xinyun/vllm
1
0
Fork 0
mirror of https://git.datalinker.icu/vllm-project/vllm.git synced 2026-04-03 00:17:11 +08:00
Code Issues Packages Projects Releases Wiki Activity

format

Browse Source 
Signed-off-by: inkcherry <mingzhi.liu@amd.com>
This commit is contained in:
inkcherry 2025-12-23 10:47:17 +00:00
parent 8ef0b5f509
commit d2a18332b7
2 changed files with 18 additions and 11 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

21
examples/online_serving/disaggregated_serving/moriio_toy_proxy_server.py
View File

@ -210,11 +210,16 @@ async def handle_request():
prefill_instance_endpoint = None
decode_instance_endpoint = None
error_msg = (
"Service Unavailable: No prefill or decode instances are registered."
)
if not prefill_instances or not decode_instances:
return await make_response(
("Service Unavailable: No prefill or decode instances are registered.",
503))
(
error_msg,
503,
)
)
pid = request_nums % len(prefill_instances)
did = request_nums % len(decode_instances)
prefill_instance_endpoint = prefill_instances[pid]
@ -297,10 +302,12 @@ async def handle_request():
return response
except Exception as e:
logger.exception("An error occurred while handling the request: %s", e)
return await make_response((
f"Internal Server Error: {e!s}",
500,
))
return await make_response(
(
f"Internal Server Error: {e!s}",
500,
)
)
if __name__ == "__main__":

8
vllm/distributed/kv_transfer/kv_connector/v1/moriio/moriio_connector.py
View File

@ -342,8 +342,8 @@ class MoRIIOConnectorScheduler:
local_block_ids = blocks.get_block_ids()[0]
self._reqs_need_save[request.request_id] = (request, local_block_ids)
if params is not None and params.get("do_remote_prefill"): #
if self.mode == MoRIIOMode.READ: #read mode decode
if params is not None and params.get("do_remote_prefill"):
if self.mode == MoRIIOMode.READ:
if remote_block_ids := params.get("remote_block_ids"):
if all(
p in params
@ -373,7 +373,7 @@ class MoRIIOConnectorScheduler:
)
else:
assert request.kv_transfer_params is not None, ( #write mode decode
assert request.kv_transfer_params is not None, (
"kv_transfer_params should not be None"
)
@ -890,7 +890,7 @@ class MoRIIOConnectorWorker:
layer_name_to_local_kv_cache_metadata: dict,
):
"""Background thread for getting new MoRIIO handshakes."""
logger.info("tmp")
encoder = msgspec.msgpack.Encoder()
encoded_data = encoder.encode(metadata)
size_in_bytes = len(encoded_data)