: nginx-flexvolume-driver-allowed spec: containers: - image: nginx name: nginx volumeMounts: - mountPath: /test name: test-volume readOnly: true volumes: - flexVolume: driver: example/lvm name: test-volume
Unzulässig
apiVersion: v1 kind: Pod metadata: labels: app: nginx-flexvolume-driver name: nginx-flexvolume-driver-disallowed spec: containers: - image: nginx name: nginx volumeMounts: - mountPath: /test name: test-volume readOnly: true volumes: - flexVolume: driver: example/testdriver name: test-volume
K8sPSPForbiddenSysctls
Forbidden Sysctls v1.1.3
Steuert das von Containern verwendete Profil sysctl. Entspricht den Feldern allowedUnsafeSysctls und forbiddenSysctls in einer PodSecurityPolicy. Wenn angegeben, wird jeder sysctl, der nicht im allowedSysctls-Parameter enthalten ist, als unzulässig betrachtet. Der Parameter forbiddenSysctls hat Vorrang vor dem Parameter allowedSysctls. Weitere Informationen finden Sie unter https://kubernetes.io/docs/tasks/administer-cluster/sysctl-cluster/
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sPSPForbiddenSysctls
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# allowedSysctls <array>: An allow-list of sysctls. `*` allows all sysctls
# not listed in the `forbiddenSysctls` parameter.
allowedSysctls:
- <string>
# forbiddenSysctls <array>: A disallow-list of sysctls. `*` forbids all
# sysctls.
forbiddenSysctls:
- <string>
Beispiele
psp-forbidden-sysctls
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sPSPForbiddenSysctls metadata: name: psp-forbidden-sysctls spec: match: kinds: - apiGroups: - "" kinds: - Pod parameters: allowedSysctls: - '*' forbiddenSysctls: - kernel.*
Zulässig
apiVersion: v1 kind: Pod metadata: labels: app: nginx-forbidden-sysctls name: nginx-forbidden-sysctls-disallowed spec: containers: - image: nginx name: nginx securityContext: sysctls: - name: net.core.somaxconn value: "1024"
Nicht zugelassen
apiVersion: v1 kind: Pod metadata: labels: app: nginx-forbidden-sysctls name: nginx-forbidden-sysctls-disallowed spec: containers: - image: nginx name: nginx securityContext: sysctls: - name: kernel.msgmax value: "65536" - name: net.core.somaxconn value: "1024"
K8sPSPHostFilesystem
Hostdateisystem v1.0.2
Steuert die Nutzung des Hostdateisystems. Entspricht dem Feld allowedHostPaths in einer PodSecurityPolicy. Weitere Informationen finden Sie unter https://kubernetes.io/docs/concepts/policy/pod-security-policy/#volumes-and-file-systems
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sPSPHostFilesystem
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# allowedHostPaths <array>: An array of hostpath objects, representing
# paths and read/write configuration.
allowedHostPaths:
- # pathPrefix <string>: The path prefix that the host volume must
# match.
pathPrefix: <string>
# readOnly <boolean>: when set to true, any container volumeMounts
# matching the pathPrefix must include `readOnly: true`.
readOnly: <boolean>
Beispiele
psp-host-filesystem
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sPSPHostFilesystem metadata: name: psp-host-filesystem spec: match: kinds: - apiGroups: - "" kinds: - Pod parameters: allowedHostPaths: - pathPrefix: /foo readOnly: true
Zulässig
apiVersion: v1 kind: Pod metadata: labels: app: nginx-host-filesystem-disallowed name: nginx-host-filesystem spec: containers: - image: nginx name: nginx volumeMounts: - mountPath: /cache name: cache-volume readOnly: true volumes: - hostPath: path: /foo/bar name: cache-volume
Nicht zugelassen
apiVersion: v1 kind: Pod metadata: labels: app: nginx-host-filesystem-disallowed name: nginx-host-filesystem spec: containers: - image: nginx name: nginx volumeMounts: - mountPath: /cache name: cache-volume readOnly: true volumes: - hostPath: path: /tmp name: cache-volume
apiVersion: v1 kind: Pod metadata: labels: app: nginx-host-filesystem-disallowed name: nginx-host-filesystem spec: ephemeralContainers: - image: nginx name: nginx volumeMounts: - mountPath: /cache name: cache-volume readOnly: true volumes: - hostPath: path: /tmp name: cache-volume
K8sPSPHostNamespace
Host Namespace v1.0.1
Verhindert die Freigabe von Host-PID- und IPC-Namespaces nach Pod-Containern. Entspricht den Feldern hostPID und hostIPC in einer PodSecurityPolicy. Weitere Informationen finden Sie unter https://kubernetes.io/docs/concepts/policy/pod-security-policy/#host-namespaces
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sPSPHostNamespace
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
<object>
Beispiele
psp-host-namespace-sample
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sPSPHostNamespace metadata: name: psp-host-namespace-sample spec: match: kinds: - apiGroups: - "" kinds: - Pod
Zulässig
apiVersion: v1 kind: Pod metadata: labels: app: nginx-host-namespace name: nginx-host-namespace-allowed spec: containers: - image: nginx name: nginx hostIPC: false hostPID: false
Unzulässig
apiVersion: v1 kind: Pod metadata: labels: app: nginx-host-namespace name: nginx-host-namespace-disallowed spec: containers: - image: nginx name: nginx hostIPC: true hostPID: true
K8sPSPHostNetworkingPorts
Hostnetzwerkports v1.0.2
Steuert die Nutzung des Hostnetzwerk-Namespace nach Pod-Containern. Bestimmte Ports müssen angegeben werden. Entspricht den Feldern hostNetwork und hostPorts in einer PodSecurityPolicy. Weitere Informationen finden Sie unter https://kubernetes.io/docs/concepts/policy/pod-security-policy/#host-namespaces
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sPSPHostNetworkingPorts
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# exemptImages <array>: Any container that uses an image that matches an
# entry in this list will be excluded from enforcement. Prefix-matching can
# be signified with `*`. For example: `my-image-*`. It is recommended that
# users use the fully-qualified Docker image name (e.g. start with a domain
# name) in order to avoid unexpectedly exempting images from an untrusted
# repository.
exemptImages:
- <string>
# hostNetwork <boolean>: Determines if the policy allows the use of
# HostNetwork in the pod spec.
hostNetwork: <boolean>
# max <integer>: The end of the allowed port range, inclusive.
max: <integer>
# min <integer>: The start of the allowed port range, inclusive.
min: <integer>
Beispiele
psp-host-network-ports-sample
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sPSPHostNetworkingPorts metadata: name: psp-host-network-ports-sample spec: match: kinds: - apiGroups: - "" kinds: - Pod parameters: hostNetwork: true max: 9000 min: 80
Zulässig
apiVersion: v1 kind: Pod metadata: labels: app: nginx-host-networking-ports name: nginx-host-networking-ports-allowed spec: containers: - image: nginx name: nginx ports: - containerPort: 9000 hostPort: 80 hostNetwork: false
Unzulässig
apiVersion: v1 kind: Pod metadata: labels: app: nginx-host-networking-ports name: nginx-host-networking-ports-disallowed spec: containers: - image: nginx name: nginx ports: - containerPort: 9001 hostPort: 9001 hostNetwork: true
apiVersion: v1 kind: Pod metadata: labels: app: nginx-host-networking-ports name: nginx-host-networking-ports-disallowed spec: ephemeralContainers: - image: nginx name: nginx ports: - containerPort: 9001 hostPort: 9001 hostNetwork: true
K8sPSPPrivilegedContainer
Privilegierter Container v1.0.1
Steuert die Fähigkeit eines Containers, den privilegierten Modus zu aktivieren. Entspricht dem Feld privileged in einer PodSecurityPolicy. Weitere Informationen finden Sie unter https://kubernetes.io/docs/concepts/policy/pod-security-policy/#privileged.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sPSPPrivilegedContainer
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# exemptImages <array>: Any container that uses an image that matches an
# entry in this list will be excluded from enforcement. Prefix-matching can
# be signified with `*`. For example: `my-image-*`. It is recommended that
# users use the fully-qualified Docker image name (e.g. start with a domain
# name) in order to avoid unexpectedly exempting images from an untrusted
# repository.
exemptImages:
- <string>
Beispiele
psp-privileged-container-sample
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sPSPPrivilegedContainer metadata: name: psp-privileged-container-sample spec: match: excludedNamespaces: - kube-system kinds: - apiGroups: - "" kinds: - Pod
Zulässig
apiVersion: v1 kind: Pod metadata: labels: app: nginx-privileged name: nginx-privileged-allowed spec: containers: - image: nginx name: nginx securityContext: privileged: false
Unzulässig
apiVersion: v1 kind: Pod metadata: labels: app: nginx-privileged name: nginx-privileged-disallowed spec: containers: - image: nginx name: nginx securityContext: privileged: true
apiVersion: v1 kind: Pod metadata: labels: app: nginx-privileged name: nginx-privileged-disallowed spec: ephemeralContainers: - image: nginx name: nginx securityContext: privileged: true
K8sPSPProcMount
Proc Mount v1.0.3
Steuert die zulässigen procMount-Typen für den Container. Entspricht dem Feld allowedProcMountTypes in einer PodSecurityPolicy. Weitere Informationen finden Sie unter https://kubernetes.io/docs/concepts/policy/pod-security-policy/#allowprocmounttypes.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sPSPProcMount
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# exemptImages <array>: Any container that uses an image that matches an
# entry in this list will be excluded from enforcement. Prefix-matching can
# be signified with `*`. For example: `my-image-*`. It is recommended that
# users use the fully-qualified Docker image name (e.g. start with a domain
# name) in order to avoid unexpectedly exempting images from an untrusted
# repository.
exemptImages:
- <string>
# procMount <string>: Defines the strategy for the security exposure of
# certain paths in `/proc` by the container runtime. Setting to `Default`
# uses the runtime defaults, where `Unmasked` bypasses the default
# behavior.
# Allowed Values: Default, Unmasked
procMount: <string>
Beispiele
psp-proc-mount
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sPSPProcMount metadata: name: psp-proc-mount spec: match: kinds: - apiGroups: - "" kinds: - Pod parameters: procMount: Default
Zulässig
apiVersion: v1 kind: Pod metadata: labels: app: nginx-proc-mount name: nginx-proc-mount-disallowed spec: containers: - image: nginx name: nginx securityContext: procMount: Default
Nicht zugelassen
apiVersion: v1 kind: Pod metadata: labels: app: nginx-proc-mount name: nginx-proc-mount-disallowed spec: containers: - image: nginx name: nginx securityContext: procMount: Unmasked
apiVersion: v1 kind: Pod metadata: labels: app: nginx-proc-mount name: nginx-proc-mount-disallowed spec: ephemeralContainers: - image: nginx name: nginx securityContext: procMount: Unmasked
K8sPSPReadOnlyRootFilesystem
Schreibgeschütztes Root-Dateisystem v1.0.1
Erfordert die Verwendung eines schreibgeschützten Root-Dateisystems durch Pod-Container. Entspricht dem Feld readOnlyRootFilesystem in einer PodSecurityPolicy. Weitere Informationen finden Sie unter https://kubernetes.io/docs/concepts/policy/pod-security-policy/#volumes-and-file-systems
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sPSPReadOnlyRootFilesystem
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# exemptImages <array>: Any container that uses an image that matches an
# entry in this list will be excluded from enforcement. Prefix-matching can
# be signified with `*`. For example: `my-image-*`. It is recommended that
# users use the fully-qualified Docker image name (e.g. start with a domain
# name) in order to avoid unexpectedly exempting images from an untrusted
# repository.
exemptImages:
- <string>
Beispiele
psp-readonlyrootfilesystem
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sPSPReadOnlyRootFilesystem metadata: name: psp-readonlyrootfilesystem spec: match: kinds: - apiGroups: - "" kinds: - Pod
Zulässig
apiVersion: v1 kind: Pod metadata: labels: app: nginx-readonlyrootfilesystem name: nginx-readonlyrootfilesystem-allowed spec: containers: - image: nginx name: nginx securityContext: readOnlyRootFilesystem: true
Unzulässig
apiVersion: v1 kind: Pod metadata: labels: app: nginx-readonlyrootfilesystem name: nginx-readonlyrootfilesystem-disallowed spec: containers: - image: nginx name: nginx securityContext: readOnlyRootFilesystem: false
apiVersion: v1 kind: Pod metadata: labels: app: nginx-readonlyrootfilesystem name: nginx-readonlyrootfilesystem-disallowed spec: ephemeralContainers: - image: nginx name: nginx securityContext: readOnlyRootFilesystem: false
K8sPSPSELinuxV2
SELinux V2 v1.0.2
Definiert eine Zulassungsliste von seLinuxOptions-Konfigurationen für Pod-Container. Entspricht einer PodSecurityPolicy, die SELinux-Konfigurationen erfordert. Weitere Informationen finden Sie unter https://kubernetes.io/docs/concepts/policy/pod-security-policy/#selinux
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sPSPSELinuxV2
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# allowedSELinuxOptions <array>: An allow-list of SELinux options
# configurations.
allowedSELinuxOptions:
# <list item: object>: An allowed configuration of SELinux options for a
# pod container.
- # level <string>: An SELinux level.
