Analyze correlated metrics

In this milestone, you examine metrics associated with insights to identify performance issues.

RCA Workbench automatically displays metrics for insights that have fired. By analyzing these metric visualizations, you can understand the magnitude of performance degradation, resource exhaustion, or unusual patterns that contributed to the incident.

To analyze correlated metrics, complete the following steps:

1. In RCA Workbench Timeline view, click an insight in the left panel to view its associated metrics.

2. Examine the metric graph that appears in the timeline for spikes, drops, or unusual patterns.

   Look for sudden spikes, sustained plateaus above normal, or drops to zero. A gradual increase over hours may indicate a resource leak, while a sharp spike suggests a sudden load change or failure.

3. Click Open chart in metrics to see the metric in a dedicated dashboard.

4. Compare metric values during the incident to baseline values before the incident.

5. Click the KPI icon next to an entity to navigate to associated dashboards for comprehensive metric views.

6. Note any metrics that show significant deviation from normal behavior.

   For example, you might see CPU usage spike to 95% when baseline is typically 30%.

You can see metric visualizations highlighting anomalies, with clear visual indicators showing when values deviated from normal patterns.

In the next milestone, you’ll review related logs to identify error messages and patterns.