No Docker image building required, no container pipeline configuration needed. Using a declarative template, you can deliver a complete and controllable OpenClaw runtime environment in just 5 minutes, truly starting your work from a "service-ready" state.
1. Why Choose This VM Automation Deployment Solution?
✅ Extremely Low Barrier to Entry, Accessible Even with Zero Container Knowledge
Avoid complex processes like Docker image building, ACR pushing, and WebSocket adaptation. Simply fill in a few basic declarative parameters to obtain a fully functional OpenClaw instance with one click. This compresses hours of environment setup work into 5 minutes, significantly lowering the barrier to entry.
✅ Full System Control, Complete Freedom for Troubleshooting and Development
The solution natively provides SSH login capabilities. You can perform granular troubleshooting, full-link log analysis, kernel-level configuration management, and in-depth secondary development directly at the system level. Say goodbye to the configuration limitations and troubleshooting blind spots of container environments, with all environmental details under your control.
✅ Full Scenario Adaptation, Meeting Diverse Delivery Needs
-
Customer Demos: Quickly set up independent, isolated demo environments for different clients, fully showcasing the OpenClaw product without the functional limitations of multi-tenant environments.
-
Internal Enterprise Provisioning: IT/development teams can use this solution to deliver standardized cloud runtime environments for internal employees uniformly. This avoids the risks of environmental fragmentation, management blind spots, and data leakage from local installations, while enabling secure and compliant "on-demand usage."
2. Core Technical Capabilities
Using automation templates, Azure converges the originally scattered processes of host creation, environment initialization, dependency installation, and security configuration into a set of standardized atomic operations. Three core capabilities underpin rapid delivery:
-
Complete Runtime Foundation: Pre-installed Ubuntu system, built-in network ingress, persistent storage, and all runtime dependencies for OpenClaw.
-
Automated Template Deployment: Full-process Infrastructure as Code (IaC), encapsulating resource creation, environment initialization, and HTTPS certificate configuration.
-
Security Management Boundary: Employs a dual-layer access logic of "SSH Key Verification + Dynamic Token," balancing delivery efficiency and entry security.
3. Pre-Deployment Preparation
(I) Mandatory Prerequisites
ssh-keygen -t ed25519 -C "openclaw-azure"
Get-Content $env:USERPROFILE\.ssh\id_ed25519.pub
macOS/Linux用户生成密钥
ssh-keygen -t ed25519 -C "openclaw-azure"
cat ~/.ssh/id_ed25519.pub
(II) Optional Prerequisites
If you need to connect to Azure OpenAI immediately after deployment, prepare in advance:
4. 5-Step Automated Deployment Process
01. Access the Deployment Entry
Click the "Deploy to Azure" button and log in to your Azure account.
02. Configure Resource Group
Name the resource group according to environment conventions (e.g., rg-openclaw-poc) for easier cost control and lifecycle management.
03. Fill in Deployment Parameters
Customize vmName, adminUsername(default: azureuser), sshPublicKey, vmSize(e.g., Standard_B2as_v2). Fill in Azure OpenAI-related parameters as needed.
04. Submit Deployment
Click "Review + create" and wait for the resources and the openclaw-bootstrapextension deployment to complete.
05. Save Output Information
On the deployment page, under the "Outputs" section, save the vmPublicFqdn(VM public fully qualified domain name).
5. Post-Deployment Access Process
SSH into the Virtual Machine
- Windows (PowerShell)
ssh -i "$env:USERPROFILE\.ssh\id_ed25519" azureuser@<您的vmPublicFqdn>
- macOS/Linux
ssh -i ~/.ssh/id_ed25519 azureuser@<您的vmPublicFqdn>
Get Web Console Address:
After login, execute openclaw-browser-urland copy the complete Dashboard link output.
Device Pairing Authorization:
If the browser prompts "pairing required", execute openclaw-approve-browserin the SSH terminal, then refresh the page to complete login.
6. Advanced Operation: Scripted Deployment via Azure CLI
az loginaz group create --name rg-openclaw-sea --location southeastasia
az deployment group create \ --name openclaw-sea-20260307 \ --resource-group rg-openclaw-sea \ --template-uri https://raw.githubusercontent.com/hanhsia/openclaw-azure-deploy/main/azuredeploy.json \ --parameters \ vmName=openclaw-vm-01 \ adminUsername=azureuser \ sshPublicKey="ssh-ed25519 AAAA..." \ vmSize=Standard_B2as_v2 \ azureOpenAiEndpoint="https://your-resource.cognitiveservices.azure.com/" \ azureOpenAiDeployment="gpt-4o" \ azureOpenAiApiKey="your-key"
az deployment group show \ --name openclaw-sea-20260307 \ --resource-group rg-openclaw-sea \ --query properties.outputs
7. Common Issue Troubleshooting (FAQ)
$Key = "$env:USERPROFILE\.ssh\id_ed25519"icacls $Key /inheritance:ricacls $Key /remove:g "Users""Authenticated Users""Everyone"icacls $Key /grant:r "${env:USERNAME}:R"
-
-
SSH Error: REMOTE HOST IDENTIFICATION HAS CHANGED!
-
Browser prompts "pairing required", unable to log in
-
Access Error 502 Bad Gateway
-
Cause: Service not fully started (wait 1-2 minutes after deployment completes) or service running abnormally.
-
Troubleshooting: After SSH login, execute sudo systemctl status openclaw-gateway caddyto check service status.
-
Virtual Machine Connection Timeout (Timed Out)
The core of the OpenClaw automated deployment solution based on Azure Virtual Machine is to hide all the complex environment initialization, dependency configuration, and security management behind the template. This allows you to skip all the tedious preparatory work and start directly from a "service-ready" state, focusing on the implementation of OpenClaw's capabilities and the exploration of its business value.