level: <string>
# role <string>: An SELinux role.
role: <string>
# type <string>: An SELinux type.
type: <string>
# user <string>: An SELinux user.
user: <string>
# exemptImages <array>: Any container that uses an image that matches an
# entry in this list will be excluded from enforcement. Prefix-matching can
# be signified with `*`. For example: `my-image-*`. It is recommended that
# users use the fully-qualified Docker image name (e.g. start with a domain
# name) in order to avoid unexpectedly exempting images from an untrusted
# repository.
exemptImages:
- <string>
Beispiele
psp-selinux-v2
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sPSPSELinuxV2 metadata: name: psp-selinux-v2 spec: match: kinds: - apiGroups: - "" kinds: - Pod parameters: allowedSELinuxOptions: - level: s0:c123,c456 role: object_r type: svirt_sandbox_file_t user: system_u
Zulässig
apiVersion: v1 kind: Pod metadata: labels: app: nginx-selinux name: nginx-selinux-allowed spec: containers: - image: nginx name: nginx securityContext: seLinuxOptions: level: s0:c123,c456 role: object_r type: svirt_sandbox_file_t user: system_u
Nicht zugelassen
apiVersion: v1 kind: Pod metadata: labels: app: nginx-selinux name: nginx-selinux-disallowed spec: containers: - image: nginx name: nginx securityContext: seLinuxOptions: level: s1:c234,c567 role: sysadm_r type: svirt_lxc_net_t user: sysadm_u
apiVersion: v1 kind: Pod metadata: labels: app: nginx-selinux name: nginx-selinux-disallowed spec: ephemeralContainers: - image: nginx name: nginx securityContext: seLinuxOptions: level: s1:c234,c567 role: sysadm_r type: svirt_lxc_net_t user: sysadm_u
K8sPSPSeccomp
Seccomp v1.0.1
Steuert das von Containern verwendete seccomp-Profil. Entspricht der Annotation seccomp.security.alpha.kubernetes.io/allowedProfileNames in einer PodSecurityPolicy. Weitere Informationen finden Sie unter https://kubernetes.io/docs/concepts/policy/pod-security-policy/#seccomp
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sPSPSeccomp
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# allowedLocalhostFiles <array>: When using securityContext naming scheme
# for seccomp and including `Localhost` this array holds the allowed
# profile JSON files. Putting a `*` in this array will allows all JSON
# files to be used. This field is required to allow `Localhost` in
# securityContext as with an empty list it will block.
allowedLocalhostFiles:
- <string>
# allowedProfiles <array>: An array of allowed profile values for seccomp
# on Pods/Containers. Can use the annotation naming scheme:
# `runtime/default`, `docker/default`, `unconfined` and/or
# `localhost/some-profile.json`. The item `localhost/*` will allow any
# localhost based profile. Can also use the securityContext naming scheme:
# `RuntimeDefault`, `Unconfined` and/or `Localhost`. For securityContext
# `Localhost`, use the parameter `allowedLocalhostProfiles` to list the
# allowed profile JSON files. The policy code will translate between the
# two schemes so it is not necessary to use both. Putting a `*` in this
# array allows all Profiles to be used. This field is required since with
# an empty list this policy will block all workloads.
allowedProfiles:
- <string>
# exemptImages <array>: Any container that uses an image that matches an
# entry in this list will be excluded from enforcement. Prefix-matching can
# be signified with `*`. For example: `my-image-*`. It is recommended that
# users use the fully-qualified Docker image name (e.g. start with a domain
# name) in order to avoid unexpectedly exempting images from an untrusted
# repository.
exemptImages:
- <string>
Beispiele
psp-seccomp
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sPSPSeccomp metadata: name: psp-seccomp spec: match: kinds: - apiGroups: - "" kinds: - Pod parameters: allowedProfiles: - runtime/default - docker/default
Zulässig
apiVersion: v1 kind: Pod metadata: annotations: container.seccomp.security.alpha.kubernetes.io/nginx: runtime/default labels: app: nginx-seccomp name: nginx-seccomp-allowed spec: containers: - image: nginx name: nginx
apiVersion: v1 kind: Pod metadata: annotations: seccomp.security.alpha.kubernetes.io/pod: runtime/default labels: app: nginx-seccomp name: nginx-seccomp-allowed2 spec: containers: - image: nginx name: nginx
Unzulässig
apiVersion: v1 kind: Pod metadata: annotations: seccomp.security.alpha.kubernetes.io/pod: unconfined labels: app: nginx-seccomp name: nginx-seccomp-disallowed2 spec: containers: - image: nginx name: nginx
apiVersion: v1 kind: Pod metadata: annotations: container.seccomp.security.alpha.kubernetes.io/nginx: unconfined labels: app: nginx-seccomp name: nginx-seccomp-disallowed spec: containers: - image: nginx name: nginx
apiVersion: v1 kind: Pod metadata: annotations: container.seccomp.security.alpha.kubernetes.io/nginx: unconfined labels: app: nginx-seccomp name: nginx-seccomp-disallowed spec: ephemeralContainers: - image: nginx name: nginx
K8sPSPVolumeTypes
Volume-Typen v1.0.2
Beschränkt die bereitstellbaren Volume-Typen auf die vom Nutzer angegebenen Volumes. Entspricht dem Feld volumes in einer PodSecurityPolicy. Weitere Informationen finden Sie unter https://kubernetes.io/docs/concepts/policy/pod-security-policy/#volumes-and-file-systems
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sPSPVolumeTypes
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# volumes <array>: `volumes` is an array of volume types. All volume types
# can be enabled using `*`.
volumes:
- <string>
Beispiele
psp-volume-types
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sPSPVolumeTypes metadata: name: psp-volume-types spec: match: kinds: - apiGroups: - "" kinds: - Pod parameters: volumes: - configMap - emptyDir - projected - secret - downwardAPI - persistentVolumeClaim - flexVolume
Zulässig
apiVersion: v1 kind: Pod metadata: labels: app: nginx-volume-types name: nginx-volume-types-allowed spec: containers: - image: nginx name: nginx volumeMounts: - mountPath: /cache name: cache-volume - image: nginx name: nginx2 volumeMounts: - mountPath: /cache2 name: demo-vol volumes: - emptyDir: {} name: cache-volume - emptyDir: {} name: demo-vol
Nicht zugelassen
apiVersion: v1 kind: Pod metadata: labels: app: nginx-volume-types name: nginx-volume-types-disallowed spec: containers: - image: nginx name: nginx volumeMounts: - mountPath: /cache name: cache-volume - image: nginx name: nginx2 volumeMounts: - mountPath: /cache2 name: demo-vol volumes: - hostPath: path: /tmp name: cache-volume - emptyDir: {} name: demo-vol
K8sPSPWindowsHostProcess
Schränkt Windows-HostProcess-Container/-Pods ein. V1.0.0
Schränkt die Ausführung von Windows HostProcess-Containern/-Pods ein. Weitere Informationen finden Sie unter https://kubernetes.io/docs/tasks/configure-pod-container/create-hostprocess-pod/.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sPSPWindowsHostProcess
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
Beispiele
restrict-windows-hostprocess
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sPSPWindowsHostProcess metadata: name: restrict-windows-hostprocess spec: match: kinds: - apiGroups: - "" kinds: - Pod
Zulässig
apiVersion: v1 kind: Pod metadata: name: nanoserver-ping-loop spec: containers: - command: - ping - -t - 127.0.0.1 image: mcr.microsoft.com/windows/nanoserver:1809 name: ping-loop nodeSelector: kubernetes.io/os: windows
Nicht zugelassen
apiVersion: v1 kind: Pod metadata: name: nanoserver-ping-loop-hostprocess-container spec: containers: - command: - ping - -t - 127.0.0.1 image: mcr.microsoft.com/windows/nanoserver:1809 name: ping-test securityContext: windowsOptions: hostProcess: true runAsUserName: NT AUTHORITY\SYSTEM hostNetwork: true nodeSelector: kubernetes.io/os: windows
apiVersion: v1 kind: Pod metadata: name: nanoserver-ping-loop-hostprocess-pod spec: containers: - command: - ping - -t - 127.0.0.1 image: mcr.microsoft.com/windows/nanoserver:1809 name: ping-test hostNetwork: true nodeSelector: kubernetes.io/os: windows securityContext: windowsOptions: hostProcess: true runAsUserName: NT AUTHORITY\SYSTEM
K8sPSSRunAsNonRoot
Erfordert, dass Container als Nutzer ohne Root-Berechtigung ausgeführt werden. V1.0.0
Erfordert, dass Container als Nutzer ohne Root-Berechtigung ausgeführt werden. Weitere Informationen finden Sie unter https://kubernetes.io/docs/concepts/security/pod-security-standards/
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sPSSRunAsNonRoot
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
Beispiele
restrict-runasnonroot
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sPSSRunAsNonRoot metadata: name: restrict-runasnonroot spec: match: kinds: - apiGroups: - "" kinds: - Pod
Zulässig
apiVersion: v1 kind: Pod metadata: name: nginx-pod-allowed spec: containers: - image: nginx name: nginx-container-allowed securityContext: runAsNonRoot: true securityContext: runAsNonRoot: true
apiVersion: v1 kind: Pod metadata: name: nginx-allowed spec: containers: - image: nginx name: nginx-allowed securityContext: runAsNonRoot: true
Nicht zugelassen
apiVersion: v1 kind: Pod metadata: name: nginx-pod-allowed spec: containers: - image: nginx name: nginx-container-disallowed securityContext: runAsNonRoot: false securityContext: runAsNonRoot: true
apiVersion: v1 kind: Pod metadata: name: nginx-pod-disallowed spec: containers: - image: nginx name: nginx-container-allowed securityContext: runAsNonRoot: true securityContext: runAsNonRoot: false
apiVersion: v1 kind: Pod metadata: name: nginx-pod-disallowed spec: containers: - image: nginx name: nginx-container-disallowed securityContext: runAsNonRoot: false
K8sPodDisruptionBudget
Budget für Pod-Störungen v1.0.3
Die folgenden Szenarien sollen bei der Bereitstellung von PodDisruptionBudgets oder Ressourcen, die die Replikatunterressource implementieren (z. B. Deployment, ReplicationController, ReplicaSet, StatefulSet), nicht zugelassen werden: 1. Bereitstellung von PodDisruptionBudgets mit .spec.maxUnavailable == 0. 2. Bereitstellung von PodDisruptionBudgets mit .spec.minAvailable == .spec.replicas der Ressource mit Replikatunterressource. Dies verhindert, dass PodDisruptionBudgets freiwillige Störungen wie das Leeren von Knoten blockieren. https://kubernetes.io/docs/concepts/workloads/Pods/Unterbrechungen/
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sPodDisruptionBudget
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
Referenzielle Einschränkung
Diese Einschränkung ist referenziell. Vor der Verwendung müssen Sie referenzielle Einschränkungen aktivieren und eine Konfiguration erstellen, die Policy Controller mitteilt, welche Arten von Objekten beobachtet werden sollen.
Ihr Policy Controller Config erfordert einen syncOnly-Eintrag ähnlich dem folgenden:
spec:
sync:
syncOnly:
- group: "policy"
version: "v1"
kind: "PodDisruptionBudget"
Beispiele
pod-distruption-budget
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sPodDisruptionBudget metadata: name: pod-distruption-budget spec: match: kinds: - apiGroups: - apps kinds: - Deployment - ReplicaSet - StatefulSet - apiGroups: - policy kinds: - PodDisruptionBudget - apiGroups: - "" kinds: - ReplicationController
Zulässig
apiVersion: policy/v1 kind: PodDisruptionBudget metadata: name: nginx-pdb-allowed namespace: default spec: maxUnavailable: 1 selector: matchLabels: foo: bar
apiVersion: apps/v1 kind: Deployment metadata: labels: app: nginx name: nginx-deployment-allowed-1 namespace: default spec: replicas: 3 selector: matchLabels: app: nginx example: allowed-deployment-1 template: metadata: labels: app: nginx example: allowed-deployment-1 spec: containers: - image: nginx:1.14.2 name: nginx ports: - containerPort: 80 --- # Referential Data apiVersion: policy/v1 kind: PodDisruptionBudget metadata: name: inventory-nginx-pdb-allowed-1 namespace: default spec: minAvailable: 2 selector: matchLabels: app: nginx example: allowed-deployment-1
apiVersion: apps/v1 kind: Deployment metadata: labels: app: nginx name: nginx-deployment-allowed-2 namespace: default spec: replicas: 3 selector: matchLabels: app: nginx example: allowed-deployment-2 template: metadata: labels: app: nginx example: allowed-deployment-2 spec: containers: - image: nginx:1.14.2 name: nginx ports: - containerPort: 80 --- # Referential Data apiVersion: policy/v1 kind: PodDisruptionBudget metadata: name: inventory-nginx-pdb-allowed-2 namespace: default spec: maxUnavailable: 1 selector: matchLabels: app: nginx example: allowed-deployment-2
apiVersion: apps/v1 kind: Deployment metadata: labels: app: nginx name: nginx-deployment-allowed-3 namespace: default spec: replicas: 3 selector: matchLabels: app: nginx example: allowed-deployment-3 template: metadata: labels: app: nginx example: allowed-deployment-3 spec: containers: - image: nginx:1.14.2 name: nginx ports: - containerPort: 80 --- # Referential Data apiVersion: policy/v1 kind: PodDisruptionBudget metadata: name: inventory-nginx-pdb-allowed-3 namespace: default spec: minAvailable: 2 selector: matchLabels: app: nginx
apiVersion: apps/v1 kind: Deployment metadata: labels: app: non-matching-nginx name: nginx-deployment-allowed-4 namespace: default spec: replicas: 1 selector: matchLabels: app: non-matching-nginx example: allowed-deployment-4 template: metadata: labels: app: non-matching-nginx example: allowed-deployment-4 spec: containers: - image: nginx:1.14.2 name: nginx ports: - containerPort: 80 --- # Referential Data apiVersion: policy/v1 kind: PodDisruptionBudget metadata: name: inventory-mongo-pdb-allowed-3 namespace: default spec: minAvailable: 2 selector: matchLabels: app: mongo example: non-matching-deployment-3
Nicht zugelassen
apiVersion: policy/v1 kind: PodDisruptionBudget metadata: name: nginx-pdb-disallowed namespace: default spec: maxUnavailable: 0 selector: matchLabels: foo: bar
apiVersion: apps/v1 kind: Deployment metadata: labels: app: nginx name: nginx-deployment-disallowed namespace: default spec: replicas: 3 selector: matchLabels: app: nginx example: disallowed-deployment template: metadata: labels: app: nginx example: disallowed-deployment spec: containers: - image: nginx:1.14.2 name: nginx ports: - containerPort: 80 --- # Referential Data apiVersion: policy/v1 kind: PodDisruptionBudget metadata: name: inventory-nginx-pdb-disallowed namespace: default spec: minAvailable: 3 selector: matchLabels: app: nginx example: disallowed-deployment
K8sPodResourcesBestPractices
Erfordert Container, die nicht auf Best-Effort-Basis und gemäß den Best Practices für Burstable v1.0.5
Erfordert, dass Container nicht auf Best-Effort-Basis (durch Festlegen von CPU- und Arbeitsspeicheranforderungen) und gemäß den Best Practices für Bursts funktionieren (die Arbeitsspeicherspeicheranfrage muss genau dem Limit entsprechen). Optional können Annotationsschlüssel konfiguriert werden, um das Überspringen der verschiedenen Validierungen zu ermöglichen.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sPodResourcesBestPractices
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# exemptImages <array>: A list of exempt Images.
