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-2026-102623 Overview
CVE-2026-102623 is a denial-of-service vulnerability in KubeVirt, the Kubernetes extension that runs virtual machines alongside container workloads. An authenticated user with permission to create Virtual Machine Instances (VMIs) can submit a virtual machine definition containing an empty ephemeral volume. The virt-controller component fails to validate the volume configuration and raises an unhandled exception during processing. Because the malformed definition persists in the cluster, the controller enters a continuous crash loop. This disrupts virtual machine lifecycle operations across the entire environment. The flaw is tracked under CWE-476 (NULL Pointer Dereference).
Critical Impact
A single low-privileged user can halt all VM lifecycle operations cluster-wide by submitting one malformed VMI definition.
Affected Products
- KubeVirt (upstream project)
- Red Hat OpenShift Virtualization (see Red Hat CVE-2026-102623)
- Kubernetes clusters running vulnerable KubeVirt virt-controller builds
Discovery Timeline
- 2026-09-29 - CVE CVE-2026-102623 published to NVD
- 2026-09-29 - Last updated in NVD database
Technical Details for CVE-2026-102623
Vulnerability Analysis
KubeVirt's virt-controller reconciles VirtualMachineInstance objects submitted through the Kubernetes API. During reconciliation, the controller inspects each volume attached to the VMI specification. When a volume declares itself as an ephemeral type but contains no backing configuration, the controller dereferences a nil field and panics. The panic terminates the controller process. Kubernetes then restarts the pod, which re-reads the same persisted VMI object and panics again. This produces a sustained crash loop that blocks scheduling, migration, and deletion of virtual machines across the cluster.
Root Cause
The root cause is missing input validation on the volume specification of a VirtualMachineInstance. The controller assumes that an ephemeral volume entry contains a populated PersistentVolumeClaim reference or source structure. When the field is empty, the controller performs a nil pointer dereference [CWE-476] instead of rejecting the object with a validation error. The malformed object is admitted to etcd before the controller processes it, so restarting the controller does not clear the poisoned state.
Attack Vector
Exploitation requires network access to the Kubernetes API server and a service account or user credential with create permission on the virtualmachineinstances resource. The attacker submits a VMI manifest that includes a volumes entry with an ephemeral type declared but no source populated. Once the object is accepted by the API server, the controller crashes on its next reconcile pass. No user interaction is required, and the impact affects the entire KubeVirt tenant, not just the attacker's namespace. See the Red Hat Bug Report #2543219 for additional technical context.
Detection Methods for CVE-2026-102623
Indicators of Compromise
- Repeated CrashLoopBackOff status on virt-controller pods in the kubevirt namespace.
- Panic stack traces in virt-controller logs referencing volume validation or nil pointer dereference.
- VirtualMachineInstance objects containing a volumes[].ephemeral entry with no populated source fields.
- Sudden failure of VM start, stop, or migration operations across all tenants.
Detection Strategies
- Audit Kubernetes API server logs for create and update requests on virtualmachineinstances from unexpected service accounts.
- Alert on virt-controller pod restart counts exceeding a baseline threshold within a short window.
- Scan existing VMI manifests in etcd for volumes with empty ephemeral source definitions.
Monitoring Recommendations
- Forward virt-controller container logs to a centralized logging platform and alert on panic signatures.
- Track Kubernetes control-plane metrics for virt-controller availability and reconcile error rates.
- Correlate VMI object creation events with subsequent controller crashes to identify the offending manifest and originating identity.
How to Mitigate CVE-2026-102623
Immediate Actions Required
- Identify and delete any VirtualMachineInstance objects containing malformed ephemeral volume definitions to break the crash loop.
- Restrict create and update permissions on virtualmachineinstances to trusted service accounts using Kubernetes RBAC.
- Apply the vendor-supplied patch for virt-controller once available for your distribution.
Patch Information
Refer to the Red Hat CVE-2026-102623 advisory for fixed package versions and errata. Upstream KubeVirt users should upgrade to the release that adds validation for empty ephemeral volume sources. Track Red Hat Bug Report #2543219 for remediation status.
Workarounds
- Deploy a Kubernetes ValidatingAdmissionWebhook or OPA/Gatekeeper policy that rejects VMI objects with empty ephemeral volume sources.
- Reduce the blast radius by limiting VMI creation privileges to a small set of automation accounts.
- Manually purge any poisoned VMI objects from etcd before restarting virt-controller to prevent recurring crashes.
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

