Merge pull request #12629 from fosterseth/task_manager_refactor_squashed

Task manager refactor

awx/api/urls/debug.py

@@ -0,0 +1,17 @@
+from django.urls import re_path
+
+from awx.api.views.debug import (
+    DebugRootView,
+    TaskManagerDebugView,
+    DependencyManagerDebugView,
+    WorkflowManagerDebugView,
+)
+
+urls = [
+    re_path(r'^$', DebugRootView.as_view(), name='debug'),
+    re_path(r'^task_manager/$', TaskManagerDebugView.as_view(), name='task_manager'),
+    re_path(r'^dependency_manager/$', DependencyManagerDebugView.as_view(), name='dependency_manager'),
+    re_path(r'^workflow_manager/$', WorkflowManagerDebugView.as_view(), name='workflow_manager'),
+]
+
+__all__ = ['urls']

awx/api/urls/urls.py

@@ -2,9 +2,9 @@
 # All Rights Reserved.
 
 from __future__ import absolute_import, unicode_literals
-from django.conf import settings
 from django.urls import include, re_path
 
+from awx import MODE
 from awx.api.generics import LoggedLoginView, LoggedLogoutView
 from awx.api.views import (
     ApiRootView,
@@ -145,7 +145,12 @@ urlpatterns = [
     re_path(r'^logout/$', LoggedLogoutView.as_view(next_page='/api/', redirect_field_name='next'), name='logout'),
     re_path(r'^o/', include(oauth2_root_urls)),
 ]
-if settings.SETTINGS_MODULE == 'awx.settings.development':
+if MODE == 'development':
+    # Only include these if we are in the development environment
     from awx.api.swagger import SwaggerSchemaView
 
     urlpatterns += [re_path(r'^swagger/$', SwaggerSchemaView.as_view(), name='swagger_view')]
+
+    from awx.api.urls.debug import urls as debug_urls
+
+    urlpatterns += [re_path(r'^debug/', include(debug_urls))]

awx/api/views/__init__.py

@@ -93,7 +93,7 @@ from awx.main.utils import (
     get_object_or_400,
     getattrd,
     get_pk_from_dict,
-    schedule_task_manager,
+    ScheduleWorkflowManager,
     ignore_inventory_computed_fields,
 )
 from awx.main.utils.encryption import encrypt_value
@@ -3391,7 +3391,7 @@ class WorkflowJobCancel(RetrieveAPIView):
         obj = self.get_object()
         if obj.can_cancel:
             obj.cancel()
-            schedule_task_manager()
+            ScheduleWorkflowManager().schedule()
             return Response(status=status.HTTP_202_ACCEPTED)
         else:
             return self.http_method_not_allowed(request, *args, **kwargs)

awx/api/views/debug.py

@@ -0,0 +1,68 @@
+from collections import OrderedDict
+
+from django.conf import settings
+
+from rest_framework.permissions import AllowAny
+from rest_framework.response import Response
+from awx.api.generics import APIView
+
+from awx.main.scheduler import TaskManager, DependencyManager, WorkflowManager
+
+
+class TaskManagerDebugView(APIView):
+    _ignore_model_permissions = True
+    exclude_from_schema = True
+    permission_classes = [AllowAny]
+    prefix = 'Task'
+
+    def get(self, request):
+        TaskManager().schedule()
+        if not settings.AWX_DISABLE_TASK_MANAGERS:
+            msg = f"Running {self.prefix} manager. To disable other triggers to the {self.prefix} manager, set AWX_DISABLE_TASK_MANAGERS to True"
+        else:
+            msg = f"AWX_DISABLE_TASK_MANAGERS is True, this view is the only way to trigger the {self.prefix} manager"
+        return Response(msg)
+
+
+class DependencyManagerDebugView(APIView):
+    _ignore_model_permissions = True
+    exclude_from_schema = True
+    permission_classes = [AllowAny]
+    prefix = 'Dependency'
+
+    def get(self, request):
+        DependencyManager().schedule()
+        if not settings.AWX_DISABLE_TASK_MANAGERS:
+            msg = f"Running {self.prefix} manager. To disable other triggers to the {self.prefix} manager, set AWX_DISABLE_TASK_MANAGERS to True"
+        else:
+            msg = f"AWX_DISABLE_TASK_MANAGERS is True, this view is the only way to trigger the {self.prefix} manager"
+        return Response(msg)
+
+
+class WorkflowManagerDebugView(APIView):
+    _ignore_model_permissions = True
+    exclude_from_schema = True
+    permission_classes = [AllowAny]
+    prefix = 'Workflow'
+
+    def get(self, request):
+        WorkflowManager().schedule()
+        if not settings.AWX_DISABLE_TASK_MANAGERS:
+            msg = f"Running {self.prefix} manager. To disable other triggers to the {self.prefix} manager, set AWX_DISABLE_TASK_MANAGERS to True"
+        else:
+            msg = f"AWX_DISABLE_TASK_MANAGERS is True, this view is the only way to trigger the {self.prefix} manager"
+        return Response(msg)
+
+
+class DebugRootView(APIView):
+    _ignore_model_permissions = True
+    exclude_from_schema = True
+    permission_classes = [AllowAny]
+
+    def get(self, request, format=None):
+        '''List of available debug urls'''
+        data = OrderedDict()
+        data['task_manager'] = '/api/debug/task_manager/'
+        data['dependency_manager'] = '/api/debug/dependency_manager/'
+        data['workflow_manager'] = '/api/debug/workflow_manager/'
+        return Response(data)

awx/main/analytics/subsystem_metrics.py

@@ -184,19 +184,28 @@ class Metrics:
             FloatM('subsystem_metrics_pipe_execute_seconds', 'Time spent saving metrics to redis'),
             IntM('subsystem_metrics_pipe_execute_calls', 'Number of calls to pipe_execute'),
             FloatM('subsystem_metrics_send_metrics_seconds', 'Time spent sending metrics to other nodes'),
-            SetFloatM('task_manager_get_tasks_seconds', 'Time spent in loading all tasks from db'),
+            SetFloatM('task_manager_get_tasks_seconds', 'Time spent in loading tasks from db'),
             SetFloatM('task_manager_start_task_seconds', 'Time spent starting task'),
             SetFloatM('task_manager_process_running_tasks_seconds', 'Time spent processing running tasks'),
             SetFloatM('task_manager_process_pending_tasks_seconds', 'Time spent processing pending tasks'),
-            SetFloatM('task_manager_generate_dependencies_seconds', 'Time spent generating dependencies for pending tasks'),
-            SetFloatM('task_manager_spawn_workflow_graph_jobs_seconds', 'Time spent spawning workflow jobs'),
             SetFloatM('task_manager__schedule_seconds', 'Time spent in running the entire _schedule'),
-            IntM('task_manager_schedule_calls', 'Number of calls to task manager schedule'),
+            IntM('task_manager__schedule_calls', 'Number of calls to _schedule, after lock is acquired'),
             SetFloatM('task_manager_recorded_timestamp', 'Unix timestamp when metrics were last recorded'),
             SetIntM('task_manager_tasks_started', 'Number of tasks started'),
             SetIntM('task_manager_running_processed', 'Number of running tasks processed'),
             SetIntM('task_manager_pending_processed', 'Number of pending tasks processed'),
             SetIntM('task_manager_tasks_blocked', 'Number of tasks blocked from running'),
+            SetFloatM('dependency_manager_get_tasks_seconds', 'Time spent loading pending tasks from db'),
+            SetFloatM('dependency_manager_generate_dependencies_seconds', 'Time spent generating dependencies for pending tasks'),
+            SetFloatM('dependency_manager__schedule_seconds', 'Time spent in running the entire _schedule'),
+            IntM('dependency_manager__schedule_calls', 'Number of calls to _schedule, after lock is acquired'),
+            SetFloatM('dependency_manager_recorded_timestamp', 'Unix timestamp when metrics were last recorded'),
+            SetIntM('dependency_manager_pending_processed', 'Number of pending tasks processed'),
+            SetFloatM('workflow_manager__schedule_seconds', 'Time spent in running the entire _schedule'),
+            IntM('workflow_manager__schedule_calls', 'Number of calls to _schedule, after lock is acquired'),
+            SetFloatM('workflow_manager_recorded_timestamp', 'Unix timestamp when metrics were last recorded'),
+            SetFloatM('workflow_manager_spawn_workflow_graph_jobs_seconds', 'Time spent spawning workflow tasks'),
+            SetFloatM('workflow_manager_get_tasks_seconds', 'Time spent loading workflow tasks from db'),
         ]
         # turn metric list into dictionary with the metric name as a key
         self.METRICS = {}

awx/main/dispatch/pool.py

@@ -401,13 +401,15 @@ class AutoscalePool(WorkerPool):
                 # the task manager to never do more work
                 current_task = w.current_task
                 if current_task and isinstance(current_task, dict):
-                    if current_task.get('task', '').endswith('tasks.run_task_manager'):
+                    endings = ['tasks.task_manager', 'tasks.dependency_manager', 'tasks.workflow_manager']
+                    current_task_name = current_task.get('task', '')
+                    if any(current_task_name.endswith(e) for e in endings):
                         if 'started' not in current_task:
                             w.managed_tasks[current_task['uuid']]['started'] = time.time()
                         age = time.time() - current_task['started']
                         w.managed_tasks[current_task['uuid']]['age'] = age
-                        if age > (60 * 5):
+                        if age > (settings.TASK_MANAGER_TIMEOUT + settings.TASK_MANAGER_TIMEOUT_GRACE_PERIOD):
-                            logger.error(f'run_task_manager has held the advisory lock for >5m, sending SIGTERM to {w.pid}')  # noqa
+                            logger.error(f'{current_task_name} has held the advisory lock for {age}, sending SIGTERM to {w.pid}')  # noqa
                             os.kill(w.pid, signal.SIGTERM)
 
         for m in orphaned:

awx/main/management/commands/inventory_import.py

@@ -862,7 +862,7 @@ class Command(BaseCommand):
                     overwrite_vars=bool(options.get('overwrite_vars', False)),
                 )
                 inventory_update = inventory_source.create_inventory_update(
-                    _eager_fields=dict(job_args=json.dumps(sys.argv), job_env=dict(os.environ.items()), job_cwd=os.getcwd())
+                    _eager_fields=dict(status='running', job_args=json.dumps(sys.argv), job_env=dict(os.environ.items()), job_cwd=os.getcwd())
                 )
 
             data = AnsibleInventoryLoader(source=source, verbosity=verbosity).load()

awx/main/migrations/0165_task_manager_refactor.py

@@ -0,0 +1,35 @@
+# Generated by Django 3.2.13 on 2022-08-10 14:03
+
+from django.db import migrations, models
+
+
+class Migration(migrations.Migration):
+
+    dependencies = [
+        ('main', '0164_remove_inventorysource_update_on_project_update'),
+    ]
+
+    operations = [
+        migrations.AddField(
+            model_name='unifiedjob',
+            name='preferred_instance_groups_cache',
+            field=models.JSONField(
+                blank=True, default=None, editable=False, help_text='A cached list with pk values from preferred instance groups.', null=True
+            ),
+        ),
+        migrations.AddField(
+            model_name='unifiedjob',
+            name='task_impact',
+            field=models.PositiveIntegerField(default=0, editable=False, help_text='Number of forks an instance consumes when running this job.'),
+        ),
+        migrations.AddField(
+            model_name='workflowapproval',
+            name='expires',
+            field=models.DateTimeField(
+                default=None,
+                editable=False,
+                help_text='The time this approval will expire. This is the created time plus timeout, used for filtering.',
+                null=True,
+            ),
+        ),
+    ]

awx/main/models/ad_hoc_commands.py

@@ -90,6 +90,9 @@ class AdHocCommand(UnifiedJob, JobNotificationMixin):
 
     extra_vars_dict = VarsDictProperty('extra_vars', True)
 
+    def _set_default_dependencies_processed(self):
+        self.dependencies_processed = True
+
     def clean_inventory(self):
         inv = self.inventory
         if not inv:
@@ -178,12 +181,12 @@ class AdHocCommand(UnifiedJob, JobNotificationMixin):
     def get_passwords_needed_to_start(self):
         return self.passwords_needed_to_start
 
-    @property
+    def _get_task_impact(self):
-    def task_impact(self):
         # NOTE: We sorta have to assume the host count matches and that forks default to 5
-        from awx.main.models.inventory import Host
+        if self.inventory:
-
+            count_hosts = self.inventory.total_hosts
-        count_hosts = Host.objects.filter(enabled=True, inventory__ad_hoc_commands__pk=self.pk).count()
+        else:
+            count_hosts = 5
         return min(count_hosts, 5 if self.forks == 0 else self.forks) + 1
 
     def copy(self):
@@ -207,10 +210,20 @@ class AdHocCommand(UnifiedJob, JobNotificationMixin):
 
     def save(self, *args, **kwargs):
         update_fields = kwargs.get('update_fields', [])
+
+        def add_to_update_fields(name):
+            if name not in update_fields:
+                update_fields.append(name)
+
+        if not self.preferred_instance_groups_cache:
+            self.preferred_instance_groups_cache = self._get_preferred_instance_group_cache()
+            add_to_update_fields("preferred_instance_groups_cache")
         if not self.name:
             self.name = Truncator(u': '.join(filter(None, (self.module_name, self.module_args)))).chars(512)
-            if 'name' not in update_fields:
+            add_to_update_fields("name")
-                update_fields.append('name')
+        if self.task_impact == 0:
+            self.task_impact = self._get_task_impact()
+            add_to_update_fields("task_impact")
         super(AdHocCommand, self).save(*args, **kwargs)
 
