New dashboard elements: heatmaps, pie charts, monitor lists, and notes

From the start, we’ve focused on making dashboards not just easy to create, but also powerful and insightful—helping you monitor your data in meaningful ways. Our latest update introduces several highly-requested elements that enhance visibility and accelerate problem resolution, all while being simple to set up. Take a look!

Heatmaps

Averages and other aggregations often hide issues like outliers or bimodal distributions. These are critical when investigating response times, latency, or other metrics affected by slowdowns or behavior shifts. Heatmaps offer a visual representation of data distributions over time, where each bucket aggregates the count of data points within specific ranges, providing an immediate, comprehensive view of distribution patterns and outliers.

For instance, a heatmap of request durations over time can reveal that while your p99 latency remains stable, and average latency has only increased slightly, certain users experience intermittent slowdowns, perhaps due to cache misses. Averages or percentile views can mask these patterns. A heatmap highlights these sporadic delays, allowing teams to see shifts in user experience and detect issues that might otherwise go unnoticed.

In another example, a heatmap of memory usage across services or instances shows memory distribution and trends in usage levels over time. Average memory consumption might look steady on a timeline plot, but a heatmap could reveal that certain services occasionally spike close to their memory limits, potentially causing crashes or performance degradation under specific workloads that could be investigated. Insights like these let teams proactively identify and address risks before they impact overall system stability.

└ Heatmap created with Axiom’s query builder.

└ Heatmap created from an APL query.

Pie charts

A pie chart help you quickly understand distributions in your event data, such as the relative number of different status codes in HTTP logs. By visualizing critical metrics, your team can detect and resolve issues faster, reducing downtime and improving service reliability. Whether you use our intuitive query builder or craft advanced APL queries, creating powerful visualizations takes seconds. See our docs page for instructions.

Pie charts can show obvious metrics, such as traffic distribution by microservice or application, to help better allocate resources. They can also spot surprises. A pie chart of segment errors by type — 5xx, 4xx, connection refused, timeout — could alert you to a network connectivity issue that had seemed to manifest itself as an application bug.

└ Pie chart created with Axiom’s query builder.

└ Pie chart created from an APL query.

Monitor list

This new element gives a visual overview of your monitors in a single table, making it easier to track the status of key systems. Customize the element to show the information that matters most—status, type, dataset, history, or notifiers. See our docs page for instructions.

It’s convenient to be able to check all monitors with one glance, but seeing them together can also lead to new insights brought by correlating pattens across seemingly unrelated services, revealing a hidden dependency chain. A cache-warming event in the catalog service might trigger a cascade of memory pressure through the auth service's token validation, causing the checkout service to spawn excess DB connections as a compensation mechanism. The ripple effect would be impossible to miss in a monitor list.

└ The new monitor list as it appears on a dashboard.

Notes

To add helpful text anywhere on the dashboard, you can now add a note formatted in plain text or GitHub-flavored markdown. See our docs page for instructions.

Notes can alert other dashboard users to things they need to know, e.g. “Connection pool is intentionally oversized until April 15th migration from legacy Oracle RAC.” They can also serve as a institutional memory: “Payment success rate 0% is due to an expired TLS cert on the payment gateway. Time to recovery 45 minutes. See #127 in automation repo.”

└ Dashboard with a note at the top, which includes a link to relevant content.

What do you think? What other dashboard elements would help you do more with Axiom? Talk to us anytime on our Discord server.