Skip to content
Origin Cloud Docs

Observability / Overview

What is Origin Cloud Observability?

Updated · 3 min read

Observability lets you query CPU, network, and disk metrics that the platform collects for your instances, volumes, and network interfaces.

On this page7 sections

Origin Cloud Observability gives you usage metrics for your own resources: how much CPU time an instance consumes, how many bytes a network interface sends and receives, and how many bytes an instance reads from and writes to a volume. The platform collects these metrics for you. You query them through the GraphQL API to graph activity, size instances, or find a busy disk.

How metrics flow

Every 30 seconds the platform collects counter values from your compute instances, and through them from their volumes and network interfaces. A metrics query (RATE or INCREASE) turns those collected values into a series of data points.

Every metric is a counter: a running total, such as total CPU seconds or total bytes received, that grows while the instance runs. A raw total is rarely useful, so a query never returns it. Instead, you pick a statistic, and the platform computes it for each period of your time window:

  • RATE returns the average per-second increase over the period, for example bytes per second.
  • INCREASE returns the total increase over the period, for example bytes transferred in the last five minutes.

The platform collects a value for each counter every 30 seconds, but only while the instance runs. Volumes and network interfaces are measured through the instance they're attached to, so a volume or interface that isn't attached to a running instance has no data points.

Features

  • Built-in metrics: Five metrics cover CPU time, network traffic, and volume reads and writes. See Metrics reference.
  • Metric queries: One query returns a series of data points per metric for one resource over a time window you choose. See Query metrics.

Ways to access Observability

  • GraphQL API: Metrics are available through the observability namespace of the API, with the metrics query. See Query metrics. For authenticating against the API, see Use the API.

The origin CLI has no command for metrics.

What Observability doesn't do

  • Metrics are collected by the platform only. You can't submit your own metrics.
  • There's no query that lists metric names. The complete list is in the Metrics reference.
  • Metrics measure your resources from the outside. They don't show memory use, processes, or file systems inside an instance.
  • There are no alarms or notifications on metric values.

Pricing

Each metrics query you run and each value the platform collects for your resources are billed. For the rates, see Observability prices in the price list reference.

  • Instances: the compute instances whose CPU time is measured.
  • Volumes: the volumes whose reads and writes are measured.
  • Network interfaces: the interfaces whose traffic is measured.

Next steps