Skip to main content

Anomaly Detection in Metric Explorer

Overview​

Anomaly detection is a powerful tool in the Metric Explorer that allows users to identify and investigate anomalous behavior or deviations from expected patterns in their metrics. By leveraging sophisticated algorithms and anomaly policies, the Metric Explorer automatically detects anomalies in metric values, highlighting them for further analysis.

The concept behind the anomaly detection feature involves analyzing historical metric data and identifying points that significantly deviate from the expected behavior. This can help users identify critical events, performance bottlenecks, or system failures that may have occurred within their infrastructure. By proactively detecting anomalies, users can swiftly address potential issues, minimize downtime, and ensure optimal performance.

Use-Case​

Let us suppose that you are comparing a metric's value with its past values to understand any infrastructural behaviour changes that you might have noticed recently. While doing so you feel the need to check if the current values of the metric is behaving normally or if it is showing anomalous behaviour, then you can use the anomaly metric explorer option to do so.

Some of the many use-cases for anomaly detection in metric explorer include the following:

  1. Network Security: Anomaly detection can help identify unusual network traffic patterns, indicating potential security breaches or malicious activities.

  2. Server Performance: Detecting anomalies in CPU, memory, or disk usage can help pinpoint performance issues, such as spikes in resource consumption or memory leaks.

  3. Application Monitoring: Monitoring application metrics, such as response time or error rate, can reveal anomalies that indicate performance degradation or anomalies caused by bugs or coding errors.

Utilizing the Metric Explorer as a playground for metric analysis empowers users to take a deep dive into the intricacies of their metrics, spot anomalies, and uncover performance trends that might have otherwise gone unnoticed. It develops a proactive and exploratory approach to infrastructure management, enabling users to harness the full potential of their metrics and extract meaningful insights that drive operational excellence.

Go to Menu, Select Metric Explorer. The Metric Explorer Screen is displayed.

Detecting Anomaly​

Add the metric trend to the metric explorer for which you want to find the anomalous behaviour. After adding the metric to the metric explorer, click on .

Select Anomaly to display the screen showing anomalous behaviour.

The Metric Explorer will apply anomaly detection to the selected metric and visually indicate any detected anomalies on the trend graph. You can hover over the anomaly points to view detailed information about the anomaly, such as the time and value of the anomalous metric value.