Quick Start

This guide walks you through the deployment process for the MCPCAN platform. We offer two deployment paths to suit different environments:

• One-Click Deployment Script: Ideal for new, clean server environments. The script automatically installs all dependencies and completes the deployment. Recommended for new users and test environments.
• Standard Helm Deployment: Suitable if you already have a running Kubernetes cluster.

---

1. Prerequisites

Before you begin, ensure your environment meets the following minimum requirements. If you use the one-click deployment script, it will automatically check and install these dependencies.

• System Requirements:
  • Kubernetes: v1.20 or higher.
  • Helm: v3.0 or higher.
  • NGINX Ingress Controller: Required for accessing the platform via a domain name.
  • Persistent Storage: For persisting platform data.
  • Hardware Resources: At least 4GB of RAM and a 2-core CPU.

---

2. Clone the Deployment Repository

First, clone the mcpcan-deploy repository, which contains all deployment scripts and configuration files.

# Clone the repository
git clone https://github.com/kymo-mcp/mcpcan-deploy.git

# Navigate to the repository directory
cd mcpcan-deploy

3. Install Runtime Dependencies

This method uses a script to automatically install K3s (a lightweight Kubernetes distribution), Helm, an Ingress controller, and finally deploys the MCPCAN platform in a new environment. If you already have K3s, Helm, and an Ingress controller installed, you can skip this step.

Execute the Environment Installation Script

The install-run-environment.sh script in the repository is the core of the deployment, handling all environment setup and application installation.

# Execute the installation script
./scripts/install-run-environment.sh

Tip: If you are running this script in mainland China, we recommend adding the --cn flag. This will use a domestic mirror source to speed up the download of dependencies.

./scripts/install-run-environment.sh --cn

After the script runs successfully, the basic environment is deployed.

4. Standard Helm Deployment

We deploy using Helm. For easier management and upgrades, the best practice is to separate custom configurations from the default ones.

1. Copy the Configuration File: Copy the default values.yaml file to values-custom.yaml. All our modifications will be made in this new file.

   cp helm/values.yaml helm/values-custom.yaml

2. Modify Custom Configuration: Open helm/values-custom.yaml with your preferred editor. The most important configurations are the domain and TLS certificate.

   # helm/values-custom.yaml
   
   # Global configuration
   global:
     # Whether to use a domestic mirror source, default is false
     cn: false
     # Set your domain here, e.g., demo.mcpcan.com
     domain: "demo.mcpcan.com"
   
   # Ingress configuration
   ingress:
     tls:
       # Enable TLS
       enabled: true
       # Configure certificate content (for self-signed or existing certificates)
       # -----BEGIN CERTIFICATE-----
       #  Your certificate content
       # -----END CERTIFICATE-----
       crt: |
         
       # -----BEGIN PRIVATE KEY-----
       #  Your private key content
       # -----END PRIVATE KEY-----
       key: |

Step 3: Deploy with Helm

After customizing the configuration, execute the helm install command to deploy MCPCAN.

helm install mcpcan ./helm -f helm/values-custom.yaml \
  --namespace mcpcan --create-namespace --timeout 600s --wait

• Command Breakdown:
  • helm install mcpcan ./helm: Installs a release named mcpcan from the chart at ./helm.
  • -f helm/values-custom.yaml: Uses our custom configuration file to override default values.
  • --namespace mcpcan: Specifies that the installation should occur in the mcpcan namespace.
  • --create-namespace: Automatically creates the mcpcan namespace if it does not exist.
  • --timeout 600s: Sets the installation timeout to 600 seconds.
  • --wait: Waits for all resources to be successfully deployed and in a Ready state before returning.

---

5. Verify the Deployment

After the deployment is complete, you can check if all Pods are running correctly with the following command:

kubectl get pods -n mcpcan

If the STATUS of all Pods is Running, the platform has started successfully.

You can now access the MCPCAN login page by visiting the domain you configured (e.g., http://demo.mcpcan.com) in your browser.

6. Uninstall the Platform

If you need to uninstall the MCPCAN platform, simply execute the following Helm command:

helm uninstall mcpcan -n mcpcan