HostMetric Cleanup task

awx/main/conf.py

@@ -822,6 +822,15 @@ register(
     category_slug='system',
 )
 
+register(
+    'CLEANUP_HOST_METRICS_LAST_TS',
+    field_class=fields.DateTimeField,
+    label=_('Last cleanup date for HostMetrics'),
+    allow_null=True,
+    category=_('System'),
+    category_slug='system',
+)
+
 
 def logging_validate(serializer, attrs):
     if not serializer.instance or not hasattr(serializer.instance, 'LOG_AGGREGATOR_HOST') or not hasattr(serializer.instance, 'LOG_AGGREGATOR_TYPE'):

awx/main/management/commands/cleanup_host_metrics.py

@@ -0,0 +1,22 @@
+from awx.main.models import HostMetric
+from django.core.management.base import BaseCommand
+from django.conf import settings
+
+
+class Command(BaseCommand):
+    """
+    Run soft-deleting of HostMetrics
+    """
+
+    help = 'Run soft-deleting of HostMetrics'
+
+    def add_arguments(self, parser):
+        parser.add_argument('--months-ago', type=int, dest='months-ago', action='store', help='Threshold in months for soft-deleting')
+
+    def handle(self, *args, **options):
+        months_ago = options.get('months-ago') or None
+
+        if not months_ago:
+            months_ago = getattr(settings, 'CLEANUP_HOST_METRICS_THRESHOLD', 12)
+
+        HostMetric.cleanup_task(months_ago)

awx/main/models/inventory.py

@@ -9,6 +9,8 @@ import re
 import copy
 import os.path
 from urllib.parse import urljoin
+
+import dateutil.relativedelta
 import yaml
 
 # Django
@@ -888,6 +890,16 @@ class HostMetric(models.Model):
             self.deleted = False
             self.save(update_fields=['deleted'])
 
+    @classmethod
+    def cleanup_task(cls, months_ago):
+        last_automation_before = now() - dateutil.relativedelta.relativedelta(months=months_ago)
+
+        logger.info(f'Cleanup [HostMetric]: soft-deleting records last automated before {last_automation_before}')
+        HostMetric.objects.filter(last_automation__lt=last_automation_before).update(
+            deleted=True, deleted_counter=models.F('deleted_counter') + 1, last_deleted=now()
+        )
+        settings.CLEANUP_HOST_METRICS_LAST_TS = now()
+
 
 class HostMetricSummaryMonthly(models.Model):
     """

awx/main/tasks/system.py

@@ -47,6 +47,7 @@ from awx.main.models import (
     Inventory,
     SmartInventoryMembership,
     Job,
+    HostMetric,
 )
 from awx.main.constants import ACTIVE_STATES
 from awx.main.dispatch.publish import task
@@ -378,6 +379,20 @@ def cleanup_images_and_files():
     _cleanup_images_and_files()
 
 
+@task(queue=get_task_queuename)
+def cleanup_host_metrics():
+    from awx.conf.models import Setting
+    from rest_framework.fields import DateTimeField
+
+    last_cleanup = Setting.objects.filter(key='CLEANUP_HOST_METRICS_LAST_TS').first()
+    last_time = DateTimeField().to_internal_value(last_cleanup.value) if last_cleanup and last_cleanup.value else None
+
+    cleanup_interval_secs = getattr(settings, 'CLEANUP_HOST_METRICS_INTERVAL', 30) * 86400
+    if not last_time or ((now() - last_time).total_seconds() > cleanup_interval_secs):
+        months_ago = getattr(settings, 'CLEANUP_HOST_METRICS_THRESHOLD', 12)
+        HostMetric.cleanup_task(months_ago)
+
+
 @task(queue=get_task_queuename)
 def cluster_node_health_check(node):
     """

awx/settings/defaults.py

@@ -475,6 +475,7 @@ CELERYBEAT_SCHEDULE = {
     'receptor_reaper': {'task': 'awx.main.tasks.system.awx_receptor_workunit_reaper', 'schedule': timedelta(seconds=60)},
     'send_subsystem_metrics': {'task': 'awx.main.analytics.analytics_tasks.send_subsystem_metrics', 'schedule': timedelta(seconds=20)},
     'cleanup_images': {'task': 'awx.main.tasks.system.cleanup_images_and_files', 'schedule': timedelta(hours=3)},
+    'cleanup_host_metrics': {'task': 'awx.main.tasks.system.cleanup_host_metrics', 'schedule': timedelta(days=1)},
 }
 
 # Django Caching Configuration
@@ -1052,3 +1053,10 @@ UI_NEXT = True
 # - '': No model - Subscription not counted from Host Metrics
 # - 'unique_managed_hosts': Compliant = automated - deleted hosts (using /api/v2/host_metrics/)
 SUBSCRIPTION_USAGE_MODEL = ''
+
+# Host metrics cleanup - last time of the cleanup run (soft-deleting records)
+CLEANUP_HOST_METRICS_LAST_TS = None
+# Host metrics cleanup - minimal interval between two cleanups in days
+CLEANUP_HOST_METRICS_INTERVAL = 30  # days
+# Host metrics cleanup - soft-delete HostMetric records with last_automation < [threshold] (in months)
+CLEANUP_HOST_METRICS_THRESHOLD = 12  # months