RESOURCE CATALOG

Autoscale

Adjusts Compute replica count based on a metric.

type: "autoscale"analogous to an auto scaling group

Definition

Scaling policy attached to a Compute. Keeps the average of the selected metric (CPU, memory, or requests per second) near the configured threshold, within the replica bounds declared on the target Compute.

When to use

Loads with predictable variation (business hours, seasonality).
Workers consuming queues with variable volume.
APIs with non-constant traffic.

When not to use

Constant loads — set replicas_min = replicas_max.
Very low volume sites — keep fixed replicas.

Configuration fields

Each field is validated at creation time. Defaults apply when fields are omitted.

config · autoscale

Campo	Tipo	Default · Opções
alvo_recurso_id VM alvo	ref → compute	—
metrica Métrica	enum	`cpumemoriarequests` Default`cpu`
limiar Limiar (%)	int	Default`75`

Examples

Full payloads for POST to /api/projetos/{id}/recursos:

Scale on CPU at 70%

json

{
  "tipo": "autoscale",
  "nome": "auto-api",
  "config": {
    "alvo_recurso_id": "<id_api>",
    "metrica": "cpu",
    "limiar": 70
  }
}

Related resources

VM (Instância)

VM instance: web, API, workers, or cron — web terminal and root SSH.

Insight

Collects logs and metrics from Applications, with optional alerts.

Gateway

Sentinel