Google Cloud metrics overview (original) (raw)

Cloud Monitoring supports the metric types that Google Cloud services define. Services write metric data to record performance information. For example, a metric might record the CPU usage for a Compute Engine instance.

Each Google Cloud metric has a type, which has the formatAPI.googleapis.com/METRIC_PATH. The field API.googleapis.com specifies a service domain and it identifies the Google Cloud service or product area that owns and emits the metric. The field METRIC_PATH provides a hierarchical path that uniquely identifies the specific metric within the service domain.

The reference pages list metric types by service domain:

For each domain, the metrics are listed alphabetically.

Find metric types

To find the metric data that is of interest to you, try the following:

Search the API and product table

The following table can help you find the available metric types for a service. For example, if you enter GKE in the filter bar, then the table provides links to your GKE metric types.

Issue an API request to a project

If your search terms don't match any table rows or the matched rows don't contain the data you need, then issue an API command. You can apply a filter to limit the command's response.

For example, to list metrics whose type includes disk, do the following:

  1. Go to the projects.metricDescriptors.list page.
  2. In the Try this method pane, do the following:
    • In the name field, enter projects/PROJECT_ID, and replacePROJECT_ID with your project ID.
    • In the filter field, enter the following:
    metric.type = has_substring("disk")  

    The hierarchical path that identifies a specific metric within a service domain uses forward slashes (/) as delimiters. You can include these delimiters in the filter. For example, the following filter matches metrics that include /disk/ in their path:

    metric.type = has_substring("/disk/")  

    There are other operators available. For example, to list all metrics whose type starts with compute, then set the filter to the following:

    metric.type = starts_with("compute")  
  3. Click Execute and complete the required authorization steps.
    The command response is a list of metric descriptors whose type field matches the specified filter.

You can also issue curl commands or use client libraries to issue API commands.

Pricing

For pricing information, see Google Cloud Observability pricing.