Select a synthetic monitor in the Synthetic monitors page to open the Synthetic details page, which provides an overview of execution results, their visualization, and monitor properties. Powered by the Dynatrace AI engine Davis, the Synthetic details page shows you at-a-glance information and graphs, with ready filters to drill right down to problem details or the waterfall for a failing action in a specific execution.
Metrics and visualizations
In the top panel, you can choose between visualizations of your monitor's availability or performance—select the respective box. The performance view is customized to show a key performance metric along with Total duration.
Gray markers above the availability or performance timelines indicate maintenance windows. Select a marker to view details of the window.
Use the filter bar at the top of the page to filter all Synthetic details geographically—by Continent, Country, Region, and monitoring Location.
- You select a key performance metric in monitor settings in edit mode.
- Whether or not you see problems and receive alert notifications during maintenance windows depends on how you configure the maintenance windows.
- Maintenance windows may be excluded from availability calculations by applying a global setting.
Select Browse in the upper-right corner to access controls to Edit, Delete, or Disable the monitor. Select Analyze synthetic sessions to see the User sessions page. You can pin a Browser monitor tile for the monitor to a dashboard (Pin to dashboard) and, optionally, view the dashboard in edit mode.
Select the availability box on the left to view a visualization of the monitor's availability for the selected timeframe. If the monitor violates an availability threshold, global or local (defined in monitor settings), the box appears in red.
The graph shows availability for each monitor location. You can sort locations by Location name, Cloud, or Availability.
Time periods with availability outages are blocked in red. Hover over any of the bars to view availability at a given point in time for that location. Click and select Analyze to see the Multidimensional analysis page for availability filtered by that location. You can also select the location name; or select Analyze availability to see data points for all locations.
Select the performance box on the right to view a visualization of the monitor's performance for the selected timeframe. If the monitor violates a performance threshold, whether for the sum of all events or individual events, the box appears in red.
In addition to Total duration, you see average key performance metric (such as Visually complete) timings for load actions and XHR actions.
As Dynatrace captures a list of key performance metrics out of the box, you can switch your selection in monitor settings and immediately have historical data available.
For each browser monitor or clickpath, you can choose a different key performance metric for load actions and XHR actions in edit mode after monitor creation.
Key performance metric values are calculated as averages. Consider a login transaction consisting of three events resulting in one load action each:
- Initial page load (1 s Visually complete)
- Click on a login button (5 s Visually complete)
- Navigate to another page (3 s Visually complete)
The Visually complete timing for these load actions is 3 seconds while the Total duration could be 9 seconds.
Note that while Total duration is calculated as a summation, key performance metric values are averages, calculated separately for load actions and XHR actions.
If you check the Synthetic events and actions card, you can see that performance is only calculated for events with timings (that is, events that generate network activity). See Number of actions consumed by browser clickpaths for more information.
The graphs below show:
- Performance split by individual events (across all locations) for the timeframe.
- Performance split by location (for all events) for the timeframe.
Events with load actions and Events with XHR actions show key performance metric graphs for individual events that contain load actions and XHR actions, respectively. Note that an event can contain a combination of several actions of different types, for example, two load actions and one XHR action. All events shows the Total duration for all events, split by individual events (events with no network activity are not included in performance graphs by definition). The location graphs adjust to match your selection.
You can sort locations by Location name, Cloud, or performance metric.
Select an event in the legend to include/exclude it from the stacked line graph. Hover over the stacked graph to see the performance (key performance metric or Total duration) of individual events at a given point in time. Then click and select Analyze to see the corresponding analysis timeframe in the Multidimensional analysis page for performance, filtered by your selected action type.
Hover over any of the location graphs to view performance at a given point in time for that location. Then click and select Analyze to see the corresponding analysis timeframe in the Multidimensional analysis page for performance, filtered by your selected location. You can also select the location name; or select Analyze performance to see data points for all locations.
If there is a violation of the performance threshold for the Total duration of all events, a solid red line appears above the All events stacked graph for the duration of the problem; the threshold appears as a dotted red line. A solid red line also appears for locations violating the threshold.
For a violation of an event-specific thresholds, the Synthetic events and actions card highlights the events in violation, their duration, and the thresholds involved.
Reference screenshots shown in the top panel of the Synthetic details page are taken if the monitor runs successfully when the monitor is created or edited, and thereafter, once every 24 hours from a random monitor location (including private locations). Screenshots are taken at the end of each script event (even those without timings). You can scroll through screenshots here or select an image to scroll through enlarged versions.
