Metrics Explorer

Overview

Metrics Explorer provides instant visibility into all metrics flowing into your SigNoz deployment. This feature automatically discovers, categorises, and visualises every metric being ingested, giving you a complete view of your metrics ecosystem without needing to build dashboards first.

Unlike traditional monitoring workflows that require you to guess metric names from dropdowns, Metrics Explorer shows you exactly what's being ingested with real-time metadata, including types, sample counts, time series cardinality, and data sources.

When to Use & Key Benefits

Metrics Explorer delivers specific benefits in these common scenarios:

• During service onboarding: Instantly verify that new integrations (AWS, Kubernetes, custom exporters) are working correctly by seeing which metrics are being ingested and from which sources.

• When troubleshooting alerts: Quickly investigate alert-related metrics by jumping directly from alerts to see recent activity, metadata, and correlate with related metrics—all without building temporary dashboards.

• For cost optimization: Identify high-volume metrics, detect unused or misconfigured metrics, and spot instrumentation redundancies to optimize your observability pipeline.

• During incident response: Use the query playground to run ad hoc analyses, compare multiple metrics side-by-side, and correlate performance issues across different services and hosts.

How It Works

Metrics Explorer leverages your existing OpenTelemetry instrumentation to automatically collect and organize metrics metadata. The feature works by analyzing:

These attributes automatically categorise metrics, calculate cardinality and volume, identify data sources, and provide searchable metadata for quick discovery.

Feature Walkthrough

Summary View

The Metrics Explorer main interface provides two primary visualisation modes:

Proportion View

The Proportion View presents a treemap visualisation that shows the relative volume of metrics being ingested. Each rectangle represents a metric, with the size proportional to its ingestion volume. You can toggle between two modes using the dropdown in the top-right corner:

• Time Series: Shows the proportion based on the number of unique time series (unique metric name + label combinations) each metric produces. This helps identify metrics with high cardinality.
• Samples: Shows the proportion based on the total number of data points ingested. This helps identify metrics that generate the most data volume.

This view helps you:

• Identify high-volume metrics that consume the most ingestion resources
• Spot anomalies in metric volume distribution
• Understand proportional impact of different metrics on your system
• Quickly locate metrics that might need optimisation

List View

The List View displays all metrics in a comprehensive table format with the following columns:

• METRIC: The full metric identifier with search capability
• DESCRIPTION: Description of what the metric measures
• TYPE: Metric type indicator (GAUGE, HISTOGRAM, etc.)
• UNIT: Measurement units for the metric
• SAMPLES: Total number of data points received (e.g., "178.6K+")
• TIME SERIES: Number of unique metric + label combinations (e.g., "1.5K+")

Navigation and Filtering

You can navigate and filter metrics using:

• Search Filter: Use the search bar at the top to filter metrics by name or keywords
• Column Sorting: Click on the METRIC, SAMPLES, or TIME SERIES column headers to sort the list by that attribute
• Time Range Selection: Adjust the time window (shown as "Last 30 minutes" in the top right)
• Pagination: Navigate through large metric lists using the pagination controls at the bottom

Exploring Individual Metrics

When you click on a specific metric, the interface displays information in three distinct sections:

1. Basic Information Section

The top section displays essential metric details:

• SAMPLES: Shows the volume of data points (e.g., "40.8M+")
• TIME SERIES: Displays total and active time series count (e.g., "531 total — 466 active")
• LAST RECEIVED: Timestamp of the most recent data ingestion

2. Metadata Section

The Metadata section provides configurable information about the metric:

• Metric Type: Displayed as a colored badge (e.g., "GAUGE")
• Description: Human-readable description of the metric's purpose (editable)
• Unit: Measurement unit specification (e.g., {cpu})
• Temporality: Specification of how the metric data is aggregated
• Monotonic: Indicates whether the metric value only increases

Dashboard usage: If the metric is used in any dashboard, a badge (e.g., "1 dashboard") appears in the detail panel. Clicking it takes you directly to that dashboard. This makes it easy to see where a metric is being used across your monitoring setup. If the metric is not referenced by any dashboard, the badge is not shown.

Open in Explorer: Click the Open in Explorer button in the top-right corner of the metric detail panel to open the selected metric directly in the Explorer query interface with the metric pre-populated.

Inspect Metric: Next to the "Open in Explorer" button, click the Inspect Metric icon to open the Metrics Inspector. The inspector provides:

• Graph View / Table View toggle: Switch between a time-series chart and a raw data table
• Temporal Aggregation guidance: Helps you align the data by selecting a Temporal Aggregation
• Individual time series breakdown: See each unique time series for the metric plotted separately
• Data point details: Hover over the chart to see exact values, timestamps, and all associated labels for each time series
• Query Builder: Filter by specific attributes (e.g., k8s_node_name) and configure aggregation intervals

3. All Attributes Section

The "All Attributes" section displays the dimensional structure of the metric:

• Unique Identifiers: Shows the specific attribute values associated with the metric
• Search Functionality: Allows filtering through attribute values

Explorer Interface

The Explorer tab provides an interactive query building environment:

Query Configuration:

• Metric: Select the metric to query — the field auto-completes as you type
• Filter: Add a filter expression to narrow results by attribute values (e.g., service.name = 'frontend')
• Aggregate within time series: Choose the time aggregation function (Avg, Sum, Min, Max, Count, Rate, etc.) and the step interval
• Aggregate across time series: Choose the space aggregation function and optional group-by dimension
• Having / Order By / Limit / Legend format: Refine results with post-aggregation filters, sorting, row limits, and custom legend labels
• Add Formula: Combine multiple queries with mathematical expressions
• F(x): Apply functions like Cut Off Min, Log2, Time Shift, etc.

Actions:

• Run Query: Execute the configured query and render the chart
• Save this view: Preserve query configurations for reuse
• Create an Alert: Set up monitoring alerts based on the query
• Add to Dashboard: Include the visualization in existing dashboards

Saved Views in Metrics Explorer

Similar to the Saved Views feature available in Logs and Traces Explorer, Metrics Explorer also supports saving your query configurations for quick access later.

How to Use Saved Views

Saving a View:

1. Configure your metric query with desired filters and parameters in the Explorer interface
2. Click the "Save this view" button
3. Enter a descriptive label for easy identification
4. Confirm to save your view - a pop-up will confirm successful saving

Accessing Saved Views:

• Use the dropdown at the bottom of the Explorer interface
• Select any saved view to instantly apply the preset configuration
• Views are preserved across sessions for quick access

Managing Views:

• Update: Modify an existing view and click "Update this View" when changes are made
• Rename: Edit view labels in the Views tab
• Delete: Remove unused views from the Views tab

Use Cases for Saved Views

• Performance Monitoring: Save views for key metrics you check regularly during daily operations
• Incident Response: Quick access to critical metrics during outages and troubleshooting sessions
• Team Collaboration: Share common metric views across team members for consistent monitoring
• Cost Analysis: Monitor high-cardinality or high-volume metrics to optimise observability costs
• Routine Health Checks: Create daily dashboard-style views without rebuilding filters each time