     @property

awx/main/models/ha.py

@@ -12,6 +12,7 @@ from django.dispatch import receiver
 from django.utils.translation import gettext_lazy as _
 from django.conf import settings
 from django.utils.timezone import now, timedelta
+from django.db.models import Sum
 
 import redis
 from solo.models import SingletonModel
@@ -149,10 +150,13 @@ class Instance(HasPolicyEditsMixin, BaseModel):
     def consumed_capacity(self):
         capacity_consumed = 0
         if self.node_type in ('hybrid', 'execution'):
-            capacity_consumed += sum(x.task_impact for x in UnifiedJob.objects.filter(execution_node=self.hostname, status__in=('running', 'waiting')))
+            capacity_consumed += (
+                UnifiedJob.objects.filter(execution_node=self.hostname, status__in=('running', 'waiting')).aggregate(Sum("task_impact"))["task_impact__sum"]
+                or 0
+            )
         if self.node_type in ('hybrid', 'control'):
-            capacity_consumed += sum(
+            capacity_consumed += (
-                settings.AWX_CONTROL_NODE_TASK_IMPACT for x in UnifiedJob.objects.filter(controller_node=self.hostname, status__in=('running', 'waiting'))
+                settings.AWX_CONTROL_NODE_TASK_IMPACT * UnifiedJob.objects.filter(controller_node=self.hostname, status__in=('running', 'waiting')).count()
             )
         return capacity_consumed
 

awx/main/models/inventory.py

@@ -337,9 +337,12 @@ class Inventory(CommonModelNameNotUnique, ResourceMixin, RelatedJobsMixin):
         else:
             active_inventory_sources = self.inventory_sources.filter(source__in=CLOUD_INVENTORY_SOURCES)
         failed_inventory_sources = active_inventory_sources.filter(last_job_failed=True)
+        total_hosts = active_hosts.count()
+        # if total_hosts has changed, set update_task_impact to True
+        update_task_impact = total_hosts != self.total_hosts
         computed_fields = {
             'has_active_failures': bool(failed_hosts.count()),
-            'total_hosts': active_hosts.count(),
+            'total_hosts': total_hosts,
             'hosts_with_active_failures': failed_hosts.count(),
             'total_groups': active_groups.count(),
             'has_inventory_sources': bool(active_inventory_sources.count()),
@@ -357,6 +360,14 @@ class Inventory(CommonModelNameNotUnique, ResourceMixin, RelatedJobsMixin):
                 computed_fields.pop(field)
         if computed_fields:
             iobj.save(update_fields=computed_fields.keys())
+        if update_task_impact:
+            # if total hosts count has changed, re-calculate task_impact for any
+            # job that is still in pending for this inventory, since task_impact
+            # is cached on task creation and used in task management system
+            tasks = self.jobs.filter(status="pending")
+            for t in tasks:
+                t.task_impact = t._get_task_impact()
+            UnifiedJob.objects.bulk_update(tasks, ['task_impact'])
         logger.debug("Finished updating inventory computed fields, pk={0}, in " "{1:.3f} seconds".format(self.pk, time.time() - start_time))
 
     def websocket_emit_status(self, status):
@@ -1220,8 +1231,7 @@ class InventoryUpdate(UnifiedJob, InventorySourceOptions, JobNotificationMixin,
             return UnpartitionedInventoryUpdateEvent
         return InventoryUpdateEvent
 
-    @property
+    def _get_task_impact(self):
-    def task_impact(self):
         return 1
 
     # InventoryUpdate credential required

awx/main/models/jobs.py

@@ -644,8 +644,7 @@ class Job(UnifiedJob, JobOptions, SurveyJobMixin, JobNotificationMixin, TaskMana
             raise ParseError(_('{status_value} is not a valid status option.').format(status_value=status))
         return self._get_hosts(**kwargs)
 
-    @property
+    def _get_task_impact(self):
-    def task_impact(self):
         if self.launch_type == 'callback':
             count_hosts = 2
         else:
@@ -1213,6 +1212,9 @@ class SystemJob(UnifiedJob, SystemJobOptions, JobNotificationMixin):
 
     extra_vars_dict = VarsDictProperty('extra_vars', True)
 
+    def _set_default_dependencies_processed(self):
+        self.dependencies_processed = True
+
     @classmethod
     def _get_parent_field_name(cls):
         return 'system_job_template'
@@ -1238,8 +1240,7 @@ class SystemJob(UnifiedJob, SystemJobOptions, JobNotificationMixin):
             return UnpartitionedSystemJobEvent
         return SystemJobEvent
 
-    @property
+    def _get_task_impact(self):
-    def task_impact(self):
         return 5
 
     @property

awx/main/models/projects.py

@@ -513,6 +513,9 @@ class ProjectUpdate(UnifiedJob, ProjectOptions, JobNotificationMixin, TaskManage
         help_text=_('The SCM Revision discovered by this update for the given project and branch.'),
     )
 
+    def _set_default_dependencies_processed(self):
+        self.dependencies_processed = True
+
     def _get_parent_field_name(self):
         return 'project'
 
@@ -560,8 +563,7 @@ class ProjectUpdate(UnifiedJob, ProjectOptions, JobNotificationMixin, TaskManage
             return UnpartitionedProjectUpdateEvent
         return ProjectUpdateEvent
 
-    @property
+    def _get_task_impact(self):
-    def task_impact(self):
         return 0 if self.job_type == 'run' else 1
 
     @property

awx/main/models/unified_jobs.py

@@ -45,7 +45,8 @@ from awx.main.utils.common import (
     get_type_for_model,
     parse_yaml_or_json,
     getattr_dne,
-    schedule_task_manager,
+    ScheduleDependencyManager,
+    ScheduleTaskManager,
     get_event_partition_epoch,
     get_capacity_type,
 )
@@ -381,6 +382,11 @@ class UnifiedJobTemplate(PolymorphicModel, CommonModelNameNotUnique, ExecutionEn
             unified_job.survey_passwords = new_job_passwords
             kwargs['survey_passwords'] = new_job_passwords  # saved in config object for relaunch
 
+        unified_job.preferred_instance_groups_cache = unified_job._get_preferred_instance_group_cache()
+
+        unified_job._set_default_dependencies_processed()
+        unified_job.task_impact = unified_job._get_task_impact()
+
         from awx.main.signals import disable_activity_stream, activity_stream_create
 
         with disable_activity_stream():
@@ -693,6 +699,14 @@ class UnifiedJob(
         on_delete=polymorphic.SET_NULL,
         help_text=_('The Instance group the job was run under'),
     )
+    preferred_instance_groups_cache = models.JSONField(
+        blank=True,
+        null=True,
+        default=None,
+        editable=False,
+        help_text=_("A cached list with pk values from preferred instance groups."),
+    )
+    task_impact = models.PositiveIntegerField(default=0, editable=False, help_text=_("Number of forks an instance consumes when running this job."))
     organization = models.ForeignKey(
         'Organization',
         blank=True,
@@ -754,6 +768,9 @@ class UnifiedJob(
     def _get_parent_field_name(self):
         return 'unified_job_template'  # Override in subclasses.
 
+    def _get_preferred_instance_group_cache(self):
+        return [ig.pk for ig in self.preferred_instance_groups]
+
     @classmethod
     def _get_unified_job_template_class(cls):
         """
@@ -808,6 +825,9 @@ class UnifiedJob(
             update_fields = self._update_parent_instance_no_save(parent_instance)
             parent_instance.save(update_fields=update_fields)
 
+    def _set_default_dependencies_processed(self):
+        pass
+
     def save(self, *args, **kwargs):
         """Save the job, with current status, to the database.
         Ensure that all data is consistent before doing so.
@@ -1026,7 +1046,6 @@ class UnifiedJob(
             event_qs = self.get_event_queryset()
         except NotImplementedError:
             return True  # Model without events, such as WFJT
-        self.log_lifecycle("event_processing_finished")
         return self.emitted_events == event_qs.count()
 
     def result_stdout_raw_handle(self, enforce_max_bytes=True):
@@ -1241,9 +1260,8 @@ class UnifiedJob(
         except JobLaunchConfig.DoesNotExist:
             return False
 
-    @property
+    def _get_task_impact(self):
-    def task_impact(self):
+        return self.task_impact  # return default, should implement in subclass.
-        raise NotImplementedError  # Implement in subclass.
 
     def websocket_emit_data(self):
         '''Return extra data that should be included when submitting data to the browser over the websocket connection'''
@@ -1358,7 +1376,10 @@ class UnifiedJob(
         self.update_fields(start_args=json.dumps(kwargs), status='pending')
         self.websocket_emit_status("pending")
 
-        schedule_task_manager()
+        if self.dependencies_processed:
+            ScheduleTaskManager().schedule()
+        else:
+            ScheduleDependencyManager().schedule()
 
         # Each type of unified job has a different Task class; get the
         # appropirate one.
@@ -1515,8 +1536,8 @@ class UnifiedJob(
             'state': state,
             'work_unit_id': self.work_unit_id,
         }
-        if self.unified_job_template:
+        if self.name:
-            extra["template_name"] = self.unified_job_template.name
+            extra["task_name"] = self.name
         if state == "blocked" and blocked_by:
             blocked_by_msg = f"{blocked_by._meta.model_name}-{blocked_by.id}"
             msg = f"{self._meta.model_name}-{self.id} blocked by {blocked_by_msg}"

awx/main/models/workflow.py

@@ -13,6 +13,7 @@ from django.db import connection, models
 from django.conf import settings
 from django.utils.translation import gettext_lazy as _
 from django.core.exceptions import ObjectDoesNotExist
+from django.utils.timezone import now, timedelta
 
 # from django import settings as tower_settings
 
@@ -40,7 +41,7 @@ from awx.main.models.mixins import (
 from awx.main.models.jobs import LaunchTimeConfigBase, LaunchTimeConfig, JobTemplate
 from awx.main.models.credential import Credential
 from awx.main.redact import REPLACE_STR
-from awx.main.utils import schedule_task_manager
+from awx.main.utils import ScheduleWorkflowManager
 
 
 __all__ = [
@@ -622,6 +623,9 @@ class WorkflowJob(UnifiedJob, WorkflowJobOptions, SurveyJobMixin, JobNotificatio
     )
     is_sliced_job = models.BooleanField(default=False)
 
+    def _set_default_dependencies_processed(self):
+        self.dependencies_processed = True
+
     @property
     def workflow_nodes(self):
         return self.workflow_job_nodes
@@ -668,8 +672,7 @@ class WorkflowJob(UnifiedJob, WorkflowJobOptions, SurveyJobMixin, JobNotificatio
         )
         return result
 
-    @property
+    def _get_task_impact(self):
-    def task_impact(self):
         return 0
 
     def get_ancestor_workflows(self):
@@ -783,6 +786,12 @@ class WorkflowApproval(UnifiedJob, JobNotificationMixin):
         default=0,
         help_text=_("The amount of time (in seconds) before the approval node expires and fails."),
     )
+    expires = models.DateTimeField(
+        default=None,
+        null=True,
+        editable=False,
+        help_text=_("The time this approval will expire. This is the created time plus timeout, used for filtering."),
+    )
     timed_out = models.BooleanField(default=False, help_text=_("Shows when an approval node (with a timeout assigned to it) has timed out."))
     approved_or_denied_by = models.ForeignKey(
         'auth.User',
@@ -793,6 +802,9 @@ class WorkflowApproval(UnifiedJob, JobNotificationMixin):
         on_delete=models.SET_NULL,
     )
 
+    def _set_default_dependencies_processed(self):
+        self.dependencies_processed = True
+
     @classmethod
     def _get_unified_job_template_class(cls):
         return WorkflowApprovalTemplate
@@ -810,13 +822,32 @@ class WorkflowApproval(UnifiedJob, JobNotificationMixin):
     def _get_parent_field_name(self):
         return 'workflow_approval_template'
 
+    def save(self, *args, **kwargs):
+        update_fields = list(kwargs.get('update_fields', []))
+        if self.timeout != 0 and ((not self.pk) or (not update_fields) or ('timeout' in update_fields)):
+            if not self.created:  # on creation, created will be set by parent class, so we fudge it here
+                created = now()
+            else:
+                created = self.created
+            new_expires = created + timedelta(seconds=self.timeout)
+            if new_expires != self.expires:
+                self.expires = new_expires
+                if update_fields and 'expires' not in update_fields:
+                    update_fields.append('expires')
+        elif self.timeout == 0 and ((not update_fields) or ('timeout' in update_fields)):
+            if self.expires:
+                self.expires = None
+                if update_fields and 'expires' not in update_fields:
+                    update_fields.append('expires')
+        super(WorkflowApproval, self).save(*args, **kwargs)
+
     def approve(self, request=None):
         self.status = 'successful'
         self.approved_or_denied_by = get_current_user()
         self.save()
         self.send_approval_notification('approved')
         self.websocket_emit_status(self.status)
-        schedule_task_manager()
+        ScheduleWorkflowManager().schedule()
         return reverse('api:workflow_approval_approve', kwargs={'pk': self.pk}, request=request)
 
     def deny(self, request=None):
@@ -825,7 +856,7 @@ class WorkflowApproval(UnifiedJob, JobNotificationMixin):
         self.save()
         self.send_approval_notification('denied')
         self.websocket_emit_status(self.status)
-        schedule_task_manager()
+        ScheduleWorkflowManager().schedule()
         return reverse('api:workflow_approval_deny', kwargs={'pk': self.pk}, request=request)
 
     def signal_start(self, **kwargs):

awx/main/scheduler/__init__.py

@@ -1,6 +1,6 @@
 # Copyright (c) 2017 Ansible, Inc.
 #
 
-from .task_manager import TaskManager
+from .task_manager import TaskManager, DependencyManager, WorkflowManager
 
-__all__ = ['TaskManager']
+__all__ = ['TaskManager', 'DependencyManager', 'WorkflowManager']

awx/main/scheduler/dependency_graph.py

@@ -7,6 +7,11 @@ from awx.main.models import (
     WorkflowJob,
 )
 
+import logging
+
+
+logger = logging.getLogger('awx.main.scheduler.dependency_graph')
+
 
 class DependencyGraph(object):
     PROJECT_UPDATES = 'project_updates'
@@ -36,6 +41,9 @@ class DependencyGraph(object):
         self.data[self.WORKFLOW_JOB_TEMPLATES_JOBS] = {}
 
     def mark_if_no_key(self, job_type, id, job):
+        if id is None:
+            logger.warning(f'Null dependency graph key from {job}, could be integrity error or bug, ignoring')
+            return
         # only mark first occurrence of a task. If 10 of JobA are launched
         # (concurrent disabled), the dependency graph should return that jobs
         # 2 through 10 are blocked by job1
@@ -66,7 +74,10 @@ class DependencyGraph(object):
         self.mark_if_no_key(self.JOB_TEMPLATE_JOBS, job.job_template_id, job)
 
     def mark_workflow_job(self, job):
-        self.mark_if_no_key(self.WORKFLOW_JOB_TEMPLATES_JOBS, job.workflow_job_template_id, job)
+        if job.workflow_job_template_id:
+            self.mark_if_no_key(self.WORKFLOW_JOB_TEMPLATES_JOBS, job.workflow_job_template_id, job)
+        elif job.unified_job_template_id:  # for sliced jobs
+            self.mark_if_no_key(self.WORKFLOW_JOB_TEMPLATES_JOBS, job.unified_job_template_id, job)
 
     def project_update_blocked_by(self, job):
         return self.get_item(self.PROJECT_UPDATES, job.project_id)
@@ -85,7 +96,13 @@ class DependencyGraph(object):
 
     def workflow_job_blocked_by(self, job):
         if job.allow_simultaneous is False:
-            return self.get_item(self.WORKFLOW_JOB_TEMPLATES_JOBS, job.workflow_job_template_id)
+            if job.workflow_job_template_id:
+                return self.get_item(self.WORKFLOW_JOB_TEMPLATES_JOBS, job.workflow_job_template_id)
+            elif job.unified_job_template_id:
+                # Sliced jobs can be either Job or WorkflowJob type, and either should block a sliced WorkflowJob
+                return self.get_item(self.WORKFLOW_JOB_TEMPLATES_JOBS, job.unified_job_template_id) or self.get_item(
+                    self.JOB_TEMPLATE_JOBS, job.unified_job_template_id
+                )
         return None
 
     def system_job_blocked_by(self, job):

awx/main/scheduler/task_manager.py

@@ -15,27 +15,31 @@ from django.db import transaction, connection
 from django.utils.translation import gettext_lazy as _, gettext_noop
 from django.utils.timezone import now as tz_now
 from django.conf import settings
+from django.contrib.contenttypes.models import ContentType
 
 # AWX
 from awx.main.dispatch.reaper import reap_job
 from awx.main.models import (
-    AdHocCommand,
     Instance,
     InventorySource,
     InventoryUpdate,
     Job,
     Project,
     ProjectUpdate,
-    SystemJob,
     UnifiedJob,
     WorkflowApproval,
     WorkflowJob,
+    WorkflowJobNode,
     WorkflowJobTemplate,
 )
 from awx.main.scheduler.dag_workflow import WorkflowDAG
 from awx.main.utils.pglock import advisory_lock
-from awx.main.utils import get_type_for_model, task_manager_bulk_reschedule, schedule_task_manager
+from awx.main.utils import (
-from awx.main.utils.common import create_partition
+    get_type_for_model,
+    ScheduleTaskManager,
+    ScheduleWorkflowManager,
+)
+from awx.main.utils.common import create_partition, task_manager_bulk_reschedule
 from awx.main.signals import disable_activity_stream
 from awx.main.constants import ACTIVE_STATES
 from awx.main.scheduler.dependency_graph import DependencyGraph
@@ -53,167 +57,97 @@ def timeit(func):
         t_now = time.perf_counter()
         result = func(*args, **kwargs)
         dur = time.perf_counter() - t_now
-        args[0].subsystem_metrics.inc("task_manager_" + func.__name__ + "_seconds", dur)
+        args[0].subsystem_metrics.inc(f"{args[0].prefix}_{func.__name__}_seconds", dur)
         return result
 
     return inner
 
 
-class TaskManager:
+class TaskBase:
-    def __init__(self):
+    def __init__(self, prefix=""):
-        """
+        self.prefix = prefix
-        Do NOT put database queries or other potentially expensive operations
-        in the task manager init. The task manager object is created every time a
-        job is created, transitions state, and every 30 seconds on each tower node.
-        More often then not, the object is destroyed quickly because the NOOP case is hit.
-
-        The NOOP case is short-circuit logic. If the task manager realizes that another instance
-        of the task manager is already running, then it short-circuits and decides not to run.
-        """
-        # start task limit indicates how many pending jobs can be started on this
-        # .schedule() run. Starting jobs is expensive, and there is code in place to reap
-        # the task manager after 5 minutes. At scale, the task manager can easily take more than
-        # 5 minutes to start pending jobs. If this limit is reached, pending jobs
-        # will no longer be started and will be started on the next task manager cycle.
-        self.start_task_limit = settings.START_TASK_LIMIT
-        self.time_delta_job_explanation = timedelta(seconds=30)
-        self.subsystem_metrics = s_metrics.Metrics(auto_pipe_execute=False)
         # initialize each metric to 0 and force metric_has_changed to true. This
         # ensures each task manager metric will be overridden when pipe_execute
         # is called later.
+        self.subsystem_metrics = s_metrics.Metrics(auto_pipe_execute=False)
+        self.start_time = time.time()
+        self.start_task_limit = settings.START_TASK_LIMIT
         for m in self.subsystem_metrics.METRICS:
-            if m.startswith("task_manager"):
+            if m.startswith(self.prefix):
                 self.subsystem_metrics.set(m, 0)
 
-    def after_lock_init(self, all_sorted_tasks):
+    def timed_out(self):
-        """
+        """Return True/False if we have met or exceeded the timeout for the task manager."""
-        Init AFTER we know this instance of the task manager will run because the lock is acquired.
+        elapsed = time.time() - self.start_time
-        """
+        if elapsed >= settings.TASK_MANAGER_TIMEOUT:
-        self.dependency_graph = DependencyGraph()
+            logger.warning(f"{self.prefix} manager has run for {elapsed} which is greater than TASK_MANAGER_TIMEOUT of {settings.TASK_MANAGER_TIMEOUT}.")
-        self.instances = TaskManagerInstances(all_sorted_tasks)
+            return True
-        self.instance_groups = TaskManagerInstanceGroups(instances_by_hostname=self.instances)
+        return False
-        self.controlplane_ig = self.instance_groups.controlplane_ig
-
-    def job_blocked_by(self, task):
-        # TODO: I'm not happy with this, I think blocking behavior should be decided outside of the dependency graph
-        #       in the old task manager this was handled as a method on each task object outside of the graph and
-        #       probably has the side effect of cutting down *a lot* of the logic from this task manager class
-        blocked_by = self.dependency_graph.task_blocked_by(task)
-        if blocked_by:
-            return blocked_by
-
-        for dep in task.dependent_jobs.all():
-            if dep.status in ACTIVE_STATES:
-                return dep
-            # if we detect a failed or error dependency, go ahead and fail this
-            # task. The errback on the dependency takes some time to trigger,
-            # and we don't want the task to enter running state if its
-            # dependency has failed or errored.
-            elif dep.status in ("error", "failed"):
-                task.status = 'failed'
-                task.job_explanation = 'Previous Task Failed: {"job_type": "%s", "job_name": "%s", "job_id": "%s"}' % (
-                    get_type_for_model(type(dep)),
-                    dep.name,
-                    dep.id,
-                )
-                task.save(update_fields=['status', 'job_explanation'])
-                task.websocket_emit_status('failed')
-                return dep
-
-        return None
 
     @timeit
-    def get_tasks(self, status_list=('pending', 'waiting', 'running')):
+    def get_tasks(self, filter_args):
-        jobs = [j for j in Job.objects.filter(status__in=status_list).prefetch_related('instance_group')]
+        wf_approval_ctype_id = ContentType.objects.get_for_model(WorkflowApproval).id
-        inventory_updates_qs = (
+        qs = (
-            InventoryUpdate.objects.filter(status__in=status_list).exclude(source='file').prefetch_related('inventory_source', 'instance_group')
+            UnifiedJob.objects.filter(**filter_args)
+            .exclude(launch_type='sync')
+            .exclude(polymorphic_ctype_id=wf_approval_ctype_id)
+            .order_by('created')
+            .prefetch_related('dependent_jobs')
         )
-        inventory_updates = [i for i in inventory_updates_qs]
+        self.all_tasks = [t for t in qs]
-        # Notice the job_type='check': we want to prevent implicit project updates from blocking our jobs.
-        project_updates = [p for p in ProjectUpdate.objects.filter(status__in=status_list, job_type='check').prefetch_related('instance_group')]
-        system_jobs = [s for s in SystemJob.objects.filter(status__in=status_list).prefetch_related('instance_group')]
-        ad_hoc_commands = [a for a in AdHocCommand.objects.filter(status__in=status_list).prefetch_related('instance_group')]
-        workflow_jobs = [w for w in WorkflowJob.objects.filter(status__in=status_list)]
-        all_tasks = sorted(jobs + project_updates + inventory_updates + system_jobs + ad_hoc_commands + workflow_jobs, key=lambda task: task.created)
-        return all_tasks
 
-    def get_running_workflow_jobs(self):
+    def record_aggregate_metrics(self, *args):
-        graph_workflow_jobs = [wf for wf in WorkflowJob.objects.filter(status='running')]
+        if not settings.IS_TESTING():
-        return graph_workflow_jobs
+            # increment task_manager_schedule_calls regardless if the other
+            # metrics are recorded
+            s_metrics.Metrics(auto_pipe_execute=True).inc(f"{self.prefix}__schedule_calls", 1)
+            # Only record metrics if the last time recording was more
+            # than SUBSYSTEM_METRICS_TASK_MANAGER_RECORD_INTERVAL ago.
+            # Prevents a short-duration task manager that runs directly after a
+            # long task manager to override useful metrics.
+            current_time = time.time()
+            time_last_recorded = current_time - self.subsystem_metrics.decode(f"{self.prefix}_recorded_timestamp")
+            if time_last_recorded > settings.SUBSYSTEM_METRICS_TASK_MANAGER_RECORD_INTERVAL:
+                logger.debug(f"recording {self.prefix} metrics, last recorded {time_last_recorded} seconds ago")
+                self.subsystem_metrics.set(f"{self.prefix}_recorded_timestamp", current_time)
+                self.subsystem_metrics.pipe_execute()
+            else:
+                logger.debug(f"skipping recording {self.prefix} metrics, last recorded {time_last_recorded} seconds ago")
 
-    def get_inventory_source_tasks(self, all_sorted_tasks):
+    def record_aggregate_metrics_and_exit(self, *args):
-        inventory_ids = set()
+        self.record_aggregate_metrics()
-        for task in all_sorted_tasks:
+        sys.exit(1)
-            if isinstance(task, Job):
+
-                inventory_ids.add(task.inventory_id)
+    def schedule(self):
-        return [invsrc for invsrc in InventorySource.objects.filter(inventory_id__in=inventory_ids, update_on_launch=True)]
+        # Lock
+        with task_manager_bulk_reschedule():
+            with advisory_lock(f"{self.prefix}_lock", wait=False) as acquired:
+                with transaction.atomic():
+                    if acquired is False:
+                        logger.debug(f"Not running {self.prefix} scheduler, another task holds lock")
+                        return
+                    logger.debug(f"Starting {self.prefix} Scheduler")
+                    # if sigterm due to timeout, still record metrics
+                    signal.signal(signal.SIGTERM, self.record_aggregate_metrics_and_exit)
+                    self._schedule()
+                    self.record_aggregate_metrics()
+                    logger.debug(f"Finishing {self.prefix} Scheduler")
+
+
+class WorkflowManager(TaskBase):
+    def __init__(self):
+        super().__init__(prefix="workflow_manager")
 
     @timeit
-    def spawn_workflow_graph_jobs(self, workflow_jobs):
+    def spawn_workflow_graph_jobs(self):
-        for workflow_job in workflow_jobs:
-            if workflow_job.cancel_flag:
-                logger.debug('Not spawning jobs for %s because it is pending cancelation.', workflow_job.log_format)
-                continue
-            dag = WorkflowDAG(workflow_job)
-            spawn_nodes = dag.bfs_nodes_to_run()
-            if spawn_nodes:
-                logger.debug('Spawning jobs for %s', workflow_job.log_format)
-            else:
-                logger.debug('No nodes to spawn for %s', workflow_job.log_format)
-            for spawn_node in spawn_nodes:
-                if spawn_node.unified_job_template is None:
-                    continue
-                kv = spawn_node.get_job_kwargs()
-                job = spawn_node.unified_job_template.create_unified_job(**kv)
-                spawn_node.job = job
-                spawn_node.save()
-                logger.debug('Spawned %s in %s for node %s', job.log_format, workflow_job.log_format, spawn_node.pk)
-                can_start = True
-                if isinstance(spawn_node.unified_job_template, WorkflowJobTemplate):
-                    workflow_ancestors = job.get_ancestor_workflows()
-                    if spawn_node.unified_job_template in set(workflow_ancestors):
-                        can_start = False
-                        logger.info(
-                            'Refusing to start recursive workflow-in-workflow id={}, wfjt={}, ancestors={}'.format(
-                                job.id, spawn_node.unified_job_template.pk, [wa.pk for wa in workflow_ancestors]
-                            )
-                        )
-                        display_list = [spawn_node.unified_job_template] + workflow_ancestors
-                        job.job_explanation = gettext_noop(
-                            "Workflow Job spawned from workflow could not start because it " "would result in recursion (spawn order, most recent first: {})"
-                        ).format(', '.join(['<{}>'.format(tmp) for tmp in display_list]))
-                    else:
-                        logger.debug(
-                            'Starting workflow-in-workflow id={}, wfjt={}, ancestors={}'.format(
-                                job.id, spawn_node.unified_job_template.pk, [wa.pk for wa in workflow_ancestors]
-                            )
-                        )
-                if not job._resources_sufficient_for_launch():
-                    can_start = False
-                    job.job_explanation = gettext_noop(
-                        "Job spawned from workflow could not start because it " "was missing a related resource such as project or inventory"
-                    )
-                if can_start:
-                    if workflow_job.start_args:
-                        start_args = json.loads(decrypt_field(workflow_job, 'start_args'))
-                    else:
-                        start_args = {}
-                    can_start = job.signal_start(**start_args)
-                    if not can_start:
-                        job.job_explanation = gettext_noop(
-                            "Job spawned from workflow could not start because it " "was not in the right state or required manual credentials"
-                        )
-                if not can_start:
-                    job.status = 'failed'
-                    job.save(update_fields=['status', 'job_explanation'])
-                    job.websocket_emit_status('failed')
-
-                # TODO: should we emit a status on the socket here similar to tasks.py awx_periodic_scheduler() ?
-                # emit_websocket_notification('/socket.io/jobs', '', dict(id=))
-
-    def process_finished_workflow_jobs(self, workflow_jobs):
         result = []
-        for workflow_job in workflow_jobs:
+        for workflow_job in self.all_tasks:
+            if self.timed_out():
+                logger.warning("Workflow manager has reached time out while processing running workflows, exiting loop early")
+                ScheduleWorkflowManager().schedule()
+                # Do not process any more workflow jobs. Stop here.
+                # Maybe we should schedule another WorkflowManager run
+                break
             dag = WorkflowDAG(workflow_job)
             status_changed = False
             if workflow_job.cancel_flag:
@@ -228,99 +162,106 @@ class TaskManager:
                     status_changed = True
             else:
                 workflow_nodes = dag.mark_dnr_nodes()
-                for n in workflow_nodes:
+                WorkflowJobNode.objects.bulk_update(workflow_nodes, ['do_not_run'])
-                    n.save(update_fields=['do_not_run'])
+                # If workflow is now done, we do special things to mark it as done.
                 is_done = dag.is_workflow_done()
-                if not is_done:
+                if is_done:
-                    continue
+                    has_failed, reason = dag.has_workflow_failed()
-                has_failed, reason = dag.has_workflow_failed()
+                    logger.debug('Marking %s as %s.', workflow_job.log_format, 'failed' if has_failed else 'successful')
-                logger.debug('Marking %s as %s.', workflow_job.log_format, 'failed' if has_failed else 'successful')
+                    result.append(workflow_job.id)
-                result.append(workflow_job.id)
+                    new_status = 'failed' if has_failed else 'successful'
-                new_status = 'failed' if has_failed else 'successful'
+                    logger.debug("Transitioning {} to {} status.".format(workflow_job.log_format, new_status))
-                logger.debug("Transitioning {} to {} status.".format(workflow_job.log_format, new_status))
+                    update_fields = ['status', 'start_args']
-                update_fields = ['status', 'start_args']
+                    workflow_job.status = new_status
-                workflow_job.status = new_status
+                    if reason:
-                if reason:
+                        logger.info(f'Workflow job {workflow_job.id} failed due to reason: {reason}')
-                    logger.info(f'Workflow job {workflow_job.id} failed due to reason: {reason}')
+                        workflow_job.job_explanation = gettext_noop("No error handling paths found, marking workflow as failed")
-                    workflow_job.job_explanation = gettext_noop("No error handling paths found, marking workflow as failed")
+                        update_fields.append('job_explanation')
-                    update_fields.append('job_explanation')
+                    workflow_job.start_args = ''  # blank field to remove encrypted passwords
-                workflow_job.start_args = ''  # blank field to remove encrypted passwords
+                    workflow_job.save(update_fields=update_fields)
-                workflow_job.save(update_fields=update_fields)
+                    status_changed = True
-                status_changed = True
+
             if status_changed:
                 if workflow_job.spawned_by_workflow:
-                    schedule_task_manager()
+                    ScheduleWorkflowManager().schedule()
                 workflow_job.websocket_emit_status(workflow_job.status)
                 # Operations whose queries rely on modifications made during the atomic scheduling session
                 workflow_job.send_notification_templates('succeeded' if workflow_job.status == 'successful' else 'failed')
+
+            if workflow_job.status == 'running':
+                spawn_nodes = dag.bfs_nodes_to_run()
+                if spawn_nodes:
+                    logger.debug('Spawning jobs for %s', workflow_job.log_format)
+                else:
+                    logger.debug('No nodes to spawn for %s', workflow_job.log_format)
+                for spawn_node in spawn_nodes:
+                    if spawn_node.unified_job_template is None:
+                        continue
+                    kv = spawn_node.get_job_kwargs()
+                    job = spawn_node.unified_job_template.create_unified_job(**kv)
+                    spawn_node.job = job
+                    spawn_node.save()
+                    logger.debug('Spawned %s in %s for node %s', job.log_format, workflow_job.log_format, spawn_node.pk)
+                    can_start = True
+                    if isinstance(spawn_node.unified_job_template, WorkflowJobTemplate):
+                        workflow_ancestors = job.get_ancestor_workflows()
+                        if spawn_node.unified_job_template in set(workflow_ancestors):
+                            can_start = False
+                            logger.info(
+                                'Refusing to start recursive workflow-in-workflow id={}, wfjt={}, ancestors={}'.format(
+                                    job.id, spawn_node.unified_job_template.pk, [wa.pk for wa in workflow_ancestors]
+                                )
+                            )
+                            display_list = [spawn_node.unified_job_template] + workflow_ancestors
+                            job.job_explanation = gettext_noop(
+                                "Workflow Job spawned from workflow could not start because it "
+                                "would result in recursion (spawn order, most recent first: {})"
+                            ).format(', '.join('<{}>'.format(tmp) for tmp in display_list))
+                        else:
+                            logger.debug(
+                                'Starting workflow-in-workflow id={}, wfjt={}, ancestors={}'.format(
+                                    job.id, spawn_node.unified_job_template.pk, [wa.pk for wa in workflow_ancestors]
+                                )
+                            )
+                    if not job._resources_sufficient_for_launch():
+                        can_start = False
+                        job.job_explanation = gettext_noop(
+                            "Job spawned from workflow could not start because it was missing a related resource such as project or inventory"
+                        )
+                    if can_start:
+                        if workflow_job.start_args:
+                            start_args = json.loads(decrypt_field(workflow_job, 'start_args'))
+                        else:
+                            start_args = {}
+                        can_start = job.signal_start(**start_args)
+                        if not can_start:
+                            job.job_explanation = gettext_noop(
+                                "Job spawned from workflow could not start because it was not in the right state or required manual credentials"
+                            )
+                    if not can_start:
+                        job.status = 'failed'
+                        job.save(update_fields=['status', 'job_explanation'])
+                        job.websocket_emit_status('failed')
+
+                    # TODO: should we emit a status on the socket here similar to tasks.py awx_periodic_scheduler() ?
+                    # emit_websocket_notification('/socket.io/jobs', '', dict(id=))
+
         return result
 
