Highlights in Real User Monitoring

Highlights is a powerful tool designed to help teams quickly understand how data is distributed across different attributes. It makes it easier to identify commonalities and outliers. The feature visualizes the relationships between data points using specific labels, enabling rapid issue detection and resolution.

For example, if system response times are rising, Highlights can accelerate your investigation by pinpointing errors concentrated within key attributes—such as user group, browser, or geographic region—and identifying outliers. Comparing these data points helps reveal the root cause as you troubleshoot.

With Highlights, you can:

• Pinpoint the root cause of slow API performance
• Identify which service or dependency is contributing to slowdowns or errors
• Uncover systemic issues in distributed systems

Jumpstart your investigation

Highlights provides insights into data distribution for key attributes, helping you narrow down potential issues. It’s available in both Explore and Real User Monitoring (RUM).

Label distribution

• Data is grouped into charts based on labels like application, user_id, browser, geolocation, or http_route.
• View full distributions by selecting a label from the drop-down menu or hovering over it for a mini-distribution tooltip.

Action menu

Click a label value to open the action menu:

• Include in query: Add the value to the query filter.
• Exclude from query: Remove the value from the query filter.
• Copy to clipboard: Copy the value.

Using Highlights in RUM

Access Highlights in Error Templates, User Sessions, and Error Analytics.

Error Templates Example

Step 1: Click the Highlights icon for the error template you want to investigate.

Step 2: The Highlights modal will open, displaying two key insights:

• A graph of error evolution over time, showing when error counts spiked or stabilized.
• A label distribution chart, grouping errors by attributes like application version, operating system, or geographic location.

For instance, the error distribution in the example above reveals that most errors stem from specific application versions and operating systems. By analyzing these outliers, you can quickly identify patterns, such as problematic deployment or compatibility issues. From here, you could refine your query to zoom in on the problematic versions or OS to further diagnose the root cause.

Tips for best results:

• Use smaller time windows to reduce variance and focus on meaningful patterns.
• Refine results by applying filters in the sidebar or using the filter command in DataPrime.