CrowdStrike Azure VM Extension

The CrowdStrike Azure VM Extension is an open-source solution that simplifies and automates the installation of the CrowdStrike Falcon sensor on Azure virtual machines at enterprise scale.

Overview

The CrowdStrike Azure VM Extension provides:

• Automated Deployment: Deploy the Falcon sensor to Azure VMs at scale using Azure's native extension framework
• Cross-Platform Support: Support for both Linux and Windows Azure virtual machines
• Enterprise Integration: Seamless integration with Azure infrastructure and CrowdStrike APIs
• Lifecycle Management: Handle sensor installation, configuration, updates, and removal
• Architecture Support: Support for both x86_64 and arm64 architectures on Linux

Architecture

The extension follows Azure VM Extension standards with a handler-based architecture:

Handler Framework

• Lifecycle Operations: Install, enable, disable, uninstall, and update operations
• Platform-Specific Implementation: Separate Linux (bash) and Windows (PowerShell) handlers
• Azure Integration: Native integration with Azure VM Extension framework
• Status Reporting: Structured JSON status updates to Azure portal

Deployment Flow

1. Azure invokes the extension handler with configuration parameters
2. Handler validates the configuration and credentials
3. CrowdStrike API is called to download the appropriate sensor package
4. Falcon installer is executed to install and configure the sensor
5. Status reporting back to Azure portal with success/failure information

Supported Platforms

[!NOTE]
For a complete list of supported platforms and versions, architectures, and compatibility, please refer to the CrowdStrike documentation.

Supported Linux Distributions

See https://falcon.crowdstrike.com/documentation/page/edd7717e/falcon-sensor-for-linux-system-requirements for detailed support information and a complete list of supported platforms:

• Ubuntu LTS
• Debian
• Red Hat Enterprise Linux
• SUSE Linux Enterprise Server

Windows Versions

See https://falcon.crowdstrike.com/documentation/page/ecc97e75/falcon-sensor-for-windows-deployment for detailed support information and a complete list of supported platforms:

• Windows Server
• Windows Desktop

Falcon API Permissions

API clients are granted one or more API scopes. Scopes allow access to specific CrowdStrike APIs and describe the actions that an API client can perform.

Ensure the following API scopes are enabled:

[!IMPORTANT]

• Sensor Download [read]
• Sensor update policies [read]
• (optional) Installation Tokens [read]

  This scope allows the installer to retrieve a provisioning token from the API, but only if installation tokens are required in your environment.

• (Optional) Sensor update policies [write]

  Required for reading the maintenance token during uninstall. This is only required for uninstall.

Installation

The extension can be deployed through:

• Azure CLI - Command-line deployment
• Azure Resource Manager templates - Infrastructure as Code
• Azure Policy - Enterprise-scale automated deployment (see Policy Templates)

Configuration

The CrowdStrike Azure VM Extension uses two types of configuration parameters:

• Settings: Non-sensitive configuration parameters passed as plain text
• Protected Settings: Sensitive parameters (credentials, tokens) that are encrypted in transit and at rest

[!IMPORTANT]
Always place sensitive information like credentials and tokens in protectedSettings to ensure they are encrypted and secure.

Protected Settings (Sensitive Parameters)

These parameters contain sensitive information and must be placed in the protectedSettings section:

*One of the following authentication methods is required:

• client_id + client_secret (OAuth2 credentials)
• access_token (direct API access token)
• azure_vault_name (Azure Key Vault integration)

[!IMPORTANT]
When specifying the Azure vault with azure_vault_name, make sure that all VMs have the appropriate permissions to list and get the Key Vault secrets.
The extension will fail to install if the VM doesn't have the required permissions to access the secrets.
Any secrets in the vault should be prefixed with FALCON- e.g. FALCON-CLIENT-ID, FALCON-CLIENT-SECRET, FALCON-ACCESS-TOKEN, etc.

When using azure_vault_name with azure_managed_identity_client_id, the extension will use the specified user-assigned managed identity to authenticate with the Key Vault instead of the VM's system-assigned managed identity. This provides more granular control over Key Vault access permissions and is useful in scenarios where you want to use a specific managed identity for Key Vault authentication.

If you choose to use Azure Key Vault with ARM templates, you do not have to specify azure_vault_name or give VMs access to the vault.
See Using Key Vault with ARM Templates for details on using Key Vault with ARM templates.

Settings (Non-Sensitive Parameters)

These configuration parameters can be placed in the settings section:

[!IMPORTANT]
As a best practice, make sure to change the sensor_update_policy from the default to match your organization's desired sensor update policy as this determins what sensor version will be installed!

Common Settings (Linux and Windows)

Windows-Specific Settings

Usage

Azure CLI

Linux

az vm extension set \
  --resource-group myResourceGroup \
  --vm-name myVM \
  --name FalconSensorLinux \
  --publisher Crowdstrike.Falcon \
  --settings '{
    "cloud": "autodiscover",
    "tags": "azure,production"
  }' \
  --protected-settings '{
    "client_id": "YOUR_CLIENT_ID",
    "client_secret": "YOUR_CLIENT_SECRET"
  }'

Windows

az vm extension set \
  --resource-group myResourceGroup \
  --vm-name myVM \
  --name FalconSensorWindows \
  --publisher Crowdstrike.Falcon \
  --settings '{
    "cloud": "autodiscover",
    "tags": "azure,production"
  }' \
  --protected-settings '{
    "client_id": "YOUR_CLIENT_ID",
    "client_secret": "YOUR_CLIENT_SECRET"
  }'

Using Azure Key Vault with ARM Templates

For enhanced security, store sensitive CrowdStrike API credentials in Azure Key Vault rather than directly in ARM templates. This ensures credentials are encrypted, access-controlled, and auditable.