exemptImages:
- <string>
# skipBestEffortValidationAnnotationKey <string>: Optional annotation key
# to skip best-effort container validation.
skipBestEffortValidationAnnotationKey: <string>
# skipBurstableValidationAnnotationKey <string>: Optional annotation key to
# skip burstable container validation.
skipBurstableValidationAnnotationKey: <string>
# skipResourcesBestPracticesValidationAnnotationKey <string>: Optional
# annotation key to skip both best-effort and burstable validation.
skipResourcesBestPracticesValidationAnnotationKey: <string>
Beispiele
gke-pod-resources-best-practices
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sPodResourcesBestPractices metadata: name: gke-pod-resources-best-practices spec: match: kinds: - apiGroups: - "" kinds: - Pod parameters: skipBestEffortValidationAnnotationKey: skip_besteffort_validation skipBurstableValidationAnnotationKey: skip_burstable_validation skipResourcesBestPracticesValidationAnnotationKey: skip_resources_best_practices_validation
Zulässig
apiVersion: v1 kind: Pod metadata: name: pod-setting-cpu-requests-memory-limits spec: containers: - image: nginx name: nginx resources: limits: memory: 500Mi requests: cpu: 250m
apiVersion: v1 kind: Pod metadata: name: pod-setting-limits-only spec: containers: - image: nginx name: nginx resources: limits: cpu: 250m memory: 100Mi
apiVersion: v1 kind: Pod metadata: name: pod-setting-requests-memory-limits spec: containers: - image: nginx name: nginx resources: limits: memory: 100Mi requests: cpu: 250m memory: 100Mi
apiVersion: v1 kind: Pod metadata: annotations: skip_besteffort_validation: "true" skip_burstable_validation: "true" skip_resources_best_practices_validation: "false" name: pod-skip-validation spec: containers: - image: nginx name: nginx
Nicht zugelassen
apiVersion: v1 kind: Pod metadata: name: pod-not-setting-cpu-burstable-on-memory spec: containers: - image: nginx name: nginx resources: limits: memory: 500Mi requests: memory: 100Mi
apiVersion: v1 kind: Pod metadata: name: pod-not-setting-requests spec: containers: - image: nginx name: nginx restartPolicy: OnFailure
apiVersion: v1 kind: Pod metadata: name: pod-setting-cpu-not-burstable-on-memory spec: containers: - image: nginx name: nginx resources: limits: memory: 500Mi requests: cpu: 250m memory: 100Mi
apiVersion: v1 kind: Pod metadata: name: pod-setting-memory-requests-cpu-limits spec: containers: - image: nginx name: nginx resources: limits: cpu: 30m requests: memory: 100Mi
apiVersion: v1 kind: Pod metadata: name: pod-setting-only-cpu-limits spec: containers: - image: nginx name: nginx resources: limits: cpu: 250m
apiVersion: v1 kind: Pod metadata: name: pod-setting-only-cpu-requests spec: containers: - image: nginx name: nginx resources: requests: cpu: 250m
apiVersion: v1 kind: Pod metadata: name: pod-setting-only-cpu spec: containers: - image: nginx name: nginx resources: limits: cpu: 500m requests: cpu: 250m
apiVersion: v1 kind: Pod metadata: name: pod-setting-only-memory-limits spec: containers: - image: nginx name: nginx resources: limits: memory: 250Mi
apiVersion: v1 kind: Pod metadata: name: pod-setting-only-memory-requests spec: containers: - image: nginx name: nginx resources: requests: memory: 100Mi
apiVersion: v1 kind: Pod metadata: name: pod-setting-only-memory spec: containers: - image: nginx name: nginx resources: limits: memory: 100Mi requests: memory: 100Mi
K8sPodsRequireSecurityContext
Pods erfordern Sicherheitskontext v1.1.1
Erfordert, dass alle Pods securityContext definieren. Erfordert, dass alle in Pods definierten Container einen SecurityContext auf Pod- oder Containerebene definieren.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sPodsRequireSecurityContext
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# exemptImages <array>: A list of exempt Images.
exemptImages:
- <string>
Beispiele
pods-require-security-context-sample
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sPodsRequireSecurityContext metadata: name: pods-require-security-context-sample spec: enforcementAction: dryrun parameters: exemptImages: - nginix-exempt - alpine*
Zulässig
apiVersion: v1 kind: Pod metadata: name: allowed-example spec: containers: - image: nginx name: nginx securityContext: runAsUser: 2000
apiVersion: v1 kind: Pod metadata: name: allowed-example-exemptImage spec: containers: - image: nginix-exempt name: nginx
apiVersion: v1 kind: Pod metadata: name: allowed-example-exemptImage-wildcard spec: containers: - image: alpine17 name: alpine
Unzulässig
apiVersion: v1 kind: Pod metadata: name: disallowed-example spec: containers: - image: nginx name: nginx
K8sProhibitRoleWildcardAccess
Platzhalterzugriff (Version 1.0.5) untersagen
Erfordert, dass Roles und ClusterRoles den Ressourcenzugriff nicht auf einen Platzhalterwert "" festlegen. Dies gilt nicht für die ausgenommenen Roles und ClusterRoles, die als Ausnahmen angegeben werden. Beschränkt den Platzhalterzugriff auf untergeordnete Ressourcen wie „/status“.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sProhibitRoleWildcardAccess
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# exemptions <object>: The list of exempted Roles and/or ClusterRoles name
# that are allowed to set resource access to a wildcard.
exemptions:
clusterRoles:
- # name <string>: The name of the ClusterRole to be exempted.
name: <string>
# regexMatch <boolean>: The flag to allow a regular expression
# based match on the name.
regexMatch: <boolean>
roles:
- # name <string>: The name of the Role to be exempted.
name: <string>
# namespace <string>: The namespace of the Role to be exempted.
namespace: <string>
Beispiele
prohibit-role-wildcard-access-sample
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sProhibitRoleWildcardAccess metadata: name: prohibit-role-wildcard-access-sample spec: enforcementAction: dryrun
Zulässig
apiVersion: rbac.authorization.k8s.io/v1 kind: ClusterRole metadata: name: cluster-role-example rules: - apiGroups: - "" resources: - pods verbs: - get
Nicht zugelassen
apiVersion: rbac.authorization.k8s.io/v1 kind: ClusterRole metadata: name: cluster-role-bad-example rules: - apiGroups: - "" resources: - pods verbs: - '*'
prohibit-wildcard-except-exempted-cluster-role
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sProhibitRoleWildcardAccess metadata: name: prohibit-wildcard-except-exempted-cluster-role spec: enforcementAction: dryrun parameters: exemptions: clusterRoles: - name: cluster-role-allowed-example roles: - name: role-allowed-example namespace: role-ns-allowed-example
Zulässig
apiVersion: rbac.authorization.k8s.io/v1 kind: ClusterRole metadata: name: cluster-role-allowed-example rules: - apiGroups: - "" resources: - pods verbs: - '*'
apiVersion: rbac.authorization.k8s.io/v1 kind: Role metadata: name: role-allowed-example namespace: role-ns-allowed-example rules: - apiGroups: - "" resources: - pods verbs: - '*'
Unzulässig
apiVersion: rbac.authorization.k8s.io/v1 kind: ClusterRole metadata: name: cluster-role-not-allowed-example rules: - apiGroups: - "" resources: - pods verbs: - '*'
apiVersion: rbac.authorization.k8s.io/v1 kind: Role metadata: name: role-not-allowed-example namespace: role-ns-not-allowed-example rules: - apiGroups: - "" resources: - pods verbs: - '*'
K8sReplicaLimits
Replikatlimits v1.0.2
Erfordert, dass Objekte mit dem Feld spec.replicas (Deployments, ReplicaSets usw.) eine Anzahl von Replikaten innerhalb definierter Bereiche angeben.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sReplicaLimits
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# ranges <array>: Allowed ranges for numbers of replicas. Values are
# inclusive.
ranges:
# <list item: object>: A range of allowed replicas. Values are
# inclusive.
- # max_replicas <integer>: The maximum number of replicas allowed,
# inclusive.
max_replicas: <integer>
# min_replicas <integer>: The minimum number of replicas allowed,
# inclusive.
min_replicas: <integer>
Beispiele
replica-limits
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sReplicaLimits metadata: name: replica-limits spec: match: kinds: - apiGroups: - apps kinds: - Deployment parameters: ranges: - max_replicas: 50 min_replicas: 3
Zulässig
apiVersion: apps/v1 kind: Deployment metadata: name: allowed-deployment spec: replicas: 3 selector: matchLabels: app: nginx template: metadata: labels: app: nginx spec: containers: - image: nginx:1.14.2 name: nginx ports: - containerPort: 80
Nicht zugelassen
apiVersion: apps/v1 kind: Deployment metadata: name: disallowed-deployment spec: replicas: 100 selector: matchLabels: app: nginx template: metadata: labels: app: nginx spec: containers: - image: nginx:1.14.2 name: nginx ports: - containerPort: 80
K8sRequireAdmissionController
Admission-Controller v1.0.0 erforderlich machen
Erfordert entweder "Pod Security Admission" oder ein externes Richtlinienkontrollsystem
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sRequireAdmissionController
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# permittedValidatingWebhooks <array>: List of permitted validating
# webhooks which are valid external policy control systems
permittedValidatingWebhooks:
- <string>
Referenzielle Einschränkung
Diese Einschränkung ist referenziell. Vor der Verwendung müssen Sie referenzielle Einschränkungen aktivieren und eine Konfiguration erstellen, die Policy Controller mitteilt, welche Arten von Objekten beobachtet werden sollen.
Ihr Policy Controller Config erfordert einen syncOnly-Eintrag ähnlich dem folgenden:
spec:
sync:
syncOnly:
- group: "admissionregistration.k8s.io"
version: "v1" OR "v1beta1"
kind: "ValidatingWebhookConfiguration"
Beispiele
require-admission-controller
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sRequireAdmissionController metadata: name: require-admission-controller spec: match: kinds: - apiGroups: - "" kinds: - Namespace
Zulässig
apiVersion: v1 kind: Namespace metadata: labels: pod-security.kubernetes.io/enforce: baseline pod-security.kubernetes.io/enforce-version: v1.28 name: allowed-namespace
Nicht zugelassen
apiVersion: v1 kind: Namespace metadata: name: disallowed-namespace
K8sRequireBinAuthZ
Erfordert Binärautorisierung v1.0.2
Erfordert validierenden Zulassungs-Webhook der Binärautorisierung. Einschränkungen mit diesem ConstraintTemplate werden unabhängig vom enforcementAction-Wert nur geprüft.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sRequireBinAuthZ
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
Referenzielle Einschränkung
Diese Einschränkung ist referenziell. Vor der Verwendung müssen Sie referenzielle Einschränkungen aktivieren und eine Konfiguration erstellen, die Policy Controller mitteilt, welche Arten von Objekten beobachtet werden sollen.
Ihr Policy Controller Config erfordert einen syncOnly-Eintrag ähnlich dem folgenden:
spec:
sync:
syncOnly:
- group: "admissionregistration.k8s.io"
version: "v1" OR "v1beta1"
kind: "ValidatingWebhookConfiguration"
Beispiele
require-binauthz
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sRequireBinAuthZ metadata: name: require-binauthz spec: enforcementAction: dryrun match: kinds: - apiGroups: - "" kinds: - Namespace
Zulässig
apiVersion: v1 kind: Namespace metadata: name: default --- # Referential Data apiVersion: admissionregistration.k8s.io/v1 kind: ValidatingWebhookConfiguration metadata: name: binauthz-admission-controller webhooks: - admissionReviewVersions: - v1 - v1beta1 clientConfig: url: https://binaryauthorization.googleapis.com/internal/projects/ap-bps-experimental-gke/policy/locations/us-central1/clusters/acm-test-cluster:admissionReview name: imagepolicywebhook.image-policy.k8s.io rules: - operations: - CREATE - UPDATE - apiVersion: - v1 sideEffects: None
Nicht zugelassen
apiVersion: v1 kind: Namespace metadata: name: default
K8sRequireCosNodeImage
COS-Knoten-Image erforderlich v1.1.1
Erzwingt die Verwendung von Container-Optimized OS von Google auf Knoten.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sRequireCosNodeImage
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# exemptOsImages <array>: A list of exempt OS Images.
exemptOsImages:
- <string>
Beispiele
nodes-have-consistent-time
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sRequireCosNodeImage metadata: name: nodes-have-consistent-time spec: enforcementAction: dryrun parameters: exemptOsImages: - Debian - Ubuntu*
Zulässig
apiVersion: v1 kind: Node metadata: name: allowed-example status: nodeInfo: osImage: Container-Optimized OS from Google
apiVersion: v1 kind: Node metadata: name: example-exempt status: nodeInfo: osImage: Debian
apiVersion: v1 kind: Node metadata: name: example-exempt-wildcard status: nodeInfo: osImage: Ubuntu 18.04.5 LTS
Nicht zugelassen
apiVersion: v1 kind: Node metadata: name: disallowed-example status: nodeInfo: osImage: Debian GNUv1.0
K8sRequireDaemonsets
Erforderliche Daemonsets v1.1.2
Erfordert die vorhandene Liste der Daemonsets.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sRequireDaemonsets
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# requiredDaemonsets <array>: A list of names and namespaces of the
# required daemonsets.
requiredDaemonsets:
- # name <string>: The name of the required daemonset.
name: <string>
# namespace <string>: The namespace for the required daemonset.
namespace: <string>
# restrictNodeSelector <boolean>: The daemonsets cannot include
# `NodeSelector`.
restrictNodeSelector: <boolean>
Referenzielle Einschränkung
Diese Einschränkung ist referenziell. Vor der Verwendung müssen Sie referenzielle Einschränkungen aktivieren und eine Konfiguration erstellen, die Policy Controller mitteilt, welche Arten von Objekten beobachtet werden sollen.
