Skip to content
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
1 change: 1 addition & 0 deletions tensorrt_llm/llmapi/disagg_utils.py
Original file line number Diff line number Diff line change
Expand Up @@ -58,6 +58,7 @@ class MetadataServerConfig():
hostname: str = "localhost"
port: int = 2379
health_check_timeout: float = 5.0
refresh_interval: float = 10.0


def parse_disagg_config_file(yaml_config_file: str):
Expand Down
14 changes: 7 additions & 7 deletions tensorrt_llm/serve/openai_disagg_server.py
Original file line number Diff line number Diff line change
Expand Up @@ -35,20 +35,20 @@
class OpenAIDisaggServer:

def __init__(self,
ctx_servers: List[str] = None,
gen_servers: List[str] = None,
ctx_servers: List[str],
gen_servers: List[str],
req_timeout_secs: int = 180,
server_start_timeout_secs: int = 180,
ctx_router_config: Optional[RouterConfig] = None,
gen_router_config: Optional[RouterConfig] = None,
conditional_disagg_config: Optional[ConditionalDisaggConfig] = None,
metadata_server_cfg: MetadataServerConfig = None):
metadata_server_cfg: Optional[MetadataServerConfig] = None):

self.ctx_servers = ctx_servers
self.gen_servers = gen_servers
self.metadata_server = create_metadata_server(metadata_server_cfg)
self.ctx_router = create_router(ctx_router_config, ctx_servers, self.metadata_server)
self.gen_router = create_router(gen_router_config, gen_servers, self.metadata_server)
self.ctx_router = create_router(ctx_router_config, ctx_servers, metadata_server_cfg, self.metadata_server)
self.gen_router = create_router(gen_router_config, gen_servers, metadata_server_cfg, self.metadata_server)
self.conditional_disagg_config = conditional_disagg_config


Expand Down Expand Up @@ -76,8 +76,8 @@ async def lifespan(app: FastAPI):

if self.metadata_server:
logger.info("Starting server monitoring via metadata service")
await self.ctx_router.start_server_monitoring()
await self.gen_router.start_server_monitoring()
await self.ctx_router.start_server_monitoring(metadata_server_cfg.refresh_interval)
await self.gen_router.start_server_monitoring(metadata_server_cfg.refresh_interval)

yield

Expand Down
Loading