Fix continuous audit fails once it fails

Currently continuous audit fails once it fails
because continous audit tries to remove job
even if job is not exists.

This patch set fixes it.

Change-Id: Ic461408c97d71e14c57e368f8436b26fe355fa4e
Closes-Bug: #1706857

watcher/tests/decision_engine/audit/test_audit_handlers.py

@@ -361,3 +361,26 @@ class TestContinuousAuditHandler(base.DbTestCase):
         self.assertTrue(is_inactive)
         is_inactive = audit_handler._is_audit_inactive(self.audits[0])
         self.assertTrue(is_inactive)
+
+    @mock.patch.object(objects.service.Service, 'list')
+    @mock.patch.object(sq_api, 'get_engine')
+    @mock.patch.object(scheduling.BackgroundSchedulerService, 'get_jobs')
+    @mock.patch.object(objects.audit.AuditStateTransitionManager,
+                       'is_inactive')
+    @mock.patch.object(continuous.ContinuousAuditHandler, 'execute')
+    def test_execute_audit_with_interval_no_job(
+            self,
+            m_execute,
+            m_is_inactive,
+            m_get_jobs,
+            m_get_engine,
+            m_service):
+        audit_handler = continuous.ContinuousAuditHandler()
+        self.audits[0].next_run_time = (datetime.datetime.now() -
+                                        datetime.timedelta(seconds=1800))
+        m_is_inactive.return_value = True
+        m_get_jobs.return_value = None
+
+        audit_handler.execute_audit(self.audits[0], self.context)
+        m_execute.assert_called_once_with(self.audits[0], self.context)
+        self.assertIsNotNone(self.audits[0].next_run_time)