     @timeit
-    def start_task(self, task, instance_group, dependent_tasks=None, instance=None):
+    def get_tasks(self, filter_args):
-        self.subsystem_metrics.inc("task_manager_tasks_started", 1)
+        self.all_tasks = [wf for wf in WorkflowJob.objects.filter(**filter_args)]
-        self.start_task_limit -= 1
-        if self.start_task_limit == 0:
-            # schedule another run immediately after this task manager
-            schedule_task_manager()
-        from awx.main.tasks.system import handle_work_error, handle_work_success
-
-        dependent_tasks = dependent_tasks or []
-
-        task_actual = {
-            'type': get_type_for_model(type(task)),
-            'id': task.id,
-        }
-        dependencies = [{'type': get_type_for_model(type(t)), 'id': t.id} for t in dependent_tasks]
-
-        task.status = 'waiting'
-
-        (start_status, opts) = task.pre_start()
-        if not start_status:
-            task.status = 'failed'
-            if task.job_explanation:
-                task.job_explanation += ' '
-            task.job_explanation += 'Task failed pre-start check.'
-            task.save()
-            # TODO: run error handler to fail sub-tasks and send notifications
-        else:
-            if type(task) is WorkflowJob:
-                task.status = 'running'
-                task.send_notification_templates('running')
-                logger.debug('Transitioning %s to running status.', task.log_format)
-                schedule_task_manager()
-            # at this point we already have control/execution nodes selected for the following cases
-            else:
-                task.instance_group = instance_group
-                execution_node_msg = f' and execution node {task.execution_node}' if task.execution_node else ''
-                logger.debug(
-                    f'Submitting job {task.log_format} controlled by {task.controller_node} to instance group {instance_group.name}{execution_node_msg}.'
-                )
-            with disable_activity_stream():
-                task.celery_task_id = str(uuid.uuid4())
-                task.save()
-                task.log_lifecycle("waiting")
-
-        def post_commit():
-            if task.status != 'failed' and type(task) is not WorkflowJob:
-                # Before task is dispatched, ensure that job_event partitions exist
-                create_partition(task.event_class._meta.db_table, start=task.created)
-                task_cls = task._get_task_class()
-                task_cls.apply_async(
-                    [task.pk],
-                    opts,
-                    queue=task.get_queue_name(),
-                    uuid=task.celery_task_id,
-                    callbacks=[{'task': handle_work_success.name, 'kwargs': {'task_actual': task_actual}}],
-                    errbacks=[{'task': handle_work_error.name, 'args': [task.celery_task_id], 'kwargs': {'subtasks': [task_actual] + dependencies}}],
-                )
-
-        task.websocket_emit_status(task.status)  # adds to on_commit
-        connection.on_commit(post_commit)
 
     @timeit
-    def process_running_tasks(self, running_tasks):
+    def _schedule(self):
-        for task in running_tasks:
+        self.get_tasks(dict(status__in=["running"], dependencies_processed=True))
-            self.dependency_graph.add_job(task)
+        if len(self.all_tasks) > 0:
+            self.spawn_workflow_graph_jobs()
+
+
+class DependencyManager(TaskBase):
+    def __init__(self):
+        super().__init__(prefix="dependency_manager")
 
     def create_project_update(self, task, project_id=None):
         if project_id is None:
@@ -341,14 +282,20 @@ class TaskManager:
         inventory_task.status = 'pending'
         inventory_task.save()
         logger.debug('Spawned {} as dependency of {}'.format(inventory_task.log_format, task.log_format))
-        # inventory_sources = self.get_inventory_source_tasks([task])
+
-        # self.process_inventory_sources(inventory_sources)
         return inventory_task
 
     def add_dependencies(self, task, dependencies):
         with disable_activity_stream():
             task.dependent_jobs.add(*dependencies)
 
+    def get_inventory_source_tasks(self):
+        inventory_ids = set()
+        for task in self.all_tasks:
+            if isinstance(task, Job):
+                inventory_ids.add(task.inventory_id)
+        self.all_inventory_sources = [invsrc for invsrc in InventorySource.objects.filter(inventory_id__in=inventory_ids, update_on_launch=True)]
+
     def get_latest_inventory_update(self, inventory_source):
         latest_inventory_update = InventoryUpdate.objects.filter(inventory_source=inventory_source).order_by("-created")
         if not latest_inventory_update.exists():
@@ -481,16 +428,166 @@ class TaskManager:
 
         return created_dependencies
 
+    def process_tasks(self):
+        deps = self.generate_dependencies(self.all_tasks)
+        self.generate_dependencies(deps)
+        self.subsystem_metrics.inc(f"{self.prefix}_pending_processed", len(self.all_tasks) + len(deps))
+
+    @timeit
+    def _schedule(self):
+        self.get_tasks(dict(status__in=["pending"], dependencies_processed=False))
+
+        if len(self.all_tasks) > 0:
+            self.get_inventory_source_tasks()
+            self.process_tasks()
+            ScheduleTaskManager().schedule()
+
+
+class TaskManager(TaskBase):
+    def __init__(self):
+        """
+        Do NOT put database queries or other potentially expensive operations
+        in the task manager init. The task manager object is created every time a
+        job is created, transitions state, and every 30 seconds on each tower node.
+        More often then not, the object is destroyed quickly because the NOOP case is hit.
+
+        The NOOP case is short-circuit logic. If the task manager realizes that another instance
+        of the task manager is already running, then it short-circuits and decides not to run.
+        """
+        # start task limit indicates how many pending jobs can be started on this
+        # .schedule() run. Starting jobs is expensive, and there is code in place to reap
+        # the task manager after 5 minutes. At scale, the task manager can easily take more than
+        # 5 minutes to start pending jobs. If this limit is reached, pending jobs
+        # will no longer be started and will be started on the next task manager cycle.
+        self.time_delta_job_explanation = timedelta(seconds=30)
+        super().__init__(prefix="task_manager")
+
+    def after_lock_init(self):
+        """
+        Init AFTER we know this instance of the task manager will run because the lock is acquired.
+        """
+        self.dependency_graph = DependencyGraph()
+        self.instances = TaskManagerInstances(self.all_tasks)
+        self.instance_groups = TaskManagerInstanceGroups(instances_by_hostname=self.instances)
+        self.controlplane_ig = self.instance_groups.controlplane_ig
+
+    def job_blocked_by(self, task):
+        # TODO: I'm not happy with this, I think blocking behavior should be decided outside of the dependency graph
+        #       in the old task manager this was handled as a method on each task object outside of the graph and
+        #       probably has the side effect of cutting down *a lot* of the logic from this task manager class
+        blocked_by = self.dependency_graph.task_blocked_by(task)
+        if blocked_by:
+            return blocked_by
+
+        for dep in task.dependent_jobs.all():
+            if dep.status in ACTIVE_STATES:
+                return dep
+            # if we detect a failed or error dependency, go ahead and fail this
+            # task. The errback on the dependency takes some time to trigger,
+            # and we don't want the task to enter running state if its
+            # dependency has failed or errored.
+            elif dep.status in ("error", "failed"):
+                task.status = 'failed'
+                task.job_explanation = 'Previous Task Failed: {"job_type": "%s", "job_name": "%s", "job_id": "%s"}' % (
+                    get_type_for_model(type(dep)),
+                    dep.name,
+                    dep.id,
+                )
+                task.save(update_fields=['status', 'job_explanation'])
+                task.websocket_emit_status('failed')
+                return dep
+
+        return None
+
+    @timeit
+    def start_task(self, task, instance_group, dependent_tasks=None, instance=None):
+        self.dependency_graph.add_job(task)
+        self.subsystem_metrics.inc(f"{self.prefix}_tasks_started", 1)
+        self.start_task_limit -= 1
+        if self.start_task_limit == 0:
+            # schedule another run immediately after this task manager
+            ScheduleTaskManager().schedule()
+        from awx.main.tasks.system import handle_work_error, handle_work_success
+
+        # update capacity for control node and execution node
+        if task.controller_node:
+            self.instances[task.controller_node].consume_capacity(settings.AWX_CONTROL_NODE_TASK_IMPACT)
+        if task.execution_node:
+            self.instances[task.execution_node].consume_capacity(task.task_impact)
+
+        dependent_tasks = dependent_tasks or []
+
+        task_actual = {
+            'type': get_type_for_model(type(task)),
+            'id': task.id,
+        }
+        dependencies = [{'type': get_type_for_model(type(t)), 'id': t.id} for t in dependent_tasks]
+
+        task.status = 'waiting'
+
+        (start_status, opts) = task.pre_start()
+        if not start_status:
+            task.status = 'failed'
+            if task.job_explanation:
+                task.job_explanation += ' '
+            task.job_explanation += 'Task failed pre-start check.'
+            task.save()
+            # TODO: run error handler to fail sub-tasks and send notifications
+        else:
+            if type(task) is WorkflowJob:
+                task.status = 'running'
+                task.send_notification_templates('running')
+                logger.debug('Transitioning %s to running status.', task.log_format)
+                # Call this to ensure Workflow nodes get spawned in timely manner
+                ScheduleWorkflowManager().schedule()
+            # at this point we already have control/execution nodes selected for the following cases
+            else:
+                task.instance_group = instance_group
+                execution_node_msg = f' and execution node {task.execution_node}' if task.execution_node else ''
+                logger.debug(
+                    f'Submitting job {task.log_format} controlled by {task.controller_node} to instance group {instance_group.name}{execution_node_msg}.'
+                )
+            with disable_activity_stream():
+                task.celery_task_id = str(uuid.uuid4())
+                task.save()
+                task.log_lifecycle("waiting")
+
+        def post_commit():
+            if task.status != 'failed' and type(task) is not WorkflowJob:
+                # Before task is dispatched, ensure that job_event partitions exist
+                create_partition(task.event_class._meta.db_table, start=task.created)
+                task_cls = task._get_task_class()
+                task_cls.apply_async(
+                    [task.pk],
+                    opts,
+                    queue=task.get_queue_name(),
+                    uuid=task.celery_task_id,
+                    callbacks=[{'task': handle_work_success.name, 'kwargs': {'task_actual': task_actual}}],
+                    errbacks=[{'task': handle_work_error.name, 'args': [task.celery_task_id], 'kwargs': {'subtasks': [task_actual] + dependencies}}],
+                )
+
+        task.websocket_emit_status(task.status)  # adds to on_commit
+        connection.on_commit(post_commit)
+
+    @timeit
+    def process_running_tasks(self, running_tasks):
+        for task in running_tasks:
+            if type(task) is WorkflowJob:
+                ScheduleWorkflowManager().schedule()
+            self.dependency_graph.add_job(task)
+
     @timeit
     def process_pending_tasks(self, pending_tasks):
-        running_workflow_templates = {wf.unified_job_template_id for wf in self.get_running_workflow_jobs()}
         tasks_to_update_job_explanation = []
         for task in pending_tasks:
             if self.start_task_limit <= 0:
                 break
+            if self.timed_out():
+                logger.warning("Task manager has reached time out while processing pending jobs, exiting loop early")
+                break
             blocked_by = self.job_blocked_by(task)
             if blocked_by:
-                self.subsystem_metrics.inc("task_manager_tasks_blocked", 1)
+                self.subsystem_metrics.inc(f"{self.prefix}_tasks_blocked", 1)
                 task.log_lifecycle("blocked", blocked_by=blocked_by)
                 job_explanation = gettext_noop(f"waiting for {blocked_by._meta.model_name}-{blocked_by.id} to finish")
                 if task.job_explanation != job_explanation:
@@ -499,19 +596,16 @@ class TaskManager:
                         tasks_to_update_job_explanation.append(task)
                 continue
 
-            found_acceptable_queue = False
-            preferred_instance_groups = task.preferred_instance_groups
-
             if isinstance(task, WorkflowJob):
-                if task.unified_job_template_id in running_workflow_templates:
+                # Previously we were tracking allow_simultaneous blocking both here and in DependencyGraph.
-                    if not task.allow_simultaneous:
+                # Double check that using just the DependencyGraph works for Workflows and Sliced Jobs.
-                        logger.debug("{} is blocked from running, workflow already running".format(task.log_format))
-                        continue
-                else:
-                    running_workflow_templates.add(task.unified_job_template_id)
                 self.start_task(task, None, task.get_jobs_fail_chain(), None)
                 continue
 
+            found_acceptable_queue = False
+
+            preferred_instance_groups = self.instance_groups.get_instance_groups_from_task_cache(task)
+
             # Determine if there is control capacity for the task
             if task.capacity_type == 'control':
                 control_impact = task.task_impact + settings.AWX_CONTROL_NODE_TASK_IMPACT
@@ -530,8 +624,6 @@ class TaskManager:
             # All task.capacity_type == 'control' jobs should run on control plane, no need to loop over instance groups
             if task.capacity_type == 'control':
                 task.execution_node = control_instance.hostname
-                control_instance.consume_capacity(control_impact)
-                self.dependency_graph.add_job(task)
                 execution_instance = self.instances[control_instance.hostname].obj
                 task.log_lifecycle("controller_node_chosen")
                 task.log_lifecycle("execution_node_chosen")
@@ -541,7 +633,6 @@ class TaskManager:
 
             for instance_group in preferred_instance_groups:
                 if instance_group.is_container_group:
-                    self.dependency_graph.add_job(task)
                     self.start_task(task, instance_group, task.get_jobs_fail_chain(), None)
                     found_acceptable_queue = True
                     break
@@ -563,9 +654,7 @@ class TaskManager:
                         control_instance = execution_instance
                         task.controller_node = execution_instance.hostname
 
-                    control_instance.consume_capacity(settings.AWX_CONTROL_NODE_TASK_IMPACT)
                     task.log_lifecycle("controller_node_chosen")
-                    execution_instance.consume_capacity(task.task_impact)
                     task.log_lifecycle("execution_node_chosen")
                     logger.debug(
                         "Starting {} in group {} instance {} (remaining_capacity={})".format(
@@ -573,7 +662,6 @@ class TaskManager:
                         )
                     )
                     execution_instance = self.instances[execution_instance.hostname].obj
-                    self.dependency_graph.add_job(task)
                     self.start_task(task, instance_group, task.get_jobs_fail_chain(), execution_instance)
                     found_acceptable_queue = True
                     break
@@ -599,25 +687,6 @@ class TaskManager:
                 tasks_to_update_job_explanation.append(task)
         logger.debug("{} couldn't be scheduled on graph, waiting for next cycle".format(task.log_format))
 
-    def timeout_approval_node(self):
-        workflow_approvals = WorkflowApproval.objects.filter(status='pending')
-        now = tz_now()
-        for task in workflow_approvals:
-            approval_timeout_seconds = timedelta(seconds=task.timeout)
-            if task.timeout == 0:
-                continue
-            if (now - task.created) >= approval_timeout_seconds:
-                timeout_message = _("The approval node {name} ({pk}) has expired after {timeout} seconds.").format(
-                    name=task.name, pk=task.pk, timeout=task.timeout
-                )
-                logger.warning(timeout_message)
-                task.timed_out = True
-                task.status = 'failed'
-                task.send_approval_notification('timed_out')
-                task.websocket_emit_status(task.status)
-                task.job_explanation = timeout_message
-                task.save(update_fields=['status', 'job_explanation', 'timed_out'])
-
     def reap_jobs_from_orphaned_instances(self):
         # discover jobs that are in running state but aren't on an execution node
         # that we know about; this is a fairly rare event, but it can occur if you,
@@ -630,92 +699,45 @@ class TaskManager:
                 logger.error(f'{j.execution_node} is not a registered instance; reaping {j.log_format}')
                 reap_job(j, 'failed')
 
-    def process_tasks(self, all_sorted_tasks):
+    def process_tasks(self):
-        running_tasks = [t for t in all_sorted_tasks if t.status in ['waiting', 'running']]
+        running_tasks = [t for t in self.all_tasks if t.status in ['waiting', 'running']]
         self.process_running_tasks(running_tasks)
-        self.subsystem_metrics.inc("task_manager_running_processed", len(running_tasks))
+        self.subsystem_metrics.inc(f"{self.prefix}_running_processed", len(running_tasks))
 
-        pending_tasks = [t for t in all_sorted_tasks if t.status == 'pending']
+        pending_tasks = [t for t in self.all_tasks if t.status == 'pending']
-
-        undeped_tasks = [t for t in pending_tasks if not t.dependencies_processed]
-        dependencies = self.generate_dependencies(undeped_tasks)
-        deps_of_deps = self.generate_dependencies(dependencies)
-        dependencies += deps_of_deps
-        self.process_pending_tasks(dependencies)
-        self.subsystem_metrics.inc("task_manager_pending_processed", len(dependencies))
 
         self.process_pending_tasks(pending_tasks)
-        self.subsystem_metrics.inc("task_manager_pending_processed", len(pending_tasks))
+        self.subsystem_metrics.inc(f"{self.prefix}_pending_processed", len(pending_tasks))
+
+    def timeout_approval_node(self, task):
+        if self.timed_out():
+            logger.warning("Task manager has reached time out while processing approval nodes, exiting loop early")
+            # Do not process any more workflow approval nodes. Stop here.
+            # Maybe we should schedule another TaskManager run
+            return
+        timeout_message = _("The approval node {name} ({pk}) has expired after {timeout} seconds.").format(name=task.name, pk=task.pk, timeout=task.timeout)
+        logger.warning(timeout_message)
+        task.timed_out = True
+        task.status = 'failed'
+        task.send_approval_notification('timed_out')
+        task.websocket_emit_status(task.status)
+        task.job_explanation = timeout_message
+        task.save(update_fields=['status', 'job_explanation', 'timed_out'])
+
+    def get_expired_workflow_approvals(self):
+        # timeout of 0 indicates that it never expires
+        qs = WorkflowApproval.objects.filter(status='pending').exclude(timeout=0).filter(expires__lt=tz_now())
+        return qs
 
     @timeit
     def _schedule(self):
-        finished_wfjs = []
+        self.get_tasks(dict(status__in=["pending", "waiting", "running"], dependencies_processed=True))
-        all_sorted_tasks = self.get_tasks()
 
-        self.after_lock_init(all_sorted_tasks)
+        self.after_lock_init()
+        self.reap_jobs_from_orphaned_instances()
 
-        if len(all_sorted_tasks) > 0:
+        if len(self.all_tasks) > 0:
-            # TODO: Deal with
+            self.process_tasks()
-            # latest_project_updates = self.get_latest_project_update_tasks(all_sorted_tasks)
-            # self.process_latest_project_updates(latest_project_updates)
 
-            # latest_inventory_updates = self.get_latest_inventory_update_tasks(all_sorted_tasks)
+        for workflow_approval in self.get_expired_workflow_approvals():
-            # self.process_latest_inventory_updates(latest_inventory_updates)
+            self.timeout_approval_node(workflow_approval)
-
-            self.all_inventory_sources = self.get_inventory_source_tasks(all_sorted_tasks)
-
-            running_workflow_tasks = self.get_running_workflow_jobs()
-            finished_wfjs = self.process_finished_workflow_jobs(running_workflow_tasks)
-
-            previously_running_workflow_tasks = running_workflow_tasks
-            running_workflow_tasks = []
-            for workflow_job in previously_running_workflow_tasks:
-                if workflow_job.status == 'running':
-                    running_workflow_tasks.append(workflow_job)
-                else:
-                    logger.debug('Removed %s from job spawning consideration.', workflow_job.log_format)
-
-            self.spawn_workflow_graph_jobs(running_workflow_tasks)
-
-            self.timeout_approval_node()
-            self.reap_jobs_from_orphaned_instances()
-
-            self.process_tasks(all_sorted_tasks)
-        return finished_wfjs
-
-    def record_aggregate_metrics(self, *args):
-        if not settings.IS_TESTING():
-            # increment task_manager_schedule_calls regardless if the other
-            # metrics are recorded
-            s_metrics.Metrics(auto_pipe_execute=True).inc("task_manager_schedule_calls", 1)
-            # Only record metrics if the last time recording was more
-            # than SUBSYSTEM_METRICS_TASK_MANAGER_RECORD_INTERVAL ago.
-            # Prevents a short-duration task manager that runs directly after a
-            # long task manager to override useful metrics.
-            current_time = time.time()
-            time_last_recorded = current_time - self.subsystem_metrics.decode("task_manager_recorded_timestamp")
-            if time_last_recorded > settings.SUBSYSTEM_METRICS_TASK_MANAGER_RECORD_INTERVAL:
-                logger.debug(f"recording metrics, last recorded {time_last_recorded} seconds ago")
-                self.subsystem_metrics.set("task_manager_recorded_timestamp", current_time)
-                self.subsystem_metrics.pipe_execute()
-            else:
-                logger.debug(f"skipping recording metrics, last recorded {time_last_recorded} seconds ago")
-
-    def record_aggregate_metrics_and_exit(self, *args):
-        self.record_aggregate_metrics()
-        sys.exit(1)
-
-    def schedule(self):
-        # Lock
-        with advisory_lock('task_manager_lock', wait=False) as acquired:
-            with transaction.atomic():
-                if acquired is False:
-                    logger.debug("Not running scheduler, another task holds lock")
-                    return
-                logger.debug("Starting Scheduler")
-                with task_manager_bulk_reschedule():
-                    # if sigterm due to timeout, still record metrics
-                    signal.signal(signal.SIGTERM, self.record_aggregate_metrics_and_exit)
-                    self._schedule()
-                    self.record_aggregate_metrics()
-                logger.debug("Finishing Scheduler")

awx/main/scheduler/task_manager_models.py

@@ -67,6 +67,7 @@ class TaskManagerInstanceGroups:
     def __init__(self, instances_by_hostname=None, instance_groups=None, instance_groups_queryset=None):
         self.instance_groups = dict()
         self.controlplane_ig = None
+        self.pk_ig_map = dict()
 
         if instance_groups is not None:  # for testing
             self.instance_groups = instance_groups
@@ -81,6 +82,7 @@ class TaskManagerInstanceGroups:
                         instances_by_hostname[instance.hostname] for instance in instance_group.instances.all() if instance.hostname in instances_by_hostname
                     ],
                 )
+                self.pk_ig_map[instance_group.pk] = instance_group
 
     def get_remaining_capacity(self, group_name):
         instances = self.instance_groups[group_name]['instances']
@@ -121,3 +123,17 @@ class TaskManagerInstanceGroups:
                 elif i.capacity > largest_instance.capacity:
                     largest_instance = i
         return largest_instance
+
+    def get_instance_groups_from_task_cache(self, task):
+        igs = []
+        if task.preferred_instance_groups_cache:
+            for pk in task.preferred_instance_groups_cache:
+                ig = self.pk_ig_map.get(pk, None)
+                if ig:
+                    igs.append(ig)
+                else:
+                    logger.warn(f"Unknown instance group with pk {pk} for task {task}")
+        if len(igs) == 0:
+            logger.warn(f"No instance groups in cache exist, defaulting to global instance groups for task {task}")
+            return task.global_instance_groups
+        return igs

awx/main/scheduler/tasks.py

@@ -1,15 +1,35 @@
 # Python
 import logging
 
+# Django
+from django.conf import settings
+
 # AWX
-from awx.main.scheduler import TaskManager
+from awx import MODE
+from awx.main.scheduler import TaskManager, DependencyManager, WorkflowManager
 from awx.main.dispatch.publish import task
 from awx.main.dispatch import get_local_queuename
 
 logger = logging.getLogger('awx.main.scheduler')
 
 
+def run_manager(manager, prefix):
+    if MODE == 'development' and settings.AWX_DISABLE_TASK_MANAGERS:
+        logger.debug(f"Not running {prefix} manager, AWX_DISABLE_TASK_MANAGERS is True. Trigger with GET to /api/debug/{prefix}_manager/")
+        return
+    manager().schedule()
+
+
 @task(queue=get_local_queuename)
-def run_task_manager():
+def task_manager():
-    logger.debug("Running task manager.")
+    run_manager(TaskManager, "task")
-    TaskManager().schedule()
+
+
+@task(queue=get_local_queuename)
+def dependency_manager():
+    run_manager(DependencyManager, "dependency")
+
+
+@task(queue=get_local_queuename)
+def workflow_manager():
+    run_manager(WorkflowManager, "workflow")

awx/main/tasks/system.py

@@ -53,7 +53,8 @@ from awx.main.dispatch import get_local_queuename, reaper
 from awx.main.utils.common import (
     ignore_inventory_computed_fields,
     ignore_inventory_group_removal,
-    schedule_task_manager,
+    ScheduleWorkflowManager,
+    ScheduleTaskManager,
 )
 
 from awx.main.utils.external_logging import reconfigure_rsyslog
@@ -657,6 +658,13 @@ def awx_periodic_scheduler():
         state.save()
 
 
+def schedule_manager_success_or_error(instance):
+    if instance.unifiedjob_blocked_jobs.exists():
+        ScheduleTaskManager().schedule()
+    if instance.spawned_by_workflow:
+        ScheduleWorkflowManager().schedule()
+
+
 @task(queue=get_local_queuename)
 def handle_work_success(task_actual):
     try:
@@ -666,8 +674,7 @@ def handle_work_success(task_actual):
         return
     if not instance:
         return
-
+    schedule_manager_success_or_error(instance)
-    schedule_task_manager()
 
 
 @task(queue=get_local_queuename)
@@ -709,8 +716,7 @@ def handle_work_error(task_id, *args, **kwargs):
     # what the job complete message handler does then we may want to send a
     # completion event for each job here.
     if first_instance:
-        schedule_task_manager()
+        schedule_manager_success_or_error(first_instance)
-        pass
 
