App performance overview. A stacked area chart showing latency activity at each layer of the application. The user can filter on time, URL, Domain, area of the stack, errors, and alerts. The context of the throughput is available as a sparkling. Selecting a layer of the stack showed individual traces allowing the users to pinpoint performance issues. This view also shows cached ram usage overtime for this app.

Request volume and application errors as a bar chart. Selectable so the user can find the exact error. Errors can be viewed in the app layers or the client side. Each error showed the layer it occurred in and the corresponding traces for investigation.

Each app overview was customizable to show the metric important to that particular server or cluster. From this view the user could create new apps in order to segment requests to their stack into different environments such as elopement, staging, and production.

Real User Monitory (RUM) broken down by browser. Average pageload times could be shown as a visualization or the raw list of traces. RUM monitoring was available by average latency, geographic, or by browser.

Geographic breakdown of real user performance monitoring. Showing average page load time by country.

Anatomy of a trace time line. Time and other statistics as a request moves through the application stack.

A heatmap of requests showing traces improving in latency as CPU usage declines in a dramatic way. This is juxtaposed with request volume for context and filtering.

A heatmap showing selecting a group of traces for investigation

Application layer breakdown of latency with call volume as a filtering context.