Ihr Policy Controller Config erfordert einen syncOnly-Eintrag ähnlich dem folgenden:
spec:
sync:
syncOnly:
- group: "extensions"
version: "v1beta1"
kind: "DaemonSet"
OR
- group: "apps"
version: "v1beta2" OR "v1"
kind: "DaemonSet"
Beispiele
require-daemonset
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sRequireDaemonsets metadata: name: require-daemonset spec: enforcementAction: dryrun match: kinds: - apiGroups: - "" kinds: - Namespace parameters: requiredDaemonsets: - name: clamav namespace: pci-dss-av restrictNodeSelector: true
Zulässig
apiVersion: v1 kind: Namespace metadata: name: pci-dss-av --- # Referential Data apiVersion: apps/v1 kind: DaemonSet metadata: name: other namespace: pci-dss-av spec: selector: matchLabels: name: other template: spec: containers: - image: us.gcr.io/{your-project-id}/other:latest name: other --- # Referential Data apiVersion: apps/v1 kind: DaemonSet metadata: labels: k8s-app: clamav-host-scanner name: clamav namespace: pci-dss-av spec: selector: matchLabels: name: clamav template: metadata: labels: name: clamav spec: containers: - image: us.gcr.io/{your-project-id}/clamav:latest livenessProbe: exec: command: - /health.sh initialDelaySeconds: 60 periodSeconds: 30 name: clamav-scanner resources: limits: memory: 3Gi requests: cpu: 500m memory: 2Gi volumeMounts: - mountPath: /data name: data-vol - mountPath: /host-fs name: host-fs readOnly: true - mountPath: /logs name: logs terminationGracePeriodSeconds: 30 tolerations: - effect: NoSchedule key: node-role.kubernetes.io/master volumes: - emptyDir: {} name: data-vol - hostPath: path: / name: host-fs - hostPath: path: /var/log/clamav name: logs
Nicht zugelassen
apiVersion: v1 kind: Namespace metadata: name: pci-dss-av
apiVersion: v1 kind: Namespace metadata: name: pci-dss-av --- # Referential Data apiVersion: apps/v1 kind: DaemonSet metadata: name: other namespace: pci-dss-av spec: selector: matchLabels: name: other template: spec: containers: - image: us.gcr.io/{your-project-id}/other:latest name: other
apiVersion: v1 kind: Namespace metadata: name: pci-dss-av --- # Referential Data apiVersion: apps/v1 kind: DaemonSet metadata: name: clamav namespace: pci-dss-av spec: selector: matchLabels: name: clamav template: spec: containers: - image: us.gcr.io/{your-project-id}/other:latest name: clamav nodeSelector: cloud.google.com/gke-spot: "true"
K8sRequireDefaultDenyEgressPolicy
Standardrichtlinie zum Ablehnen von ausgehendem Traffic erforderlich machen v1.0.3
Erfordert, dass jeder im Cluster definierte Namespace eine standardmäßige NetworkPolicy für ausgehenden Traffic hat.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sRequireDefaultDenyEgressPolicy
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
Referenzielle Einschränkung
Diese Einschränkung ist referenziell. Vor der Verwendung müssen Sie referenzielle Einschränkungen aktivieren und eine Konfiguration erstellen, die Policy Controller mitteilt, welche Arten von Objekten beobachtet werden sollen.
Ihr Policy Controller Config erfordert einen syncOnly-Eintrag ähnlich dem folgenden:
spec:
sync:
syncOnly:
- group: "extensions"
version: "v1beta1"
kind: "NetworkPolicy"
OR
- group: "networking.k8s.io"
version: "v1"
kind: "NetworkPolicy"
Beispiele
require-default-deny-network-policies
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sRequireDefaultDenyEgressPolicy metadata: name: require-default-deny-network-policies spec: enforcementAction: dryrun
Zulässig
apiVersion: v1 kind: Namespace metadata: name: example-namespace --- # Referential Data apiVersion: networking.k8s.io/v1 kind: NetworkPolicy metadata: name: default-deny-egress namespace: example-namespace spec: podSelector: {} policyTypes: - Egress
Unzulässig
apiVersion: v1 kind: Namespace metadata: name: example-namespace
apiVersion: v1 kind: Namespace metadata: name: example-namespace2 --- # Referential Data apiVersion: networking.k8s.io/v1 kind: NetworkPolicy metadata: name: default-deny-egress namespace: example-namespace spec: podSelector: {} policyTypes: - Egress
K8sRequireNamespaceNetworkPolicies
Namespace-Netzwerkrichtlinien 1.0.6 erforderlich
Erfordert, dass jeder im Cluster definierte Namespace eine NetworkPolicy hat.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sRequireNamespaceNetworkPolicies
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
Referenzielle Einschränkung
Diese Einschränkung ist referenziell. Vor der Verwendung müssen Sie referenzielle Einschränkungen aktivieren und eine Konfiguration erstellen, die Policy Controller mitteilt, welche Arten von Objekten beobachtet werden sollen.
Ihr Policy Controller Config erfordert einen syncOnly-Eintrag ähnlich dem folgenden:
spec:
sync:
syncOnly:
- group: "extensions"
version: "v1beta1"
kind: "NetworkPolicy"
OR
- group: "networking.k8s.io"
version: "v1"
kind: "NetworkPolicy"
Beispiele
require-namespace-network-policies-sample
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sRequireNamespaceNetworkPolicies metadata: name: require-namespace-network-policies-sample spec: enforcementAction: dryrun
Zulässig
apiVersion: v1 kind: Namespace metadata: name: require-namespace-network-policies-example --- # Referential Data apiVersion: networking.k8s.io/v1 kind: NetworkPolicy metadata: name: test-network-policy namespace: require-namespace-network-policies-example
Nicht zugelassen
apiVersion: v1 kind: Namespace metadata: name: require-namespace-network-policies-example
K8sRequireValidRangesForNetworks
Gültige Bereiche für Netzwerke erforderlich machen v1.0.2
Erzwingt, welche CIDR-Blöcke für eingehenden und ausgehenden Netzwerktraffic zulässig sind.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sRequireValidRangesForNetworks
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# allowedEgress <array>: IP ranges in CIDR format (0.0.0.0/32) that are
# allowed for egress.
allowedEgress:
- <string>
# allowedIngress <array>: IP ranges in CIDR format (0.0.0.0/32) that are
# allowed for ingress.
allowedIngress:
- <string>
Beispiele
require-valid-network-ranges
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sRequireValidRangesForNetworks metadata: name: require-valid-network-ranges spec: enforcementAction: dryrun parameters: allowedEgress: - 10.0.0.0/32 allowedIngress: - 10.0.0.0/24
Zulässig
apiVersion: networking.k8s.io/v1 kind: NetworkPolicy metadata: name: test-network-policy namespace: default spec: egress: - ports: - port: 5978 protocol: TCP to: - ipBlock: cidr: 10.0.0.0/32 ingress: - from: - ipBlock: cidr: 10.0.0.0/29 - ipBlock: cidr: 10.0.0.100/29 - namespaceSelector: matchLabels: project: myproject - podSelector: matchLabels: role: frontend ports: - port: 6379 protocol: TCP podSelector: matchLabels: role: db policyTypes: - Ingress - Egress
Unzulässig
apiVersion: networking.k8s.io/v1 kind: NetworkPolicy metadata: name: test-network-policy-disallowed namespace: default spec: egress: - ports: - port: 5978 protocol: TCP to: - ipBlock: cidr: 1.1.2.0/31 ingress: - from: - ipBlock: cidr: 1.1.2.0/24 - ipBlock: cidr: 2.1.2.0/24 - namespaceSelector: matchLabels: project: myproject - podSelector: matchLabels: role: frontend ports: - port: 6379 protocol: TCP podSelector: matchLabels: role: db policyTypes: - Ingress - Egress
K8sRequiredAnnotations
Erforderliche Annotationen v1.0.1
Erfordert Ressourcen, die bestimmte Annotationen enthalten, wobei Werte mit den regulären Ausdrücken übereinstimmen.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sRequiredAnnotations
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# annotations <array>: A list of annotations and values the object must
# specify.
annotations:
- # allowedRegex <string>: If specified, a regular expression the
# annotation's value must match. The value must contain at least one
# match for the regular expression.
allowedRegex: <string>
# key <string>: The required annotation.
key: <string>
message: <string>
Beispiele
all-must-have-certain-set-of-annotations
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sRequiredAnnotations metadata: name: all-must-have-certain-set-of-annotations spec: match: kinds: - apiGroups: - "" kinds: - Service parameters: annotations: - allowedRegex: ^([A-Za-z0-9._%+-]+@[A-Za-z0-9.-]+\.[A-Za-z]{2,6}|[a-z]{1,39})$ key: a8r.io/owner - allowedRegex: ^(http:\/\/www\.|https:\/\/www\.|http:\/\/|https:\/\/)?[a-z0-9]+([\-\.]{1}[a-z0-9]+)*\.[a-z]{2,5}(:[0-9]{1,5})?(\/.*)?$ key: a8r.io/runbook message: All services must have a `a8r.io/owner` and `a8r.io/runbook` annotations.
Zulässig
apiVersion: v1 kind: Service metadata: annotations: a8r.io/owner: dev-team-alfa@contoso.com a8r.io/runbook: https://confluence.contoso.com/dev-team-alfa/runbooks name: allowed-service spec: ports: - name: http port: 80 targetPort: 8080 selector: app: foo
Unzulässig
apiVersion: v1 kind: Service metadata: name: disallowed-service spec: ports: - name: http port: 80 targetPort: 8080 selector: app: foo
K8sRequiredLabels
Erforderliche Labels v1.0.1
Erfordert, dass Ressourcen bestimmte Labels enthalten, wobei Werte mit den regulären Ausdrücken übereinstimmen.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sRequiredLabels
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# labels <array>: A list of labels and values the object must specify.
labels:
- # allowedRegex <string>: If specified, a regular expression the
# annotation's value must match. The value must contain at least one
# match for the regular expression.
allowedRegex: <string>
# key <string>: The required label.
key: <string>
message: <string>
Beispiele
all-must-have-owner
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sRequiredLabels metadata: name: all-must-have-owner spec: match: kinds: - apiGroups: - "" kinds: - Namespace parameters: labels: - allowedRegex: ^[a-zA-Z]+.agilebank.demo$ key: owner message: All namespaces must have an `owner` label that points to your company username
Zulässig
apiVersion: v1 kind: Namespace metadata: labels: owner: user.agilebank.demo name: allowed-namespace
Unzulässig
apiVersion: v1 kind: Namespace metadata: name: disallowed-namespace
K8sRequiredProbes
Erforderliche Prüfungen v1.0.1
Erfordert, dass Pods über Bereitschafts- und/oder Aktivitätsprüfungen verfügen.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sRequiredProbes
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# probeTypes <array>: The probe must define a field listed in `probeType`
# in order to satisfy the constraint (ex. `tcpSocket` satisfies
# `['tcpSocket', 'exec']`)
probeTypes:
- <string>
# probes <array>: A list of probes that are required (ex: `readinessProbe`)
probes:
- <string>
Beispiele
must-have-probes
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sRequiredProbes metadata: name: must-have-probes spec: match: kinds: - apiGroups: - "" kinds: - Pod parameters: probeTypes: - tcpSocket - httpGet - exec probes: - readinessProbe - livenessProbe
Zulässig
apiVersion: v1 kind: Pod metadata: name: test-pod1 spec: containers: - image: tomcat livenessProbe: initialDelaySeconds: 5 periodSeconds: 10 tcpSocket: port: 80 name: tomcat ports: - containerPort: 8080 readinessProbe: initialDelaySeconds: 5 periodSeconds: 10 tcpSocket: port: 8080 volumes: - emptyDir: {} name: cache-volume
Nicht zugelassen
apiVersion: v1 kind: Pod metadata: name: test-pod1 spec: containers: - image: nginx:1.7.9 name: nginx-1 ports: - containerPort: 80 volumeMounts: - mountPath: /tmp/cache name: cache-volume - image: tomcat name: tomcat ports: - containerPort: 8080 readinessProbe: initialDelaySeconds: 5 periodSeconds: 10 tcpSocket: port: 8080 volumes: - emptyDir: {} name: cache-volume
apiVersion: v1 kind: Pod metadata: name: test-pod2 spec: containers: - image: nginx:1.7.9 livenessProbe: initialDelaySeconds: 5 periodSeconds: 10 tcpSocket: port: 80 name: nginx-1 ports: - containerPort: 80 volumeMounts: - mountPath: /tmp/cache name: cache-volume - image: tomcat name: tomcat ports: - containerPort: 8080 readinessProbe: initialDelaySeconds: 5 periodSeconds: 10 tcpSocket: port: 8080 volumes: - emptyDir: {} name: cache-volume
K8sRequiredResources
Erforderliche Ressourcen v1.0.1
Erfordert, dass Container definierte Ressourcen festlegen. https://kubernetes.io/docs/concepts/configuration/manage-resources-containers/
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sRequiredResources
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# exemptImages <array>: Any container that uses an image that matches an
