diff --git a/doc/source/cluster/k8s-operator.rst b/doc/source/cluster/k8s-operator.rst
index d17443e55..f993e4f26 100644
--- a/doc/source/cluster/k8s-operator.rst
+++ b/doc/source/cluster/k8s-operator.rst
@@ -19,6 +19,10 @@ The rest of this document explains step-by-step how to use the Ray Kubernetes Op
 .. role:: bash(code)
    :language: bash
 
+.. warning::
+   The Ray Kubernetes Operator requires Kubernetes version at least ``v1.17.0``. Check Kubernetes version info with the command
+   :bash:`kubectl version`.
+
 .. note::
    The example commands in this document launch six Kubernetes pods, using a total of 6 CPU and 3.5Gi memory.   
    If you are experimenting using a test Kubernetes environment such as `minikube`_, make sure to provision sufficient resources, e.g.
diff --git a/python/ray/operator/operator.py b/python/ray/operator/operator.py
index 9771810c1..cf83eaa24 100644
--- a/python/ray/operator/operator.py
+++ b/python/ray/operator/operator.py
@@ -1,23 +1,3 @@
-"""
-Ray operator for Kubernetes.
-
-Reads ray cluster config from a k8s ConfigMap, starts a ray head node pod using
-create_or_update_cluster(), then runs an autoscaling loop in the operator pod
-executing this script. Writes autoscaling logs to the directory
-/root/ray-operator-logs.
-
-In this setup, the ray head node does not run an autoscaler. It is important
-NOT to supply an --autoscaling-config argument to head node's ray start command
-in the cluster config when using this operator.
-
-To run, first create a ConfigMap named ray-operator-configmap from a ray
-cluster config. Then apply the manifest at python/ray/autoscaler/kubernetes/operator_configs/operator_config.yaml
-
-For example:
-kubectl create namespace raytest
-kubectl -n raytest create configmap ray-operator-configmap --from-file=python/ray/autoscaler/kubernetes/operator_configs/test_cluster_config.yaml
-kubectl -n raytest apply -f python/ray/autoscaler/kubernetes/operator_configs/operator_config.yaml
-""" # noqa
 import logging
 import multiprocessing as mp
 import os