change when we send job notifications to avoid a race condition

success/failure notifications for *playbooks* include summary data about
the hosts in based on the contents of the playbook_on_stats event

the current implementation suffers from a number of race conditions that
sometimes can cause that data to be missing or incomplete; this change
makes it so that for *playbooks* we build (and send) the notification in
response to the playbook_on_stats event, not the EOF event

awx/main/tasks.py

@@ -602,6 +602,23 @@ def handle_work_error(task_id, *args, **kwargs):
         pass
 
 
+@task(queue=get_local_queuename)
+def handle_success_and_failure_notifications(job_id):
+    uj = UnifiedJob.objects.get(pk=job_id)
+    retries = 0
+    while retries < 5:
+        if uj.finished:
+            uj.send_notification_templates('succeeded' if uj.status == 'successful' else 'failed')
+            break
+        else:
+            # wait a few seconds to avoid a race where the
+            # events are persisted _before_ the UJ.status
+            # changes from running -> successful
+            retries += 1
+            time.sleep(1)
+            uj = UnifiedJob.objects.get(pk=job_id)
+
+
 @task(queue=get_local_queuename)
 def update_inventory_computed_fields(inventory_id):
     '''