if dependency fails, fail job in task manager

awx/main/models/mixins.py

@@ -407,18 +407,11 @@ class TaskManagerUnifiedJobMixin(models.Model):
     def get_jobs_fail_chain(self):
         return []
 
-    def dependent_jobs_finished(self):
-        return True
-
 
 class TaskManagerJobMixin(TaskManagerUnifiedJobMixin):
     class Meta:
         abstract = True
 
-    def dependent_jobs_finished(self):
-        # if any dependent jobs are pending, waiting, or running, return False
-        return not any(j.status in ACTIVE_STATES for j in self.dependent_jobs.all())
-
 
 class TaskManagerUpdateOnLaunchMixin(TaskManagerUnifiedJobMixin):
     class Meta:
@@ -430,7 +423,17 @@ class TaskManagerProjectUpdateMixin(TaskManagerUpdateOnLaunchMixin):
         abstract = True
 
     def get_jobs_fail_chain(self):
-        return list(self.unifiedjob_blocked_jobs.all())
+        # project update can be a dependency of an inventory update, in which
+        # case we need to fail the job that may have spawned the inventory
+        # update.
+        # The inventory update will fail, but since it is not running it will
+        # not cascade fail to the job from the errback logic in apply_async. As
+        # such we should capture it here.
+        blocked_jobs = list(self.unifiedjob_blocked_jobs.all().prefetch_related("unifiedjob_blocked_jobs"))
+        other_tasks = []
+        for b in blocked_jobs:
+            other_tasks += list(b.unifiedjob_blocked_jobs.all())
+        return blocked_jobs + other_tasks
 
 
 class TaskManagerInventoryUpdateMixin(TaskManagerUpdateOnLaunchMixin):
@@ -452,10 +455,6 @@ class TaskManagerInventoryUpdateMixin(TaskManagerUpdateOnLaunchMixin):
                     other_updates.append(dep)
         return blocked_jobs + other_updates
 
-    def dependent_jobs_finished(self):
-        # if any dependent jobs are pending, waiting, or running, return False
-        return not any(j.status in ACTIVE_STATES for j in self.dependent_jobs.all())
-
 
 class ExecutionEnvironmentMixin(models.Model):
     class Meta:

awx/main/scheduler/task_manager.py

@@ -34,6 +34,7 @@ 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.common import create_partition
 from awx.main.signals import disable_activity_stream
+from awx.main.constants import ACTIVE_STATES
 from awx.main.scheduler.dependency_graph import DependencyGraph
 from awx.main.scheduler.task_manager_models import TaskManagerInstances
 from awx.main.scheduler.task_manager_models import TaskManagerInstanceGroups
@@ -79,10 +80,23 @@ class TaskManager:
         if blocked_by:
             return blocked_by
 
-        if not task.dependent_jobs_finished():
-            blocked_by = task.dependent_jobs.first()
-            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
 

awx/main/tasks/system.py

@@ -695,7 +695,7 @@ def handle_work_error(task_id, *args, **kwargs):
                 first_instance = instance
                 first_instance_type = each_task['type']
 
-            if instance.celery_task_id != task_id and not instance.cancel_flag and not instance.status == 'successful':
+            if instance.celery_task_id != task_id and not instance.cancel_flag and not instance.status in ('successful', 'failed'):
                 instance.status = 'failed'
                 instance.failed = True
                 if not instance.job_explanation: