detecting-privilege-escalation-in-kubernetes-pods

检测 Kubernetes Pod 中的权限提升

概述

Kubernetes 中的权限提升是指 Pod 或容器获得超出预期范围的权限。这包括以 root 身份运行、使用特权模式、挂载主机文件系统、启用危险的 Linux 功能，或利用内核漏洞。检测需要结合准入控制（预防）、运行时监控（检测）和审计日志（调查）。

适用场景

• 调查需要检测 Kubernetes Pod 中权限提升的安全事件时
• 为此领域编写检测规则或威胁狩猎查询时
• SOC 分析师需要针对这类分析制定结构化流程时
• 验证对相关攻击技术的安全监控覆盖情况时

前提条件

• Kubernetes 集群 v1.25+（支持 Pod Security Admission）
• 具有 cluster-admin 访问权限的 kubectl
• Falco 或类似的运行时安全工具
• 用于准入策略的 OPA Gatekeeper 或 Kyverno

Kubernetes 中的权限提升途径

途径	风险	检测方法
privileged: true	完全访问主机	准入控制 + 审计
hostPID: true	访问主机进程	准入控制
hostNetwork: true	访问主机网络堆栈	准入控制
hostPath volumes	读写主机文件系统	准入控制
SYS_ADMIN capability	近似特权访问	准入控制 + 运行时
allowPrivilegeEscalation: true	setuid/setgid 利用	准入控制
runAsUser: 0	容器 root	准入控制
automountServiceAccountToken	窃取令牌以访问 API	准入控制
可写的 /proc 或 /sys	操纵内核参数	运行时监控

使用准入控制进行检测

Pod Security Admission（内置）

# Enforce restricted policy on namespace
apiVersion: v1
kind: Namespace
metadata:
  name: production
  labels:
    pod-security.kubernetes.io/enforce: restricted
    pod-security.kubernetes.io/enforce-version: latest
    pod-security.kubernetes.io/audit: restricted
    pod-security.kubernetes.io/warn: restricted

OPA Gatekeeper 策略

# Block dangerous capabilities
apiVersion: templates.gatekeeper.sh/v1
kind: ConstraintTemplate
metadata:
  name: k8sdangerouspriv
spec:
  crd:
    spec:
      names:
        kind: K8sDangerousPriv
  targets:
    - target: admission.k8s.gatekeeper.sh
      rego: |
        package k8sdangerouspriv

        dangerous_caps := {"SYS_ADMIN", "SYS_PTRACE", "SYS_MODULE", "DAC_OVERRIDE", "NET_ADMIN", "NET_RAW"}

        violation[{"msg": msg}] {
          container := input.review.object.spec.containers[_]
          cap := container.securityContext.capabilities.add[_]
          dangerous_caps[cap]
          msg := sprintf("Container %v adds dangerous capability: %v", [container.name, cap])
        }

        violation[{"msg": msg}] {
          container := input.review.object.spec.containers[_]
          container.securityContext.privileged == true
          msg := sprintf("Container %v runs in privileged mode", [container.name])
        }

        violation[{"msg": msg}] {
          container := input.review.object.spec.containers[_]
          container.securityContext.allowPrivilegeEscalation == true
          msg := sprintf("Container %v allows privilege escalation", [container.name])
        }

        violation[{"msg": msg}] {
          input.review.object.spec.hostPID == true
          msg := "Pod uses host PID namespace"
        }

        violation[{"msg": msg}] {
          input.review.object.spec.hostNetwork == true
          msg := "Pod uses host network"
        }

使用 Falco 进行运行时检测

# /etc/falco/rules.d/privesc-detection.yaml
- rule: Setuid Binary Execution in Container
  desc: Detect execution of setuid/setgid binaries in a container
  condition: >
    spawned_process and container and
    (proc.name in (su, sudo, newgrp, chsh, passwd) or
     proc.is_exe_upper_layer=true)
  output: >
    Setuid/setgid binary executed in container
    (user=%user.name container=%container.name image=%container.image.repository
     command=%proc.cmdline parent=%proc.pname)
  priority: WARNING
  tags: [container, privilege-escalation, T1548]

