Prepare the Operation Environment

This guide helps you set up an operation environment to install or upgrade CloudPilot AI.

Prerequisites

Cloud provider account credentials or an authenticated provider CLI session
Docker or Kubernetes access
Basic familiarity with command-line operations

Step 1: Launch the Container Environment

Option 1: Docker Environment

Launch a container with all required tools pre-installed:


docker run -it --restart=unless-stopped --name cloudpilotai \
  public.ecr.aws/cloudpilotai/installationenv:latest \
  bash

Option 2: Kubernetes Pod

For environments without Docker:


kubectl run helper --image=public.ecr.aws/cloudpilotai/installationenv:latest \
  -it --rm --restart=Never -- /bin/bash

Step 2: Configure Cloud Provider Credentials

AWS Configuration

Configure AWS credentials:


aws configure

Provide your credentials when prompted:


AWS Access Key ID: <your_access_key>
AWS Secret Access Key: <your_secret_key>
Default region name: us-east-2
Default output format: json

Update kubeconfig for EKS:


mkdir -p ~/.kube
export KUBECONFIG=~/.kube/config
aws eks update-kubeconfig --name <cluster_name> --region us-east-2

Verify setup:


aws sts get-caller-identity
kubectl get nodes

Alibaba Cloud Configuration

Configure Alibaba Cloud credentials:


aliyun configure

Follow the prompts:


Access Key ID: <your_access_key>
Access Key Secret: <your_secret_key>
Default Region ID: cn-shenzhen
Default Output Format: json
Default Language: en

Setup kubeconfig for ACK:


mkdir -p ~/.kube
export KUBECONFIG=~/.kube/config
export CLUSTER_NAME=<cluster_name>
export CLUSTER_ID=$(aliyun cs GET /clusters | jq -r --arg CLUSTER_NAME "$CLUSTER_NAME" '.[] | select(.name == $CLUSTER_NAME) | .cluster_id')
aliyun cs GET /k8s/$CLUSTER_ID/user_config | jq -r '.config' > $KUBECONFIG

Verify setup:


kubectl get nodes

GCP Configuration

Authenticate with Google Cloud and select the target project:


gcloud auth login
export GCP_PROJECT_ID=<your_project_id>
gcloud config set project "${GCP_PROJECT_ID}"

Prepare kubeconfig for GKE:


mkdir -p ~/.kube
export KUBECONFIG=~/.kube/config
export CLUSTER_NAME=<cluster_name>
export CLUSTER_LOCATION=<cluster_location> # e.g. us-central1-a or us-central1
 
gcloud container clusters get-credentials "${CLUSTER_NAME}" \
  --location "${CLUSTER_LOCATION}" \
  --project "${GCP_PROJECT_ID}"

Verify setup:


gcloud auth list
kubectl get nodes
helm version
jq --version

GKE note: enable Workload Identity on the cluster before you run the CloudPilot AI Phase 2 install or upgrade flows.

Alternative Setup Using AWS CloudShell

If you don’t have access to a terminal environment for AWS clusters, you can use AWS CloudShell to launch the operational environment. Execute the following command to set up your containerized workspace:


docker run -it --restart=unless-stopped --network host --name cloudpilotai \
  -v ~/.kube:/root/.kube/ \
  -e AWS_CONTAINER_CREDENTIALS_FULL_URI=http://localhost:1338/latest/meta-data/container/security-credentials \
  -e AWS_CONTAINER_AUTHORIZATION_TOKEN="$AWS_CONTAINER_AUTHORIZATION_TOKEN" \
  -e AWS_REGION="$AWS_REGION" \
  public.ecr.aws/cloudpilotai/installationenv:latest \
  bash

It will look like: aws_cloudshell

Verify setup:


kubectl get nodes

Next Steps

Your environment is now configured and ready for CloudPilot AI installation or upgrade operations. Continue with the Quickstart flow in the CloudPilot AI console.