Automatic Anomaly Detection
  • 19 Mar 2024
  • 1 Minute to read
  • Dark
    Light
  • PDF

Automatic Anomaly Detection

  • Dark
    Light
  • PDF

Article Summary

Introduction

The Cost Analyzer offers real-time visibility into your Azure expenses and utilizes your historical data to spot any unusual cost occurrences. It automatically identifies any spikes in subscription costs within the specified time intervals, like monthly or daily, at the group level. This feature includes the benefit of alerting the user whenever anomalies are detected.

Anomaly detection is automatically activated upon creating a cost management group, alerting the user in case of of any anomalies detected.

Configuration

Follow the below steps to manage anomaly detection configuration at the group level:

  1. Go to the Monitoring section within a Cost group.

The Automatic anomaly detection widget displays a chart of the monthly usage trend, with red points denoting any anomalies.

anomaly detection cost trend chart.PNG

  1. Click the Manage option and specify the preferred interval in the Frequency settings.
  2. Users have the flexibility to choose either one or both intervals according to their reporting preferences and needs.
  3. In the Alert settings section, select the desired channel for receiving anomaly alerts.
  4. Finally, click on Save to apply changes.

manage the cost anomaly detection trend .gif

Cost anomaly report

Below is the Cost anomaly report received via Turbo360 mail. This comprehensive report outlines potential root causes and the overall cost variation detected on the specified date.

The cost anomaly report contains a navigation link that directs users to explore the cost expenditure within the cost analysis section of the corresponding group.

anomaly report.PNG

Alert incidents

When anomalies are detected within the specified interval, alerts are logged as incidents. Users can review these incidents by clicking the View Incidents option located in the anomaly detection widget.

view incidents.png

cost anomaly alerts.PNG


Was this article helpful?