 
 @task(queue=get_local_queuename)

awx/main/tests/functional/api/test_workflow_node.py

@@ -13,7 +13,10 @@ from awx.main.models.workflow import (
     WorkflowJobTemplateNode,
 )
 from awx.main.models.credential import Credential
-from awx.main.scheduler import TaskManager
+from awx.main.scheduler import TaskManager, WorkflowManager, DependencyManager
+
+# Django
+from django.utils.timezone import now, timedelta
 
 
 @pytest.fixture
@@ -137,8 +140,9 @@ class TestApprovalNodes:
         post(url, {'name': 'Approve Test', 'description': '', 'timeout': 0}, user=admin_user, expect=201)
         post(reverse('api:workflow_job_template_launch', kwargs={'pk': wfjt.pk}), user=admin_user, expect=201)
         wf_job = WorkflowJob.objects.first()
+        DependencyManager().schedule()  # TODO: exclude workflows from this and delete line
         TaskManager().schedule()
-        TaskManager().schedule()
+        WorkflowManager().schedule()
         wfj_node = wf_job.workflow_nodes.first()
         approval = wfj_node.job
         assert approval.name == 'Approve Test'
@@ -162,8 +166,9 @@ class TestApprovalNodes:
         post(url, {'name': 'Deny Test', 'description': '', 'timeout': 0}, user=admin_user, expect=201)
         post(reverse('api:workflow_job_template_launch', kwargs={'pk': wfjt.pk}), user=admin_user, expect=201)
         wf_job = WorkflowJob.objects.first()
+        DependencyManager().schedule()  # TODO: exclude workflows from this and delete line
         TaskManager().schedule()
-        TaskManager().schedule()
+        WorkflowManager().schedule()
         wfj_node = wf_job.workflow_nodes.first()
         approval = wfj_node.job
         assert approval.name == 'Deny Test'
@@ -216,6 +221,37 @@ class TestApprovalNodes:
         approval.refresh_from_db()
         assert approval.status == 'failed'
 
+    def test_expires_time_on_creation(self):
+        now_time = now()
+        wa = WorkflowApproval.objects.create(timeout=34)
+        # this is fudged, so we assert that the expires time is in reasonable range
+        assert timedelta(seconds=33) < (wa.expires - now_time) < timedelta(seconds=35)
+
+    @pytest.mark.parametrize('with_update_fields', [True, False])
+    def test_expires_time_update(self, with_update_fields):
+        wa = WorkflowApproval.objects.create()
+        assert wa.timeout == 0
+        assert wa.expires is None
+        wa.timeout = 1234
+        if with_update_fields:
+            wa.save(update_fields=['timeout'])
+        else:
+            wa.save()
+        assert wa.created + timedelta(seconds=1234) == wa.expires
+
+    @pytest.mark.parametrize('with_update_fields', [True, False])
+    def test_reset_timeout_and_expires(self, with_update_fields):
+        wa = WorkflowApproval.objects.create()
+        wa.timeout = 1234
+        wa.save()
+        assert wa.expires
+        wa.timeout = 0
+        if with_update_fields:
+            wa.save(update_fields=['timeout'])
+        else:
+            wa.save()
+        assert wa.expires is None
+
 
 @pytest.mark.django_db
 class TestExclusiveRelationshipEnforcement:

awx/main/tests/functional/models/test_unified_job.py

@@ -252,12 +252,14 @@ class TestTaskImpact:
     def test_limit_task_impact(self, job_host_limit, run_computed_fields_right_away):
         job = job_host_limit(5, 2)
         job.inventory.update_computed_fields()
+        job.task_impact = job._get_task_impact()
         assert job.inventory.total_hosts == 5
         assert job.task_impact == 2 + 1  # forks becomes constraint
 
     def test_host_task_impact(self, job_host_limit, run_computed_fields_right_away):
         job = job_host_limit(3, 5)
         job.inventory.update_computed_fields()
+        job.task_impact = job._get_task_impact()
         assert job.task_impact == 3 + 1  # hosts becomes constraint
 
     def test_shard_task_impact(self, slice_job_factory, run_computed_fields_right_away):
@@ -270,9 +272,13 @@ class TestTaskImpact:
         # Even distribution - all jobs run on 1 host
         assert [len(jobs[0].inventory.get_script_data(slice_number=i + 1, slice_count=3)['all']['hosts']) for i in range(3)] == [1, 1, 1]
         jobs[0].inventory.update_computed_fields()
+        for j in jobs:
+            j.task_impact = j._get_task_impact()
         assert [job.task_impact for job in jobs] == [2, 2, 2]  # plus one base task impact
         # Uneven distribution - first job takes the extra host
         jobs[0].inventory.hosts.create(name='remainder_foo')
         assert [len(jobs[0].inventory.get_script_data(slice_number=i + 1, slice_count=3)['all']['hosts']) for i in range(3)] == [2, 1, 1]
         jobs[0].inventory.update_computed_fields()
+        # recalculate task_impact
+        jobs[0].task_impact = jobs[0]._get_task_impact()
         assert [job.task_impact for job in jobs] == [3, 2, 2]

awx/main/tests/functional/task_management/__init__.py

@@ -0,0 +1,6 @@
+def create_job(jt, dependencies_processed=True):
+    job = jt.create_unified_job()
+    job.status = "pending"
+    job.dependencies_processed = dependencies_processed
+    job.save()
+    return job

awx/main/tests/functional/task_management/test_rampart_groups.py

@@ -1,9 +1,10 @@
 import pytest
 from unittest import mock
 from datetime import timedelta
-from awx.main.scheduler import TaskManager
+from awx.main.scheduler import TaskManager, DependencyManager
-from awx.main.models import InstanceGroup, WorkflowJob
+from awx.main.models import InstanceGroup
 from awx.main.tasks.system import apply_cluster_membership_policies
+from . import create_job
 
 
 @pytest.mark.django_db
@@ -12,16 +13,12 @@ def test_multi_group_basic_job_launch(instance_factory, controlplane_instance_gr
     i2 = instance_factory("i2")
     ig1 = instance_group_factory("ig1", instances=[i1])
     ig2 = instance_group_factory("ig2", instances=[i2])
-    objects1 = job_template_factory('jt1', organization='org1', project='proj1', inventory='inv1', credential='cred1', jobs=["job_should_start"])
+    objects1 = job_template_factory('jt1', organization='org1', project='proj1', inventory='inv1', credential='cred1')
     objects1.job_template.instance_groups.add(ig1)
-    j1 = objects1.jobs['job_should_start']
+    j1 = create_job(objects1.job_template)
-    j1.status = 'pending'
+    objects2 = job_template_factory('jt2', organization='org2', project='proj2', inventory='inv2', credential='cred2')
-    j1.save()
-    objects2 = job_template_factory('jt2', organization='org2', project='proj2', inventory='inv2', credential='cred2', jobs=["job_should_still_start"])
     objects2.job_template.instance_groups.add(ig2)
-    j2 = objects2.jobs['job_should_still_start']
+    j2 = create_job(objects2.job_template)
-    j2.status = 'pending'
-    j2.save()
     with mock.patch('awx.main.models.Job.task_impact', new_callable=mock.PropertyMock) as mock_task_impact:
         mock_task_impact.return_value = 500
         with mocker.patch("awx.main.scheduler.TaskManager.start_task"):
@@ -35,23 +32,26 @@ def test_multi_group_with_shared_dependency(instance_factory, controlplane_insta
     i2 = instance_factory("i2")
     ig1 = instance_group_factory("ig1", instances=[i1])
     ig2 = instance_group_factory("ig2", instances=[i2])
-    objects1 = job_template_factory('jt1', organization='org1', project='proj1', inventory='inv1', credential='cred1', jobs=["job_should_start"])
+    objects1 = job_template_factory(
+        'jt1',
+        organization='org1',
+        project='proj1',
+        inventory='inv1',
+        credential='cred1',
+    )
     objects1.job_template.instance_groups.add(ig1)
+    j1 = create_job(objects1.job_template, dependencies_processed=False)
     p = objects1.project
     p.scm_update_on_launch = True
     p.scm_update_cache_timeout = 0
     p.scm_type = "git"
     p.scm_url = "http://github.com/ansible/ansible.git"
     p.save()
-    j1 = objects1.jobs['job_should_start']
+    objects2 = job_template_factory('jt2', organization=objects1.organization, project=p, inventory='inv2', credential='cred2')
-    j1.status = 'pending'
-    j1.save()
-    objects2 = job_template_factory('jt2', organization=objects1.organization, project=p, inventory='inv2', credential='cred2', jobs=["job_should_still_start"])
     objects2.job_template.instance_groups.add(ig2)
-    j2 = objects2.jobs['job_should_still_start']
+    j2 = create_job(objects2.job_template, dependencies_processed=False)
-    j2.status = 'pending'
-    j2.save()
     with mocker.patch("awx.main.scheduler.TaskManager.start_task"):
+        DependencyManager().schedule()
         TaskManager().schedule()
         pu = p.project_updates.first()
         TaskManager.start_task.assert_called_once_with(pu, controlplane_instance_group, [j1, j2], controlplane_instance_group.instances.all()[0])
@@ -59,6 +59,7 @@ def test_multi_group_with_shared_dependency(instance_factory, controlplane_insta
         pu.status = "successful"
         pu.save()
     with mock.patch("awx.main.scheduler.TaskManager.start_task"):
+        DependencyManager().schedule()
         TaskManager().schedule()
 
         TaskManager.start_task.assert_any_call(j1, ig1, [], i1)
@@ -69,7 +70,7 @@ def test_multi_group_with_shared_dependency(instance_factory, controlplane_insta
 @pytest.mark.django_db
 def test_workflow_job_no_instancegroup(workflow_job_template_factory, controlplane_instance_group, mocker):
     wfjt = workflow_job_template_factory('anicedayforawalk').workflow_job_template
-    wfj = WorkflowJob.objects.create(workflow_job_template=wfjt)
+    wfj = wfjt.create_unified_job()
     wfj.status = "pending"
     wfj.save()
     with mocker.patch("awx.main.scheduler.TaskManager.start_task"):
@@ -85,39 +86,50 @@ def test_overcapacity_blocking_other_groups_unaffected(instance_factory, control
     i1.capacity = 1020
     i1.save()
     i2 = instance_factory("i2")
+    i2.capacity = 1020
+    i2.save()
     ig1 = instance_group_factory("ig1", instances=[i1])
     ig2 = instance_group_factory("ig2", instances=[i2])
-    objects1 = job_template_factory('jt1', organization='org1', project='proj1', inventory='inv1', credential='cred1', jobs=["job_should_start"])
+    objects1 = job_template_factory('jt1', organization='org1', project='proj1', inventory='inv1', credential='cred1')
     objects1.job_template.instance_groups.add(ig1)
-    j1 = objects1.jobs['job_should_start']
+    j1 = create_job(objects1.job_template)
-    j1.status = 'pending'
+    objects2 = job_template_factory('jt2', organization=objects1.organization, project='proj2', inventory='inv2', credential='cred2')
-    j1.save()
-    objects2 = job_template_factory(
-        'jt2', organization=objects1.organization, project='proj2', inventory='inv2', credential='cred2', jobs=["job_should_start", "job_should_also_start"]
-    )
     objects2.job_template.instance_groups.add(ig1)
-    j1_1 = objects2.jobs['job_should_also_start']
+    j1_1 = create_job(objects2.job_template)
-    j1_1.status = 'pending'
+    objects3 = job_template_factory('jt3', organization='org2', project='proj3', inventory='inv3', credential='cred3')
-    j1_1.save()
-    objects3 = job_template_factory('jt3', organization='org2', project='proj3', inventory='inv3', credential='cred3', jobs=["job_should_still_start"])
     objects3.job_template.instance_groups.add(ig2)
-    j2 = objects3.jobs['job_should_still_start']
+    j2 = create_job(objects3.job_template)
-    j2.status = 'pending'
+    objects4 = job_template_factory('jt4', organization=objects3.organization, project='proj4', inventory='inv4', credential='cred4')
-    j2.save()
-    objects4 = job_template_factory(
-        'jt4', organization=objects3.organization, project='proj4', inventory='inv4', credential='cred4', jobs=["job_should_not_start"]
-    )
     objects4.job_template.instance_groups.add(ig2)
-    j2_1 = objects4.jobs['job_should_not_start']
+    j2_1 = create_job(objects4.job_template)
-    j2_1.status = 'pending'
+
-    j2_1.save()
-    tm = TaskManager()
     with mock.patch('awx.main.models.Job.task_impact', new_callable=mock.PropertyMock) as mock_task_impact:
         mock_task_impact.return_value = 500
-        with mock.patch.object(TaskManager, "start_task", wraps=tm.start_task) as mock_job:
+        TaskManager().schedule()
-            tm.schedule()
+
-            mock_job.assert_has_calls([mock.call(j1, ig1, [], i1), mock.call(j1_1, ig1, [], i1), mock.call(j2, ig2, [], i2)])
+        # all jobs should be able to run, plenty of capacity across both instances
-            assert mock_job.call_count == 3
+        for j in [j1, j1_1, j2, j2_1]:
+            j.refresh_from_db()
+            assert j.status == "waiting"
+
+        # reset to pending
+        for j in [j1, j1_1, j2, j2_1]:
+            j.status = "pending"
+            j.save()
+
+        # make i2 can only be able to fit 1 job
+        i2.capacity = 510
+        i2.save()
+
+        TaskManager().schedule()
+
+        for j in [j1, j1_1, j2]:
+            j.refresh_from_db()
+            assert j.status == "waiting"
+
+        j2_1.refresh_from_db()
+        # could not run because i2 is full
+        assert j2_1.status == "pending"
 
 
 @pytest.mark.django_db
@@ -126,19 +138,13 @@ def test_failover_group_run(instance_factory, controlplane_instance_group, mocke
     i2 = instance_factory("i2")
     ig1 = instance_group_factory("ig1", instances=[i1])
     ig2 = instance_group_factory("ig2", instances=[i2])
-    objects1 = job_template_factory('jt1', organization='org1', project='proj1', inventory='inv1', credential='cred1', jobs=["job_should_start"])
+    objects1 = job_template_factory('jt1', organization='org1', project='proj1', inventory='inv1', credential='cred1')
     objects1.job_template.instance_groups.add(ig1)
-    j1 = objects1.jobs['job_should_start']
+    j1 = create_job(objects1.job_template)
-    j1.status = 'pending'
+    objects2 = job_template_factory('jt2', organization=objects1.organization, project='proj2', inventory='inv2', credential='cred2')
-    j1.save()
-    objects2 = job_template_factory(
-        'jt2', organization=objects1.organization, project='proj2', inventory='inv2', credential='cred2', jobs=["job_should_start", "job_should_also_start"]
-    )
     objects2.job_template.instance_groups.add(ig1)
     objects2.job_template.instance_groups.add(ig2)
-    j1_1 = objects2.jobs['job_should_also_start']
+    j1_1 = create_job(objects2.job_template)
-    j1_1.status = 'pending'
-    j1_1.save()
     tm = TaskManager()
     with mock.patch('awx.main.models.Job.task_impact', new_callable=mock.PropertyMock) as mock_task_impact:
         mock_task_impact.return_value = 500

awx/main/tests/functional/task_management/test_scheduler.py

@@ -3,21 +3,19 @@ from unittest import mock
 import json
 from datetime import timedelta
 
