Not a U.S. government website. NDD is an independent vulnerability database by Volerion and is not affiliated with or endorsed by NIST or NVD.

CVE-2026-94627 Details

Description

vLLM Mooncake connector through 0.29.0 fails to properly manage GPU KV cache block ownership when concurrent child requests share a single transfer ID in prefill/decode disaggregated deployments. Attackers can trigger GPU memory exhaustion by submitting completion requests with multiple prompts, causing orphaned KV cache blocks to accumulate until process restart and eventually preventing legitimate requests from executing.

Metrics

CVSS 3.x Severity and Vector Strings:

CNA: [email protected]CVSS-B:7.5 HIGHVector:CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H

References to Advisories, Solutions, and Tools

By selecting these links, you will be leaving this site. These are references gathered from the official CVE record and are not endorsed by Volerion.

URL	Source(s)	Tag(s)
https://github.com/vllm-project/vllm	[email protected]
https://github.com/vllm-project/vllm/blob/v0.29.0/vllm/distributed/kv_transfer/kv_connector/v1/mooncake/mooncake_connector.py#L1978-L1989	[email protected]
https://github.com/vllm-project/vllm/pull/49796	[email protected]
https://www.vulncheck.com/advisories/vllm-through-0.29.0-gpu-kv-cache-leak-via-mooncake-transfer-id-collision	[email protected]

Weakness Enumeration

CWE-ID	CWE Name	Source
CWE-401	Missing Release of Memory after Effective Lifetime	[email protected]

Affected Products

No affected product data is available for this CVE.

Change History

2 change records found show changes

Date	Action	Recorded By
Sep 22, 2026	CVE Modified	CISA-ADP
Sep 21, 2026	New CVE Received	[email protected]