[Misc] Remove orphaned env vars and stale env-var references (#44749)

Signed-off-by: Daoyuan Li <[email protected]>

docs/configuration/env_vars.md

@@ -5,7 +5,7 @@ vLLM uses the following environment variables to configure the system:
 !!! warning
     Please note that `VLLM_PORT` and `VLLM_HOST_IP` set the port and ip for vLLM's **internal usage**. It is not the port and ip for the API server. If you use `--host $VLLM_HOST_IP` and `--port $VLLM_PORT` to start the API server, it will not work.
 
-    All environment variables used by vLLM are prefixed with `VLLM_`. **Special care should be taken for Kubernetes users**: please do not name the service as `vllm`, otherwise environment variables set by Kubernetes might conflict with vLLM's environment variables, because [Kubernetes sets environment variables for each service with the capitalized service name as the prefix](https://kubernetes.io/docs/concepts/services-networking/service/#environment-variables).
+    Most vLLM-specific environment variables are prefixed with `VLLM_` (a handful of standard names — for example `CUDA_VISIBLE_DEVICES`, `MAX_JOBS`, `S3_ACCESS_KEY_ID`/`S3_SECRET_ACCESS_KEY`/`S3_ENDPOINT_URL`, `DO_NOT_TRACK`, `NO_COLOR` — are also read directly when set). **Special care should be taken for Kubernetes users**: please do not name the service as `vllm`, otherwise environment variables set by Kubernetes might conflict with vLLM's environment variables, because [Kubernetes sets environment variables for each service with the capitalized service name as the prefix](https://kubernetes.io/docs/concepts/services-networking/service/#environment-variables).
 
 ```python
 --8<-- "vllm/envs.py:env-vars-definition"

docs/serving/expert_parallel_deployment.md

@@ -220,8 +220,6 @@ For multi-node deployment, add these EPLB flags to each node's command. We recom
 
 - Use simulator flags `VLLM_MOE_ROUTING_SIMULATION_STRATEGY=uniform_random` and `VLLM_RANDOMIZE_DP_DUMMY_INPUTS=1` so token routing is balanced across EP ranks.
 
-- Increasing `VLLM_MOE_DP_CHUNK_SIZE` may increase throughput by increasing the maximum batch size for inter-rank token transfers. This may cause DeepEP  to throw `assert self.nvshmem_qp_depth >= (num_max_dispatch_tokens_per_rank + 1) * 2`, which can be fixed by increasing environment variable `NVSHMEM_QP_DEPTH`.
-
 ## Disaggregated Serving (Prefill/Decode Split)
 
 For production deployments requiring strict SLA guarantees for time-to-first-token and inter-token latency, disaggregated serving allows independent scaling of prefill and decode operations.

vllm/envs.py

@@ -127,7 +127,6 @@ if TYPE_CHECKING:
     VLLM_MXFP8_EMULATION_DEQUANT_AT_LOAD: bool = True
     VLLM_ROCM_USE_AITER: bool = False
     VLLM_ROCM_USE_AITER_CUSTOM_AR: bool = True
-    VLLM_ROCM_USE_AITER_PAGED_ATTN: bool = False
     VLLM_ROCM_USE_AITER_LINEAR: bool = True
     VLLM_ROCM_USE_AITER_LINEAR_HIPBMM: bool = False
     VLLM_ROCM_USE_AITER_MOE: bool = True
@@ -182,8 +181,6 @@ if TYPE_CHECKING:
     VLLM_HUMMING_MOE_GEMM_TYPE: Literal["indexed", "grouped", "auto"] | None = None
     VLLM_DEEPEPLL_NVFP4_DISPATCH: bool = False
     VLLM_V1_USE_OUTLINES_CACHE: bool = False