-from awx.main.scheduler import TaskManager
+from awx.main.scheduler import TaskManager, DependencyManager, WorkflowManager
-from awx.main.scheduler.dependency_graph import DependencyGraph
 from awx.main.utils import encrypt_field
 from awx.main.models import WorkflowJobTemplate, JobTemplate, Job
 from awx.main.models.ha import Instance
+from . import create_job
 from django.conf import settings
 
 
 @pytest.mark.django_db
 def test_single_job_scheduler_launch(hybrid_instance, controlplane_instance_group, job_template_factory, mocker):
     instance = controlplane_instance_group.instances.all()[0]
-    objects = job_template_factory('jt', organization='org1', project='proj', inventory='inv', credential='cred', jobs=["job_should_start"])
+    objects = job_template_factory('jt', organization='org1', project='proj', inventory='inv', credential='cred')
-    j = objects.jobs["job_should_start"]
+    j = create_job(objects.job_template)
-    j.status = 'pending'
-    j.save()
     with mocker.patch("awx.main.scheduler.TaskManager.start_task"):
         TaskManager().schedule()
         TaskManager.start_task.assert_called_once_with(j, controlplane_instance_group, [], instance)
@@ -32,10 +30,8 @@ class TestJobLifeCycle:
         expect_commit - list of expected on_commit calls
         If any of these are None, then the assertion is not made.
         """
-        if expect_schedule and len(expect_schedule) > 1:
-            raise RuntimeError('Task manager should reschedule itself one time, at most.')
         with mock.patch('awx.main.models.unified_jobs.UnifiedJob.websocket_emit_status') as mock_channel:
-            with mock.patch('awx.main.utils.common._schedule_task_manager') as tm_sch:
+            with mock.patch('awx.main.utils.common.ScheduleManager._schedule') as tm_sch:
                 # Job are ultimately submitted in on_commit hook, but this will not
                 # actually run, because it waits until outer transaction, which is the test
                 # itself in this case
@@ -56,22 +52,21 @@ class TestJobLifeCycle:
         wj = wfjt.create_unified_job()
         assert wj.workflow_nodes.count() == 2
         wj.signal_start()
-        tm = TaskManager()
 
         # Transitions workflow job to running
         # needs to re-schedule so it spawns jobs next round
-        self.run_tm(tm, [mock.call('running')], [mock.call()])
+        self.run_tm(TaskManager(), [mock.call('running')])
 
         # Spawns jobs
         # needs re-schedule to submit jobs next round
-        self.run_tm(tm, [mock.call('pending'), mock.call('pending')], [mock.call()])
+        self.run_tm(WorkflowManager(), [mock.call('pending'), mock.call('pending')])
 
         assert jt.jobs.count() == 2  # task manager spawned jobs
 
         # Submits jobs
         # intermission - jobs will run and reschedule TM when finished
-        self.run_tm(tm, [mock.call('waiting'), mock.call('waiting')], [])
+        self.run_tm(DependencyManager())  # flip dependencies_processed to True
-
+        self.run_tm(TaskManager(), [mock.call('waiting'), mock.call('waiting')])
         # I am the job runner
         for job in jt.jobs.all():
             job.status = 'successful'
@@ -79,7 +74,7 @@ class TestJobLifeCycle:
 
         # Finishes workflow
         # no further action is necessary, so rescheduling should not happen
-        self.run_tm(tm, [mock.call('successful')], [])
+        self.run_tm(WorkflowManager(), [mock.call('successful')])
 
     def test_task_manager_workflow_workflow_rescheduling(self, controlplane_instance_group):
         wfjts = [WorkflowJobTemplate.objects.create(name='foo')]
@@ -90,16 +85,13 @@ class TestJobLifeCycle:
 
         wj = wfjts[0].create_unified_job()
         wj.signal_start()
-        tm = TaskManager()
 
-        while wfjts[0].status != 'successful':
+        attempts = 10
-            wfjts[1].refresh_from_db()
+        while wfjts[0].status != 'successful' and attempts > 0:
-            if wfjts[1].status == 'successful':
+            self.run_tm(TaskManager())
-                # final run, no more work to do
+            self.run_tm(WorkflowManager())
-                self.run_tm(tm, expect_schedule=[])
-            else:
-                self.run_tm(tm, expect_schedule=[mock.call()])
             wfjts[0].refresh_from_db()
+            attempts -= 1
 
     def test_control_and_execution_instance(self, project, system_job_template, job_template, inventory_source, control_instance, execution_instance):
         assert Instance.objects.count() == 2
@@ -113,6 +105,7 @@ class TestJobLifeCycle:
         for uj in all_ujs:
             uj.signal_start()
 
+        DependencyManager().schedule()
         tm = TaskManager()
         self.run_tm(tm)
 
@@ -135,6 +128,7 @@ class TestJobLifeCycle:
         for uj in all_ujs:
             uj.signal_start()
 
+        DependencyManager().schedule()
         # There is only enough control capacity to run one of the jobs so one should end up in pending and the other in waiting
         tm = TaskManager()
         self.run_tm(tm)
@@ -157,6 +151,7 @@ class TestJobLifeCycle:
         for uj in all_ujs:
             uj.signal_start()
 
+        DependencyManager().schedule()
         # There is only enough control capacity to run one of the jobs so one should end up in pending and the other in waiting
         tm = TaskManager()
         self.run_tm(tm)
@@ -197,63 +192,49 @@ class TestJobLifeCycle:
 
 
 @pytest.mark.django_db
-def test_single_jt_multi_job_launch_blocks_last(controlplane_instance_group, job_template_factory, mocker):
+def test_single_jt_multi_job_launch_blocks_last(job_template_factory):
-    instance = controlplane_instance_group.instances.all()[0]
+    objects = job_template_factory('jt', organization='org1', project='proj', inventory='inv', credential='cred')
-    objects = job_template_factory(
+    j1 = create_job(objects.job_template)
-        'jt', organization='org1', project='proj', inventory='inv', credential='cred', jobs=["job_should_start", "job_should_not_start"]
+    j2 = create_job(objects.job_template)
-    )
+
-    j1 = objects.jobs["job_should_start"]
+    TaskManager().schedule()
-    j1.status = 'pending'
+    j1.refresh_from_db()
+    j2.refresh_from_db()
+    assert j1.status == "waiting"
+    assert j2.status == "pending"
+
+    # mimic running j1 to unblock j2
+    j1.status = "successful"
     j1.save()
-    j2 = objects.jobs["job_should_not_start"]
+    TaskManager().schedule()
-    j2.status = 'pending'
+
-    j2.save()
+    j2.refresh_from_db()
-    with mock.patch("awx.main.scheduler.TaskManager.start_task"):
+    assert j2.status == "waiting"
-        TaskManager().schedule()
-        TaskManager.start_task.assert_called_once_with(j1, controlplane_instance_group, [], instance)
-        j1.status = "successful"
-        j1.save()
-    with mocker.patch("awx.main.scheduler.TaskManager.start_task"):
-        TaskManager().schedule()
-        TaskManager.start_task.assert_called_once_with(j2, controlplane_instance_group, [], instance)
 
 
 @pytest.mark.django_db
-def test_single_jt_multi_job_launch_allow_simul_allowed(controlplane_instance_group, job_template_factory, mocker):
+def test_single_jt_multi_job_launch_allow_simul_allowed(job_template_factory):
-    instance = controlplane_instance_group.instances.all()[0]
+    objects = job_template_factory('jt', organization='org1', project='proj', inventory='inv', credential='cred')
-    objects = job_template_factory(
-        'jt', organization='org1', project='proj', inventory='inv', credential='cred', jobs=["job_should_start", "job_should_not_start"]
-    )
     jt = objects.job_template
+    jt.allow_simultaneous = True
     jt.save()
-
+    j1 = create_job(objects.job_template)
-    j1 = objects.jobs["job_should_start"]
+    j2 = create_job(objects.job_template)
-    j1.allow_simultaneous = True
+    TaskManager().schedule()
-    j1.status = 'pending'
+    j1.refresh_from_db()
-    j1.save()
+    j2.refresh_from_db()
-    j2 = objects.jobs["job_should_not_start"]
+    assert j1.status == "waiting"
-    j2.allow_simultaneous = True
+    assert j2.status == "waiting"
-    j2.status = 'pending'
-    j2.save()
-    with mock.patch("awx.main.scheduler.TaskManager.start_task"):
-        TaskManager().schedule()
-        TaskManager.start_task.assert_has_calls(
-            [mock.call(j1, controlplane_instance_group, [], instance), mock.call(j2, controlplane_instance_group, [], instance)]
-        )
 
 
 @pytest.mark.django_db
 def test_multi_jt_capacity_blocking(hybrid_instance, job_template_factory, mocker):
     instance = hybrid_instance
     controlplane_instance_group = instance.rampart_groups.first()
-    objects1 = job_template_factory('jt1', organization='org1', project='proj1', inventory='inv1', credential='cred1', jobs=["job_should_start"])
+    objects1 = job_template_factory('jt1', organization='org1', project='proj1', inventory='inv1', credential='cred1')
-    objects2 = job_template_factory('jt2', organization='org2', project='proj2', inventory='inv2', credential='cred2', jobs=["job_should_not_start"])
+    objects2 = job_template_factory('jt2', organization='org2', project='proj2', inventory='inv2', credential='cred2')
-    j1 = objects1.jobs["job_should_start"]
+    j1 = create_job(objects1.job_template)
-    j1.status = 'pending'
+    j2 = create_job(objects2.job_template)
-    j1.save()
-    j2 = objects2.jobs["job_should_not_start"]
-    j2.status = 'pending'
-    j2.save()
     tm = TaskManager()
     with mock.patch('awx.main.models.Job.task_impact', new_callable=mock.PropertyMock) as mock_task_impact:
         mock_task_impact.return_value = 505
@@ -269,11 +250,9 @@ def test_multi_jt_capacity_blocking(hybrid_instance, job_template_factory, mocke
 
 @pytest.mark.django_db
 def test_single_job_dependencies_project_launch(controlplane_instance_group, job_template_factory, mocker):
-    objects = job_template_factory('jt', organization='org1', project='proj', inventory='inv', credential='cred', jobs=["job_should_start"])
+    objects = job_template_factory('jt', organization='org1', project='proj', inventory='inv', credential='cred')
     instance = controlplane_instance_group.instances.all()[0]
-    j = objects.jobs["job_should_start"]
+    j = create_job(objects.job_template, dependencies_processed=False)
-    j.status = 'pending'
-    j.save()
     p = objects.project
     p.scm_update_on_launch = True
     p.scm_update_cache_timeout = 0
@@ -281,12 +260,13 @@ def test_single_job_dependencies_project_launch(controlplane_instance_group, job
     p.scm_url = "http://github.com/ansible/ansible.git"
     p.save(skip_update=True)
     with mock.patch("awx.main.scheduler.TaskManager.start_task"):
-        tm = TaskManager()
+        dm = DependencyManager()
-        with mock.patch.object(TaskManager, "create_project_update", wraps=tm.create_project_update) as mock_pu:
+        with mock.patch.object(DependencyManager, "create_project_update", wraps=dm.create_project_update) as mock_pu:
-            tm.schedule()
+            dm.schedule()
             mock_pu.assert_called_once_with(j)
             pu = [x for x in p.project_updates.all()]
             assert len(pu) == 1
+            TaskManager().schedule()
             TaskManager.start_task.assert_called_once_with(pu[0], controlplane_instance_group, [j], instance)
             pu[0].status = "successful"
             pu[0].save()
@@ -297,11 +277,9 @@ def test_single_job_dependencies_project_launch(controlplane_instance_group, job
 
 @pytest.mark.django_db
 def test_single_job_dependencies_inventory_update_launch(controlplane_instance_group, job_template_factory, mocker, inventory_source_factory):
-    objects = job_template_factory('jt', organization='org1', project='proj', inventory='inv', credential='cred', jobs=["job_should_start"])
+    objects = job_template_factory('jt', organization='org1', project='proj', inventory='inv', credential='cred')
     instance = controlplane_instance_group.instances.all()[0]
-    j = objects.jobs["job_should_start"]
+    j = create_job(objects.job_template, dependencies_processed=False)
-    j.status = 'pending'
-    j.save()
     i = objects.inventory
     ii = inventory_source_factory("ec2")
     ii.source = "ec2"
@@ -310,12 +288,13 @@ def test_single_job_dependencies_inventory_update_launch(controlplane_instance_g
     ii.save()
     i.inventory_sources.add(ii)
     with mock.patch("awx.main.scheduler.TaskManager.start_task"):
-        tm = TaskManager()
+        dm = DependencyManager()
-        with mock.patch.object(TaskManager, "create_inventory_update", wraps=tm.create_inventory_update) as mock_iu:
+        with mock.patch.object(DependencyManager, "create_inventory_update", wraps=dm.create_inventory_update) as mock_iu:
-            tm.schedule()
+            dm.schedule()
             mock_iu.assert_called_once_with(j, ii)
             iu = [x for x in ii.inventory_updates.all()]
             assert len(iu) == 1
+            TaskManager().schedule()
             TaskManager.start_task.assert_called_once_with(iu[0], controlplane_instance_group, [j], instance)
             iu[0].status = "successful"
             iu[0].save()
@@ -334,19 +313,17 @@ def test_inventory_update_launches_project_update(controlplane_instance_group, s
     iu.status = "pending"
     iu.save()
     with mock.patch("awx.main.scheduler.TaskManager.start_task"):
-        tm = TaskManager()
+        dm = DependencyManager()
-        with mock.patch.object(TaskManager, "create_project_update", wraps=tm.create_project_update) as mock_pu:
+        with mock.patch.object(DependencyManager, "create_project_update", wraps=dm.create_project_update) as mock_pu:
-            tm.schedule()
+            dm.schedule()
             mock_pu.assert_called_with(iu, project_id=project.id)
 
