Critical Red Hat OpenShift CVE-2023-39418 Exposes Internal Services

Key Takeaways

• A high-severity Server-Side Request Forgery (SSRF) vulnerability, CVE-2026-66794, has been disclosed by Red Hat.
• The flaw affects the cluster-proxy-addon component in Multicluster Engine for Kubernetes.
• Unauthenticated remote attackers could exploit this vulnerability to access internal services across managed Kubernetes clusters.
• There is no patch available at the time of reporting; immediate network access restrictions are recommended.

Red Hat has issued a warning regarding a critical Server-Side Request Forgery (SSRF) vulnerability, identified as CVE-2026-66794, impacting the cluster-proxy-addon component within its Multicluster Engine for Kubernetes. This flaw, which carries a CVSS v3.1 score of 9.3, could enable unauthenticated remote attackers to reach services that are typically isolated within managed Kubernetes environments.

The vulnerability stems from an exposed user-facing route in the cluster proxy add-on. According to Red Hat, this route fails to adequately enforce authentication and authorization protocols before processing and forwarding requests. Consequently, an attacker gaining access to this endpoint can manipulate URL path segments, compelling the proxy to relay requests to arbitrary internal services across managed clusters.

This behavior is categorized under CWE-918 (Server-Side Request Forgery), where the attacker leverages the vulnerable proxy as an intermediary rather than directly targeting an internal service from external networks. The proxy’s inherent trust within the internal infrastructure allows it to bypass network controls, service exposure limitations, and cluster segmentation that would otherwise prevent unauthorized access.

Exploiting the Multicluster Kubernetes Flaw

In a multicluster Kubernetes setup, the implications of such a vulnerability are particularly severe. Management-plane components frequently maintain connectivity with services distributed across numerous managed clusters. If exposed through a compromised route, attackers could potentially access internal APIs, extract sensitive application data, map reachable internal endpoints, or interact with services never intended for external exposure.

For instance, an attacker might craft a request path to direct the affected proxy towards an internal service within a managed cluster. Should that service return configuration details, authentication tokens, debugging information, or application responses without requiring additional authentication, the attacker could acquire valuable data for subsequent intrusion stages.

The precise impact hinges on the range of services the proxy can reach and the security measures enforced by those services. Despite the high CVSS score of 9.3, Red Hat rated the issue as “Important” rather than “Critical.” Their assessment confirms the impact as unauthenticated SSRF, not remote code execution. The vendor noted a high confidentiality impact, a low integrity impact, and no demonstrated availability impact.

The affected components specifically include multicluster-engine/cluster-proxy-addon-rhel9 and multicluster-engine/cluster-proxy-rhel9 within Multicluster Engine for Kubernetes. As of this report, Red Hat has listed both as affected and has not yet released associated security errata.

What You Should Do

• Immediately restrict network access to the user-facing cluster-proxy-addon route.
• Apply firewall rules or equivalent network controls to ensure that only trusted networks and authorized sources can connect to the exposed port.
• Administrators should review route exposure configurations and inspect proxy access logs for any unusual path patterns.
• Identify and assess any sensitive internal services that are reachable from the management environment.

Disclaimer: HackersRadar reports on cybersecurity threats and incidents for informational and awareness purposes only. We do not engage in hacking activities, data exfiltration, or the hosting or distribution of stolen or leaked information. All content is based on publicly available sources.