CyberRota Analysis
AI-GeneratedVersions of vLLM up to 0.29.0 are vulnerable to a denial of service due to improper handling of metadata in the NIXL connector, which can be exploited by sending malformed requests. This flaw leads to an uncaught KeyError that causes the decode engine to crash, resulting in service disruption until a manual restart is performed. Organizations utilizing vLLM should prioritize patching this vulnerability to maintain service availability and prevent potential exploitation.
Public Exploit Signal
A public exploit, PoC, GitHub repository or Metasploit reference was detected for this CVE.
Note: these links are listed for security research and verification purposes only.
Original NVD Description
vLLM versions through 0.29.0 contain a denial of service vulnerability in the NIXL connector's metadata handling for prefill/decode disaggregated deployments. Attackers can send requests with incomplete kv_transfer_params dictionary entries to trigger an uncaught KeyError in EngineCore scheduling, causing the decode engine to terminate and making all routed requests fail until manual restart.
Related CVEs
Other vulnerabilities affecting the same vendor(s)