Header background

Elevate your dashboards with the new Dynatrace metrics framework

Dynatrace leverages high-fidelity data to fuel Davis®, our AI-driven causation engine for automatic monitoring insights. If you’re already using Davis as a foundation for custom drill-downs or dashboards to answer business-specific questions, you’ll be able to do this even more extensively with Dynatrace version 1.172.

We’ve introduced a new framework for metrics that provides (1) a logical tree structure, (2) globally unique metric keys that ease integration between multiple Dynatrace environments, and (3) more flexibility to extend Dynatrace so it better fits your specific business use cases.

Going forward, the new metrics framework will be at the core of everything that you can do with metrics in Dynatrace. Custom charting, the REST API, custom alerts, and many more features—everything now builds from a common understanding of what constitutes a customer-facing metric in Dynatrace. As an example of just how expansive this change is, the number of metrics available for custom charting has more than doubled.

To see the impact of the new metrics framework on the public REST API, be sure to check out my colleague Wolfgang Beer’s post Integrate Dynatrace more easily using the new Metrics REST API.

Read on to see how this enhancement allows you to get even more out of your dashboards.

Find metrics more quickly with metric categories

Providing a larger number of metrics wouldn’t necessarily be an improvement because more isn’t always better. So we’ve placed an emphasis on taking you more quickly to the metrics you’re looking for. First and foremost, we’ve added a category picker to metric selection so you can search for available metrics faster.

New metric categories

The new metrics framework provides more contextual information than just a flat list of categories. When opening the metric picker, once you choose a category, you’ll see the metrics structured in a logical hierarchy. As you can see below, we’ve tried to structure metrics meaningfully so you can quickly slice and dice the list of metrics.

Metrics structured in a logical hierarchy

Both the Category and Metric selectors support full-text search, which is a handy shortcut if you already know the name of the metric you’re looking for.

Dimension splitting for many more metrics to empower your reports

As with life in general, it’s often helpful to shift your perspective on a problem in order to see the best solution. Metric dimension splitting allows you to shift your perspective on how you view your metrics. And from now on, you’ll find many more metrics that you can split along multiple dimensions. The image below shows the Action count metric for web applications, split by Apdex category.

Metric splitting by dimension

Although this is just one example that shows off the user experience of such slicing and dicing of monitoring data, all our newly equipped metrics will empower you to build more elaborate charts for your reports.

Extension metrics are now first-class citizens

Dynatrace has long strived to integrate seamlessly with your existing IT landscape. Dynatrace OneAgent provides unrivaled capabilities that automatically pick up business-critical monitoring data without any configuration. This means that you can capture roughly 95% of what you need out of the box in order to monitor your applications efficiently. But what about the remaining 5%? With Dynatrace, you’ve always had the option to utilize one of many extensions that add to the value of our monitoring platform or to write simple scripts that send metric data using our REST API.

These extension metrics have received an upgrade in that they’re now indistinguishable from metrics that Dynatrace provides out of the box. The image below shows just such an extension metric for Apache Hadoop. As you can see, it looks just like every other metric in the list.

Selecting a OneAgent extension metric

By promoting extension metrics to first-class citizenship in Dynatrace, we’ve further opened up our platform for individualization by partners and customers to support long-tail use cases.

Limitations

  • Display names for metrics in existing charts will likely change, the underlying data won’t.
  • While most metrics have a 1:1 successor in the new framework, the rest will be mapped to the most reasonable alternative.
  • A small number of metrics had to be split into separate ones, it was necessary to chose one of the resulting metrics as the default mapping.

All the above only applies to custom charts on existing dashboards. Davis, custom alerts and product screens aren’t affected.

What’s next?

In support of more sophisticated and convenient data export for custom charts, we’ll bridge the gap by allowing the copying and pasting of cURL commands that invoke the public REST API and fetch metric data with the configuration of the current chart. With this approach, we’ll empower users to do more with Dynatrace, out of the box.