Cleanup after playing with EKS and kubeflow

cleanup.sh

#!/bin/bash

# delete the jupyter deployment
kubectl delete -f /tmp/deploy.yaml

# delete the FSx filesystems
kubectl delete -f /tmp/claim.yaml
kubectl delete -f /tmp/claim-kf.yaml
kubectl delete -f /tmp/storage-class.yaml
kubectl delete -k github.com/kubernetes-sigs/aws-fsx-csi-driver/deploy/kubernetes/overlays/stable/?ref=master

aws ec2 delete-security-group --group-id ${SECURITY_GROUP_ID}
aws iam detach-role-policy --role-name ${INSTANCE_ROLE_NAME} --policy-arn ${POLICY_ARN}
aws iam delete-policy --policy-arn $POLICY_ARN

# delete kubeflow
cd ${KF_DIR}
kfctl delete -f ${CONFIG_FILE} --delete_storage

# delete the cluster
# remember to delete all services with an external ip
# refer to: https://docs.aws.amazon.com/eks/latest/userguide/delete-cluster.html
cd ~/eks_getting_started
eksctl delete cluster -f ./configs/clusters/minimal.yaml