# entry in this list will be excluded from enforcement. Prefix-matching can
# be signified with `*`. For example: `my-image-*`. It is recommended that
# users use the fully-qualified Docker image name (e.g. start with a domain
# name) in order to avoid unexpectedly exempting images from an untrusted
# repository.
exemptImages:
- <string>
# limits <array>: A list of limits that should be enforced (`cpu`,
# `memory`, or both).
limits:
# Allowed Values: cpu, memory
- <string>
# requests <array>: A list of requests that should be enforced (`cpu`,
# `memory`, or both).
requests:
# Allowed Values: cpu, memory
- <string>
Beispiele
container-must-have-limits-and-requests
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sRequiredResources metadata: name: container-must-have-limits-and-requests spec: match: kinds: - apiGroups: - "" kinds: - Pod parameters: limits: - cpu - memory requests: - cpu - memory
Zulässig
apiVersion: v1 kind: Pod metadata: labels: owner: me.agilebank.demo name: opa-allowed spec: containers: - args: - run - --server - --addr=localhost:8080 image: openpolicyagent/opa:0.9.2 name: opa resources: limits: cpu: 100m memory: 1Gi requests: cpu: 100m memory: 1Gi
Nicht zugelassen
apiVersion: v1 kind: Pod metadata: labels: owner: me.agilebank.demo name: opa-disallowed spec: containers: - args: - run - --server - --addr=localhost:8080 image: openpolicyagent/opa:0.9.2 name: opa resources: requests: cpu: 100m memory: 2Gi
apiVersion: v1 kind: Pod metadata: labels: owner: me.agilebank.demo name: opa-disallowed spec: containers: - args: - run - --server - --addr=localhost:8080 image: openpolicyagent/opa:0.9.2 name: opa resources: limits: memory: 2Gi requests: cpu: 100m
apiVersion: v1 kind: Pod metadata: labels: owner: me.agilebank.demo name: opa-disallowed spec: containers: - args: - run - --server - --addr=localhost:8080 image: openpolicyagent/opa:0.9.2 name: opa resources: limits: memory: 2Gi
container-must-have-cpu-requests-memory-limits-and-requests
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sRequiredResources metadata: name: container-must-have-cpu-requests-memory-limits-and-requests spec: match: kinds: - apiGroups: - "" kinds: - Pod parameters: limits: - memory requests: - cpu - memory
Zulässig
apiVersion: v1 kind: Pod metadata: labels: owner: me.agilebank.demo name: opa-allowed spec: containers: - args: - run - --server - --addr=localhost:8080 image: openpolicyagent/opa:0.9.2 name: opa resources: limits: cpu: 100m memory: 1Gi requests: cpu: 100m memory: 1Gi
apiVersion: v1 kind: Pod metadata: labels: owner: me.agilebank.demo name: opa-disallowed spec: containers: - args: - run - --server - --addr=localhost:8080 image: openpolicyagent/opa:0.9.2 name: opa resources: limits: memory: 2Gi requests: cpu: 100m memory: 2Gi
Nicht zugelassen
apiVersion: v1 kind: Pod metadata: labels: owner: me.agilebank.demo name: opa-disallowed spec: containers: - args: - run - --server - --addr=localhost:8080 image: openpolicyagent/opa:0.9.2 name: opa resources: requests: cpu: 100m memory: 2Gi
apiVersion: v1 kind: Pod metadata: labels: owner: me.agilebank.demo name: opa-disallowed spec: containers: - args: - run - --server - --addr=localhost:8080 image: openpolicyagent/opa:0.9.2 name: opa resources: limits: memory: 2Gi
apiVersion: v1 kind: Pod metadata: labels: owner: me.agilebank.demo name: opa-disallowed spec: containers: - args: - run - --server - --addr=localhost:8080 image: openpolicyagent/opa:0.9.2 name: opa resources: {}
no-enforcements
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sRequiredResources metadata: name: no-enforcements spec: match: kinds: - apiGroups: - "" kinds: - Pod
Zugelassen
apiVersion: v1 kind: Pod metadata: labels: owner: me.agilebank.demo name: opa-allowed spec: containers: - args: - run - --server - --addr=localhost:8080 image: openpolicyagent/opa:0.9.2 name: opa resources: limits: cpu: 100m memory: 1Gi requests: cpu: 100m memory: 1Gi
apiVersion: v1 kind: Pod metadata: labels: owner: me.agilebank.demo name: opa-disallowed spec: containers: - args: - run - --server - --addr=localhost:8080 image: openpolicyagent/opa:0.9.2 name: opa resources: requests: cpu: 100m memory: 2Gi
apiVersion: v1 kind: Pod metadata: labels: owner: me.agilebank.demo name: opa-disallowed spec: containers: - args: - run - --server - --addr=localhost:8080 image: openpolicyagent/opa:0.9.2 name: opa resources: limits: memory: 2Gi requests: cpu: 100m
apiVersion: v1 kind: Pod metadata: labels: owner: me.agilebank.demo name: opa-disallowed spec: containers: - args: - run - --server - --addr=localhost:8080 image: openpolicyagent/opa:0.9.2 name: opa resources: {}
K8sRestrictAdmissionController
Admission-Controller v1.0.0 einschränken
Dynamische Zulassungs-Controller auf zulässige Zulassungs-Controller beschränken
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sRestrictAdmissionController
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# permittedMutatingWebhooks <array>: List of permitted mutating webhooks
# (mutating admission controllers)
permittedMutatingWebhooks:
- <string>
# permittedValidatingWebhooks <array>: List of permitted validating
# webhooks (validating admission controllers)
permittedValidatingWebhooks:
- <string>
Beispiele
Admission-Controller einschränken
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sRestrictAdmissionController metadata: name: restrict-admission-controller spec: match: kinds: - apiGroups: - admissionregistration.k8s.io kinds: - MutatingWebhookConfiguration - ValidatingWebhookConfiguration parameters: permittedMutatingWebhooks: - allowed-mutating-webhook permittedValidatingWebhooks: - allowed-validating-webhook
Zulässig
apiVersion: admissionregistration.k8s.io/v1 kind: ValidatingWebhookConfiguration metadata: name: allowed-validating-webhook
Nicht zugelassen
apiVersion: admissionregistration.k8s.io/v1 kind: ValidatingWebhookConfiguration metadata: name: disallowed-validating-webhook
K8sRestrictAutomountServiceAccountTokens
Dienstkonto-Tokens v1.0.1 einschränken
Beschränkt die Verwendung von Dienstkonto-Tokens.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sRestrictAutomountServiceAccountTokens
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
Beispiele
restrict-serviceaccounttokens
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sRestrictAutomountServiceAccountTokens metadata: name: restrict-serviceaccounttokens spec: enforcementAction: dryrun match: kinds: - apiGroups: - "" kinds: - Pod - ServiceAccount
Zulässig
apiVersion: v1 kind: Pod metadata: name: allowed-example-pod spec: containers: - image: nginx name: nginx
apiVersion: v1 kind: ServiceAccount metadata: name: disallowed-example-serviceaccount
Nicht zugelassen
apiVersion: v1 kind: Pod metadata: name: disallowed-example-pod spec: automountServiceAccountToken: true containers: - image: nginx name: nginx
apiVersion: v1 automountServiceAccountToken: true kind: ServiceAccount metadata: name: allowed-example-serviceaccount
K8sRestrictLabels
Labels einschränken v1.0.2
Verhindert, dass Ressourcen bestimmte Labels enthalten, es sei denn, es gibt eine Ausnahme für die jeweilige Ressource.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sRestrictLabels
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# exceptions <array>: Objects listed here are exempt from enforcement of
# this constraint. All fields must be provided.
exceptions:
# <list item: object>: A single object's identification, based on group,
# kind, namespace, and name.
- # group <string>: The Kubernetes group of the exempt object.
group: <string>
# kind <string>: The Kubernetes kind of the exempt object.
kind: <string>
# name <string>: The name of the exempt object.
name: <string>
# namespace <string>: The namespace of the exempt object. For
# cluster-scoped resources, use the empty string `""`.
namespace: <string>
# restrictedLabels <array>: A list of label keys strings.
restrictedLabels:
- <string>
Beispiele
restrict-label-example
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sRestrictLabels metadata: name: restrict-label-example spec: enforcementAction: dryrun parameters: exceptions: - group: "" kind: Pod name: allowed-example namespace: default restrictedLabels: - label-example
Zulässig
apiVersion: v1 kind: Pod metadata: labels: label-example: example name: allowed-example namespace: default spec: containers: - image: nginx name: nginx
Unzulässig
apiVersion: v1 kind: Pod metadata: labels: label-example: example name: disallowed-example namespace: default spec: containers: - image: nginx name: nginx
K8sRestrictNamespaces
Namespaces einschränken v1.0.1
Schränkt die Verwendung von Namespaces durch Ressourcen ein, die unter dem Parameter restrictedNamespaces aufgeführt sind.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sRestrictNamespaces
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# restrictedNamespaces <array>: A list of Namespaces to restrict.
restrictedNamespaces:
- <string>
Beispiele
restrict-default-namespace-sample
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sRestrictNamespaces metadata: name: restrict-default-namespace-sample spec: enforcementAction: dryrun parameters: restrictedNamespaces: - default
Zulässig
apiVersion: v1 kind: Pod metadata: name: allowed-example namespace: test-namespace spec: containers: - image: nginx name: nginx
Nicht zugelassen
apiVersion: v1 kind: Pod metadata: name: disallowed-example namespace: default spec: containers: - image: nginx name: nginx
K8sRestrictNfsUrls
NFS-URLs v1.0.1 einschränken
Verhindert, dass Ressourcen NFS-URLs enthalten, sofern nicht anders angegeben.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sRestrictNfsUrls
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# allowedNfsUrls <array>: A list of allowed NFS URLs
allowedNfsUrls:
- <string>
Beispiele
restrict-label-example
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sRestrictNfsUrls metadata: name: restrict-label-example spec: enforcementAction: dryrun parameters: allowedNfsUrls: - my-nfs-server.example.com/my-nfs-volume - my-nfs-server.example.com/my-wildcard-nfs-volume/*
Zulässig
apiVersion: v1 kind: Pod metadata: labels: label-example: example name: allowed-example namespace: default spec: containers: - image: nginx name: nginx
apiVersion: v1 kind: Pod metadata: labels: label-example: example name: allowed-example-nfs namespace: default spec: containers: - image: nginx name: nginx - name: test-volume nfs: path: /my-nfs-volume server: my-nfs-server.example.com
apiVersion: v1 kind: Pod metadata: labels: label-example: example name: allowed-example-nfs-wildcard namespace: default spec: containers: - image: nginx name: nginx - name: test-volume nfs: path: /my-nfs-volume/my-wildcard-nfs-volume/wildcard_matched_path server: my-nfs-server.example.com
Nicht zugelassen
apiVersion: v1 kind: Pod metadata: labels: label-example: example name: disallowed-example-nfs namespace: default spec: containers: - image: nginx name: nginx volumes: - name: test-volume nfs: path: /my-nfs-volume server: disallowed-nfs-server.example.com
apiVersion: v1 kind: Pod metadata: labels: label-example: example name: disallowed-example-nfs-mixed namespace: default spec: containers: - image: nginx name: nginx volumes: - name: test-volume-allowed nfs: path: /my-nfs-volume server: my-nfs-server.example.com - name: test-volume-disallowed nfs: path: /my-nfs-volume server: disallowed-nfs-server.example.com
K8sRestrictRbacSubjects
RBAC Subjects v1.0.3 einschränken
Schränkt die Verwendung von Namen in RBAC auf zulässige Werte ein.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sRestrictRbacSubjects
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# allowedSubjects <array>: The list of names permitted in RBAC subjects.
allowedSubjects:
- # name <string>: The exact-name or the pattern of the allowed subject
name: <string>
# regexMatch <boolean>: The flag to allow a regular expression based
# match on the name.
regexMatch: <boolean>
Beispiele
restrict-rbac-subjects
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sRestrictRbacSubjects metadata: name: restrict-rbac-subjects spec: enforcementAction: dryrun match: kinds: - apiGroups: - rbac.authorization.k8s.io kinds: - RoleBinding - ClusterRoleBinding parameters: allowedSubjects: - name: system:masters - name: ^.+@gcp-sa-[a-z-]+.iam.gserviceaccount.com$ regexMatch: true - name: ^.+@system.gserviceaccount.com$ regexMatch: true - name: ^.+@google.com$ regexMatch: true
Zulässig
apiVersion: rbac.authorization.k8s.io/v1 kind: ClusterRoleBinding metadata: name: good-clusterrolebinding roleRef: apiGroup: rbac.authorization.k8s.io kind: ClusterRole name: cluster-admin subjects: - apiGroup: rbac.authorization.k8s.io kind: User name: user@google.com - apiGroup: rbac.authorization.k8s.io kind: Group name: system:masters - apiGroup: rbac.authorization.k8s.io kind: User name: service-1234567890@gcp-sa-ktd-control.iam.gserviceaccount.com
Unzulässig
apiVersion: rbac.authorization.k8s.io/v1 kind: ClusterRoleBinding metadata: name: bad-clusterrolebinding roleRef: apiGroup: rbac.authorization.k8s.io kind: ClusterRole name: cluster-admin subjects: - apiGroup: rbac.authorization.k8s.io kind: User name: user1@example.com - apiGroup: rbac.authorization.k8s.io kind: User name: user2@example.com
K8sRestrictRoleBindings
Rollenbindungen einschränken v1.0.3
Beschränkt die in ClusterRoleBindings und RoleBindings angegebenen Subjekte auf eine Liste zulässiger Subjekte.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sRestrictRoleBindings
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# allowedSubjects <array>: The list of subjects that are allowed to bind to
# the restricted role.
allowedSubjects:
- # apiGroup <string>: The Kubernetes API group of the subject.
