Diagnostics for web applications

Diagnostics apply artificial intelligence (AI) powered algorithms to generate correlation-based insights for the following properties:

  • Key Page

  • Entity

  • Network adapter type

The diagnostics are automatically enabled on the Page loads widget of the Speed dashboard and the Reliability dashboard. In the Reliability dashboard, Diagnostics analyzes the properties separately for each error type.

Diagnostics accepts filters to scope the issues according to the user-selected conditions. Please note that the Page loads filters (Good, Average, Frustrating) do not affect the diagnostics, as they need to compare slow and fast page loads to calculate the results.

Insights

Insights are available only if the system finds significant correlations. The significance is labeled with the following level of confidence: High, Medium, and Low. High confidence means a strong correlation exists between some values of the analyzed dimensions and the issue. Low confidence means that there is a weak correlation. If there is no correlation between the property and the issue, the insight is not available.

Each insight can be expanded, revealing the details of the analysis in a table.

Troubleshoot application slowness

Diagnostics considers a single page load time slow when it is above the second (highest) threshold, which is customizable in the application’s configuration. Refer to the Thresholds documentation for more information.

Troubleshoot application slowness insights in the Page loads tab contain the following:

  • A banner showing the summary KPI where the system calculates Employees with issues as the number of employees who experienced frustrating navigations at least 30% of the time.

  • Slow navigations: an insight with a total number of slow navigations broken down into:

    • Min page load time

    • Median page load time: the median page load time indicates that 50% of loaded pages were slower than the median value. This information allows Nexthink to detect situations where outliers impact the page load time when in reality, most of the page loads were quick.

    • Maximum page load time

    • Average page load time

  • Slowness ratio: the percentage of the number of slow page loads divided by the total number of page loads.

  • Employees with issues: the number of employees that experienced slow page loads at least 30% of the time during the analyzed time period. The system calculates the percentage of employees with slow performance only for the specific property value. It shows what percentage of those users experienced issues.

Troubleshoot application reliability

Troubleshoot application reliability insights show the following information for the Reliability tab:

  • A summary banner for which the system calculates Employees with issues as the number of employees who experienced frustrating navigations at least 30% of the time.

  • Severely impacted users: an insight with a total number of employees who experienced navigation errors at least 10% of the time in a given time period. This metric is broken down into:

    • Mean errors count

    • Max errors count

  • Error ratio - the percentage of the number of navigations with errors (unsuccessful page loads) / total number of page loads

  • Employees with errors - the number of employees that experienced at least one error during the analyzed time period. The system calculates the percentage of employees with errors only for the particular property value. It shows what percentage of those users experienced errors.

The value listed in the first row of the table is the one that correlates the most with the specific error type occurrence.


RELATED TASKS

Last updated