[autoscaler][kubernetes] Example Cluster Config Returns KeyError #13667
Description
What is the problem?
Python version: 3.8
Kubernetes version: Client Version: v1.18.2 Server Version: v1.18.9-eks-d1db3c
ray-operator-pod returns an error when deploying an example ray cluster.
Traceback (most recent call last):
File "/home/ray/anaconda3/bin/ray-operator", line 8, in <module>
sys.exit(main())
File "/home/ray/anaconda3/lib/python3.7/site-packages/ray/operator/operator.py", line 123, in main
cluster_config = operator_utils.cr_to_config(cluster_cr)
File "/home/ray/anaconda3/lib/python3.7/site-packages/ray/operator/operator_utils.py", line 62, in cr_to_config
config["available_node_types" = get_node_types(cluster_resource)
File "/home/ray/anaconda3/lib/python3.7/site-packages/ray/operator/operator_utils.py", line 76, in get_node_types
pod_type_copy, dictionary=NODE_TYPE_FIELDS)
File "/home/ray/anaconda3/lib/python3.7/site-packages/ray/operator/operator_utils.py", line 98, in translate
return {dictionary[field: configuration[field for field in dictionary}
File "/home/ray/anaconda3/lib/python3.7/site-packages/ray/operator/operator_utils.py", line 98, in <dictcomp>
return {dictionary[field: configuration[field for field in dictionary}
KeyError: 'minWorkers'
stream closed
Reproduction (REQUIRED)
$ kubectl apply -f ray/python/ray/autoscaler/kubernetes/operator_configs/cluster_crd.yaml
$ kubectl create namespace ray
$ kubectl -n ray apply -f ray/python/ray/autoscaler/kubernetes/operator_configs/operator.yaml
$ kubectl -n ray apply -f ray/python/ray/autoscaler/kubernetes/operator_configs/example_cluster.yamlThis causes KeyError in the Ray operator.
If you add
minWorkers: 1
maxWorkers: 1
rayResources: {}
setupCommands: []these four specs to example_cluster.yaml, then you can successfully deploy the example ray cluster on k8s with k8s operator.
- I have verified my script runs in a clean environment and reproduces the issue.
- I have verified the issue also occurs with the latest wheels.