A synthetic monitor creates a problem when one of the following is true:
A monitored system is not available. In this case, the monitor reports an availability problem. The problem is created after a specified number of consecutive executions fail.
A monitored system is considered not available if:
A specified performance threshold is violated, which means that a monitored system is available but slow. In this case, the monitor reports a performance problem.
To set up outage handling and specify performance thresholds, go to Synthetic, select your monitor, and then select Edit.
Be sure to Save to monitor settings.
In the Outage handling section, on the Outage and performance tab, you can set up conditions for creating a problem. Two options are available:
Generate a problem and send an alert when the monitor is unavailable at all configured locations (global outage). You can specify how many consecutive executions must fail so the problem is generated.
Generate a problem and send an alert when the monitor is unavailable for one or more consecutive runs at any location (local outage). You can specify:
The option is available only when at least two locations are assigned.
For browser monitors, you can additionally select Automatic retry on error—it configures the monitor to avoid false positives by making one more execution after the first one failed. The error is thrown only after the second execution failed.
In the Performance thresholds section, on the Outage and performance tab, you can specify a performance threshold for a whole monitor. The monitor creates a problem when the sum of the durations of all steps violates this threshold.
Dynatrace generates a new problem if a synthetic monitor exceeds any of the total duration performance thresholds in three of the five most recent executions at a given location, unless there is an open maintenance window for the synthetic monitor. Multiple locations with three such violations can be included in a problem. The problem is closed if no performance threshold is violated in the five most recent executions at each of the previously affected locations.
On the Steps tab, you can specify performance thresholds for separate steps. The monitor creates a problem when at least one step with a specified threshold violates this threshold.
You can monitor problems in: