CVE-2026-92983
InternLM LMDeploy through 0.17.0 in DistServe prefill/decode disaggregation mode fails to release scheduler sessions because the proxy uses user-facing session IDs instead of internal scheduler keys. Unauthenticated attackers can send completion requests to the proxy endpoint that accumulate unreleased scheduler metadata and memory until the prefill worker is out-of-memory killed.
취약점 유형
InternLM LMDeploy through 0.17.0 in DistServe prefill/decode disaggregation mode fails to release scheduler sessions because the proxy uses user-facing session IDs instead of internal scheduler keys. Unauthenticated attackers can send completion requests to the proxy endpoint that accumulate unreleased scheduler metadata and memory until the prefill worker is out-of-memory killed.
참고 자료
- https://github.com/InternLM/lmdeploy
- https://github.com/InternLM/lmdeploy/blob/v0.17.0/lmdeploy/pytorch/disagg/conn/engine_conn.py#L93-L99
- https://github.com/InternLM/lmdeploy/issues/4967
- https://www.vulncheck.com/advisories/internlm-lmdeploy-through-0.17.0-memory-exhaustion-via-session-id-mismatch
- https://github.com/InternLM/lmdeploy/issues/4967
공격자보다 먼저 취약점을 찾으세요.
GitHub로 로그인하고 1분 안에 첫 감사를 실행하세요. 무료 플랜은 신용카드가 필요 없습니다.