Azure Costs Analysis preview with Anomaly detection

Microsoft has released new functionality in public preview to manage the costs of Azure subscriptions. This new functionality gives excellent insights into the costs and lets you detect any anomaly of the expenses on those subscriptions.

New Cost Analysis Preview

The new cost analysis preview blade shows the Azure resources in a different overview than before. This new overview gives you more insight into your data and your costs.

Azure Cost Analysis Preview

To use the cost analysis preview, go to the following link. On this blade, select the appropriate scope.

Cost Analysis Preview

The overview offers the option to report on and analyze your cloud costs and review critical insights to understand better and control spending patterns. This is accomplished on four levels:

  • Resource Groups
  • Subscriptions
  • Services
  • Reservations

All four overviews offer great insights. I love the services overview where you see the costs per service and can view the expenses underneath it. For example, bandwidth costs and VPN gateway costs.

Services costs analysis

Anomaly detection

The costs analysis preview now also includes anomaly detection. To better understand the functionality, we should also know what an anomaly is. If we look into a dictionary, we will find something like:

  • Something different, abnormal, peculiar, or not easily classifiedsomething anomalous.
  • Deviation from the common ruleirregularity

It means that a cost increase of 5000 euros could be an anomaly regarding cost management. But that would not be the case if the billing was raised monthly, for example.

After selecting a specific view for the first time, the preview is enabled, and you will get more insights regarding the Azure costs. When using the preview for the first time, you will see the below screenshot.
Opening up the blade after a day will show any anomaly, if there are any.

Leave a Reply

Your email address will not be published.

This site uses Akismet to reduce spam. Learn how your comment data is processed.