This commit is contained in:
Your Name
2026-04-17 14:07:50 +08:00
parent dc34a4e294
commit 1921318c25
11 changed files with 2415 additions and 19 deletions

View File

@@ -4,12 +4,77 @@ import logging
import time
from app.core.config import settings
from app.services.debug_event_service import push_debug_event
from app.services.detect_job_service import get_active_detect_job_summary, get_detect_queue_health, list_recent_detect_run_events
from app.services.sync_push_service import pull_detect_task_batch_now, push_runtime_projection_now
logger = logging.getLogger("domaincheck.sync_agent")
def _emit_structured_tick(
*,
base_event_type: str,
ok: bool,
message: str,
data: dict | None = None,
) -> None:
payload = {"ok": ok, "data": data or {}}
event_type = f"{base_event_type}_failed"
level = "warning"
if ok:
event_type = f"{base_event_type}_success"
level = "info"
if "但远端确认失败" in str(message or ""):
event_type = f"{base_event_type}_partial"
level = "warning"
results = list((data or {}).get("results") or [])
if results and any(not bool(item.get("ok")) for item in results):
event_type = f"{base_event_type}_partial"
level = "warning"
push_debug_event(
service="sync-agent",
event_type=event_type,
level=level,
message=message,
payload=payload,
)
def _emit_sync_result_breakdown(data: dict | None) -> None:
results = list((data or {}).get("results") or [])
for item in results:
sync_type = str(item.get("sync_type") or "").strip() or "unknown"
ok = bool(item.get("ok"))
message = str(item.get("message") or "").strip() or f"{sync_type} sync result"
result_data = item.get("data") or {}
event_type = f"{sync_type}_sync_failed"
level = "warning"
if ok:
event_type = f"{sync_type}_sync_success"
level = "info"
if isinstance(result_data, dict) and result_data.get("success_count") is not None:
batch_count = int(result_data.get("batch_count") or 0)
success_count = int(result_data.get("success_count") or 0)
if batch_count > 0 and success_count < batch_count:
event_type = f"{sync_type}_sync_partial"
level = "warning"
if "成功 1/" in message or "" in message:
event_type = f"{sync_type}_sync_partial"
level = "warning"
push_debug_event(
service="sync-agent",
event_type=event_type,
level=level,
message=message,
payload={
"sync_type": sync_type,
"ok": ok,
"data": result_data,
},
)
def main() -> None:
logging.basicConfig(
level=logging.INFO,
@@ -28,10 +93,85 @@ def main() -> None:
try:
ok, message, data = push_runtime_projection_now()
logger.info("sync tick: ok=%s message=%s data=%s", ok, message, data)
push_debug_event(
service="sync-agent",
event_type="sync_tick",
level="info" if ok else "warning",
message=message,
payload={"ok": ok, "data": data},
)
_emit_structured_tick(base_event_type="sync_push", ok=ok, message=message, data=data)
_emit_sync_result_breakdown(data)
pull_ok, pull_message, pull_data = pull_detect_task_batch_now()
logger.info("task pull tick: ok=%s message=%s data=%s", pull_ok, pull_message, pull_data)
push_debug_event(
service="sync-agent",
event_type="task_pull_tick",
level="info" if pull_ok else "warning",
message=pull_message,
payload={"ok": pull_ok, "data": pull_data},
)
_emit_structured_tick(base_event_type="task_pull", ok=pull_ok, message=pull_message, data=pull_data)
active_job = get_active_detect_job_summary(event_limit=10)
if active_job:
queue_health = get_detect_queue_health(window_minutes=15)
recent_events = list_recent_detect_run_events(limit=8)
push_debug_event(
service="detect-runtime",
event_type="active_job_snapshot",
level="info",
message=f"active job {active_job.get('job_code', '')} status={active_job.get('status', '')}",
payload={
"job": {
"job_id": active_job.get("job_id"),
"job_code": active_job.get("job_code", ""),
"status": active_job.get("status", ""),
"items_total": active_job.get("items_total", 0),
"items_pending": active_job.get("items_pending", 0),
"items_claimed": active_job.get("items_claimed", 0),
"items_running": active_job.get("items_running", 0),
"items_completed": active_job.get("items_completed", 0),
"items_failed": active_job.get("items_failed", 0),
"progress_percent": active_job.get("progress_percent", 0),
"node_stats": list(active_job.get("node_stats") or []),
},
"queue_health": queue_health,
"recent_events": recent_events,
},
)
if queue_health.get("queue", {}).get("overdue_leases", 0):
push_debug_event(
service="detect-runtime",
event_type="queue_overdue_leases",
level="warning",
message=f"检测队列存在过期租约 {queue_health.get('queue', {}).get('overdue_leases', 0)}",
payload=queue_health,
)
for event in recent_events:
event_type = str(event.get("event_type") or "").strip()
if event_type not in {"domain_started", "domain_completed", "domain_failed", "domain_blacklisted"}:
continue
push_debug_event(
service="worker-event",
event_type=event_type,
level=str(event.get("level") or "info"),
message=str(event.get("message") or "").strip(),
payload={
"job_id": event.get("job_id"),
"node_code": event.get("node_code", ""),
"created_at": event.get("created_at", ""),
**(event.get("payload") or {}),
},
)
except Exception as exc:
logger.exception("sync tick failed: %s", exc)
push_debug_event(
service="sync-agent",
event_type="sync_tick_failed",
level="error",
message=str(exc),
payload={},
)
time.sleep(interval)