Join us at the Dash conference! July 16-17, NYC

Metrics Summary

The Metrics Summary page displays a list of all the metrics reported to Datadog under a specified time frame: the past hour, day, or week. This list can be filtered by name. Clicking on a metric brings up a panel with more detailed information.


Metric panel

The metric panel displays an overview for a given metric:

Key information about your metric can be seen:

  • Metric name: The name of your metric to be used in the metric explorer or in dashboard widgets.
  • Number of distinct metrics: A metric name can correspond to multiple distinct metrics depending of its associated tags, Consult the custom metrics documentation to learn more.
  • Number of host: The total number of hosts reporting this metric.
  • Number of tags: The total number of tags attached to this metric. Read more about tagging and how to assign tags.
  • Metrics metadata: All metadata attached to your metric:
    • the metric description
    • the metric unit
    • the metric type
    • the integration name, if this metric is coming from an integration
    • the interval of collection of this metric.

Metric metadata

Every piece of metric metadata can be manually edited:

  • Edit the metric description to help understand what a metric does.

    If a metric is coming from an integration and you notice a wrong description, open an issue in the Datadog documentation GitHub repository for a fix.

  • Edit the metric unit or add a custom unit for a custom metrics.

    When submitting custom metrics to Datadog, it is possible to change the unit of measurement which shows up when hovering over a certain metric in your graph. Do this by selecting your custom metric from the list and then selecting the unit of measurement you would like to use as depicted below:

    Metrics Metadata

    Note: This does not change the way that a metric graph is displayed (only the units of measurement when hovering over a metric)

  • Edit the metric type to match the real metric type send. Warning: this changes your metric behavior in ALL your analytics and monitors; do this at your own risk.

Further reading

Additional helpful documentation, links, and articles: