diff --git a/sky/backends/cloud_vm_ray_backend.py b/sky/backends/cloud_vm_ray_backend.py
index a0f746a7098..6d2447fe89b 100644
--- a/sky/backends/cloud_vm_ray_backend.py
+++ b/sky/backends/cloud_vm_ray_backend.py
@@ -2011,10 +2011,10 @@ def provision_with_retries(
                     cloud_user = to_provision.cloud.get_current_user_identity()
 
                 requested_features = self._requested_features.copy()
-                # Skip stop feature for Kubernetes jobs controller.
+                # Skip stop feature for Kubernetes controllers.
                 if (isinstance(to_provision.cloud, clouds.Kubernetes) and
                         controller_utils.Controllers.from_name(cluster_name)
-                        == controller_utils.Controllers.JOBS_CONTROLLER):
+                        is not None):
                     assert (clouds.CloudImplementationFeatures.STOP
                             in requested_features), requested_features
                     requested_features.remove(
@@ -4152,11 +4152,10 @@ def set_autostop(self,
             # Skip auto-stop for Kubernetes clusters.
             if (isinstance(handle.launched_resources.cloud, clouds.Kubernetes)
                     and not down and idle_minutes_to_autostop >= 0):
-                # We should hit this code path only for the jobs controller on
+                # We should hit this code path only for the controllers on
                 # Kubernetes clusters.
                 assert (controller_utils.Controllers.from_name(
-                    handle.cluster_name) == controller_utils.Controllers.
-                        JOBS_CONTROLLER), handle.cluster_name
+                    handle.cluster_name) is not None), handle.cluster_name
                 logger.info('Auto-stop is not supported for Kubernetes '
                             'clusters. Skipping.')
                 return
diff --git a/sky/serve/core.py b/sky/serve/core.py
index f193a85285b..0252437d0e2 100644
--- a/sky/serve/core.py
+++ b/sky/serve/core.py
@@ -186,14 +186,13 @@ def up(
         # whether the service is already running. If the id is the same
         # with the current job id, we know the service is up and running
         # for the first time; otherwise it is a name conflict.
-        idle_minutes_to_autodown = constants.CONTROLLER_IDLE_MINUTES_TO_AUTOSTOP
+        idle_minutes_to_autostop = constants.CONTROLLER_IDLE_MINUTES_TO_AUTOSTOP
         controller_job_id, controller_handle = sky.launch(
             task=controller_task,
             stream_logs=False,
             cluster_name=controller_name,
             detach_run=True,
-            idle_minutes_to_autostop=idle_minutes_to_autodown,
-            down=True,
+            idle_minutes_to_autostop=idle_minutes_to_autostop,
             retry_until_up=True,
             _disable_controller_check=True,
         )