Metrics Usage

Metrics Usage provides insights into your metrics count, helping you understand which custom metrics are responsible for your time series usage and which are being scraped across your entire environment. With this information, you can make informed decisions about optimizing your resources and gain control over your monitoring infrastructure.

Overview

With Metrics Usage, you can:

  • Get a comprehensive view of time series cardinality and time series ingestion distribution across your infrastructure.

    This information, coupled with the ability to search for a metric within your entire infrastructure or within a scope, lets you quickly access the time series information you need when you need it.

  • Obtain the time series count for each metric and an overview of the time series ingested in the last 20 minutes.

    This functionality equips you with valuable data regarding metric influx and soon, usage.

  • Download per-metric time series count as a CSV or JSON file.

    This provides a tangible, offline reference point that can be helpful in trend identification and reporting.

  • Discover which dashboards and alerts are using a given metric.

Usage

To access Metrics Usage, select Explore > Metrics Usage in the Sysdig Monitor UI.

Search for Metrics

On the Metrics Usage tab, open the infrastructure filtering menu and select the infrastructure object you want to search. You can simultaneously use all the filtering mechanisms to drill down to find what you are looking for.

CriteriaPreview

Search a Metric

On the Search field, enter the full (such as container_cpu_usage_seconds_total) or partial (such as container_cpu) metric name. The metric list and panels will instantly be updated to show the metrics that match your search criteria.

Search by Source

Select one or more sources from the All sources drop-down to filter metrics coming from different sources (agent, prometheus_remote_write, azure, gcp, aws).

Search by Jobs

Select one or more jobs from the All jobs drop-down to filter all the metrics scraped by that job. Refer to this topic, for a list of Agent jobs.

View Metrics Details

You can select a metric and access additional information and actions in the slider window.

  • Description: A short description of the metric.
  • Type: Specifies if the metric is gauge or counter.
  • Unit: The unit of the metric. For example, Kilobyte.
  • Manage Integrations: If the metric originates from a Sysdig Monitoring Integration, this link lets you view and manage the integration.
  • Open in Metrics Explorer: Takes you to Metrics Explorer where you can further explore this metric in the context of others.
  • Open in PromQL Query: Takes you to PromQL Query Explorer where you can craft complex PromQL queries.
  • Dashboards: Shows the dashboards built from the selected metric. See View Dashboards.
  • Alerts: Helps you explore the alerts built from the selected metric. See View Alerts.

View Detailed Time Series Usage

The Time Series Usage tab offers two main capabilities:

  • Time Series Count Over Time: Comparing the count of timeseries over different periods, gives you the ability to identify metric churn over time.

  • Label Usage: Taking into account the active grouping and scope, this enables you to have a deeper understanding of which labels are part of a selected metric. Hovering each label reveals three actions:

    • Label Values Preview: A preview of label values.
    • Open in Metrics Explorer: Takes you to Metrics Explorer where you can further explore this metric and label combination over time.
    • Open in PromQL Query: Takes you to PromQL Query where you can further query this metric and label combination over time. An initial PromQL query such as count(storage_operation_duration_seconds_bucket{_sysdig_datasource="agent",job="k8s-storage-default"}) by (_sysdig_datasource, job, host_hostname) will be given as a starting point.

View Dashboards

The Dashboards tab provides the ability for you to explore which dashboards are using the currently selected metric. Results are presented in three categories.

  • Dashboards in current Team: The dashboards that are currently accessible and visible to the current team.
  • Library Dashboards: The read-only dashboards available in the Library.
  • Dashboards in other Teams: The dashboards that are created by other teams but are not shared with the current team.
  • Other dashboards: The dashboards that are created privately by other users or created by the teams that you are not a part of.

View Alerts

The Alerts tab provides you with the ability to explore which alerts are using the selected metric. The results are presented in the following categories:

  • Alerts in current Team: The alerts that are currently accessible and visible to the current active team.
  • Alerts in other Teams: The alerts created by other teams.
  • Other alerts: The alerts created by the teams that you are not a part of.

Export to CSV or JSON

For further processing of metrics usage with a spreadsheet (CSV) or programming language of your choice (JSON), use the Export all results link at the bottom of the table showing all metrics.