Analyzing a synthetic transaction

For applications with synthetic transaction monitoring enabled, you as the application operator can identify and diagnose application issues by examining the behavior of a single transaction and the locations where it ran. You can use the transaction details to help identify the source of application issues.

This topic contains the following information:

Before you begin

Before you can examine the behavior of a single transaction and the locations where it ran, ensure that your system meets the following conditions:

To access the Synthetic Health tab, you must be logged in with Application Operator-level access, or higher.
(Optional) Synthetic metric rules are defined, usually by an application specialist in your organization.
If you do not define your own synthetic metric rules, only the global metric rules for performance, availability, accuracy, and execution events are evaluated.

To examine the synthetic health of a transaction

You can examine a single synthetic transaction and its behavior in different locations and its threshold violations, performance graph, and relevant executions over 24 hours.

From the navigation pane in the TrueSight console, select Monitoring > Applications.
Select one application that you want to examine in detail.
From the Synthetic Health tab, click a transaction.

The Synthetic Health > Analysis view is displayed.

Evaluating the sections of the Synthetic Health > Analysis view

The Synthetic Health > Analysis view of the Application Monitoring page displays the status of a selected transaction over 24 hours using a series of tiles. Each tile represents one hour so you can see the hour-by-hour status of your transaction as a whole, and the status of the transaction in each location where it ran.

Examine the transaction health in the following sections:

Transaction—The Transaction section, at the top of the Synthetic Health > Analysis view, displays the most severe event status for every hour, during a 24-hour period.
To the right of the transaction tiles is displayed the current number and severity of open events for the transaction.
Locations—The Locations section displays each location where the transaction ran during the selected 24-hour period. Each tile represents the most severe violation for that hour.
To the right of the location tiles you can see the percentage of transaction executions that were impacted by at least one metric breach. Next to the percentage is the ratio of impacted execution to all executions during the 24 hours.
Locations are ordered by the highest violation percentage to the lowest.

The tabbed pages near the lower part screen display data according to the selected locations.

Violations—Displays a list of violations
Performance Graph—Displays a a line graph of performance metrics
Executions—Displays details of each transaction execution

What is the difference between an event and a violation?

A violation occurs when the value of a metric breaches the defined minor or critical threshold (as defined in the rule).

An event is generated based on all the definitions in a synthetic metric rule:

The metrics breached the defined threshold (according to the AND or OR logical operator).
The breaches occurred over the defined number of locations.
The breaches occurred at the defined violation frequency.

Example of the Synthetic Health > Analysis view

To filter locations and corresponding violations

Locations are ordered by the highest violation percentage. You can filter locations and the filters are reflected on the Violation tab.

You can perform the following actions to filter locations:

Show or hide locations with no violations—By default, all locations are displayed, including those that had no violations. Clear the Show locations with no violations check box to hide locations with no violations.
Filter locations by violation type—The filters buttons above the Locations section display the number of locations that have the indicated violation type: Availability, Performance & Timers (which includes custom timers), Accuracy, and Execution. Click a button to filter locations with that type of violation, or filter that type of violation from locations with more than one violation type.

Filter buttons for the Locations section and the Violations tab
Filter specific locations—Select one or more locations to further filter the Violations tab.

The locations are ordered by the most impacted according to the violation frequency ratio: impactedViolations:totalViolations. When all else is equal, the locations are in alphabetical order.

The first four locations are displayed; click Show more to examine the rest. When all the locations are displayed, click Show less to collapse down to four locations.

To examine transaction violations, performance graph, and executions

Below the tiles, data is displayed in tabbed pages. The data reflects the entire 24 hours if no hour is selected, or the selected hour.

In the Violations tab you see data about metrics that breached defined thresholds in selected locations. By default, the metrics are sorted by the time that the transaction ends. To sort the data, click a column heading in the table.

Description of the Violations table

Column	Description
Action menu	For each violation, you can perform the following actions: Download Execution Log —Downloads the Execution Log, if available Custom Timers—Displays all custom timers defined for this transaction and their latency Page Timers—Displays all page timers for the transaction and their latency Error Messages—Displays all error messages received while running the transaction. The error messages are sorted by category (Availability, Accuracy, and Execution).
Severity	Indicates if the violation is Minor or Critical
Location	Name of the location where the violation occurred
Metric	Name of the measured metric For predefined metrics, the name of the metric is displayed. It can be one of the following: Availability Performance Accuracy Execution For timers, the name of the timer, as defined in your script is displayed.
Threshold	Defined threshold for the metric
Value	Actual value that breached the threshold
Evaluation Start Time	Time that the execution cycle started An execution cycle comprises a single run of an Execution Plan in all of its defined locations.
Transaction End Time	Time that the transaction ended
Metric Rule	Name of the metric rule that was breached

The Performance Graph tab shows the details of performance metrics for the selected transaction during the selected time period. One line is shown for each metric-location combination. By default, the graph shows the overall Performance metric of the transaction, for all selected locations. You can add any timers that are defined for the transaction as well.

