Skip to content
Helicarrier Developers
Support Dashboard
DOCUMENTATION

Metrics & resource usage

Inspect CPU, memory, network traffic, and instance counts to understand your service and choose its capacity.

Open a service’s metrics view to understand how it uses its resources over time. Combine metrics with runtime logs and deployment history when diagnosing a change.

Application metrics

Section titled “Application metrics”

Application metrics include CPU, memory, network usage, and instance count. Available windows are 1h, 6h, 24h, and 7d. Choose a longer window to distinguish a recurring pattern from a momentary spike.

For services with multiple instances, CPU and memory report per-instance averages, not the sum across replicas. The instance-count chart shows capacity changes alongside utilization.

Database metrics

Section titled “Database metrics”

Managed databases provide a separate metrics view with resource and engine-native information where supported, such as connection counts, cache behavior, and disk usage relative to provisioned capacity.

A nearly full database may need storage rather than more compute. Check both the workload and the resource approaching its limit before changing a plan.

Use metrics to make a change

Section titled “Use metrics to make a change”
  1. Identify the time the behavior started.
  2. Check whether a deployment, traffic change, or instance-count change happened then.
  3. Compare CPU and memory with application logs.
  4. Adjust the plan or scaling configuration for the resource that is under pressure.
  5. Observe the result over a comparable workload window.

Autoscaling is available for eligible application services and uses the targets and bounds you configure. Each running instance is billed. See scaling for the supported service types and behavior.

Programmatic access

Section titled “Programmatic access”

A scoped key needs the appropriate allowed route and permission. Service-key metric access does not automatically authorize database operations.