apiGroup: <string>
# kind <string>: The Kubernetes kind of the subject.
kind: <string>
# name <string>: The name of the subject which is matched exactly as
# provided as well as based on a regular expression.
name: <string>
# regexMatch <boolean>: The flag to allow a regular expression based
# match on the name.
regexMatch: <boolean>
# restrictedRole <object>: The role that cannot be bound to unless
# expressly allowed.
restrictedRole:
# apiGroup <string>: The Kubernetes API group of the role.
apiGroup: <string>
# kind <string>: The Kubernetes kind of the role.
kind: <string>
# name <string>: The name of the role.
name: <string>
Beispiele
restrict-clusteradmin-rolebindings-sample
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sRestrictRoleBindings metadata: name: restrict-clusteradmin-rolebindings-sample spec: enforcementAction: dryrun parameters: allowedSubjects: - apiGroup: rbac.authorization.k8s.io kind: Group name: system:masters restrictedRole: apiGroup: rbac.authorization.k8s.io kind: ClusterRole name: cluster-admin
Zulässig
apiVersion: rbac.authorization.k8s.io/v1 kind: ClusterRoleBinding metadata: name: good-clusterrolebinding roleRef: apiGroup: rbac.authorization.k8s.io kind: ClusterRole name: cluster-admin subjects: - apiGroup: rbac.authorization.k8s.io kind: Group name: system:masters
Nicht zugelassen
apiVersion: rbac.authorization.k8s.io/v1 kind: ClusterRoleBinding metadata: name: bad-clusterrolebinding roleRef: apiGroup: rbac.authorization.k8s.io kind: ClusterRole name: cluster-admin subjects: - apiGroup: rbac.authorization.k8s.io kind: Group name: system:unauthenticated
restrict-clusteradmin-rolebindings-regex
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sRestrictRoleBindings metadata: name: restrict-clusteradmin-rolebindings-regex spec: enforcementAction: dryrun parameters: allowedSubjects: - apiGroup: rbac.authorization.k8s.io kind: User name: ^service-[0-9]+@gcp-sa-anthosconfigmanagement.iam.gserviceaccount.com$ regexMatch: true restrictedRole: apiGroup: rbac.authorization.k8s.io kind: ClusterRole name: cluster-admin
Zulässig
apiVersion: rbac.authorization.k8s.io/v1 kind: ClusterRoleBinding metadata: name: good-clusterrolebinding roleRef: apiGroup: rbac.authorization.k8s.io kind: ClusterRole name: cluster-admin subjects: - apiGroup: rbac.authorization.k8s.io kind: User name: service-123456789@gcp-sa-anthosconfigmanagement.iam.gserviceaccount.com
Nicht zugelassen
apiVersion: rbac.authorization.k8s.io/v1 kind: ClusterRoleBinding metadata: name: bad-clusterrolebinding roleRef: apiGroup: rbac.authorization.k8s.io kind: ClusterRole name: cluster-admin subjects: - apiGroup: rbac.authorization.k8s.io kind: User name: someotherservice-123456789@gcp-sa-anthosconfigmanagement.iam.gserviceaccount.com
K8sRestrictRoleRules
Schränken Sie Role- und ClusterRole-Regeln ein. V1.0.4
Schränkt Regeln ein, die für Role- und ClusterRole-Objekte festgelegt werden können.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sRestrictRoleRules
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# allowedRules <array>: AllowedRules is the list of rules that are allowed
# on Role or ClusterRole objects. If set, any item off this list will be
# rejected.
allowedRules:
- # apiGroups <array>: APIGroups is the name of the APIGroup that
# contains the resources. If multiple API groups are specified, any
# action requested against one of the enumerated resources in any API
# group will be allowed. "" represents the core API group and "*"
# represents all API groups.
apiGroups:
- <string>
# resources <array>: Resources is a list of resources this rule
# applies to. '*' represents all resources.
resources:
- <string>
# verbs <array>: Verbs is a list of Verbs that apply to ALL the
# ResourceKinds contained in this rule. '*' represents all verbs.
verbs:
- <string>
# disallowedRules <array>: DisallowedRules is the list of rules that are
# NOT allowed on Role or ClusterRole objects. If set, any item on this list
# will be rejected.
disallowedRules:
- # apiGroups <array>: APIGroups is the name of the APIGroup that
# contains the resources. If multiple API groups are specified, any
# action requested against one of the enumerated resources in any API
# group will be disallowed. "" represents the core API group and "*"
# represents all API groups.
apiGroups:
- <string>
# resources <array>: Resources is a list of resources this rule
# applies to. '*' represents all resources.
resources:
- <string>
# verbs <array>: Verbs is a list of Verbs that apply to ALL the
# ResourceKinds contained in this rule. '*' represents all verbs.
verbs:
- <string>
# exemptions <object>: Exemptions is the list of Roles and/or ClusterRoles
# names that are allowed to violate this policy.
exemptions:
clusterRoles:
- # name <string>: Name is the name or a pattern of the ClusterRole
# to be exempted.
name: <string>
# regexMatch <boolean>: RegexMatch is the flag to toggle exact vs
# regex match of the ClusterRole name.
regexMatch: <boolean>
roles:
- # name <string>: Name is the name of the Role to be exempted.
name: <string>
# namespace <string>: Namespace is the namespace of the Role to be
# exempted.
namespace: <string>
Beispiele
restrict-pods-exec
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sRestrictRoleRules metadata: name: restrict-pods-exec spec: enforcementAction: dryrun match: kinds: - apiGroups: - rbac.authorization.k8s.io kinds: - Role - ClusterRole parameters: disallowedRules: - apiGroups: - "" resources: - pods/exec verbs: - create
Zulässig
apiVersion: rbac.authorization.k8s.io/v1 kind: Role metadata: name: allowed-role-example rules: - apiGroups: - "" resources: - pods verbs: - get - list - watch
Nicht zugelassen
apiVersion: rbac.authorization.k8s.io/v1 kind: ClusterRole metadata: name: disallowed-cluster-role-example rules: - apiGroups: - "" resources: - pods/exec verbs: - '*'
K8sStorageClass
Speicherklasse v1.1.2
Erfordert die Angabe von Speicherklassen, wenn verwendet. Nur Gatekeeper 3.9+ und nicht sitzungsspezifische Container werden unterstützt.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sStorageClass
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# allowedStorageClasses <array>: An optional allow-list of storage classes.
# If specified, any storage class not in the `allowedStorageClasses`
# parameter is disallowed.
allowedStorageClasses:
- <string>
includeStorageClassesInMessage: <boolean>
Referenzielle Einschränkung
Diese Einschränkung ist referenziell. Vor der Verwendung müssen Sie referenzielle Einschränkungen aktivieren und eine Konfiguration erstellen, die Policy Controller mitteilt, welche Arten von Objekten beobachtet werden sollen.
Ihr Policy Controller Config erfordert einen syncOnly-Eintrag ähnlich dem folgenden:
spec:
sync:
syncOnly:
- group: "storage.k8s.io"
version: "v1"
kind: "StorageClass"
Beispiele
storageclass
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sStorageClass metadata: name: storageclass spec: match: kinds: - apiGroups: - "" kinds: - PersistentVolumeClaim - apiGroups: - apps kinds: - StatefulSet parameters: includeStorageClassesInMessage: true
Zulässig
apiVersion: v1 kind: PersistentVolumeClaim metadata: name: ok spec: accessModes: - ReadWriteOnce resources: requests: storage: 8Gi storageClassName: somestorageclass volumeMode: Filesystem --- # Referential Data allowVolumeExpansion: true apiVersion: storage.k8s.io/v1 kind: StorageClass metadata: name: somestorageclass provisioner: foo
apiVersion: apps/v1 kind: StatefulSet metadata: name: volumeclaimstorageclass spec: replicas: 1 selector: matchLabels: app: volumeclaimstorageclass serviceName: volumeclaimstorageclass template: metadata: labels: app: volumeclaimstorageclass spec: containers: - image: registry.k8s.io/nginx-slim:0.8 name: main volumeMounts: - mountPath: /usr/share/nginx/html name: data volumeClaimTemplates: - metadata: name: data spec: accessModes: - ReadWriteOnce resources: requests: storage: 1Gi storageClassName: somestorageclass --- # Referential Data allowVolumeExpansion: true apiVersion: storage.k8s.io/v1 kind: StorageClass metadata: name: somestorageclass provisioner: foo
Nicht zugelassen
apiVersion: v1 kind: PersistentVolumeClaim metadata: name: badstorageclass spec: accessModes: - ReadWriteOnce resources: requests: storage: 8Gi storageClassName: badstorageclass volumeMode: Filesystem
apiVersion: apps/v1 kind: StatefulSet metadata: name: badvolumeclaimstorageclass spec: replicas: 1 selector: matchLabels: app: badvolumeclaimstorageclass serviceName: badvolumeclaimstorageclass template: metadata: labels: app: badvolumeclaimstorageclass spec: containers: - image: registry.k8s.io/nginx-slim:0.8 name: main volumeMounts: - mountPath: /usr/share/nginx/html name: data volumeClaimTemplates: - metadata: name: data spec: accessModes: - ReadWriteOnce resources: requests: storage: 1Gi storageClassName: badstorageclass
apiVersion: v1 kind: PersistentVolumeClaim metadata: name: nostorageclass spec: accessModes: - ReadWriteOnce resources: requests: storage: 8Gi volumeMode: Filesystem
apiVersion: apps/v1 kind: StatefulSet metadata: name: novolumeclaimstorageclass spec: replicas: 1 selector: matchLabels: app: novolumeclaimstorageclass serviceName: novolumeclaimstorageclass template: metadata: labels: app: novolumeclaimstorageclass spec: containers: - image: registry.k8s.io/nginx-slim:0.8 name: main volumeMounts: - mountPath: /usr/share/nginx/html name: data volumeClaimTemplates: - metadata: name: data spec: accessModes: - ReadWriteOnce resources: requests: storage: 1Gi
allowed-storageclass
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sStorageClass metadata: name: allowed-storageclass spec: match: kinds: - apiGroups: - "" kinds: - PersistentVolumeClaim - apiGroups: - apps kinds: - StatefulSet parameters: allowedStorageClasses: - allowed-storage-class includeStorageClassesInMessage: true
Zulässig
apiVersion: v1 kind: PersistentVolumeClaim metadata: name: allowed-storage-class-pvc spec: accessModes: - ReadWriteOnce resources: requests: storage: 8Gi storageClassName: allowed-storage-class volumeMode: Filesystem --- # Referential Data allowVolumeExpansion: true apiVersion: storage.k8s.io/v1 kind: StorageClass metadata: name: allowed-storage-class provisioner: foo
Unzulässig
apiVersion: v1 kind: PersistentVolumeClaim metadata: name: disallowed-storage-class-pvc spec: accessModes: - ReadWriteOnce resources: requests: storage: 8Gi storageClassName: disallowed-storage-class volumeMode: Filesystem --- # Referential Data allowVolumeExpansion: true apiVersion: storage.k8s.io/v1 kind: StorageClass metadata: name: allowed-storage-class provisioner: foo
K8sUniqueIngressHost
Eindeutiger Ingress-Host v1.0.4
Alle Ingress-Regelhosts müssen eindeutig sein. Unterstützt keine Hostnamen-Platzhalter: https://kubernetes.io/docs/concepts/services-networking/ingress/
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sUniqueIngressHost
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
Referenzielle Einschränkung
Diese Einschränkung ist referenziell. Vor der Verwendung müssen Sie referenzielle Einschränkungen aktivieren und eine Konfiguration erstellen, die Policy Controller mitteilt, welche Arten von Objekten beobachtet werden sollen.
Ihr Policy Controller Config erfordert einen syncOnly-Eintrag ähnlich dem folgenden:
spec:
sync:
syncOnly:
- group: "extensions"
version: "v1beta1"
kind: "Ingress"
OR
- group: "networking.k8s.io"
version: "v1beta1" OR "v1"
kind: "Ingress"
Beispiele
unique-ingress-host
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sUniqueIngressHost metadata: name: unique-ingress-host spec: match: kinds: - apiGroups: - extensions - networking.k8s.io kinds: - Ingress
Zulässig
apiVersion: networking.k8s.io/v1 kind: Ingress metadata: name: ingress-host-allowed namespace: default spec: rules: - host: example-allowed-host.example.com http: paths: - backend: service: name: nginx port: number: 80 path: / pathType: Prefix - host: example-allowed-host1.example.com http: paths: - backend: service: name: nginx2 port: number: 80 path: / pathType: Prefix
Nicht zugelassen
apiVersion: networking.k8s.io/v1 kind: Ingress metadata: name: ingress-host-disallowed namespace: default spec: rules: - host: example-host.example.com http: paths: - backend: service: name: nginx port: number: 80 path: / pathType: Prefix --- # Referential Data apiVersion: networking.k8s.io/v1 kind: Ingress metadata: name: ingress-host-example namespace: default spec: rules: - host: example-host.example.com http: paths: - backend: service: name: nginx port: number: 80 path: / pathType: Prefix
apiVersion: networking.k8s.io/v1 kind: Ingress metadata: name: ingress-host-disallowed2 namespace: default spec: rules: - host: example-host2.example.com http: paths: - backend: service: name: nginx port: number: 80 path: / pathType: Prefix - host: example-host3.example.com http: paths: - backend: service: name: nginx2 port: number: 80 path: / pathType: Prefix --- # Referential Data apiVersion: networking.k8s.io/v1 kind: Ingress metadata: name: ingress-host-example2 namespace: default spec: rules: - host: example-host2.example.com http: paths: - backend: service: name: nginx port: number: 80 path: / pathType: Prefix
K8sUniqueServiceSelector
Eindeutiger Service Selector v1.0.2
Dienste müssen einen eindeutigen Selektor innerhalb eines Namespace haben. Selektoren werden als identisch betrachtet, wenn sie identische Schlüssel und Werte haben. Selektoren können ein Schlüssel/Wert-Paar verwenden, solange sich mindestens ein separates Schlüssel/Wert-Paar zwischen ihnen befindet. https://kubernetes.io/docs/concepts/services-networking/service/#defining-a-service
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: K8sUniqueServiceSelector
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
Referenzielle Einschränkung
Diese Einschränkung ist referenziell. Vor der Verwendung müssen Sie referenzielle Einschränkungen aktivieren und eine Konfiguration erstellen, die Policy Controller mitteilt, welche Arten von Objekten beobachtet werden sollen.
