This commit is contained in:
Your Name
2026-04-22 14:13:21 +08:00
parent e0406b5d0e
commit 7cbde2aa78
145 changed files with 23086 additions and 2243 deletions

View File

@@ -4,6 +4,7 @@ from uuid import uuid4
from fastapi import APIRouter
from app.core.config import settings
from app.schemas.common import ApiResponse
from app.services.detect_job_service import (
append_detect_job_event,
@@ -12,6 +13,7 @@ from app.services.detect_job_service import (
get_detect_job_summary,
get_detect_queue_health,
list_detect_jobs,
normalize_detect_step_code,
)
from app.services.detect_service import get_detect_status
from app.services.detect_run_service import create_detect_run_snapshot, finalize_detect_run, mark_detect_run_stopping
@@ -22,6 +24,13 @@ from app.services.worker_control_service import send_worker_command, start_worke
router = APIRouter(tags=["detect"])
def _local_worker_expected_on_this_node() -> bool:
return not (
str(settings.node_region or "").strip() == "overseas"
and str(settings.node_role or "").strip() == "control"
)
def _build_detect_action_result(
*,
action: str,
@@ -255,6 +264,11 @@ def detect_queue_summary(window_minutes: int = 15) -> ApiResponse:
return ApiResponse(data=get_detect_queue_health(window_minutes=window_minutes))
@router.get("/detect/queue-health", response_model=ApiResponse)
def detect_queue_health_alias(window_minutes: int = 15) -> ApiResponse:
return ApiResponse(data=get_detect_queue_health(window_minutes=window_minutes))
@router.get("/detect/jobs/{job_id}", response_model=ApiResponse)
def detect_job_detail(job_id: int) -> ApiResponse:
data = get_detect_job_summary(job_id, event_limit=100)
@@ -264,8 +278,17 @@ def detect_job_detail(job_id: int) -> ApiResponse:
@router.post("/detect/start", response_model=ApiResponse)
def start_detect() -> ApiResponse:
job_summary = create_detect_job_if_needed(limit=1000, created_by="api")
def start_detect(step_code: str | None = None) -> ApiResponse:
normalized_step_code = normalize_detect_step_code(step_code)
if step_code and not normalized_step_code:
result = _build_detect_action_result(
action="start",
ok=False,
message=f"暂不支持步骤任务: {step_code}",
data={"job": None, "step_code": str(step_code or "").strip()},
)
return ApiResponse(code=1, message=f"暂不支持步骤任务: {step_code}", data=result)
job_summary = create_detect_job_if_needed(limit=10000, created_by="api", step_code=step_code)
if not job_summary:
result = _build_detect_action_result(
action="start",
@@ -292,42 +315,58 @@ def start_detect() -> ApiResponse:
},
)
ok, message = start_worker()
if not ok:
result = _build_detect_action_result(
action="start",
ok=False,
message=message,
data={"job": job_summary},
local_worker_expected = _local_worker_expected_on_this_node()
if local_worker_expected:
ok, message = start_worker()
if not ok:
result = _build_detect_action_result(
action="start",
ok=False,
message=message,
data={"job": job_summary},
)
append_detect_job_event(
job_summary["job_id"],
event_type="job_dispatch_failed",
level="error",
message=f"启动 Worker 失败: {message}",
payload={"cycle_token": cycle_token},
)
return ApiResponse(
code=1,
message=message,
data=result,
)
command_ok, command_message = send_worker_command(
"start_detection",
payload={
"cycle_token": cycle_token,
"job_id": job_summary["job_id"],
"job_code": job_summary["job_code"],
"task_mode": job_summary.get("task_mode", ""),
"step_code": job_summary.get("step_code", ""),
},
)
append_detect_job_event(
job_summary["job_id"],
event_type="job_dispatch_failed",
level="error",
message=f"启动 Worker 失败: {message}",
event_type="job_dispatch_sent" if command_ok else "job_dispatch_rejected",
level="info" if command_ok else "error",
message=command_message,
payload={"cycle_token": cycle_token},
)
return ApiResponse(
code=1,
message=message,
data=result,
else:
ok = True
message = "当前节点为海外控制面,仅负责派单,不启动本机 Worker"
command_ok = True
command_message = "已跳过本机 Worker 启动,转为仅向大陆执行节点派发检测动作"
append_detect_job_event(
job_summary["job_id"],
event_type="job_dispatch_skipped_local",
level="info",
message=command_message,
payload={"cycle_token": cycle_token},
)
command_ok, command_message = send_worker_command(
"start_detection",
payload={
"cycle_token": cycle_token,
"job_id": job_summary["job_id"],
"job_code": job_summary["job_code"],
},
)
append_detect_job_event(
job_summary["job_id"],
event_type="job_dispatch_sent" if command_ok else "job_dispatch_rejected",
level="info" if command_ok else "error",
message=command_message,
payload={"cycle_token": cycle_token},
)
snapshot = get_detect_status()
settings_payload = get_settings_payload()
settings_summary = _build_settings_summary(settings_payload)

