Search and explore data with the Metric Explorer

Use Metric Explorer to discover, validate, and visualize the data available in your environment.

Metric Explorer allows you to search for data based on the specific metrics you want to measure or by the entities (systems, services, and resources) that emit them. You can explore metrics and entities by categories and then filter them by dimensions such as deployment environments, cloud providers, services, namespaces, and more.

This feature helps centralized monitoring teams and SREs identify the correct data for their monitoring assets, even when naming conventions vary across different services or development teams.

How to use entities in the Metric Explorer

Entities represent the foundational sources of telemetry data, providing a resource-centric approach to discovering and managing metrics within the Metric Explorer.

What are entities?

In the Metric Explorer, entities are the physical or logical building blocks of an infrastructure or application. These include systems, services, and resources such as Kubernetes pods or specific microservices. While metrics provide the specific measurements of performance, entities identify the specific sources from which that data originates.

Importance of an entity-centric model

Focusing on entities aligns the platform with the natural mental model of users who need to ensure the health of a specific service or resource. This approach addresses several challenges inherent in large-scale enterprise environments:

  • Inconsistent naming standards across different teams or integrations, such as variations between AWS CloudWatch and OpenTelemetry naming conventions.
  • The loss of domain knowledge about service-specific metrics as environments scale.
  • Ensuring that a detector covers all relevant resources or only a subset with a specific naming pattern.
Understanding a few key terms related to entities and used in the Metric Explorer will help you navigate the tool and find the metrics or entities you're looking for.
Entity type
A classification or category used to define the specific nature of a system, service, or resource being monitored within the observability platform. Examples include Host, Kubernetes (with sub-types like Pod), and APM (with sub-types like Service). The entity type allows users to group and filter their infrastructure based on the role or technology of the resource.
Dimensions
Key-value pairs associated with metrics or entities that provide essential context. Dimensions allow users to categorize, filter, and group data for analysis. For example, a metric might have dimensions like region, host.name, or environment. In the Metric Explorer, users can refine search results and view the cardinality (the number of unique values) for each dimension to better understand their data.
Population percentage
A metric used to indicate the coverage of a specific metric across a selected group of entities. The population percentage represents the ratio of entities that are successfully emitting a particular metric compared to the total number of selected entities.

Benefits of using entities with the Metric Explorer

The integration of the Entity Platform into the Metric Explorer provides several advantages:

  • Metric search and discovery: Instead of remembering exact metric names that are often duplicated or similar across platforms and environments, you can start searching through categorized entities such as service, pod, or container, and then drill down to specific metrics or discover new ones.
  • Monitoring coverage: You can view the population percentage for metrics across selected entities. This enables you to instantly identify what types and how many entities are reporting critical metrics.
  • Data validity through visualization: You can quickly plot metrics on a time series chart to validate assumptions and avoid false positives or negatives before creating charts or detectors.

Overview of the Metric Explorer

Understanding the context of a metric is critical for selecting the right data source.

Before you start your search of metric and deep dive into entities, let's navigate through Metric Explorer and learn how to best and most efficiently use the tool.

The Metric Explorer is divided into two primary modes: Metrics and Entities. Users can toggle between these tabs to change the primary search focus from individual metrics to system-level resources. From either tab, you can search (Search field), filter (Add filters), select time ranges for both metrics and entities, and group by a dimension (Group by).
Figure 1.
Metric Explorer page on the Metrics tab. We have a filter for a container and app, and we are grouping the metrics by pods.

Access the Metric Explorer

To get to the Metric Explorer, navigate to Metrics > Metric Explorer. From there, you can select the Metrics or Entities tabs.

To get started, take a look at the following:

  1. Basic usage and functionality

  2. Find metrics by entity

  3. Find metrics by name

  4. Group metrics in the Metric Explorer

  5. Create charts and detectors

Basic usage and functionality

This table provides instructions to perform basic functionality.
Function Instruction
Select mode Select the Metrics or Entities tab based on your investigation goal.
Search Enter the terms in the search bar to filter entities by name.
Filter

  • Select Add filters to narrow results by specific dimensions.

  • Select from Custom Categories to add the category as a dimension filter.
Validate Select Apply or Search to update the results table.
Visualize Select metrics to plot them on the Metrics clipboard for immediate visualization and validation.
Create charts or detectors After plotting metrics to the Metrics clipboard, select Create detector or Create chart to help you monitor changes.

Find metrics by name

Search the catalog directly if you know the name or partial name of the data you need.

If you know the name or partial name of the data you need, use the Metrics tab to search the catalog directly.

