[kv_offload] Add req_id to ReqContext for per-request tracking (#42507)
Signed-off-by: Ronen Schaffer <ronen.schaffer@ibm.com>
This commit is contained in:
@@ -627,7 +627,7 @@ def _make_scheduler_with_lookup(
|
||||
return scheduler
|
||||
|
||||
|
||||
_EMPTY_REQ_CTX = ReqContext()
|
||||
_EMPTY_REQ_CTX = ReqContext(req_id="")
|
||||
|
||||
|
||||
class TestMaximalPrefixLookup:
|
||||
|
||||
@@ -19,9 +19,11 @@ from vllm.v1.kv_offload.cpu.manager import CPUOffloadingManager
|
||||
from vllm.v1.kv_offload.cpu.policies.arc import ARCCachePolicy
|
||||
|
||||
|
||||
def make_req_context(kv_transfer_params: dict | None = None) -> ReqContext:
|
||||
def make_req_context(
|
||||
req_id: str = "", kv_transfer_params: dict | None = None
|
||||
) -> ReqContext:
|
||||
"""Create a ReqContext as production code would, from a request's params."""
|
||||
return ReqContext(kv_transfer_params=kv_transfer_params)
|
||||
return ReqContext(req_id=req_id, kv_transfer_params=kv_transfer_params)
|
||||
|
||||
|
||||
_EMPTY_REQ_CTX = make_req_context()
|
||||
|
||||
Reference in New Issue
Block a user