Although the total daily cost for the specific service did not meet the required threshold, the alert was still triggered.
The policy is evaluated and alerts are generated based on the available cost data.
Specifically, for policies that monitor periodic cost differences, the evaluation engine accounts for typical delays in data reporting from one or more subscriptions or accounts.
To manage this, the policy engine adjusts the evaluation date backward until it identifies a day where complete data from all accounts and subscriptions is present, or up to a maximum of five days prior, whichever is sooner.
This approach ensures that the policy, when evaluating cost differences, uses the most recent and comprehensive dataset available from the cloud provider covering all subscriptions and accounts.
The assessment relies on the most complete dataset, leveraging the backward adjustment mechanism to maintain data integrity.
This ensures reliable and accurate cost anomaly detection in the presence of partial or delayed data reporting.