xinyun/vllm
1
0
Fork 0
mirror of https://git.datalinker.icu/vllm-project/vllm.git synced 2026-01-09 06:25:16 +08:00
Code Issues Packages Projects Releases Wiki Activity

[Frontend] Add request_id to the Request object so they can be controlled better via external load balancers (#21009)

Browse Source 
Signed-off-by: Kourosh Hakhamaneshi <kourosh@anyscale.com>
This commit is contained in:
kourosh hakhamaneshi 2025-07-25 06:49:11 -07:00 committed by GitHub
parent 29c6fbe58c
commit 9fe98d4250
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
3 changed files with 27 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

21
vllm/entrypoints/openai/protocol.py
View File

@ -1007,6 +1007,13 @@ class CompletionRequest(OpenAIBaseModel):
"default: 0). Any priority other than 0 will raise an error "
"if the served model does not use priority scheduling."),
)
request_id: str = Field(
default_factory=lambda: f"{random_uuid()}",
description=(
"The request_id related to this request. If the caller does "
"not set it, a random_uuid will be generated. This id is used "
"through out the inference process and return in response."),
)
logits_processors: Optional[LogitsProcessors] = Field(
default=None,
description=(
@ -1251,6 +1258,13 @@ class EmbeddingCompletionRequest(OpenAIBaseModel):
"default: 0). Any priority other than 0 will raise an error "
"if the served model does not use priority scheduling."),
)
request_id: str = Field(
default_factory=lambda: f"{random_uuid()}",
description=(
"The request_id related to this request. If the caller does "
"not set it, a random_uuid will be generated. This id is used "
"through out the inference process and return in response."),
)
# --8<-- [end:embedding-extra-params]
@ -1302,6 +1316,13 @@ class EmbeddingChatRequest(OpenAIBaseModel):
"default: 0). Any priority other than 0 will raise an error "
"if the served model does not use priority scheduling."),
)
request_id: str = Field(
default_factory=lambda: f"{random_uuid()}",
description=(
"The request_id related to this request. If the caller does "
"not set it, a random_uuid will be generated. This id is used "
"through out the inference process and return in response."),
)
# --8<-- [end:chat-embedding-extra-params]
@model_validator(mode="before")

4
vllm/entrypoints/openai/serving_completion.py
View File

@ -113,7 +113,9 @@ class OpenAIServingCompletion(OpenAIServing):
return self.create_error_response(
"Echo is unsupported with prompt embeds.")
request_id = f"cmpl-{self._base_request_id(raw_request)}"
request_id = (
f"cmpl-"
f"{self._base_request_id(raw_request, request.request_id)}")
created_time = int(time.time())
request_metadata = RequestResponseMetadata(request_id=request_id)

5
vllm/entrypoints/openai/serving_embedding.py
View File

@ -163,8 +163,9 @@ class OpenAIServingEmbedding(EmbeddingMixin):
for the API specification. This API mimics the OpenAI Embedding API.
"""
model_name = self._get_model_name(request.model)
request_id = (f"{self.request_id_prefix}-"
f"{self._base_request_id(raw_request)}")
request_id = (
f"{self.request_id_prefix}-"
f"{self._base_request_id(raw_request, request.request_id)}")
ctx = EmbeddingServeContext(
request=request,