- rule: Capability Gained in Container
  desc: Detect when a process gains elevated capabilities
  condition: >
    evt.type = capset and container and
    evt.arg.cap != ""
  output: >
    Process gained capabilities in container
    (container=%container.name image=%container.image.repository
     capabilities=%evt.arg.cap command=%proc.cmdline)
  priority: WARNING
  tags: [container, privilege-escalation, T1548.001]

- rule: Container with Dangerous Capabilities Started
  desc: Detect container launched with dangerous capabilities
  condition: >
    container_started and container and
    (container.image.repository != "registry.k8s.io/pause") and
    (container.cap_effective contains SYS_ADMIN or
     container.cap_effective contains SYS_PTRACE or
     container.cap_effective contains SYS_MODULE)
  output: >
    Container with dangerous capabilities
    (container=%container.name image=%container.image.repository
     caps=%container.cap_effective)
  priority: CRITICAL
  tags: [container, privilege-escalation, T1068]

- rule: Write to /etc/passwd in Container
  desc: Detect writes to /etc/passwd inside container
  condition: >
    open_write and container and fd.name = /etc/passwd
  output: >
    Write to /etc/passwd in container
    (container=%container.name image=%container.image.repository
     command=%proc.cmdline user=%user.name)
  priority: CRITICAL
  tags: [container, privilege-escalation, T1136]

Kubernetes 审计日志检测

# audit-policy.yaml - Capture privilege escalation events
apiVersion: audit.k8s.io/v1
kind: Policy
rules:
  # Log pod creation with security context details
  - level: RequestResponse
    resources:
      - group: ""
        resources: ["pods"]
    verbs: ["create", "update", "patch"]

  # Log privilege escalation attempts
  - level: RequestResponse
    resources:
      - group: "rbac.authorization.k8s.io"
        resources: ["clusterroles", "clusterrolebindings", "roles", "rolebindings"]
    verbs: ["create", "update", "patch", "bind", "escalate"]

  # Log service account token requests
  - level: Metadata
    resources:
      - group: ""
        resources: ["serviceaccounts/token"]
    verbs: ["create"]

查询审计日志以检测权限提升

# Find pods created with privileged security context
kubectl logs -n kube-system kube-apiserver-* | \
  jq 'select(.verb == "create" and .objectRef.resource == "pods") |
  select(.requestObject.spec.containers[].securityContext.privileged == true)'

# Find RBAC escalation attempts
kubectl logs -n kube-system kube-apiserver-* | \
  jq 'select(.objectRef.resource == "clusterrolebindings" and .verb == "create")'

调查手册

# Check pod security context
kubectl get pod <pod-name> -n <ns> -o jsonpath='{.spec.containers[*].securityContext}'

# Check effective capabilities
kubectl exec <pod-name> -n <ns> -- cat /proc/1/status | grep -i cap

# List pods running as root
kubectl get pods --all-namespaces -o json | \
  jq '.items[] | select(.spec.containers[].securityContext.runAsUser == 0 or .spec.containers[].securityContext.privileged == true) | {name: .metadata.name, ns: .metadata.namespace}'

# Check for hostPath volumes
kubectl get pods --all-namespaces -o json | \
  jq '.items[] | select(.spec.volumes[]?.hostPath != null) | {name: .metadata.name, ns: .metadata.namespace, paths: [.spec.volumes[].hostPath.path]}'

最佳实践

1. 在生产命名空间中，以 restricted 级别启用 Pod Security Admission
2. 移除 ALL 功能，仅添加所需功能
3. 在所有容器上设置 allowPrivilegeEscalation: false
4. 以非 root 身份运行（runAsNonRoot: true，runAsUser > 0）
5. 除非需要访问 API，否则禁用 automountServiceAccountToken
6. 使用 Falco 监控运行时权限提升尝试
7. 通过 Kubernetes 审计日志审计 RBAC 变更
8. 使用 seccomp 配置文件限制系统调用