How to Deploy Virtual Machines

Deploying a Virtual Machine

This guide walks you through creating a Virtual Machines in the NetActuate portal and connecting to it with kubectl.

Common Use Cases

Global Edge Workloads

Launch virtual machines across distributed global locations to bring applications closer to users and reduce latency for critical services.

Web Application Hosting

Deploy scalable front-end and backend services on isolated virtual infrastructure with predictable compute performance and flexible networking.

Development and Testing Environments

Spin up temporary virtual machines to test code, validate software builds, and replicate production environments without impacting live services.

Disaster Recovery Infrastructure

Maintain standby virtual machines in alternate regions to enable rapid failover, business continuity, and resilient application recover.

‍Prerequisites

• A NetActuate account with an active subscription
• At least one SSH key added to your account (recommended — see SSH Keys)

Step 1: Navigate to Virtual Machines

Go to Infrastructure → Virtual Machines. This page lists all VMs on your account with their hostname, IP address, bandwidth, and status. You can filter by operating system, plan size, location, tags, or status (running, stopped, rebooting).

Click the + Add button in the top-right corner to open the deployment wizard.

Note: If you don't have any virtual machines yet, you'll be prompted to add one as a first-time user.

Step 2: CPU Type / Hardware Pool

Select a CPU type. This determines the underlying hardware pool your VM is provisioned on:

• General Compute — Intel-based hardware
• AMD EPYC — AMD EPYC-based hardware
• Custom pools — if you have contracted custom hardware with NetActuate, additional pools may appear here

Include Accelerators

Toggle Include Accelerators to filter available locations to only those with accelerator hardware. When enabled, a Category dropdown appears where you can select the type of accelerator:

• VPU — Video Processing Unit (e.g., NETINT Quadra)
• GPU — Graphics Processing Unit (e.g., NVIDIA)

The location list updates to show only data centers that have your selected accelerator type in stock. For more about accelerator types, see Accelerators.

Step 3: Deploy Location

Select a region first, then choose a specific POP within that region.

Available regions:

• North America
• South America
• Europe
• Middle East and Africa
• Asia and Oceania
• Private Cloud (if you have a private cloud deployment with custom locations)

Step 4: Billing and Server Sizing

Select Accelerator (if enabled)

If you toggled Include Accelerators in Step 2, a Select Accelerator panel appears at the top of the Billing tab. Configure the accelerator:

1. Type — select the accelerator model (e.g., NETINT Quadra T1A).
2. Attachment Type — choose how the card connects to your VM:
   • Physical — direct PCI passthrough. The card is passed through to the VM, but some portal metrics may be limited.
   • Virtual Function (recommended) — SR-IOV virtual function. Provides full portal metrics and monitoring while maintaining near-native performance.
3. Firmware — select a firmware version for the card. Click the release notes icon next to each version to review what changed before selecting.
4. Click Add to bind the accelerator to the VM.

Note: Virtual Function attachment is recommended for most workloads. It enables the full set of real-time metrics in the portal (utilization, temperature, memory) while providing near-native hardware performance.

Note: If you plan to use NETINT's Bitstreams management software, Physical attachment is recommended for best performance and compatibility.

Billing cycle

Select your billing cycle. The options available depend on your contract and may include:

• Monthly billing
• Resource-based billing
• Usage-based bundles

Server size

Configure the server's resources:

• RAM — select from 1 GB up to 1 TB of memory
• CPU — select from 1 up to 128 vCPUs
• Disk — selectable in 25 GB increments, up to 250 GB on standard plans (custom plans may offer more)

Pricing Summary

The Pricing Summary at the bottom of the Billing tab shows the cost breakdown. When an accelerator is attached, you will see:

• Virtual Machine — the hourly or monthly cost for the VM's CPU, RAM, and disk
• Accelerator — the per-hour cost for the attached accelerator card
• Total — combined cost

Note: Outbound transfer is based on your contract. Inbound transfer is unlimited at no additional cost. Virtual machines in the public cloud have a 10 Gbps port speed by default. Each VM receives a dedicated IPv4 and IPv6 address.

Step 5: Operating System

Select an OS image from the available categories. There are over 100 operating system images available, and you can also call the API to list them all. See OS Images for more detail.

You have three options:

1. Public images — select a category (e.g., Ubuntu, CentOS, Debian, Windows Server) and then a specific version (e.g., Ubuntu 24.04 LTS).
2. Custom ISO — under Private, select Custom OS ISO Installation and enter an HTTPS URL to a custom ISO. The ISO will be mounted and the server will boot to it.
3. Private images — if you have uploaded and manage your own OS images under Platform → OS Images, they will appear in this list.

Note: For VPU workloads, Ubuntu 24.04 LTS is the recommended operating system. It provides the best compatibility with the NETINT software stack and FFmpeg integration.

Step 6: Initialization Script

Optionally add a cloud-init or bash script that runs when the VM first boots. This is commonly used with Terraform, Ansible, or direct API deployments to automate post-deploy configuration.

Click Add Initialization Script to open the script editor. You can paste a cloud-init YAML or bash script directly.

Using secrets

Click Show Secrets to view secrets stored in the Secrets Manager. You can insert secret shortcodes into your initialization script to inject secret values at boot time, making them available to the VM during bootstrapping without hardcoding sensitive data.

Step 7: Networking

Configure the network interface for your VM. By default, you will see the Primary Network Interface with a publicly routed IPv4 and IPv6 static IP assignment.

Depending on your account and contracted services, additional options may be available:

• VLAN attachment — connect to a Cloud VLAN, which may link to a cloud connect port, hyperscaler, IP transit, transport, or other NetActuate services

The networking configuration can be more advanced depending on the services available on your account.

Firewall Set

If firewall functionality is enabled on your account, you can select a Firewall Set to associate with the VM during deployment. The selected rule set is applied automatically once the VM is built. See Managing Firewall Rule Sets for details on creating and configuring rule sets.

Note: You can also apply firewall sets after deployment from the firewall management page, or set an account-wide default under Account > Settings > Security. See VM Interface Firewall for all options.

Step 8: Authentication

Choose how you will authenticate to the VM:

SSH key (recommended)

Select one or more SSH keys from your account. Keys are managed under Account → SSH Keys, but you can also add a new key directly from the deployment wizard. Keys added during deployment are saved to your account for future use.

See SSH Keys for details on managing your keys.

Password

Enter a strong password. The password must contain:

• A number
• A lowercase letter
• An uppercase letter
• A special character

A strength indicator shows whether your password meets the requirements. You can also click the generate button to create a random password.

Note: SSH key authentication is the recommended method. If you choose password authentication, store the password securely — it is not recoverable from the portal.

Step 9: Finalize

Enter a fully qualified domain name (FQDN) for the server (e.g., test-server.atlanta.example.com).

Optionally assign one or more tags to organize the VM. You can select an existing tag or create a new one. The Manage Tags link takes you to the tags section under Platform → Tags.

Click Submit to begin the build.

Step 10: Build Progress

After submitting, the screen changes to a build progress view. The deployment process includes:

1. IPv4 and IPv6 address assignment
2. Location confirmation
3. Cloud instance creation
4. Initial boot
5. File system configuration
6. Build-specific configuration (cloud-init, networking)
7. Instance optimization
8. Final boot

Note: A virtual machine typically builds within 15 seconds.

Once the build completes, click View Instance to open the VM's management panel.

Next Steps

• Managing Virtual Machines — control panel, networking, tools, and settings for your running VM

‍