fix(security): Add VLLM_MAX_N_SEQUENCES environment variable and enforce limit (#37952)
Signed-off-by: jperezde <jperezde@redhat.com> Signed-off-by: Russell Bryant <rbryant@redhat.com> Co-authored-by: Russell Bryant <rbryant@redhat.com>
This commit is contained in:
co-authored by
Russell Bryant
parent
497e234d38
commit
b111f8a61f
@@ -86,6 +86,7 @@ if TYPE_CHECKING:
|
||||
VLLM_ALLOW_LONG_MAX_MODEL_LEN: bool = False
|
||||
VLLM_RPC_TIMEOUT: int = 10000 # ms
|
||||
VLLM_HTTP_TIMEOUT_KEEP_ALIVE: int = 5 # seconds
|
||||
VLLM_MAX_N_SEQUENCES: int = 16384
|
||||
VLLM_PLUGINS: list[str] | None = None
|
||||
VLLM_LORA_RESOLVER_CACHE_DIR: str | None = None
|
||||
VLLM_LORA_RESOLVER_HF_REPO_LIST: str | None = None
|
||||
@@ -870,6 +871,12 @@ environment_variables: dict[str, Callable[[], Any]] = {
|
||||
"VLLM_HTTP_TIMEOUT_KEEP_ALIVE": lambda: int(
|
||||
os.environ.get("VLLM_HTTP_TIMEOUT_KEEP_ALIVE", "5")
|
||||
),
|
||||
# Maximum allowed value for the `n` sampling parameter (number of output
|
||||
# sequences per request). Limits resource consumption to prevent
|
||||
# denial-of-service via excessively large fan-out. Default: 16384.
|
||||
"VLLM_MAX_N_SEQUENCES": lambda: int(
|
||||
os.environ.get("VLLM_MAX_N_SEQUENCES", "16384")
|
||||
),
|
||||
# a list of plugin names to load, separated by commas.
|
||||
# if this is not set, it means all plugins will be loaded
|
||||
# if this is set to an empty string, no plugins will be loaded
|
||||
|
||||
Reference in New Issue
Block a user