Hot-Plugging CPU and Memory Resources Using Ansible Automation Platform

Lab Overview

In this lab, you will learn how to hot-plug CPU and memory resources into a running Virtual Machine (VM) using Ansible Automation Platform and the redhat.openshift_virtualization collection.

Hot-plugging is the ability to add or remove hardware resources, such as CPU or memory, to a running VM without requiring a reboot. This capability is critical for dynamic workloads, allowing you to scale resources based on demand while minimizing downtime.

This lab focuses on using instance types, which are reusable objects in OpenShift Virtualization that define the resources and characteristics for VMs. Instance types simplify resource management by enabling consistent configurations across VMs.

What Are Instance Types?

An instance type is a reusable configuration object where you define resources (like CPU and memory) and characteristics for new VMs. OpenShift Virtualization provides two types of instance types:

  1. VirtualMachineInstancetype: A namespaced object for instance types limited to a specific namespace.

  2. VirtualMachineClusterInstancetype: A cluster-wide object for instance types available across all namespaces.

Both types share the same VirtualMachineInstancetypeSpec, allowing you to define custom configurations or use the variety of instance types included by default when OpenShift Virtualization is installed.

By using instance types, you can simplify VM configuration management and ensure consistency, making them the recommended approach for hot-plugging resources.

In this lab, you will primarily focus on using the instance type method while also learning about the legacy approach of directly modifying the VM specification for context.

Legacy method only works when creating VMs that do not use an Instance Type.

How to Identify if a VM Uses Instance Types or Not?

To determine whether a VM is created with an instance type or not, follow these steps:

  1. Navigate to the Overview tab of the VM (e.g. rhel9-vm1) in the OpenShift Virtualization dashboard.

  2. In the Details section, look for the following:

    • Instance Type: If the VM uses an instance type, this field will display the name of the instance type applied to the VM (e.g. u1.small).

    • Template: If no instance type is used, this field will display either None or the name of the template used to create the VM.

You can use these visual cues to identify whether the VM relies on an instance type or a traditional template.

Below is an example image illustrating both scenarios:

  • One VM shows an assigned instance type.

  • Another VM indicates rhel9-server-small under the template field, indicating that the rhel9-server-small template was used.

    example instance type and template
    Figure 1. Instance Type vs. Template

Create the hot_plug.yml File

  1. Within your VSCode editor, right click tasks of the vm_management collection and create a New File labeled hot_plug.yml

    new file
    Figure 2. New File Creation

  2. Add the following content to the hot_plug.yml relating to the instance_type section below.

Using the instance_type (Recommended)

The instance_type method is the recommended approach for hot-plugging resources into a VM. It ensures consistent and reusable resource configurations across multiple VMs while leveraging the powerful features of OpenShift Virtualization.

- name: Swap Instance Type to add more Resources
  redhat.openshift_virtualization.kubevirt_vm:
    name: "rhel9-vm1"
    namespace: "{{ vm_namespace }}"
    state: present
    instancetype:
      name: "{{ instance_type }}"
      revisionName: ""

Explanation of the Task:

  • redhat.openshift_virtualization.kubevirt_vm: Specifies the module used to manage VMs in OpenShift Virtualization.

  • name: The name of the VM to which the new resources will be applied.

  • namespace: The namespace in which the VM resides.

  • state: Ensures the VM is present and available.

  • instancetype: Defines the instance type for the VM, allowing you to use pre-configured or custom resource settings.

  • name: The name of the instance type to be applied.

  • revisionName: Optionally specifies the exact revision of the instance type, ensuring compatibility with the VM. Typically auto-generated, thus left empty.

VMs must be created using Instance Types for this task method to work. Otherwise use Legacy method.

After making and saving the changes, ensure you commit and push them to your Gitea repository. For detailed instructions, refer to Appendix: Using VS Code or Terminal to Commit and Push Changes.

Legacy Method: Modifying the Resources via Spec

The legacy method involves directly modifying the VM’s spec file to update CPU and memory resources. While this approach is flexible, it lacks the reusability and consistency offered by instance types, making it less ideal for managing resources across multiple VMs.

- name: Modify CPU & Memory Resources
  redhat.openshift_virtualization.kubevirt_vm:
    name: "rhel9-vm2"
    namespace: "{{ vm_namespace }}"
    state: present
    spec:
      domain:
        cpu:
          sockets: 2
        memory:
          guest: 4Gi

Explanation of the Task:

  • redhat.openshift_virtualization.kubevirt_vm: Specifies the module used to manage VMs in OpenShift Virtualization.

  • name: The name of the VM being modified.

  • namespace: The namespace in which the VM resides.

  • state: Ensures the VM is in the desired state, in this case, present.

  • spec: Directly modifies the VM’s specification.

  • domain: Contains settings related to the VM’s virtualized environment.

  • cpu: Specifies the number of CPU sockets for the VM (e.g., 2).

  • memory: Defines the memory allocated to the VM, (e.g., 4Gi).

Non-Legacy VMs are not part of this lab exercise and the Legacy Method is for informational purposes only.

Create and Run the Hot-Plug Job Template

  1. Within the AAP UI Dashboard, navigate to Automation Execution → Templates.

  2. Click Create Template and select Create job template.

  3. Fill in the following details:

Parameter Value

Name

Hot Plug VMs

Job Type

Run

Inventory

OpenShift Virtual Machines

Project

Workshop Project

Playbook

manage_vm_playbook.yml

Execution Environment

Day 2 EE

Credentials

OpenShift Credential

Extra variables

vm_namespace: vms-aap-day2
task_file: hot_plug.yml
instance_type: u1.2xmedium

  1. Click Create Job Template.

  2. Launch the job by selecting Launch Template from the top-right corner.

  3. When the job completes, head to the OpenShift UI dashboard and view the details of the rhel9-vm1 Virtual Machine. You should see that the new size u1.2xmedium is now being used.

Conclusion

This lab has demonstrated how to effectively hot-plug CPU and memory resources into a running Virtual Machine (VM) using Ansible Automation Platform and the redhat.openshift_virtualization collection. By leveraging the recommended instance type method, you can ensure consistent, reusable, and scalable configurations across multiple VMs, simplifying resource management and aligning with best practices.

We also explored the legacy approach of directly modifying the VM specification, which, while flexible, is less ideal for modern environments due to its lack of reusability and standardization. Understanding both methods equips you with the knowledge to handle scenarios where instance types may not be available.

Through this lab, you have learned how to:

  • Identify whether a VM uses an instance type or a template.

  • Use the instance type method to dynamically add resources to VMs.

  • Understand the limitations and applications of the legacy method.

By applying these skills, you can efficiently manage resources in OpenShift Virtualization, adapting to the demands of dynamic workloads while minimizing downtime.

← Previous Exercise | Next Exercise →