 
 @pytest.mark.django_db
 def test_job_dependency_with_already_updated(controlplane_instance_group, job_template_factory, mocker, inventory_source_factory):
-    objects = job_template_factory('jt', organization='org1', project='proj', inventory='inv', credential='cred', jobs=["job_should_start"])
+    objects = job_template_factory('jt', organization='org1', project='proj', inventory='inv', credential='cred')
     instance = controlplane_instance_group.instances.all()[0]
-    j = objects.jobs["job_should_start"]
+    j = create_job(objects.job_template, dependencies_processed=False)
-    j.status = 'pending'
-    j.save()
     i = objects.inventory
     ii = inventory_source_factory("ec2")
     ii.source = "ec2"
@@ -359,9 +336,9 @@ def test_job_dependency_with_already_updated(controlplane_instance_group, job_te
     j.start_args = encrypt_field(j, field_name="start_args")
     j.save()
     with mock.patch("awx.main.scheduler.TaskManager.start_task"):
-        tm = TaskManager()
+        dm = DependencyManager()
-        with mock.patch.object(TaskManager, "create_inventory_update", wraps=tm.create_inventory_update) as mock_iu:
+        with mock.patch.object(DependencyManager, "create_inventory_update", wraps=dm.create_inventory_update) as mock_iu:
-            tm.schedule()
+            dm.schedule()
             mock_iu.assert_not_called()
     with mock.patch("awx.main.scheduler.TaskManager.start_task"):
         TaskManager().schedule()
@@ -371,13 +348,11 @@ def test_job_dependency_with_already_updated(controlplane_instance_group, job_te
 @pytest.mark.django_db
 def test_shared_dependencies_launch(controlplane_instance_group, job_template_factory, mocker, inventory_source_factory):
     instance = controlplane_instance_group.instances.all()[0]
-    objects = job_template_factory('jt', organization='org1', project='proj', inventory='inv', credential='cred', jobs=["first_job", "second_job"])
+    objects = job_template_factory('jt', organization='org1', project='proj', inventory='inv', credential='cred')
-    j1 = objects.jobs["first_job"]
+    objects.job_template.allow_simultaneous = True
-    j1.status = 'pending'
+    objects.job_template.save()
-    j1.save()
+    j1 = create_job(objects.job_template, dependencies_processed=False)
-    j2 = objects.jobs["second_job"]
+    j2 = create_job(objects.job_template, dependencies_processed=False)
-    j2.status = 'pending'
-    j2.save()
     p = objects.project
     p.scm_update_on_launch = True
     p.scm_update_cache_timeout = 300
@@ -392,8 +367,8 @@ def test_shared_dependencies_launch(controlplane_instance_group, job_template_fa
     ii.update_cache_timeout = 300
     ii.save()
     i.inventory_sources.add(ii)
-
     with mock.patch("awx.main.scheduler.TaskManager.start_task"):
+        DependencyManager().schedule()
         TaskManager().schedule()
         pu = p.project_updates.first()
         iu = ii.inventory_updates.first()
@@ -408,12 +383,9 @@ def test_shared_dependencies_launch(controlplane_instance_group, job_template_fa
         iu.save()
     with mock.patch("awx.main.scheduler.TaskManager.start_task"):
         TaskManager().schedule()
-        TaskManager.start_task.assert_called_once_with(j1, controlplane_instance_group, [], instance)
+        TaskManager.start_task.assert_has_calls(
-        j1.status = "successful"
+            [mock.call(j1, controlplane_instance_group, [], instance), mock.call(j2, controlplane_instance_group, [], instance)]
-        j1.save()
+        )
-    with mock.patch("awx.main.scheduler.TaskManager.start_task"):
-        TaskManager().schedule()
-        TaskManager.start_task.assert_called_once_with(j2, controlplane_instance_group, [], instance)
     pu = [x for x in p.project_updates.all()]
     iu = [x for x in ii.inventory_updates.all()]
     assert len(pu) == 1
@@ -422,30 +394,27 @@ def test_shared_dependencies_launch(controlplane_instance_group, job_template_fa
 
 @pytest.mark.django_db
 def test_job_not_blocking_project_update(controlplane_instance_group, job_template_factory):
-    objects = job_template_factory('jt', organization='org1', project='proj', inventory='inv', credential='cred', jobs=["job"])
+    instance = controlplane_instance_group.instances.all()[0]
-    job = objects.jobs["job"]
+    objects = job_template_factory('jt', organization='org1', project='proj', inventory='inv', credential='cred')
+    job = objects.job_template.create_unified_job()
     job.instance_group = controlplane_instance_group
+    job.dependencies_process = True
     job.status = "running"
     job.save()
 
     with mock.patch("awx.main.scheduler.TaskManager.start_task"):
-        task_manager = TaskManager()
-        task_manager._schedule()
-
         proj = objects.project
         project_update = proj.create_project_update()
         project_update.instance_group = controlplane_instance_group
         project_update.status = "pending"
         project_update.save()
-        assert not task_manager.job_blocked_by(project_update)
+        TaskManager().schedule()
-
+        TaskManager.start_task.assert_called_once_with(project_update, controlplane_instance_group, [], instance)
-        dependency_graph = DependencyGraph()
-        dependency_graph.add_job(job)
-        assert not dependency_graph.task_blocked_by(project_update)
 
 
 @pytest.mark.django_db
 def test_job_not_blocking_inventory_update(controlplane_instance_group, job_template_factory, inventory_source_factory):
+    instance = controlplane_instance_group.instances.all()[0]
     objects = job_template_factory('jt', organization='org1', project='proj', inventory='inv', credential='cred', jobs=["job"])
     job = objects.jobs["job"]
     job.instance_group = controlplane_instance_group
@@ -453,9 +422,6 @@ def test_job_not_blocking_inventory_update(controlplane_instance_group, job_temp
     job.save()
 
     with mock.patch("awx.main.scheduler.TaskManager.start_task"):
-        task_manager = TaskManager()
-        task_manager._schedule()
-
         inv = objects.inventory
         inv_source = inventory_source_factory("ec2")
         inv_source.source = "ec2"
@@ -465,11 +431,9 @@ def test_job_not_blocking_inventory_update(controlplane_instance_group, job_temp
         inventory_update.status = "pending"
         inventory_update.save()
 
-        assert not task_manager.job_blocked_by(inventory_update)
+        DependencyManager().schedule()
-
+        TaskManager().schedule()
-        dependency_graph = DependencyGraph()
+        TaskManager.start_task.assert_called_once_with(inventory_update, controlplane_instance_group, [], instance)
-        dependency_graph.add_job(job)
-        assert not dependency_graph.task_blocked_by(inventory_update)
 
 
 @pytest.mark.django_db
@@ -484,7 +448,7 @@ def test_generate_dependencies_only_once(job_template_factory):
         # job starts with dependencies_processed as False
         assert not job.dependencies_processed
         # run one cycle of ._schedule() to generate dependencies
-        TaskManager()._schedule()
+        DependencyManager().schedule()
 
         # make sure dependencies_processed is now True
         job = Job.objects.filter(name="job_gen_dep")[0]
@@ -492,7 +456,7 @@ def test_generate_dependencies_only_once(job_template_factory):
 
         # Run ._schedule() again, but make sure .generate_dependencies() is not
         # called with job in the argument list
-        tm = TaskManager()
+        dm = DependencyManager()
-        tm.generate_dependencies = mock.MagicMock(return_value=[])
+        dm.generate_dependencies = mock.MagicMock(return_value=[])
-        tm._schedule()
+        dm.schedule()
-        tm.generate_dependencies.assert_has_calls([mock.call([]), mock.call([])])
+        dm.generate_dependencies.assert_not_called()

awx/main/utils/common.py

@@ -78,8 +78,9 @@ __all__ = [
     'IllegalArgumentError',
     'get_custom_venv_choices',
     'get_external_account',
-    'task_manager_bulk_reschedule',
+    'ScheduleTaskManager',
-    'schedule_task_manager',
+    'ScheduleDependencyManager',
+    'ScheduleWorkflowManager',
     'classproperty',
     'create_temporary_fifo',
     'truncate_stdout',
@@ -846,6 +847,66 @@ def get_mem_effective_capacity(mem_bytes):
 
 _inventory_updates = threading.local()
 _task_manager = threading.local()
+_dependency_manager = threading.local()
+_workflow_manager = threading.local()
+
+
+@contextlib.contextmanager
+def task_manager_bulk_reschedule():
+    managers = [ScheduleTaskManager(), ScheduleWorkflowManager(), ScheduleDependencyManager()]
+    """Context manager to avoid submitting task multiple times."""
+    try:
+        for m in managers:
+            m.previous_flag = getattr(m.manager_threading_local, 'bulk_reschedule', False)
+            m.previous_value = getattr(m.manager_threading_local, 'needs_scheduling', False)
+            m.manager_threading_local.bulk_reschedule = True
+            m.manager_threading_local.needs_scheduling = False
+        yield
+    finally:
+        for m in managers:
+            m.manager_threading_local.bulk_reschedule = m.previous_flag
+            if m.manager_threading_local.needs_scheduling:
+                m.schedule()
+            m.manager_threading_local.needs_scheduling = m.previous_value
+
+
+class ScheduleManager:
+    def __init__(self, manager, manager_threading_local):
+        self.manager = manager
+        self.manager_threading_local = manager_threading_local
+
+    def _schedule(self):
+        from django.db import connection
+
+        # runs right away if not in transaction
+        connection.on_commit(lambda: self.manager.delay())
+
+    def schedule(self):
+        if getattr(self.manager_threading_local, 'bulk_reschedule', False):
+            self.manager_threading_local.needs_scheduling = True
+            return
+        self._schedule()
+
+
+class ScheduleTaskManager(ScheduleManager):
+    def __init__(self):
+        from awx.main.scheduler.tasks import task_manager
+
+        super().__init__(task_manager, _task_manager)
+
+
+class ScheduleDependencyManager(ScheduleManager):
+    def __init__(self):
+        from awx.main.scheduler.tasks import dependency_manager
+
+        super().__init__(dependency_manager, _dependency_manager)
+
+
+class ScheduleWorkflowManager(ScheduleManager):
+    def __init__(self):
+        from awx.main.scheduler.tasks import workflow_manager
+
+        super().__init__(workflow_manager, _workflow_manager)
 
 
 @contextlib.contextmanager
@@ -861,37 +922,6 @@ def ignore_inventory_computed_fields():
         _inventory_updates.is_updating = previous_value
 
 
-def _schedule_task_manager():
-    from awx.main.scheduler.tasks import run_task_manager
-    from django.db import connection
-
-    # runs right away if not in transaction
-    connection.on_commit(lambda: run_task_manager.delay())
-
-
-@contextlib.contextmanager
-def task_manager_bulk_reschedule():
-    """Context manager to avoid submitting task multiple times."""
-    try:
-        previous_flag = getattr(_task_manager, 'bulk_reschedule', False)
-        previous_value = getattr(_task_manager, 'needs_scheduling', False)
-        _task_manager.bulk_reschedule = True
-        _task_manager.needs_scheduling = False
-        yield
-    finally:
-        _task_manager.bulk_reschedule = previous_flag
-        if _task_manager.needs_scheduling:
-            _schedule_task_manager()
-        _task_manager.needs_scheduling = previous_value
-
-
-def schedule_task_manager():
-    if getattr(_task_manager, 'bulk_reschedule', False):
-        _task_manager.needs_scheduling = True
-        return
-    _schedule_task_manager()
-
-
 @contextlib.contextmanager
 def ignore_inventory_group_removal():
     """

awx/settings/defaults.py

@@ -248,6 +248,11 @@ SUBSYSTEM_METRICS_TASK_MANAGER_RECORD_INTERVAL = 15
 # The maximum allowed jobs to start on a given task manager cycle
 START_TASK_LIMIT = 100
 
+# Time out task managers if they take longer than this many seconds, plus TASK_MANAGER_TIMEOUT_GRACE_PERIOD
+# We have the grace period so the task manager can bail out before the timeout.
+TASK_MANAGER_TIMEOUT = 300
+TASK_MANAGER_TIMEOUT_GRACE_PERIOD = 60
+
 # Disallow sending session cookies over insecure connections
 SESSION_COOKIE_SECURE = True
 
@@ -442,7 +447,8 @@ CELERYBEAT_SCHEDULE = {
         'options': {'expires': 50},
     },
     'gather_analytics': {'task': 'awx.main.tasks.system.gather_analytics', 'schedule': timedelta(minutes=5)},
-    'task_manager': {'task': 'awx.main.scheduler.tasks.run_task_manager', 'schedule': timedelta(seconds=20), 'options': {'expires': 20}},
+    'task_manager': {'task': 'awx.main.scheduler.tasks.task_manager', 'schedule': timedelta(seconds=20), 'options': {'expires': 20}},
+    'dependency_manager': {'task': 'awx.main.scheduler.tasks.dependency_manager', 'schedule': timedelta(seconds=20), 'options': {'expires': 20}},
     'k8s_reaper': {'task': 'awx.main.tasks.system.awx_k8s_reaper', 'schedule': timedelta(seconds=60), 'options': {'expires': 50}},
     '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)},

awx/settings/development.py

@@ -78,18 +78,6 @@ include(optional('/etc/tower/conf.d/*.py'), scope=locals())
 BASE_VENV_PATH = "/var/lib/awx/venv/"
 AWX_VENV_PATH = os.path.join(BASE_VENV_PATH, "awx")
 
-# If any local_*.py files are present in awx/settings/, use them to override
-# default settings for development.  If not present, we can still run using
-# only the defaults.
-try:
-    if os.getenv('AWX_KUBE_DEVEL', False):
-        include(optional('minikube.py'), scope=locals())
-    else:
-        include(optional('local_*.py'), scope=locals())
-except ImportError:
-    traceback.print_exc()
-    sys.exit(1)
-
 # Use SQLite for unit tests instead of PostgreSQL.  If the lines below are
 # commented out, Django will create the test_awx-dev database in PostgreSQL to
 # run unit tests.
@@ -110,5 +98,26 @@ CLUSTER_HOST_ID = socket.gethostname()
 
 AWX_CALLBACK_PROFILE = True
 
+# ======================!!!!!!! FOR DEVELOPMENT ONLY !!!!!!!=================================
+# Disable normal scheduled/triggered task managers (DependencyManager, TaskManager, WorkflowManager).
+# Allows user to trigger task managers directly for debugging and profiling purposes.
+# Only works in combination with settings.SETTINGS_MODULE == 'awx.settings.development'
+AWX_DISABLE_TASK_MANAGERS = False
+# ======================!!!!!!! FOR DEVELOPMENT ONLY !!!!!!!=================================
+
 if 'sqlite3' not in DATABASES['default']['ENGINE']:  # noqa
     DATABASES['default'].setdefault('OPTIONS', dict()).setdefault('application_name', f'{CLUSTER_HOST_ID}-{os.getpid()}-{" ".join(sys.argv)}'[:63])  # noqa
+
+
+# If any local_*.py files are present in awx/settings/, use them to override
+# default settings for development.  If not present, we can still run using
+# only the defaults.
+# this needs to stay at the bottom of this file
+try:
+    if os.getenv('AWX_KUBE_DEVEL', False):
+        include(optional('minikube.py'), scope=locals())
+    else:
+        include(optional('local_*.py'), scope=locals())
+except ImportError:
+    traceback.print_exc()
+    sys.exit(1)