Skip to main content
Vulnerability Database/CVE-2024-13484

CVE-2024-13484: OpenShift GitOps Privilege Escalation

CVE-2024-13484 is a privilege escalation vulnerability in OpenShift GitOps Operator that allows namespaces to create rogue PrometheusRules affecting cluster monitoring. This post covers technical details, impact, and mitigation.

Published:

Disclaimer: This content was generated using AI. While we strive for accuracy, please verify critical information with official sources.

Sources: National Vulnerability Database, CISA KEV, Exploit-DB

CVE-2024-13484 Overview

CVE-2024-13484 affects the Red Hat OpenShift GitOps Operator container. The operator applies the openshift.io/cluster-monitoring label to any namespace that deploys an ArgoCD custom resource (CR) instance. Once that label is present, a user with permission in the namespace can create a PrometheusRule object that is ingested cluster-wide by the platform monitoring stack. An attacker with namespace-level access can therefore inject rogue alerting or recording rules that affect the entire cluster monitoring pipeline [CWE-668].

Critical Impact

A tenant with privileges in an ArgoCD-enabled namespace can push a rogue PrometheusRule that is rolled out across the cluster, corrupting platform monitoring and alerting integrity.

Affected Products

  • Red Hat OpenShift GitOps Operator (openshift-gitops-operator-container)
  • Red Hat OpenShift GitOps 1.14 and earlier (addressed in RHSA-2025:7753, RHSA-2025:8274, RHSA-2025:9506)
  • OpenShift clusters running the upstream redhat-developer/gitops-operator

Discovery Timeline

  • 2025-01-28 - CVE-2024-13484 published to the National Vulnerability Database (NVD)
  • 2026-06-26 - Last updated in NVD database

Technical Details for CVE-2024-13484

Vulnerability Analysis

The OpenShift GitOps Operator provisions an ArgoCD instance by reconciling an ArgoCD CR inside a user-controlled namespace. During reconciliation the operator stamps the hosting namespace with the openshift.io/cluster-monitoring=true label so that Argo CD metrics are scraped by platform Prometheus.

That label is also the trigger OpenShift uses to decide whether PrometheusRule resources defined in a namespace should be aggregated into the cluster-wide monitoring stack. Any PrometheusRule created in a labeled namespace is treated as a trusted platform rule.

A user with permission to create PrometheusRule objects in the ArgoCD namespace can therefore inject arbitrary recording or alerting rules into platform Prometheus. The rules can silence legitimate alerts, generate misleading signals, or exhaust monitoring resources, undermining the integrity and availability of cluster observability.

Root Cause

The root cause is improper exposure of a trusted resource to an unauthorized actor [CWE-668]. The operator grants a tenant namespace the cluster-monitoring privilege boundary without isolating PrometheusRule ingestion. The label conflates workload scrape eligibility with the authority to publish cluster-wide monitoring rules.

Attack Vector

Exploitation requires local access to the cluster with high privileges inside the ArgoCD namespace, but no user interaction. The attacker creates a PrometheusRule object in the labeled namespace. The platform Prometheus Operator picks it up and merges it into the cluster monitoring configuration, giving the attacker influence over alerting and recording rules consumed by cluster administrators.

No verified public exploit code is available. See the Red Hat CVE analysis for CVE-2024-13484 and the upstream fixes in gitops-operator pull request #853, #867, #868, and #869 for technical details.

Detection Methods for CVE-2024-13484

Indicators of Compromise

  • Namespaces hosting an ArgoCD CR that carry the openshift.io/cluster-monitoring=true label alongside unexpected PrometheusRule objects.
  • PrometheusRule resources authored outside the openshift-monitoring or openshift-user-workload-monitoring namespaces that reference platform metrics.
  • Alerts in platform Prometheus with rule names, labels, or expressions that do not match the organization's approved rule catalog.

Detection Strategies

  • Audit Kubernetes API server logs for create and update events on monitoring.coreos.com/v1 PrometheusRule originating from ArgoCD-managed namespaces.
  • Compare the live set of PrometheusRule objects against a GitOps-controlled baseline and alert on drift.
  • Review namespaces for the openshift.io/cluster-monitoring label and correlate against the inventory of sanctioned platform namespaces.

Monitoring Recommendations

  • Forward OpenShift audit logs and Prometheus Operator events to a centralized analytics pipeline for continuous review.
  • Alert when a PrometheusRule is created by a service account tied to a tenant workload rather than a cluster administrator.
  • Track changes to namespace labels, especially additions of openshift.io/cluster-monitoring, as a privileged configuration event.

How to Mitigate CVE-2024-13484

Immediate Actions Required

  • Apply the Red Hat OpenShift GitOps updates shipped in RHSA-2025:7753, RHSA-2025:8274, and RHSA-2025:9506.
  • Inventory every namespace hosting an ArgoCD CR and review existing PrometheusRule objects for unauthorized entries.
  • Restrict create, update, and patch permissions on PrometheusRule resources in tenant namespaces via RBAC.

Patch Information

Red Hat released fixes for the OpenShift GitOps Operator through the advisories referenced above. The upstream remediation is tracked in gitops-operator PR #853 with follow-up changes in PR #867, PR #868, and PR #869. Patched builds decouple ArgoCD metric scraping from the cluster-monitoring label scope that enables tenant-authored PrometheusRule ingestion. See Red Hat Bugzilla #2269376 for the engineering record.

Workarounds

  • Remove the openshift.io/cluster-monitoring label from ArgoCD tenant namespaces where cluster-wide rule ingestion is not required.
  • Deploy an admission policy (for example, OPA Gatekeeper or Kyverno) that denies creation of PrometheusRule resources outside sanctioned monitoring namespaces.
  • Place ArgoCD instances in dedicated, administrator-controlled namespaces and prevent tenants from obtaining write access to those namespaces.

Disclaimer: This content was generated using AI. While we strive for accuracy, please verify critical information with official sources.

Sources: National Vulnerability Database, CISA KEV, Exploit-DB

Experience the Most Advanced Cybersecurity Platform

See how the world’s most intelligent, autonomous cybersecurity platform can protect your organization today and into the future.