perf: Optimize average_time_cost calculation in job model

apps/ops/models/job.py

@@ -9,7 +9,8 @@ from datetime import timedelta
 from celery import current_task
 from django.conf import settings
 from django.db import models
-from django.db.models import Q
+from django.db.models import Q, Avg, F, FloatField
+from django.db.models.functions import Coalesce, Now, Extract
 from django.utils import timezone
 from django.utils.translation import gettext_lazy as _
 
@@ -188,11 +189,17 @@ class Job(JMSOrgBaseModel, PeriodTaskModelMixin):
 
     @property
     def average_time_cost(self):
-        total_cost = 0
-        finished_count = self.executions.filter(status__in=['success', 'failed']).count()
-        for execution in self.executions.filter(status__in=['success', 'failed']).all():
-            total_cost += execution.time_cost
-        return total_cost / finished_count if finished_count else 0
+        executions = self.executions.filter(status__in=['success', 'failed'])
+        if not executions.exists():
+            return 0
+        average = executions.annotate(
+            time_cost=Coalesce(F('date_finished'), Now()) - F('date_start')
+        ).annotate(
+            time_cost_seconds=Extract('time_cost', 'epoch')
+        ).aggregate(
+            avg_time=Avg('time_cost_seconds')
+        )['avg_time']
+        return average if average is not None else 0
 
     def get_register_task(self):
         from ..tasks import run_ops_job