Merge branch 'master-dev' into 'master'
This commit is contained in:
commit
5436719bfb
39 files changed
+2395
-291
No files matched your search
@@ -22,7 +22,7 @@ from fastapi import APIRouter, Query, Request
|
||||
from pydantic import BaseModel
|
||||
from sqlalchemy import select
|
||||
|
||||
from app.gateway.deps import get_concurrency_sample_store, get_run_manager
|
||||
from app.gateway.deps import get_concurrency_gate, get_concurrency_sample_store, get_run_manager
|
||||
from app.gateway.routers.admin_users import _require_admin
|
||||
from deerflow.config.system_settings import load_system_settings, save_system_settings
|
||||
from deerflow.persistence.agents.model import AgentRow
|
||||
@@ -149,6 +149,15 @@ def _elapsed_seconds(created_at: str, now: datetime) -> int:
|
||||
return max(0, int(delta))
|
||||
|
||||
|
||||
def _elapsed_seconds_from_epoch(created_at: float, now: datetime) -> int:
|
||||
try:
|
||||
if created_at <= 0:
|
||||
return 0
|
||||
return max(0, int(now.timestamp() - float(created_at)))
|
||||
except Exception:
|
||||
return 0
|
||||
|
||||
|
||||
def _kind_for(metadata: dict | None) -> str:
|
||||
"""Coarse human label for the run's origin, from its thread metadata."""
|
||||
meta = metadata or {}
|
||||
@@ -218,14 +227,46 @@ async def list_active_runs(request: Request) -> ActiveRunsResponse:
|
||||
"""List every in-flight run with who/what is driving it (admin only)."""
|
||||
await _require_admin(request)
|
||||
run_mgr = get_run_manager(request)
|
||||
records = await run_mgr.list_active()
|
||||
gate = get_concurrency_gate(request)
|
||||
leases = await gate.list_active() if gate is not None else None
|
||||
records = [] if leases is not None else await run_mgr.list_active()
|
||||
now = datetime.now(UTC)
|
||||
maps = await _enrich(records)
|
||||
enrich_source = leases if leases is not None else records
|
||||
maps = await _enrich(enrich_source)
|
||||
thread_meta = maps["thread_meta"]
|
||||
user_emails = maps["user_emails"]
|
||||
agent_names = maps["agent_names"]
|
||||
|
||||
items: list[ActiveRunItem] = []
|
||||
if leases is not None:
|
||||
for r in leases:
|
||||
meta_row = thread_meta.get(r.thread_id)
|
||||
user_id = r.user_id or (meta_row.user_id if meta_row else "") or ""
|
||||
metadata = (meta_row.metadata_json if meta_row else None) or {}
|
||||
assistant_id = r.assistant_id or ""
|
||||
if assistant_id in _DEFAULT_ASSISTANT_IDS:
|
||||
agent_name = "榛樿鍔╂墜"
|
||||
else:
|
||||
agent_name = agent_names.get(assistant_id) or assistant_id
|
||||
items.append(
|
||||
ActiveRunItem(
|
||||
run_id=r.run_id,
|
||||
thread_id=r.thread_id,
|
||||
status="running",
|
||||
created_at=datetime.fromtimestamp(r.created_at, tz=UTC).isoformat() if r.created_at else "",
|
||||
elapsed_seconds=_elapsed_seconds_from_epoch(r.created_at, now),
|
||||
user_id=user_id,
|
||||
user_name=user_emails.get(user_id) or user_id or "鏈煡鐢ㄦ埛",
|
||||
agent_id=assistant_id,
|
||||
agent_name=agent_name,
|
||||
model_name=r.model_name,
|
||||
thread_title=(meta_row.display_name if meta_row else "") or "",
|
||||
kind=_kind_for(metadata),
|
||||
multitask_strategy="reject",
|
||||
)
|
||||
)
|
||||
return ActiveRunsResponse(count=len(items), server_time=now.isoformat(), runs=items)
|
||||
|
||||
for r in records:
|
||||
meta_row = thread_meta.get(r.thread_id)
|
||||
user_id = (meta_row.user_id if meta_row else "") or ""
|
||||
|
||||
@@ -65,6 +65,7 @@ from langgraph.checkpoint.base import empty_checkpoint
|
||||
from app.gateway.roundtable_diag import record_foreground_diag
|
||||
from app.gateway.roundtable_model_fallback import fallback_model_chain, is_llm_error_message, reason_text
|
||||
from app.gateway.roundtable_run_policy import merge_skill_stop_names, roundtable_run_policy
|
||||
from app.gateway.services import RUN_SCOPE_ORCHESTRATION_CHILD, internal_run_scope_headers
|
||||
from deerflow.agents.roundtable_orchestrator.extraction_spec import build_summary_directive
|
||||
|
||||
from app.gateway.roundtable_seat_skills import append_seat_skill_directive
|
||||
@@ -325,7 +326,11 @@ def _auth_headers(request: Request) -> dict[str, str]:
|
||||
token = request.headers.get("authorization")
|
||||
if not token:
