Anomalies - Understanding anomaly warnings and root cause

Performance problems are often the result of your code, configuration changes, upgrades, issues with data input and external systems.

AIMS uses its out-of-the-box self-learning capabilities, allowing AIMS to learn normal behavior of your infrastructure, applications and services and enabling it to detect anomalies across all components and performance parameters (in thousands of performance counters). AIMS identifies where anomalies occur first and cross-correlated them with other anomalies to identify cause and effect.

This information may identify that the root cause of the critical performance issue, allowing you to efficiently troubleshoot. Use anomaly detection insights as an early warning system and be able to prevent problems instead of solving them.

Understanding anomaly causes

1. From your selected environment, go to ‘Events’ in the top menu bar.

2. Search for anomalies. Anomalies can be filtered by:

  • Date range; ‘All’, ‘Today’, ‘Yesterday’, ‘Last 7 days’, ‘Last 30 days’ or a ‘Specific date range’,
  • Event level; An anomaly is specified as a Warning,
  • Event status; An anomaly can be ‘Active’, ‘Resolved’ or ‘Ignored’,
  • Node type; The anomalies can be filtered per BizTalk node.

3. Select an anomaly by clicking on the Anomaly detected warning. A detailed anomaly page will appear, showing all different parameters associated with the anomaly.

4. Go to ‘Timeline bars’, which shows the timeline of the anomaly: When did it first start and when was it triggered. It contains a lot of individual deviations, giving a complete view based on time on when the different deviations started, how long it took and when something was impacted by the different deviations.

5. Sort deviations by the different sections as seen in the header. It gives you a different view on the situation and shows how it escalated. Deviations can be sorted by:

  • Node,
  • Deviation,
  • Started descending or ascending,
  • Ended descending or ascending, and
  • Duration descending or ascending.

6. From the Anomaly page, go to ‘See in analytics’ to select parameters from an anomaly warning and generate a chart in Analytics.

7. Create a chart in Analytics by selecting parameters (node name, Deviation, Started/Ended or duration).

Enabling the ‘Select all’ button selects all parameters.

When clicking the ‘Select’ button, a popup appears asking you to create a new report.

All selected parameters will be shown in one chart.

Use ‘Multiple chart’ to create a chart for each selected deviation parameter.

Filtering and sorting in Anomalies can be used for easier root cause analysis and troubleshooting.