CVE-2026-18618 Red Hat CVE debrief

Who should care

Kubernetes administrators and security teams using ml-metadata should assess their exposure and verify if their environments are vulnerable to this issue. They should review pipeline runs for potential disruptions and monitor for specially crafted HTTP/2 requests. Defenders should prioritize verifying exposure in Kubernetes environments using ml-metadata and assessing the feasibility of in-cluster attacks.

Why it matters

CVE-2026-18618 is a high-severity vulnerability in ml-metadata that could lead to denial of service in Kubernetes environments. Defenders should verify exposure, assess in-cluster attack feasibility, and prioritize remediation based on the provided CVE and NVD details.

• Potential disruption of pipeline runs in affected namespaces
• Denial of service through crashing of the MLMD pod
• Need for verification of ml-metadata usage in Kubernetes environments
• Possible impact on Red Hat products using ml-metadata

Technical summary

The ml-metadata library uses a statically-linked gRPC stack that is outdated, making it vulnerable to known HTTP/2 denial of service (DoS) issues. An attacker with network access to the MLMD pod in a Kubernetes cluster could exploit this by sending specially crafted HTTP/2 requests, potentially causing a denial of service by crashing the MLMD pod. This could disrupt pipeline runs in the affected namespace. Defenders should prioritize verifying exposure in Kubernetes environments using ml-metadata and assessing the feasibility of in-cluster attacks.

Defensive priority

Defenders should prioritize verifying exposure in Kubernetes environments using ml-metadata and assessing the feasibility of in-cluster attacks.

Recommended defensive actions

• Verify if ml-metadata is used in your Kubernetes environment
• Assess network access controls to the MLMD pod
• Review pipeline runs for potential disruptions
• Monitor for specially crafted HTTP/2 requests
• Confirm whether affected product deployments exist in managed environments and assign an owner for follow-up
• Review the supplied official advisory or CVE record to validate affected scope, severity, and vendor guidance
• Plan vendor-supported updates or mitigations through normal change control where exposure is confirmed

Evidence notes

The CVE record and NVD entry provide details on the vulnerability in ml-metadata. Red Hat has released errata related to this issue, indicating potential impact on certain Red Hat products. The statically-linked gRPC stack in ml-metadata is outdated, making it vulnerable to known HTTP/2 denial of service (DoS) issues. Defenders should verify exposure in Kubernetes environments using ml-metadata and assess the feasibility of in-cluster attacks.

Sources and references

Methodology and review provenance

AI-assisted synthesis based on stored public vulnerability evidence. System validation, approval state, and publication status do not by themselves establish human review of this revision. PatchSiren helps prioritize defensive review and does not prove exposure or remediation on any system.