|
||||
raise HTTPException(status_code=401, detail="Missing Authorization header")
|
||||
return {"Authorization": token, "Content-Type": "application/json"}
|
||||
return {
|
||||
"Authorization": token,
|
||||
"Content-Type": "application/json",
|
||||
**internal_run_scope_headers(),
|
||||
}
|
||||
|
||||
|
||||
async def _create_thread(
|
||||
@@ -694,6 +699,7 @@ def _run_payload(
|
||||
}
|
||||
]
|
||||
},
|
||||
"metadata": {"run_scope": RUN_SCOPE_ORCHESTRATION_CHILD},
|
||||
"config": {"recursion_limit": 1000},
|
||||
"context": {
|
||||
"agent_name": agent_name,
|
||||
|
||||
@@ -24,7 +24,6 @@ from deerflow.config.system_settings import (
|
||||
PublishSquare,
|
||||
PublishSquaresSettings,
|
||||
SkillCompressionSettings,
|
||||
SystemSettings,
|
||||
get_module_tag_ids,
|
||||
list_publish_squares,
|
||||
load_system_settings,
|
||||
@@ -226,12 +225,16 @@ class ModelConcurrencyResponse(BaseModel):
|
||||
enabled: bool = Field(..., description="Whether per-model concurrency limiting is active")
|
||||
default_limit: int = Field(..., description="Fallback max concurrent runs for models without an explicit override")
|
||||
model_limits: dict[str, int] = Field(default_factory=dict, description="Per-model overrides: model_name -> max concurrent runs")
|
||||
user_default_limit: int = Field(..., description="Fallback max concurrent runs per user and model. 0 disables the per-user fallback")
|
||||
user_model_limits: dict[str, int] = Field(default_factory=dict, description="Per-user per-model overrides: model_name -> max concurrent runs for one user")
|
||||
|
||||
|
||||
class ModelConcurrencyUpdateRequest(BaseModel):
|
||||
enabled: bool | None = None
|
||||
default_limit: int | None = Field(default=None, ge=1, le=10_000)
|
||||
model_limits: dict[str, int] | None = None
|
||||
user_default_limit: int | None = Field(default=None, ge=0, le=10_000)
|
||||
user_model_limits: dict[str, int] | None = None
|
||||
|
||||
|
||||
def _skill_compression_to_response(settings: SkillCompressionSettings) -> SkillCompressionResponse:
|
||||
@@ -255,10 +258,24 @@ def _model_concurrency_to_response(settings: ModelConcurrencySettings) -> ModelC
|
||||
if normalized_limit < 1:
|
||||
continue
|
||||
cleaned[model_name] = min(normalized_limit, 10_000)
|
||||
cleaned_user: dict[str, int] = {}
|
||||
for name, limit in (settings.user_model_limits or {}).items():
|
||||
model_name = str(name).strip()
|
||||
if not model_name:
|
||||
continue
|
||||
try:
|
||||
normalized_limit = int(limit)
|
||||
except (TypeError, ValueError):
|
||||
continue
|
||||
if normalized_limit < 1:
|
||||
continue
|
||||
cleaned_user[model_name] = min(normalized_limit, 10_000)
|
||||
return ModelConcurrencyResponse(
|
||||
enabled=settings.enabled,
|
||||
default_limit=max(1, min(int(settings.default_limit), 10_000)),
|
||||
model_limits=cleaned,
|
||||
user_default_limit=max(0, min(int(getattr(settings, "user_default_limit", 0) or 0), 10_000)),
|
||||
user_model_limits=cleaned_user,
|
||||
)
|
||||
|
||||
|
||||
@@ -336,6 +353,22 @@ async def update_model_concurrency(
|
||||
continue
|
||||
cleaned[model_name] = min(normalized_limit, 10_000)
|
||||
update["model_limits"] = cleaned
|
||||
if "user_default_limit" in fields and fields["user_default_limit"] is not None:
|
||||
update["user_default_limit"] = max(0, min(int(fields["user_default_limit"]), 10_000))
|
||||
if "user_model_limits" in fields and fields["user_model_limits"] is not None:
|
||||
cleaned_user: dict[str, int] = {}
|
||||
for name, limit in (body.user_model_limits or {}).items():
|
||||
model_name = str(name).strip()
|
||||
if not model_name:
|
||||
continue
|
||||
try:
|
||||
normalized_limit = int(limit)
|
||||
except (TypeError, ValueError):
|
||||
continue
|
||||
if normalized_limit < 1:
|
||||
continue
|
||||
cleaned_user[model_name] = min(normalized_limit, 10_000)
|
||||
update["user_model_limits"] = cleaned_user
|
||||
settings = current.model_concurrency.model_copy(update=update)
|
||||
saved = save_system_settings(current.model_copy(update={"model_concurrency": settings}))
|
||||
return _model_concurrency_to_response(saved.model_concurrency)
|
||||
|
||||
Reference in new issue
Block a user