  1. Navigate to Metric Explorer.
  2. Select the Metrics tab.
  3. Enter a search term in the search bar.
    You can search by:
    • Metric name (e.g., cpu.utilization).
    • Dimension key or value (e.g., us-east-1).
    • Description text.
  4. Refine your results using Add filters.
    Hover over a dimension value and select it to refine the list.
  5. Select Apply to view the filtered results.
  6. Select a metric to view its details in the Details drawer.

Find metrics by entity

Scope your search to a specific resource to view all metrics associated with it.

If you know which system or service you want to monitor but are unsure of the specific metric names, use the Entities tab. This workflow allows you to scope your search to a specific resource, such as a Kubernetes Pod.

  1. Navigate to Metric Explorer.
  2. Select the Entities tab.
  3. Use the filter bar to define the scope of your entities.
    You can filter by:
    • Entity Category: (e.g., Infrastructure, Kubernetes, APM).
    • Entity Type: (e.g., Service Instance, Pod).
    • Dimensions/Tags: Specific attributes such as region, env, or service_name.
  4. Review the list of entities that match your filters.
  5. Select one or more entities from the list.
    • The Details drawer updates to show only the metrics emitted by the selected entities.
    • Population: Review the coverage column to see how many of the selected entities report a specific metric (e.g., "100% (5/5)" indicates all selected hosts report this metric).
      Note: The population is only shown when more than one entity is selected to view metrics
  6. Select one or more metric names to view the metadata and view it on the Metrics clipboard.

Group metrics or entities

Group by type, metric name, or entity by specific labels or dimensions to aggregate data and identify trends across different resources in the Metric Explorer.

You can group metrics or entities based on dimensions or labels to understand how they are distributed to draw comparisons and identify differences. For example, you could group a general CPU utilization metric by instance_type to compare performance across different hardware configurations. When you group metrics, you typically apply an aggregation function to combine the data points within each distinct group.

  1. From the Metrics or Entity tab, select a filter such as app and select all possible values.
  2. From the Group By field, select a dimension such as host.
    Note: For entities, is also possible to group by Type (entity type).
  3. Expand the grouped metrics by your chosen filter and notice the number of entity and entity types for each metric as well as the dimensions from one group to the next.
  4. Select a metric to view the Details drawer to view the information such as the Utility score that shows the number of charts created with the metric.
  5. From the Details drawer, you can navigate to the View detailed usage analytics to view active charts.
  6. Check the checkbox next to one or more metrics to plot the metric on the Metric clipboard.
  7. You can create a permanent chart by selecting Create chart or a detector by selecting Create detector.

The Metric Explorer enables you to group entities and metrics and then compare them across dimensions and labels, visualize that comparison, understand the usage, dimensions, as well as create charts and detectors.

Validate data with the Metric clipboard

Before creating a chart or detector, use the Metrics clipboard to verify that the selected metrics contain the expected data.

Use the Metrics clipboard to ensure your data is reporting correctly before proceeding with configuration.

  1. In the results table, select the checkboxes next to the metrics you want to verify.
    The selected metrics are automatically plotted on the timeline at the bottom of the screen.
  2. Adjust the time window using the time picker (for example, Last 24 hours).
    This helps identify metrics that have stopped reporting or entities that have recently become inactive.
  3. Optional: To view metrics no longer reporting data, toggle Show Inactive.

Metrics Clipboard Reference

Visualize Selection
As you select checkboxes next to metrics in the results table, they are automatically plotted on the timeline at the bottom of the screen.
Time Range
The time picker adjusts the window...
Metric State
By default, the explorer shows active metrics...

Create charts and detectors

Transition directly from exploring metrics to creating monitoring assets.

Once you have identified and validated your metrics, you can transition directly to creating monitoring assets.

  1. Select the metrics you wish to use from the results table.
  2. Ensure the Metrics clipboard reflects the data you expect.
  3. Select Create chart or Create detector.
    • Create chart: Opens the Chart Builder with the selected metrics and filters pre-populated.
    • Create detector: Opens the Detector Builder to configure alerts based on the selected metrics.

Metric Explorer reference

The Metric Explorer is a data visualization tool used to investigate, filter, and validate metric data before using it in charts or detectors.

Use the following sections to understand metrics and the syntax required to perform advanced searches within the Metric Explorer.

Metric details

This reference provides detailed information about various metrics, including their types, statuses, sources, and utility scores.

Metrics details view

You can select a metric name to view the details shown below.