Note

The graph can show a maximum of 10 metric-location combinations simultaneously. This means that the number of selected locations multiplied by the number of selected metrics must be equal to or less than 10.

For example, if you select 6 locations and 10 metrics, there are 60 combinations, and they cannot be displayed on the graph. If you select 2 locations and 5 metrics, there are 10 combinations.

Performance Graph Pearl.png

Select a different tile to see the performance graph for that time period.

You can choose not to display any of the metric-location combinations by clicking on the appropriate caption in the legend, and you can click them again to add them back to the graph. The graph adjusts automatically to show data for the selected metrics.

Place the mouse over any line on the graph to see a tooltip displaying the value of the metric at that specific time.

Drag the cursor across the graph to zoom-in to a smaller time range and see the results at a more granular level.

Performance Graph Selection Pearl.png

Click Show all in the upper-right corner to return to seeing the graph for the full time range.

Note

If you are looking at a dataset ranging less than 12 hours, each data point on the graph represents the actual metric value at that specific time. However, if you are looking at a dataset ranging more than 12 hours, each data point on the graph represents an aggregated metric value, grouped by buckets of five minutes each on an average.

To export a performance graph

Click the export icon in the upper right corner to export a graphic of the performance graph or the performance data used to generate the graph.

The following options are available:

Menu option	Description	File formats
Download as	Save the graph as a graphic file	PNG JPG SVG PDF*
Save as	Save the performance data as a data file	CSV XLSX JSON
Annotate	Add your own markings to the graph before downloading or printing it. See To annotate the performance graphfor more details.
Print	Print the graph.

^{*In Chrome and Safari on iOS 10.1.1, the performance graphs are not visible in PDF format.}

To annotate the performance graph

You can add your own markings to the graph before downloading or printing it.

If you select Annotate, the menu changes:

Click and drag the mouse cursor on the graph to add lines and markings.
Select Add to add text or shapes to the graph.
Note
To move or resize a shape, double-click the shape.
Select Change to change aspects of the markings, shapes, and text that you add to the graph, such as color or opacity.
Note
The change effects only the selected annotation element, and new elements added after the change takes effect.
Select Download as to download the graph as a graphic file. The same file formats as listed in the table are supported.

In the Executions tab you see data about executions of your selected transaction.

By default, executions are sorted by start time. To find the information you need, you can sort by any of the columns in the table, or filter the columns by entering text in the column filters.

Notes

There may be Executions listed here that are not yet included in the calculation of the Synthetic Health. This is due to a necessary data-processing delay.

The Executions tab is limited to the last 2000 executions.

The following details are displayed in the Executions tab:

Column	Description
Action menu	For each execution, you can perform the following actions: Download Execution Log —Downloads the Execution Log, if available Custom Timers—Displays all custom timers defined for this transaction and their latency Page Timers—Displays all page timers for the transaction and their latency Error Messages—Displays all error messages received while running the transaction. The error messages are sorted by category (Availability, Accuracy, and Execution).
Transaction Start Time	Start time of the transaction
Location	Location of the TEA Agent where the transaction execution ran
E2E Time	Amount of time (in milliseconds) the transaction took for completion from end to end
Availability Errors	Number of availability errors in the transaction Availability errors occur when the application does not respond, or when the application responds with an error that indicates it is unavailable.
Accuracy Errors	Number of accuracy errors in the transaction Accuracy is calculated with the assumption that monitored applications are working as designed and that the information transmitted to clients is correct. Functions that can be evaluated to determine accuracy include link checking, content validation, title validation, and response data verification.
Execution Errors	Number of execution errors in the transaction (this number is unlikely to exceed 1) An execution error indicates that there were problems with the script execution. It does not indicate anything about the health of the application.
Execution Start Time	Start time of this execution

Where to go from here

Analyze synthetic health event details.

Investigate application issues reported by synthetic health events.

If necessary, adjust synthetic metric rules to reflect better the critical data you need for your environment.