Azure Vault Setup

Follow the Azure Key Vault documentation to create a Key Vault and store your CrowdStrike API credentials as secrets.

ARM Template Integration

Store your CrowdStrike API credentials in Key Vault as secrets. You can use any secret names you prefer - the examples below use:

• FalconClientId - Your CrowdStrike API Client ID
• FalconClientSecret - Your CrowdStrike API Client Secret

Example Using Key Vault References in an ARM template

Example of parameters file with Key Vault references:

{
    "parameters": {
        "falconClientId": {
            "reference": {
                "keyVault": {
                    "id": "/subscriptions/{subscription-id}/resourceGroups/{rg-name}/providers/Microsoft.KeyVault/vaults/{vault-name}"
                },
                "secretName": "FalconClientId"
            }
        },
        "falconClientSecret": {
            "reference": {
                "keyVault": {
                    "id": "/subscriptions/{subscription-id}/resourceGroups/{rg-name}/providers/Microsoft.KeyVault/vaults/{vault-name}"
                },
                "secretName": "FalconClientSecret"
            }
        }
    }
}

Example of ARM template deployment with inline parameters:

{
    "type": "Microsoft.Compute/virtualMachines/extensions",
    "apiVersion": "2021-07-01",
    "name": "[concat(parameters('vmName'), '/', variables('extensionName'))]",
    "location": "[parameters('location')]",
    "dependsOn": [
        "[resourceId('Microsoft.Compute/virtualMachines', parameters('vmName'))]"
    ],
    "properties": {
        "publisher": "[variables('extensionPublisher')]",
        "type": "[variables('extensionType')]",
        "typeHandlerVersion": "[variables('extensionTypeHandlerVersion')]",
        "autoUpgradeMinorVersion": true,
        "protectedSettings": {
            "client_id": "[parameters('falconClientId')]",
            "client_secret": "[parameters('falconClientSecret')]"
        }
    }
}

Make sure to enable Azure Resource Manager for template deployment in your Azure Vault settings to allow the ARM template to access your Vault secrets.

Azure Policy

For automated enterprise-scale deployment using Azure Policy, see the Policy Templates documentation for detailed instructions on deploying CrowdStrike Falcon at scale using Azure Policy. The templates support both subscription and management group level assignments, with automatic detection of Windows and Linux VMs.

VM Scale Sets (VMSS)

[!IMPORTANT]
When deploying the CrowdStrike Falcon extension to Azure VM Scale Sets, there are important networking considerations to ensure proper sensor installation and operation.
Make sure that the VM Scale Sets outbound rules for the load balancer are properly configured to enable connectivity to CrowdStrike APIs for sensor download and installation; otherwise, the sensor installation will fail.

Example Adding the CrowdStrike Falcon Extension to a VM Scale Set

"extensionProfile": {
  "extensions": [
    {
      "name": "CrowdstrikeFalconSensor",
      "properties": {
        "publisher": "[variables('extensionPublisher')]",
        "type": "[variables('extensionName')]",
        "typeHandlerVersion": "[parameters('extensionVersion')]",
        "autoUpgradeMinorVersion": true,
        "protectedSettings": {
          "client_id": "[parameters('falconClientId')]",
          "client_secret": "[parameters('falconClientSecret')]",
          "tags": "[parameters('sensorTags')]",
          "sensor_update_policy": "[parameters('sensorUpdatePolicy')]"
        }
      }
    }
  ]
}

Documentation

• FAQ - Frequently asked questions and troubleshooting
• Testing Guide - Information about the testing framework
• Policy Templates - Azure Policy deployment guide for enterprise-scale deployment

Troubleshooting

If you encounter any issues during the extension deployment process, the following logs will be generated on the VMs:

• cshandler.log: This log file captures the standard output and error information from the extension handler operations (install, enable, disable, uninstall, update).
• CommandExecution.log: This log file contains detailed information about command execution during the extension operations.
• falcon-installer.log: This log file contains detailed information about the Falcon sensor installation process. It includes messages about the progress of the installation, any errors encountered, and other relevant details.

These logs can be found under /var/log/azure/Crowdstrike.Falcon.FalconSensorLinux/ for Linux VMs and C:\WindowsAzure\Logs\Plugins\Crowdstrike.Falcon.FalconSensorWindows\<version>\ for Windows VMs.

The extension handler working directory (which includes downloaded installer files, configuration, and status files) can be found under /var/lib/waagent/Crowdstrike.Falcon.FalconSensorLinux-<version>/ for Linux VMs and C:\Packages\Plugins\Crowdstrike.Falcon.FalconSensorWindows\<version>\ for Windows VMs.

In addition to the VM logs, you can also check the extension status and configuration through the Azure portal or Azure CLI:

• Check extension status in Azure portal under VM → Extensions + applications
• Use Azure CLI: az vm extension show --resource-group <rg> --vm-name <vm> --name FalconSensorLinux (or FalconSensorWindows)

Review these logs for failures as to why the installation and deployment failed. When contacting CrowdStrike support or creating a GitHub issue, these logs should be provided.

[!IMPORTANT]
When providing logs, please make sure to sanitize the output and do not provide any credentials or sensitive information.

Contributing

We welcome contributions that improve the installation and distribution processes of the Falcon Sensor. Please ensure that your contributions align with our coding standards and pass all CI/CD checks.

Support

The CrowdStrike Azure VM Extension is a community-driven, open source project designed to streamline the deployment and use of the CrowdStrike Falcon sensor. While not a formal CrowdStrike product, Falcon Installer is maintained by CrowdStrike and supported in partnership with the open source developer community.

For additional support, please see the SUPPORT.md file.

License

See LICENSE