199df03524
### What this PR does / why we need it?
Conditional Import for Mooncake: The import of
mooncake.engine.TransferEngine was moved into a try-except block within
the GlobalTE class's constructor. This ensures that mooncake is only
imported when needed and provides a clear error message with
installation instructions if it's missing.
### Does this PR introduce _any_ user-facing change?
The error message "ascend_transport.so: cannot open shared object file:
No such file or directory" in the CI is fixed to ensure the normal
running of the CI.
- vLLM version: v0.17.0
- vLLM main:
4034c3d32e
---------
Signed-off-by: 房建伟 <fangjianwei@fangjianweideMacBook-Air.local>
Co-authored-by: 房建伟 <fangjianwei@fangjianweideMacBook-Air.local>
44 lines
2.0 KiB
Python
44 lines
2.0 KiB
Python
import threading
|
|
|
|
|
|
class GlobalTE:
|
|
def __init__(self):
|
|
self.transfer_engine = None
|
|
self.is_register_buffer: bool = False
|
|
self.transfer_engine_lock = threading.Lock()
|
|
self.register_buffer_lock = threading.Lock()
|
|
|
|
def get_transfer_engine(self, hostname: str, device_name: str | None):
|
|
if self.transfer_engine is None:
|
|
with self.transfer_engine_lock:
|
|
# Double-Checked Locking
|
|
if self.transfer_engine is None:
|
|
try:
|
|
from mooncake.engine import TransferEngine # type: ignore
|
|
except ImportError as e:
|
|
raise ImportError(
|
|
"Please install mooncake by following the instructions at "
|
|
"https://github.com/kvcache-ai/Mooncake/blob/main/doc/en/build.md " # noqa: E501
|
|
"to run vLLM with MooncakeConnector."
|
|
) from e
|
|
self.transfer_engine = TransferEngine()
|
|
device_name = device_name if device_name is not None else ""
|
|
ret_value = self.transfer_engine.initialize(hostname, "P2PHANDSHAKE", "ascend", device_name)
|
|
if ret_value != 0:
|
|
raise RuntimeError(f"TransferEngine initialization failed with ret_value: {ret_value}")
|
|
return self.transfer_engine
|
|
|
|
def register_buffer(self, ptrs: list[int], sizes: list[int]):
|
|
with self.register_buffer_lock:
|
|
assert self.transfer_engine is not None, "Transfer engine must be initialized"
|
|
if self.is_register_buffer:
|
|
return
|
|
for ptr, size in zip(ptrs, sizes):
|
|
ret_value = self.transfer_engine.register_memory(ptr, size)
|
|
if ret_value != 0:
|
|
raise RuntimeError("Mooncake memory registration failed.")
|
|
self.is_register_buffer = True
|
|
|
|
|
|
global_te = GlobalTE()
|