Executive brief
A security vulnerability exists in KubeVirt, a tool used to run virtual machines within Kubernetes clusters. An attacker with permissions to manage virtual machines can trick the system into connecting to unauthorized internal network locations. This could allow an attacker to bypass network security rules and access sensitive internal services or data that should otherwise be isolated.
Technical details
A server-side request forgery (SSRF) vulnerability exists in the KubeVirt virt-api port-forward handler. The 'getTargetInterfaceIP' function in 'pkg/virt-api/rest/dialers.go' retrieves the target IP address from 'vmi.Status.Interfaces[0].IP' and passes it to 'net.Dial()' without validation. For VirtualMachineInstances (VMIs) using bridge or secondary-only network bindings, this IP is supplied by the QEMU guest agent, which is controllable by the VM owner. An attacker with 'kubevirt.io:edit' permissions can deploy a VM with a malicious guest agent that reports an arbitrary IP address. By then initiating a port-forward request, the attacker can establish a bidirectional TCP tunnel from the virt-api's network context to any routable destination, effectively bypassing Kubernetes NetworkPolicies. The default masquerade binding is not affected as it uses cluster-assigned pod IPs.
Affected products
- KubeVirt KubeVirt unspecified
- Red Hat Red Hat OpenShift Virtualization 4 4
Timeline
- 2026-06-25: disclosed: Vulnerability reported by Red Hat
- 2026-06-26: advisory: NVD publication date