Autoscaling of VM

Amazon AWS state:

You should use AWS Auto Scaling if you have an application that uses one or more scalable resources and experiences variable load. A good example would be an e-commerce web application that receives variable traffic through the day.

Autoscaling isn’t just for workloads running in a public cloud or Hyperscalers. We can achieve the same functionality running in your data center on premises within VMware.

We need to run some tools to achieve this. I am sure you can accomplish the same with configuration management tools.

Let us look at how to autoscale when the VM in question hits a threshold of 95% consumption over a 10-minute duration. In this example we will perform horizontal scaling, meaning adding a second VM.

Prerequisites:

  • VMware Aria Automation is fully set up and integrated with VMware Aria Operations (formerly vRealize Operations) for monitoring and alerting.
  • VMware Aria Orchestrator workflows are available for automation tasks.
  • The necessary permissions to create and modify policies, blueprints, and workflows.

Step-by-Step Configuration

1. Configure Monitoring in VMware Aria Operations

  • Set Up Metric Collection:
    • Ensure that the VMs are being monitored by VMware Aria Operations and that CPU usage metrics are being collected.
  • Create a Custom Alert Definition:
    • Go to the Alerts tab in VMware Aria Operations.
    • Create a new alert definition that triggers when the CPU usage of a VM exceeds 95% for 10 minutes.
    • Criteria:
    • Base Object Type: Virtual Machine.
    • Condition: CPU usage (%) > 95%.
    • Duration: 10 minutes.
    • Alert Impact: Set the impact to Performance.
    • Notification: Enable notifications to send alerts via email or other channels.

2. Create a Scaling Workflow in VMware Aria Orchestrator

  • Define a Workflow for Scaling:
    • In VMware Aria Orchestrator, create a workflow that will perform the scaling action (e.g., add more vCPUs or clone a new VM).
    • Scaling Options:
      • Horizontal Scaling: Clone a new instance of the VM or deploy a new VM from a blueprint.
      • Vertical Scaling: Increase the number of vCPUs or memory for the existing VM.
    • Input Parameters: The workflow should accept inputs such as the VM name, resource settings (vCPU, memory), etc.
    • Action: Implement the logic to scale the VM (e.g., via a PowerCLI script or vSphere API calls).

3. Integrate VMware Aria Operations with Aria Automation

  • Create a Subscription:
    • In VMware Aria Automation, navigate to the Subscriptions section.
    • Create a new subscription that listens to the alert from VMware Aria Operations.
    • Filter: Set the filter criteria to trigger when the alert (CPU > 95% for 10 minutes) is received.
  • Attach Workflow to Subscription:
    • Attach the scaling workflow you created in VMware Aria Orchestrator to this subscription.
    • This will trigger the workflow whenever the alert conditions are met.

4. Create and Apply Auto-Scaling Policies

  • Define Auto-Scaling Policies:
    • In VMware Aria Automation, define a policy for auto-scaling that uses the subscription and workflow created earlier.
    • Set the policy to monitor CPU usage and take the scaling action automatically.
  • Apply Policy to VMs:
    • Apply this auto-scaling policy to the specific VMs or VM groups that need auto-scaling based on CPU usage.

5. Test the Configuration

6. Monitor and Adjust

  • Monitor Scaling Actions:
    • Use VMware Aria Operations to monitor the scaling actions and ensure they are executed as expected.
  • Fine-Tune Alert Thresholds:
    • If necessary, adjust the alert thresholds or scaling logic based on the observed performance.

Summary

Following these steps, you can configure VMware Aria Automation to automatically scale a VM when its CPU usage exceeds 95% for 10 minutes. This involves setting up alerts in VMware Aria Operations, creating scaling workflows in VMware Aria Orchestrator, and integrating these with VMware Aria Automation policies.

Leave a Reply

Your email address will not be published. Required fields are marked *

Share on Social Media