---
id: "GHSA-8qpc-6rxx-4pxf"
kind: "vulnerability"
aliases: ["GHSA-8qpc-6rxx-4pxf"]
title: "Duplicate Advisory: vllm p2p never connected zmq_exhaustion"
severity: "high"
cvss: 7.5
epss: null
exploited: false
status: "withdrawn"
category: "ai-tooling"
products: [{"name": "vllm", "fixed": ["0.30.0"], "affected": ">= 0, < 0.30.0", "ecosystem": "pypi"}]
ecosystems: ["pypi"]
tags: ["ai-tooling", "cwe-770", "llm-serving", "python", "self-hosted", "vllm"]
not_affected_if: ["default-config", "no-p2p-offloading"]
published_at: "2026-09-22T00:30:55+00:00"
first_seen_at: "2026-10-27T14:47:35+00:00"
updated_at: "2026-10-27T14:49:40+00:00"
sources: [{"source": "ghsa", "url": "https://github.com/advisories/GHSA-8qpc-6rxx-4pxf", "first_seen_at": "2026-10-27T14:47:35+00:00"}]
---

# Duplicate Advisory: vllm p2p never connected zmq_exhaustion

A crafted kv_transfer_params request with an arbitrary remote host and port makes vLLM keep ZeroMQ sockets open for unreachable peers until the context quota runs out. The resulting uncaught ZMQError crashes EngineCore and stops all inference.

_Summary, check and fix written by a model from the sources; labelled as such._

## Affected

- pypi:vllm affected >= 0, < 0.30.0 fixed 0.30.0

## How to check

- Run `pip show vllm` and confirm the installed version is below `0.30.0`.
- Check whether the server is configured with OffloadingConnector and TieringOffloadingSpec with a peer-to-peer secondary tier.
- unknown for anything beyond that; the sources do not give further checks.

## Fix

- Upgrade vllm to `0.30.0` or later.
- The advisory was withdrawn as a duplicate of GHSA-frvg-qw88-hww9; check that advisory for the authoritative fix guidance.

## Timeline

- 2026-10-27T14:47:35+00:00 ghsa https://github.com/advisories/GHSA-8qpc-6rxx-4pxf

## From the sources

## Duplicate Advisory

This advisory has been withdrawn because it is a duplicate of GHSA-frvg-qw88-hww9. This link is maintained to preserve external references.

## Original Description
vLLM through 0.29.0 contains a denial of service vulnerability in P2P KV offloading when OffloadingConnector is configured with TieringOffloadingSpec and a peer-to-peer secondary tier. Attackers can supply arbitrary remote host and port values in kv_transfer_params to create unreachable peer sessions that retain ZeroMQ sockets until the context quota is exhausted, causing an uncaught ZMQError that crashes EngineCore and stops all inference.

---

_Free to read at a human pace. An API with keys and an MCP server are in preparation; early access: hello@infrastack.ai_