Ihr Policy Controller Config erfordert einen syncOnly-Eintrag ähnlich dem folgenden:
spec:
sync:
syncOnly:
- group: ""
version: "v1"
kind: "Service"
Beispiele
unique-service-selector
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: K8sUniqueServiceSelector metadata: labels: owner: admin.agilebank.demo name: unique-service-selector
Zulässig
apiVersion: v1 kind: Service metadata: name: gatekeeper-test-service-disallowed namespace: default spec: ports: - port: 443 selector: key: other-value
Nicht zugelassen
apiVersion: v1 kind: Service metadata: name: gatekeeper-test-service-disallowed namespace: default spec: ports: - port: 443 selector: key: value --- # Referential Data apiVersion: v1 kind: Service metadata: name: gatekeeper-test-service-example namespace: default spec: ports: - port: 443 selector: key: value
NoUpdateServiceAccount
Aktualisierung des Dienstkontos v1.0.1 blockieren
Blockiert das Aktualisieren des Dienstkontos für Ressourcen, die über Pods abstrahiert werden. Diese Richtlinie wird im Prüfmodus ignoriert.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: NoUpdateServiceAccount
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# allowedGroups <array>: Groups that should be allowed to bypass the
# policy.
allowedGroups:
- <string>
# allowedUsers <array>: Users that should be allowed to bypass the policy.
allowedUsers:
- <string>
Beispiele
no-update-kube-system-service-account
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: NoUpdateServiceAccount metadata: name: no-update-kube-system-service-account spec: match: kinds: - apiGroups: - "" kinds: - ReplicationController - apiGroups: - apps kinds: - ReplicaSet - Deployment - StatefulSet - DaemonSet - apiGroups: - batch kinds: - CronJob namespaces: - kube-system parameters: allowedGroups: [] allowedUsers: []
Zulässig
apiVersion: apps/v1 kind: Deployment metadata: labels: app: policy-test name: policy-test namespace: kube-system spec: replicas: 1 selector: matchLabels: app: policy-test-deploy template: metadata: labels: app: policy-test-deploy spec: containers: - command: - /bin/bash - -c - sleep 99999 image: ubuntu name: policy-test serviceAccountName: policy-test-sa-1
PolicyStrictOnly
STRICT Istio mTLS-Richtlinie v1.0.4 erforderlich machen
Erfordert, dass die STRICT gegenseitige Istio-TLS-Authentifizierung immer angegeben wird, wenn PeerAuthentication verwendet wird. Diese Einschränkung sorgt auch dafür, dass die verworfenen Policy-- und MeshPolicy-Ressourcen STRICT die gegenseitige TLS-Authentifizierung erzwingen. Weitere Informationen finden Sie unter https://istio.io/latest/docs/tasks/security/authentication/mtls-migration/#lock-down-mutual-tls-for-the-entire-mesh
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: PolicyStrictOnly
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
Beispiele
peerauthentication-strict-constraint
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: PolicyStrictOnly metadata: name: peerauthentication-strict-constraint spec: enforcementAction: dryrun match: kinds: - apiGroups: - security.istio.io kinds: - PeerAuthentication namespaces: - default
Zulässig
apiVersion: security.istio.io/v1beta1 kind: PeerAuthentication metadata: name: mode-strict namespace: default spec: mtls: mode: STRICT
apiVersion: security.istio.io/v1beta1 kind: PeerAuthentication metadata: name: mode-strict-port-level namespace: default spec: mtls: mode: STRICT portLevelMtls: "8080": mode: STRICT
apiVersion: security.istio.io/v1beta1 kind: PeerAuthentication metadata: name: mode-strict-port-unset namespace: default spec: mtls: mode: STRICT portLevelMtls: "8080": mode: UNSET
Nicht zugelassen
apiVersion: security.istio.io/v1beta1 kind: PeerAuthentication metadata: name: empty-mtls namespace: default spec: mtls: {}
apiVersion: security.istio.io/v1beta1 kind: PeerAuthentication metadata: name: unspecified-mtls namespace: default
apiVersion: security.istio.io/v1beta1 kind: PeerAuthentication metadata: name: mode-null namespace: default spec: mtls: mode: null
apiVersion: security.istio.io/v1beta1 kind: PeerAuthentication metadata: name: mtls-null namespace: default spec: mtls: null
apiVersion: security.istio.io/v1beta1 kind: PeerAuthentication metadata: name: mode-permissive namespace: default spec: mtls: mode: PERMISSIVE
apiVersion: security.istio.io/v1beta1 kind: PeerAuthentication metadata: name: mode-strict-port-permissive namespace: default spec: mtls: mode: STRICT portLevelMtls: "8080": mode: PERMISSIVE
apiVersion: security.istio.io/v1beta1 kind: PeerAuthentication metadata: name: mode-strict-port-permissive namespace: default spec: mtls: mode: STRICT portLevelMtls: "8080": mode: PERMISSIVE "8081": mode: STRICT
deprecated-policy-strict-constraint
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: PolicyStrictOnly metadata: name: deprecated-policy-strict-constraint spec: enforcementAction: dryrun match: kinds: - apiGroups: - authentication.istio.io kinds: - Policy namespaces: - default
Zulässig
apiVersion: authentication.istio.io/v1alpha1 kind: Policy metadata: name: default-mode-strict namespace: default spec: peers: - mtls: mode: STRICT
Nicht zugelassen
apiVersion: authentication.istio.io/v1alpha1 kind: Policy metadata: name: default-mtls-empty namespace: default spec: peers: - mtls: {}
apiVersion: authentication.istio.io/v1alpha1 kind: Policy metadata: name: default-mtls-null namespace: default spec: peers: - mtls: null
apiVersion: authentication.istio.io/v1alpha1 kind: Policy metadata: name: peers-empty namespace: default spec: peers: []
apiVersion: authentication.istio.io/v1alpha1 kind: Policy metadata: name: policy-no-peers namespace: default spec: targets: - name: httpbin
apiVersion: authentication.istio.io/v1alpha1 kind: Policy metadata: name: policy-permissive namespace: default spec: peers: - mtls: mode: PERMISSIVE
RestrictNetworkExclusions
Restrict Network Exclusions v1.0.2
Steuert, welche Eingangs-Ports, Ausgangs-Ports und Ausgangs-IP-Bereiche von der Istio-Netzwerkerfassung ausgeschlossen werden können. Ports und IP-Bereiche, die die Istio-Netzwerkerfassung umgehen, werden vom Istio-Proxy nicht verarbeitet und unterliegen nicht der Authentifizierung, der Autorisierungsrichtlinie und anderen Istio-Features von Istio mTLS. Mithilfe dieser Einschränkung können Einschränkungen auf die Verwendung der folgenden Annotationen angewendet werden:
traffic.sidecar.istio.io/excludeInboundPortstraffic.sidecar.istio.io/excludeOutboundPortstraffic.sidecar.istio.io/excludeOutboundIPRanges
Weitere Informationen finden Sie unter https://istio.io/latest/docs/reference/config/annotations/.
Beim Einschränken von Ausgangs-IP-Bereichen berechnet die Einschränkung, ob ausgeschlossene IP-Bereiche übereinstimmen oder ein Teil der zulässigen IP-Bereichsausschlüsse sind.
Wenn Sie diese Einschränkung verwenden, müssen alle Eingangs- und Ausgangs-IP-Bereiche immer einbezogen werden. Setzen Sie dazu entweder die entsprechenden "include"-Annotationen auf "*" oder legen Sie sie nicht fest. Das Festlegen einer der folgenden Anmerkungen auf etwas anderes als "*" ist nicht zulässig:
traffic.sidecar.istio.io/includeInboundPortstraffic.sidecar.istio.io/includeOutboundPortstraffic.sidecar.istio.io/includeOutboundIPRanges
Mit dieser Einschränkung kann Port 15020 immer ausgeschlossen werden, da der Istio-Sidecar-Injektor immer die Annotation traffic.sidecar.istio.io/excludeInboundPorts hinzufügt, damit er für Systemdiagnosen verwendet werden kann.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: RestrictNetworkExclusions
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# allowedInboundPortExclusions <array>: A list of ports that this
# constraint will allow in the
# `traffic.sidecar.istio.io/excludeInboundPorts` annotation.
allowedInboundPortExclusions:
- <string>
# allowedOutboundIPRangeExclusions <array>: A list of IP ranges that this
# constraint will allow in the
# `traffic.sidecar.istio.io/excludeOutboundIPRanges` annotation. The
# constraint calculates whether excluded IP ranges match or are a subset of
# the ranges in this list.
allowedOutboundIPRangeExclusions:
- <string>
# allowedOutboundPortExclusions <array>: A list of ports that this
# constraint will allow in the
# `traffic.sidecar.istio.io/excludeOutboundPorts` annotation.
allowedOutboundPortExclusions:
- <string>
Beispiele
restrict-network-exclusions
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: RestrictNetworkExclusions metadata: name: restrict-network-exclusions spec: enforcementAction: deny match: kinds: - apiGroups: - "" kinds: - Pod parameters: allowedInboundPortExclusions: - "80" allowedOutboundIPRangeExclusions: - 169.254.169.254/32 allowedOutboundPortExclusions: - "8888"
Zulässig
apiVersion: v1 kind: Pod metadata: labels: app: nginx name: nothing-excluded spec: containers: - image: nginx name: nginx ports: - containerPort: 80
apiVersion: v1 kind: Pod metadata: annotations: traffic.sidecar.istio.io/excludeInboundPorts: "80" traffic.sidecar.istio.io/excludeOutboundIPRanges: 169.254.169.254/32 traffic.sidecar.istio.io/excludeOutboundPorts: "8888" labels: app: nginx name: allowed-port-and-ip-exclusions spec: containers: - image: nginx name: nginx ports: - containerPort: 80
apiVersion: v1 kind: Pod metadata: annotations: traffic.sidecar.istio.io/excludeOutboundIPRanges: 169.254.169.254/32 traffic.sidecar.istio.io/includeOutboundIPRanges: '*' labels: app: nginx name: all-ip-ranges-included-with-one-allowed-ip-excluded spec: containers: - image: nginx name: nginx ports: - containerPort: 80
apiVersion: v1 kind: Pod metadata: annotations: traffic.sidecar.istio.io/includeInboundPorts: '*' traffic.sidecar.istio.io/includeOutboundIPRanges: '*' traffic.sidecar.istio.io/includeOutboundPorts: '*' labels: app: nginx name: everything-included-with-no-exclusions spec: containers: - image: nginx name: nginx ports: - containerPort: 80
Nicht zugelassen
apiVersion: v1 kind: Pod metadata: annotations: traffic.sidecar.istio.io/excludeOutboundIPRanges: 1.1.2.0/24 labels: app: nginx name: disallowed-ip-range-exclusion spec: containers: - image: nginx name: nginx ports: - containerPort: 80 - containerPort: 443
apiVersion: v1 kind: Pod metadata: annotations: traffic.sidecar.istio.io/excludeOutboundIPRanges: 169.254.169.254/32,1.1.2.0/24 labels: app: nginx name: one-disallowed-ip-exclusion-and-one-allowed-exclusion spec: containers: - image: nginx name: nginx ports: - containerPort: 80 - containerPort: 443
apiVersion: v1 kind: Pod metadata: annotations: traffic.sidecar.istio.io/includeInboundPorts: 80,443 traffic.sidecar.istio.io/includeOutboundIPRanges: 169.254.169.254/32 traffic.sidecar.istio.io/includeOutboundPorts: "8888" labels: app: nginx name: disallowed-specific-port-and-ip-inclusions spec: containers: - image: nginx name: nginx ports: - containerPort: 80
SourceNotAllAuthz
Istio AuthorizationPolicy-Quelle ist nicht für alle v1.0.1 erforderlich
Erfordert, dass die AuthorizationPolicy-Regeln von Istio Quellhauptkonten auf etwas anderes als "*" gesetzt haben. https://istio.io/latest/docs/reference/config/security/authorization-policy/
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: SourceNotAllAuthz
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
Beispiele
sourcenotall-authz-constraint
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: SourceNotAllAuthz metadata: name: sourcenotall-authz-constraint spec: enforcementAction: dryrun match: kinds: - apiGroups: - security.istio.io kinds: - AuthorizationPolicy
Zulässig
apiVersion: security.istio.io/v1beta1 kind: AuthorizationPolicy metadata: name: source-principals-good namespace: foo spec: rules: - from: - source: principals: - cluster.local/ns/default/sa/sleep - source: namespaces: - test to: - operation: methods: - GET paths: - /info* - operation: methods: - POST paths: - /data when: - key: request.auth.claims[iss] values: - https://accounts.google.com selector: matchLabels: app: httpbin version: v1
Nicht zugelassen
apiVersion: security.istio.io/v1beta1 kind: AuthorizationPolicy metadata: name: source-principals-dne namespace: foo spec: rules: - from: - source: namespaces: - test to: - operation: methods: - GET paths: - /info* - operation: methods: - POST paths: - /data when: - key: request.auth.claims[iss] values: - https://accounts.google.com selector: matchLabels: app: httpbin version: v1
apiVersion: security.istio.io/v1beta1 kind: AuthorizationPolicy metadata: name: source-principals-all namespace: foo spec: rules: - from: - source: principals: - '*' - source: namespaces: - test to: - operation: methods: - GET paths: - /info* - operation: methods: - POST paths: - /data when: - key: request.auth.claims[iss] values: - https://accounts.google.com selector: matchLabels: app: httpbin version: v1
apiVersion: security.istio.io/v1beta1 kind: AuthorizationPolicy metadata: name: source-principals-someall namespace: foo spec: rules: - from: - source: principals: - cluster.local/ns/default/sa/sleep - '*' - source: namespaces: - test to: - operation: methods: - GET paths: - /info* - operation: methods: - POST paths: - /data when: - key: request.auth.claims[iss] values: - https://accounts.google.com selector: matchLabels: app: httpbin version: v1
VerifyDeprecatedAPI
Verworfene APIs Version 1.0.0 prüfen
Verifiziert verworfene Kubernetes APIs, um sicherzustellen, dass alle API-Versionen aktuell sind. Diese Vorlage gilt nicht für Audits, da Audits die Ressourcen betrachten, die bereits im Cluster mit nicht verworfenen API-Versionen vorhanden sind.