Note that reference screenshots are always the most current, even when viewing Synthetic details for a historical timeframe.
When a monitor fails, screenshots for every failing execution (SCoE) are available on the Multidimensional analysis page.
To view screenshots from private browser monitors, your Synthetic-enabled ActiveGate or the browser you access Dynatrace from must be able to access the Amazon S3 service—ensure that your firewall allows connections to
*.s3-accelerate.amazonaws.com on port
Synthetic events and actions
The Synthetic events and actions card helps you distinguish between script events with and without timings. Use the Show events with timings only toggle to view only events that captured actions or all events. For events with timings, you see the average action timing of your selected key performance metric, that is, if an event has more than one associated action, the average timing of those actions is the event timing.
Events are all the interactions you record and are therefore simulated during the playback of a browser clickpath. Not all events necessarily trigger network requests, for example, clicking an input field or entering text into a form. These events are important from a functional perspective but don’t generate any performance data. Events with timings, or events that trigger actions, are the basis of performance data.
If you expand an event from the event list, you can compare all key performance metrics in one chart. Click a metric in the legend to deselect/select it from the action performance graph. Select Edit clickpath to go to monitor settings from this card.
Dynatrace uses exactly the same technology to capture real user and synthetic data, so you can validate your synthetic results against real users easily. When a browser monitor has an associated application, Dynatrace automatically shows the corresponding real user action for each synthetic event with timings—select the Contributing load actions link. You are taken to the user action page, where you can filter for only synthetic users. Note that you can also see real user locations on the World map.
When an event is in violation of its performance threshold, it is highlighted. Expand the event to see the event timing and the threshold violated. A solid red line appears above the stacked graph for the duration of the problem; the threshold appears as a dotted red line.
The Problems card shows performance (threshold violation) and availability (local or global outage) problems when you enable the respective thresholds in monitor settings. Expand the card to see active as well as resolved problems for the selected timeframe.
See Configure browser monitors for information on how to define performance and availability thresholds. See Synthetic calculations for how availability and performance are calculated and how problems are generated and dismissed. See the Synthetic alerting overview for alerting workflow and concepts, including setting up notification profiles and templates.
There are three main problem types for browser monitors:
- Global outage (availability)
- Local outage (availability)
- Performance threshold violation (performance)
Select a problem to view the problem overview page. Note that performance problems may combine threshold violations at the monitor as well as the event levels. Drill further into the problem to see Synthetic details filtered by the problem duration.
The Events card shows all events that can lead to problems. Custom events you ingest into Dynatrace also show up in this card. Events for active as well as resolved problems show up in the list and timeline.
Hover over a time slot in the event timeline to see the type and number of events generated in that interval. Click to select the time slot and display the events in it.
Select an event type, for example, Synthetic slowdown, to see the list of events. There is always one slowdown event created per location where your monitor violates event- or monitor-level performance thresholds. Select an individual event to see details.
The Errors card displays the timeline of all failed executions of the monitor with the corresponding error codes, providing an easy way to quickly find the main reason for monitor failures.
A monitor failure counts towards availability outages (global or local threshold violations). You might see errors with no problems if you haven't enabled availability thresholds or if the errors don't cause a violation of your availability thresholds. A performance threshold violation is not necessarily an error unless the run also fails.
Select Analyze errors at the bottom of the card to view these error runs in the Multidimensional analysis page. Select Analyze next to an individual error to view Multidimensional analysis filtered by that error.
Hover over any time slot in the error timeline to see the count of errors by type during that interval. Select Analyze errors in the tooltip to view the Multidimensional analysis page showing that time slot.
The Properties card summarizes key properties of your monitor and displays the number of events and the consumption for the selected timeframe. For single-URL browser monitors, the number of events is always 1.
The World map shows if your monitoring locations are online or have an outage, helping you distinguish between global and local outages. Assign monitor to application (see the first image below) takes you to monitor settings where you can associate the monitor with one of your monitored web applications.
If Real User Monitoring (RUM) is enabled for the applications your synthetic monitor runs against, Dynatrace automatically links the RUM applications to the monitor, and the Monitored applications card is displayed (see the second image below). You can see the key metrics of the application and can jump directly to RUM data from here.
If a RUM application is linked to the monitor, you’ll see a toggle to augment the world map with RUM data. When there's an outage at a synthetic location, you can immediately compare the RUM traffic. The RUM data is also a good indicator of which other locations you should run your synthetic monitor from.