View File

@@ -9,6 +9,7 @@ from app.services.ops_agent_service import (
agent_heartbeat,
agent_mark_job_started,
agent_pull_jobs,
agent_pull_runtime_config,
agent_register,
build_node_agent_bootstrap_plan,
issue_node_agent_token,
@@ -72,6 +73,12 @@ def ops_agent_pull(payload: dict, limit: int = 1, x_domaincheck_agent_token: Opt
return _build_agent_response(ok, message, data)
@router.post("/ops/agent/runtime-config", response_model=ApiResponse)
def ops_agent_runtime_config(payload: dict, x_domaincheck_agent_token: Optional[str] = Header(default=None)) -> ApiResponse:
ok, message, data = agent_pull_runtime_config(payload, token=_resolve_agent_token(x_domaincheck_agent_token))
return _build_agent_response(ok, message, data)
@router.post("/ops/agent/jobs/{job_id}/start", response_model=ApiResponse)
def ops_agent_job_start(job_id: int, payload: dict, x_domaincheck_agent_token: Optional[str] = Header(default=None)) -> ApiResponse:
ok, message, data = agent_mark_job_started(job_id, payload, token=_resolve_agent_token(x_domaincheck_agent_token))

View File

@@ -1,11 +1,12 @@
from typing import Optional
from fastapi import APIRouter, Header
from fastapi import APIRouter, Body, Header
from app.schemas.common import ApiResponse
from app.services.build_info_service import get_runtime_build_info
from app.services.cluster_runtime_service import get_cluster_snapshot
from app.services.debug_event_service import get_debug_diagnosis, get_debug_event_overview, get_debug_handoff_report, ingest_debug_event, list_debug_events
from app.services.detect_job_service import get_detect_queue_health
from app.services.runtime_control_service import runtime_action
from app.services.runtime_status_service import get_runtime_preflight, get_runtime_status
from app.services.sync_push_service import (
@@ -103,6 +104,36 @@ def runtime_debug_handoff(window_minutes: int = 10, source_region: Optional[str]
)
@router.get("/runtime/queue-health", response_model=ApiResponse)
def runtime_queue_health(window_minutes: int = 15) -> ApiResponse:
return ApiResponse(data=get_detect_queue_health(window_minutes=window_minutes))
@router.get("/runtime/health-handover", response_model=ApiResponse)
def runtime_health_handover(
node_code: Optional[str] = None,
window_minutes: int = 10,
source_region: Optional[str] = None,
) -> ApiResponse:
sync_summary = get_sync_summary()
runtime_status_payload = get_runtime_status()
data = get_debug_handoff_report(
window_minutes=window_minutes,
source_region=source_region,
sync_summary=sync_summary,
readiness=runtime_status_payload.get("readiness") or {},
)
normalized_node_code = str(node_code or "").strip()
if normalized_node_code:
data = dict(data)
data["nodes"] = [
item
for item in list(data.get("nodes") or [])
if str(item.get("node_code") or "").strip() == normalized_node_code
]
return ApiResponse(data=data)
@router.post("/runtime/sync-ingest", response_model=ApiResponse)
def runtime_sync_ingest(payload: dict, x_domaincheck_sync_token: Optional[str] = Header(default=None)) -> ApiResponse:
ok, message, data = ingest_runtime_projection(payload, shared_token=x_domaincheck_sync_token)
@@ -110,7 +141,7 @@ def runtime_sync_ingest(payload: dict, x_domaincheck_sync_token: Optional[str] =
@router.get("/runtime/task-export", response_model=ApiResponse)
def runtime_task_export(limit: int = 200, x_domaincheck_sync_token: Optional[str] = Header(default=None)) -> ApiResponse:
def runtime_task_export(limit: int = 1000, x_domaincheck_sync_token: Optional[str] = Header(default=None)) -> ApiResponse:
ok, message, data = export_detect_task_projection(limit=limit, shared_token=x_domaincheck_sync_token)
return ApiResponse(code=0 if ok else 1, message=message, data=data)
@@ -128,6 +159,6 @@ def runtime_debug_ingest(payload: dict, x_domaincheck_sync_token: Optional[str]
@router.post("/runtime/actions/{action}", response_model=ApiResponse)
def runtime_action_trigger(action: str) -> ApiResponse:
ok, message, data = runtime_action(action)
def runtime_action_trigger(action: str, payload: Optional[dict] = Body(default=None)) -> ApiResponse:
ok, message, data = runtime_action(action, payload=payload)
return ApiResponse(code=0 if ok else 1, message=message, data=data)