Einschränkungsschema
apiVersion: constraints.gatekeeper.sh/v1beta1
kind: VerifyDeprecatedAPI
metadata:
name: example
spec:
# match <object>: lets you configure which resources are in scope for this
# constraint. For more information, see the Policy Controller Constraint
# match documentation:
# https://cloud.google.com/kubernetes-engine/enterprise/policy-controller/docs/reference/match
match:
[match schema]
parameters:
# k8sVersion <number>: kubernetes version
k8sVersion: <number>
# kvs <array>: Deprecated api versions and corresponding kinds
kvs:
- # deprecatedAPI <string>: deprecated api
deprecatedAPI: <string>
# kinds <array>: impacted list of kinds
kinds:
- <string>
# targetAPI <string>: target api
targetAPI: <string>
Beispiele
verify-1.16
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: VerifyDeprecatedAPI metadata: name: verify-1.16 spec: match: kinds: - apiGroups: - apps kinds: - Deployment - ReplicaSet - StatefulSet - DaemonSet - apiGroups: - extensions kinds: - PodSecurityPolicy - ReplicaSet - Deployment - DaemonSet - NetworkPolicy parameters: k8sVersion: 1.16 kvs: - deprecatedAPI: apps/v1beta1 kinds: - Deployment - ReplicaSet - StatefulSet targetAPI: apps/v1 - deprecatedAPI: extensions/v1beta1 kinds: - ReplicaSet - Deployment - DaemonSet targetAPI: apps/v1 - deprecatedAPI: extensions/v1beta1 kinds: - PodSecurityPolicy targetAPI: policy/v1beta1 - deprecatedAPI: apps/v1beta2 kinds: - ReplicaSet - StatefulSet - Deployment - DaemonSet targetAPI: apps/v1 - deprecatedAPI: extensions/v1beta1 kinds: - NetworkPolicy targetAPI: networking.k8s.io/v1
Zulässig
apiVersion: apps/v1 kind: Deployment metadata: labels: app: nginx name: allowed-deployment spec: replicas: 3 selector: matchLabels: app: nginx template: metadata: labels: app: nginx spec: containers: - image: nginx:1.14.2 name: nginx ports: - containerPort: 80
Nicht zugelassen
apiVersion: apps/v1beta1 kind: Deployment metadata: labels: app: nginx name: disallowed-deployment spec: replicas: 3 selector: matchLabels: app: nginx template: metadata: labels: app: nginx spec: containers: - image: nginx:1.14.2 name: nginx ports: - containerPort: 80
verify-1.22
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: VerifyDeprecatedAPI metadata: name: verify-1.22 spec: match: kinds: - apiGroups: - admissionregistration.k8s.io kinds: - MutatingWebhookConfiguration - ValidatingWebhookConfiguration - apiGroups: - apiextensions.k8s.io kinds: - CustomResourceDefinition - apiGroups: - apiregistration.k8s.io kinds: - APIService - apiGroups: - authentication.k8s.io kinds: - TokenReview - apiGroups: - authorization.k8s.io kinds: - SubjectAccessReview - apiGroups: - certificates.k8s.io kinds: - CertificateSigningRequest - apiGroups: - coordination.k8s.io kinds: - Lease - apiGroups: - extensions - networking.k8s.io kinds: - Ingress - apiGroups: - networking.k8s.io kinds: - IngressClass - apiGroups: - rbac.authorization.k8s.io kinds: - ClusterRole - ClusterRoleBinding - Role - RoleBinding - apiGroups: - scheduling.k8s.io kinds: - PriorityClass - apiGroups: - storage.k8s.io kinds: - CSIDriver - CSINode - StorageClass - VolumeAttachment parameters: k8sVersion: 1.22 kvs: - deprecatedAPI: admissionregistration.k8s.io/v1beta1 kinds: - MutatingWebhookConfiguration - ValidatingWebhookConfiguration targetAPI: admissionregistration.k8s.io/v1 - deprecatedAPI: apiextensions.k8s.io/v1beta1 kinds: - CustomResourceDefinition targetAPI: apiextensions.k8s.io/v1 - deprecatedAPI: apiregistration.k8s.io/v1beta1 kinds: - APIService targetAPI: apiregistration.k8s.io/v1 - deprecatedAPI: authentication.k8s.io/v1beta1 kinds: - TokenReview targetAPI: authentication.k8s.io/v1 - deprecatedAPI: authorization.k8s.io/v1beta1 kinds: - SubjectAccessReview targetAPI: authorization.k8s.io/v1 - deprecatedAPI: certificates.k8s.io/v1beta1 kinds: - CertificateSigningRequest targetAPI: certificates.k8s.io/v1 - deprecatedAPI: coordination.k8s.io/v1beta1 kinds: - Lease targetAPI: coordination.k8s.io/v1 - deprecatedAPI: extensions/v1beta1 kinds: - Ingress targetAPI: networking.k8s.io/v1 - deprecatedAPI: networking.k8s.io/v1beta1 kinds: - Ingress - IngressClass targetAPI: networking.k8s.io/v1 - deprecatedAPI: rbac.authorization.k8s.io/v1beta1 kinds: - ClusterRole - ClusterRoleBinding - Role - RoleBinding targetAPI: rbac.authorization.k8s.io/v1 - deprecatedAPI: scheduling.k8s.io/v1beta1 kinds: - PriorityClass targetAPI: scheduling.k8s.io/v1 - deprecatedAPI: storage.k8s.io/v1beta1 kinds: - CSIDriver - CSINode - StorageClass - VolumeAttachment targetAPI: storage.k8s.io/v1
Zulässig
apiVersion: networking.k8s.io/v1 kind: Ingress metadata: annotations: nginx.ingress.kubernetes.io/rewrite-target: / name: allowed-ingress spec: ingressClassName: nginx-example rules: - http: paths: - backend: service: name: test port: number: 80 path: /testpath pathType: Prefix
Nicht zugelassen
apiVersion: networking.k8s.io/v1beta1 kind: Ingress metadata: annotations: nginx.ingress.kubernetes.io/rewrite-target: / name: disallowed-ingress spec: ingressClassName: nginx-example rules: - http: paths: - backend: service: name: test port: number: 80 path: /testpath pathType: Prefix
verify-1.25
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: VerifyDeprecatedAPI metadata: name: verify-1.25 spec: match: kinds: - apiGroups: - batch kinds: - CronJob - apiGroups: - discovery.k8s.io kinds: - EndpointSlice - apiGroups: - events.k8s.io kinds: - Event - apiGroups: - autoscaling kinds: - HorizontalPodAutoscaler - apiGroups: - policy kinds: - PodDisruptionBudget - PodSecurityPolicy - apiGroups: - node.k8s.io kinds: - RuntimeClass parameters: k8sVersion: 1.25 kvs: - deprecatedAPI: batch/v1beta1 kinds: - CronJob targetAPI: batch/v1 - deprecatedAPI: discovery.k8s.io/v1beta1 kinds: - EndpointSlice targetAPI: discovery.k8s.io/v1 - deprecatedAPI: events.k8s.io/v1beta1 kinds: - Event targetAPI: events.k8s.io/v1 - deprecatedAPI: autoscaling/v2beta1 kinds: - HorizontalPodAutoscaler targetAPI: autoscaling/v2 - deprecatedAPI: policy/v1beta1 kinds: - PodDisruptionBudget targetAPI: policy/v1 - deprecatedAPI: policy/v1beta1 kinds: - PodSecurityPolicy targetAPI: None - deprecatedAPI: node.k8s.io/v1beta1 kinds: - RuntimeClass targetAPI: node.k8s.io/v1
Zulässig
apiVersion: batch/v1 kind: CronJob metadata: name: allowed-cronjob namespace: default spec: jobTemplate: spec: template: spec: containers: - command: - /bin/sh - -c - date; echo Hello from the Kubernetes cluster image: busybox:1.28 imagePullPolicy: IfNotPresent name: hello restartPolicy: OnFailure schedule: '* * * * *'
Nicht zugelassen
apiVersion: batch/v1beta1 kind: CronJob metadata: name: disallowed-cronjob namespace: default spec: jobTemplate: spec: template: spec: containers: - command: - /bin/sh - -c - date; echo Hello from the Kubernetes cluster image: busybox:1.28 imagePullPolicy: IfNotPresent name: hello restartPolicy: OnFailure schedule: '* * * * *'
verify-1.26
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: VerifyDeprecatedAPI metadata: name: verify-1.26 spec: match: kinds: - apiGroups: - flowcontrol.apiserver.k8s.io kinds: - FlowSchema - PriorityLevelConfiguration - apiGroups: - autoscaling kinds: - HorizontalPodAutoscaler parameters: k8sVersion: 1.26 kvs: - deprecatedAPI: flowcontrol.apiserver.k8s.io/v1beta1 kinds: - FlowSchema - PriorityLevelConfiguration targetAPI: flowcontrol.apiserver.k8s.io/v1beta3 - deprecatedAPI: autoscaling/v2beta2 kinds: - HorizontalPodAutoscaler targetAPI: autoscaling/v2
Zulässig
apiVersion: flowcontrol.apiserver.k8s.io/v1beta3 kind: FlowSchema metadata: name: allowed-flowcontrol namespace: default spec: matchingPrecedence: 1000 priorityLevelConfiguration: name: exempt rules: - nonResourceRules: - nonResourceURLs: - /healthz - /livez - /readyz verbs: - '*' subjects: - group: name: system:unauthenticated kind: Group
Nicht zugelassen
apiVersion: flowcontrol.apiserver.k8s.io/v1beta1 kind: FlowSchema metadata: name: disallowed-flowcontrol namespace: default spec: matchingPrecedence: 1000 priorityLevelConfiguration: name: exempt rules: - nonResourceRules: - nonResourceURLs: - /healthz - /livez - /readyz verbs: - '*' subjects: - group: name: system:unauthenticated kind: Group
verify-1.27
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: VerifyDeprecatedAPI metadata: name: verify-1.27 spec: match: kinds: - apiGroups: - storage.k8s.io kinds: - CSIStorageCapacity parameters: k8sVersion: 1.27 kvs: - deprecatedAPI: storage.k8s.io/v1beta1 kinds: - CSIStorageCapacity targetAPI: storage.k8s.io/v1
Zulässig
apiVersion: storage.k8s.io/v1 kind: CSIStorageCapacity metadata: name: allowed-csistoragecapacity storageClassName: standard
Nicht zugelassen
apiVersion: storage.k8s.io/v1beta1 kind: CSIStorageCapacity metadata: name: allowed-csistoragecapacity namespace: default storageClassName: standard
verify-1.29
Einschränkung
apiVersion: constraints.gatekeeper.sh/v1beta1 kind: VerifyDeprecatedAPI metadata: name: verify-1.29 spec: match: kinds: - apiGroups: - flowcontrol.apiserver.k8s.io kinds: - FlowSchema - PriorityLevelConfiguration parameters: k8sVersion: 1.29 kvs: - deprecatedAPI: flowcontrol.apiserver.k8s.io/v1beta2 kinds: - FlowSchema - PriorityLevelConfiguration targetAPI: flowcontrol.apiserver.k8s.io/v1beta3
Zulässig
apiVersion: flowcontrol.apiserver.k8s.io/v1beta3 kind: FlowSchema metadata: name: allowed-flowcontrol namespace: default spec: matchingPrecedence: 1000 priorityLevelConfiguration: name: exempt rules: - nonResourceRules: - nonResourceURLs: - /healthz - /livez - /readyz verbs: - '*' subjects: - group: name: system:unauthenticated kind: Group
Nicht zugelassen
apiVersion: flowcontrol.apiserver.k8s.io/v1beta2 kind: FlowSchema metadata: name: disallowed-flowcontrol namespace: default spec: matchingPrecedence: 1000 priorityLevelConfiguration: name: exempt rules: - nonResourceRules: - nonResourceURLs: - /healthz - /livez - /readyz verbs: - '*' subjects: - group: name: system:unauthenticated kind: Group
Nächste Schritte
- Mehr über Policy Controller erfahren
- Policy Controller installieren
- Einschränkungen anstelle von PodSecurityPolicies verwenden
- Open-Source-Bibliothek mit Einschränkungsvorlagen im gatekeeper-library-Repository aufrufen