Description

vLLM, an inference and serving engine for large language models (LLMs), has an issue in versions 0.6.5 through 0.8.4 that ONLY impacts environments using the `PyNcclPipe` KV cache transfer integration with the V0 engine. No other configurations are affected. vLLM supports the use of the `PyNcclPipe` class to establish a peer-to-peer communication domain for data transmission between distributed nodes. The GPU-side KV-Cache transmission is implemented through the `PyNcclCommunicator` class, while CPU-side control message passing is handled via the `send_obj` and `recv_obj` methods on the CPU side.​ The intention was that this interface should only be exposed to a private network using the IP address specified by the `--kv-ip` CLI parameter. The vLLM documentation covers how this must be limited to a secured network. The default and intentional behavior from PyTorch is that the `TCPStore` interface listens on ALL interfaces, regardless of what IP address is provided. The IP address given was only used as a client-side address to use. vLLM was fixed to use a workaround to force the `TCPStore` instance to bind its socket to a specified private interface. As of version 0.8.5, vLLM limits the `TCPStore` socket to the private interface as configured.

Metrics

 

CVSS Version 4.0 CVSS Version 3.x CVSS Version 2.0

NVD enrichment efforts reference publicly available information to associate vector strings. CVSS information contributed by other sources is also displayed.

CVSS 4.0 Severity and Vector Strings:

NIST: NVD

N/A

NVD assessment not yet provided.

CVSS 3.x Severity and Vector Strings:

NIST: NVD

Base Score:  N/A

NVD assessment not yet provided.

CNA:  GitHub, Inc.

Base Score:  9.8 CRITICAL

Vector:  CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H

CVSS 2.0 Severity and Vector Strings:

NIST: NVD

Base Score: N/A

NVD assessment not yet provided.

vLLM, an inference and serving engine for large language models (LLMs), has an issue in versions 0.6.5 through 0.8.4 that ONLY impacts environments using the `PyNcclPipe` KV cache transfer integration with the V0 engine. No other configurations are affected. vLLM supports the use of the `PyNcclPipe` class to establish a peer-to-peer communication domain for data transmission between distributed nodes. The GPU-side KV-Cache transmission is implemented through the `PyNcclCommunicator` class, while

AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H

CWE-502

https://docs.vllm.ai/en/latest/deployment/security.html

https://github.com/vllm-project/vllm/commit/0d6e187e88874c39cda7409cf673f9e6546893e7

https://github.com/vllm-project/vllm/pull/15988

https://github.com/vllm-project/vllm/security/advisories/GHSA-hjq4-87xh-g4fv