Description
A human-readable explanation of what the metric measures.
Metric type
Defines the nature of the data being reported. In your case, "Gauge" is a standard metric type used to represent a value at a specific point in time. See Metric types for more information.
Metric status
Indicates whether the metric is currently "Active" (sending data) or "Inactive" (no longer actively sending data).
Source
The origin of the data (e.g., OpenTelemetry, AWS CloudWatch, Azure, API).
Receiver
The specific collector configuration used (e.g., hostmetrics, nginx, otlp).
Units
The unit of measure (e.g., Bytes, Percent, Seconds).
Entity coverage
The coverage statistics represent the scope of your infrastructure reporting this specific metric that includes the entity types and number of entities.
Population
This indicates the ratio of entities that are successfully emitting the metric compared to the total number of selected entities. This is useful for identifying if a metric is reported by all hosts in a cluster or only a subset.
Utility Score
This metric quantifies how heavily a specific metric is utilized within your organization. The score also shows how many detectors and charts are using the metric.
  • R4: Referenced by detectors (highest priority).
  • R3: Referenced by actively viewed charts.
  • R2: Referenced by API queries or unsaved charts.
  • R1: Referenced by charts that are not actively viewed.
  • R0: Not referenced at all.
Tags and Dimensions
These are key-value pairs associated with metrics or entities that provide essential context. They allow you to categorize, filter, and group data for analysis (e.g., filtering by region, host.name, or environment).

Search guidance and syntax

Metric Explorer supports advanced search syntax to help you refine your results using Boolean operators and wildcards.

Syntax Description Example
AND Limits results to items that match both terms. cpu AND region:us-east-1
OR Expands results to items that match either term. env:prod OR env:stage
= or != Include or exclude items from the results. app != ai-assistant

aws_architecture = arm64

* (Wildcard) Matches any character sequence. service:payment-*
: (Key-Value) Searches for a specific dimension key and value. cloud.provider:aws

Metric Explorer use case examples

The use case examples show you how you would use Metric Explorer to investigate and troubleshoot common problems.

The following use case examples will show you how to use the Metric Explorer to perform common tasks, improve your monitoring, and troubleshoot. For the best results, see the Overview of the Metric Explorer and learn its basic functionality.

View metrics for a specific entity

Learn how to isolate and view performance metrics for specific services, hosts, or Kubernetes pods.

Use this procedure when you need to troubleshoot a specific component, such as a slow-performing checkout service or a Kubernetes pod that is failing to restart. This helps you isolate metrics to that specific entity.
  1. Navigate to the Entities tab.
  2. Locate the entity using the search bar, dimension filters, or by selecting the Entity Type.
  3. Click the entity name to view its specific reported metrics.

Categorize metrics by AWS namespace

Group and analyze infrastructure metrics by AWS namespace for auditing and cost management.

Use this when you need to perform resource auditing or cost-allocation analysis across your AWS infrastructure. Grouping by namespace allows you to see how different AWS services are performing relative to each other.
  1. Navigate to the Metrics tab.
  2. Click Add filters, select Namespace, and apply the filter.
  3. Select Group by and choose namespace to organize the data.

Create complex filters

Use Boolean logic and filters to isolate specific subsets of data in your environment.

Use this when you need to monitor a specific subset of your environment, such as "all production services in Namespace A" or "nodes assigned to the SRE support group." This is essential for multi-tenant environments where you only want to view data relevant to a specific team or project.
  1. Use the filter bar to drag and drop your required conditions.
  2. Toggle the Boolean logic between AND and OR to define the relationship between your filters.

Analyze entities by availability zone

Break down entity performance by cloud availability zone to verify high-availability configurations.

Use this to verify high-availability configurations. If you suspect an outage in one region, this view helps you confirm if your entities are distributed correctly across availability zones and if any specific zone is reporting errors.
  1. Navigate to the Entities tab.
  2. Filter by your specific environment (e.g., sf_environment).
  3. Click Group By and select cloud.availability_zone.

Verify data coverage

Identify data gaps and ensure full coverage across your entities before creating alerts.

Use this before creating a new alert or detector. By checking the Population tab, you ensure that all relevant entities are actually reporting data, which prevents "false negatives" where an alert fails to trigger simply because data is missing.
  1. Search for and select your target entities in the Entities tab.
  2. Click Show Metrics and navigate to the Population tab.
  3. Review the Population score to identify any entities failing to report the metric.

Access metric metadata

Retrieve detailed descriptions and metadata for metrics to better understand their purpose and units.

Use this when you are unsure what a specific metric measures or how it is calculated. This is helpful for team members onboarding to a new service or when interpreting unfamiliar custom metrics.
  1. Click on the metric name to open the details view.
  2. Review the description, metric type, and units to understand the data context.

Build charts and detectors with the Metric clipboard

Use the Metric clipboard to collect multiple metrics across different searches for building charts and detectors.

Use this when you are building a dashboard or an alert policy that requires multiple metrics. The clipboard allows you to collect metrics from different searches and contexts, ensuring you don't lose your progress while navigating the UI.
  1. Select metrics using the checkbox to add them to the Metric clipboard.
  2. Navigate through the application as needed; the clipboard will persist your selections.
  3. Click Create chart or Create detector from the clipboard panel when ready.