Scaling

mctl_scale_service changes HPA autoscaling through the deploy-service update-config workflow. It requires the developer role on the tenant.

Configure autoscaling

mctl_scale_service(
  team_name="staging",
  component_name="my-app",
  autoscaling_enabled="true",
  min_replicas="2",
  max_replicas="5",
  cpu_threshold="80"
)

Parameters are strings. The required autoscaling_enabled value is "true" or "false". Optional defaults are 1 minimum replica, 5 maximum replicas and an 80% CPU threshold. Keep the returned workflow name and check workflow and service status. The HPA needs resource requests and metrics to make scaling decisions.

This tool has no direct replicas parameter. Disabling HPA does not itself select a new fixed replica count; fixed replicas are configured in the service Helm values.

Check Resource Usage

Show me resource usage for staging

mctl_get_resource_usage(team="staging") reports namespace resource quota usage against allocation. It is not per-pod CPU/memory telemetry. Check application metrics separately before choosing autoscaling thresholds.