Skip to main content

Autoscaling Groups

Overview​

Autoscaling Groups automatically adjust the number of instances in your infrastructure based on real-time resource utilization. Define scaling policies with CPU or memory thresholds, and the platform handles provisioning and deprovisioning instances to match demand.

Creating an Autoscaling Group​

  1. Navigate to Compute > Autoscaling Groups in the sidebar.
  2. Click Create Autoscaling Group.
  3. Configure the group:
    • Name -- A descriptive name for the group.
    • Location -- The data center where instances will be created.
    • Plan -- The instance plan (CPU, RAM, storage) for scaled instances.
    • Image -- The OS image to deploy on new instances.
    • Min Instances -- Minimum number of instances to keep running.
    • Max Instances -- Maximum number of instances allowed.
  4. Click Create.

Scaling Policies​

Add policies to control when the group scales up or down.

  • Metric -- CPU or memory utilization.
  • Threshold -- The percentage that triggers scaling (e.g., scale up when CPU > 80%).
  • Cooldown -- Minimum time between scaling actions to prevent thrashing.
  • Action -- Scale up (add instances) or scale down (remove instances).

How It Works​

  1. The platform monitors resource utilization across all instances in the group.
  2. When a policy threshold is breached for the configured duration, a scaling action triggers.
  3. Scale-up creates a new instance from the configured image and plan.
  4. Scale-down removes the least loaded instance (respecting the minimum count).
  5. Cooldown prevents rapid oscillation between scale-up and scale-down.

Permissions​

With Team Management enabled:

  • Viewer can view autoscaling groups and policies.
  • Operator can create and modify groups and policies.
  • Admin can also delete groups and policies.