-    VLLM_TPU_BUCKET_PADDING_GAP: int = 0
-    VLLM_TPU_MOST_MODEL_LEN: int | None = None
     VLLM_TPU_USING_PATHWAYS: bool = False
     VLLM_USE_DEEP_GEMM: bool = True
     VLLM_MOE_USE_DEEP_GEMM: bool = True
@@ -1194,11 +1191,6 @@ environment_variables: dict[str, Callable[[], Any]] = {
     "VLLM_ROCM_USE_AITER_CUSTOM_AR": lambda: (
         os.getenv("VLLM_ROCM_USE_AITER_CUSTOM_AR", "True").lower() in ("true", "1")
     ),
-    # Whether to use aiter paged attention.
-    # By default is disabled.
-    "VLLM_ROCM_USE_AITER_PAGED_ATTN": lambda: (
-        os.getenv("VLLM_ROCM_USE_AITER_PAGED_ATTN", "False").lower() in ("true", "1")
-    ),
     # use aiter linear op if aiter ops are enabled
     # The following list of related ops
     # - scaled_mm (per-tensor / rowwise)
@@ -1432,8 +1424,6 @@ environment_variables: dict[str, Callable[[], Any]] = {
     "VLLM_RAY_EXTRA_ENV_VARS_TO_COPY": lambda: os.getenv(
         "VLLM_RAY_EXTRA_ENV_VARS_TO_COPY", ""
     ),
-    # Whether to use S3 path for model loading in CI via RunAI Streamer
-    "VLLM_CI_USE_S3": lambda: os.environ.get("VLLM_CI_USE_S3", "0") == "1",
     # Use model_redirect to redirect the model name to a local folder.
     # `model_redirect` can be a json file mapping the model between
     # repo_id and local folder:
@@ -1482,16 +1472,6 @@ environment_variables: dict[str, Callable[[], Any]] = {
     "VLLM_V1_USE_OUTLINES_CACHE": lambda: (
         os.environ.get("VLLM_V1_USE_OUTLINES_CACHE", "0") == "1"
     ),
-    # Gap between padding buckets for the forward pass. So we have
-    # 8, we will run forward pass with [16, 24, 32, ...].
-    "VLLM_TPU_BUCKET_PADDING_GAP": lambda: (
-        int(os.environ["VLLM_TPU_BUCKET_PADDING_GAP"])
-        if "VLLM_TPU_BUCKET_PADDING_GAP" in os.environ
-        else 0
-    ),
-    "VLLM_TPU_MOST_MODEL_LEN": lambda: maybe_convert_int(
-        os.environ.get("VLLM_TPU_MOST_MODEL_LEN", None)
-    ),
     # Whether using Pathways
     "VLLM_TPU_USING_PATHWAYS": lambda: bool(
         "proxy" in os.getenv("JAX_PLATFORMS", "").lower()
@@ -1667,15 +1647,6 @@ environment_variables: dict[str, Callable[[], Any]] = {
     "VLLM_MAX_TOKENS_PER_EXPERT_FP4_MOE": lambda: int(
         os.getenv("VLLM_MAX_TOKENS_PER_EXPERT_FP4_MOE", "163840")
     ),
-    # Specifies the thresholds of the communicated tensor sizes under which
-    # vllm should use flashinfer fused allreduce. The variable should be a
-    # JSON with the following format:
-    #     { <world size>: <max size in mb> }
-    # Unspecified world sizes will fall back to
-    #     { 2: 64, 4: 1, <everything else>: 0.5 }
-    "VLLM_FLASHINFER_ALLREDUCE_FUSION_THRESHOLDS_MB": lambda: json.loads(
-        os.getenv("VLLM_FLASHINFER_ALLREDUCE_FUSION_THRESHOLDS_MB", "{}")
-    ),
     # MoE routing strategy selector.
     # See `RoutingSimulator.get_available_strategies()` # for available
     # strategies.
@@ -2147,7 +2118,6 @@ def compile_factors() -> dict[str, object]:
         "VLLM_DP_MASTER_PORT",
         "VLLM_NIXL_SIDE_CHANNEL_HOST",
         "VLLM_RANDOMIZE_DP_DUMMY_INPUTS",
-        "VLLM_CI_USE_S3",
         "VLLM_MODEL_REDIRECT_PATH",
         "VLLM_HOST_IP",
         "VLLM_FORCE_AOT_LOAD",