azurerm_backup_policy_vm: duration_type in archived_restore_point is ignored/nonexistant, always shown as Months in Azure Portal

[Alxtp]

Is there an existing issue for this?

• I have searched the existing issues

Community Note

• Please vote on this issue by adding a 👍 reaction to the original issue to help the community and maintainers prioritize this request
• Please do not leave comments along the lines of "+1", "me too" or "any updates", they generate extra noise for issue followers and do not help prioritize the request
• If you are interested in working on this issue or have submitted a pull request, please leave a comment and review the contribution guide to help.

Terraform Version

v1.15.9

AzureRM Provider Version

v5.2.0

Affected Resource(s)/Data Source(s)

azurerm_backup_policy_vm

Affected Location(s)

switzerlandnorth

Terraform Configuration Files

resource "azurerm_resource_group" "example" {
  name     = "tfex-recovery_vault"
  location = "switzerlandnorth"
}

resource "azurerm_recovery_services_vault" "example" {
  name                = "tfex-recovery-vault"
  location            = azurerm_resource_group.example.location
  resource_group_name = azurerm_resource_group.example.name
  sku                 = "Standard"
}

resource "azurerm_backup_policy_vm" "example" {
  name                = "tfex-recovery-vault-policy"
  resource_group_name = azurerm_resource_group.example.name
  recovery_vault_name = azurerm_recovery_services_vault.example.name

  timezone = "UTC"

  backup {
    frequency = "Daily"
    time      = "23:00"
  }

  retention_yearly {
    count    = 77
    weekdays = ["Sunday"]
    weeks    = ["Last"]
    months   = ["January"]
  }

  tiering_policy {
    archived_restore_point {
      mode          = "TierAfter"
      duration      = 90
      duration_type = "Days"
    }
  }
}

Debug Output/Panic Output

Runs successfully

Expected Behaviour

Setting duration_type = "Days" with duration = 90 should configure the tiering policy to move recovery points to archive tier after 90 days. If it's not supported by the API, it should throw a error.

Actual Behaviour

Regardless of the duration_type configured (Days, Weeks, Months, Years), the Azure Portal always displays and applies the value as Months. With duration = 90, this results in the portal showing "90 Months" and throwing a validation error:

Steps to Reproduce

1. Create an azurerm_backup_policy_vm resource with a tiering_policy block using duration_type = "Days" and duration = 90.
2. Run terraform apply

Important Factoids

No response

References

No response

[v-yhyeo0202]

@Alxtp it is verified that after applying the Terraform configurations, the portal shows 90 months instead of days when trying to modify the backup policies (though in my case, there is no validation error). By checking the HTTP traffic via proxy, seems that the REST API requests sent by Terraform are correct during resource creation. Running terraform refresh also yields the correct response which matches the number of days set in Terraform configurations. If updating "move eligible recovery points after" to a certain number of months in portal and run terraform refresh, the response obtained will have duration_type property equal to Months and duration property equal to the correct number, showing that there is no issue with REST API used in this repository. The mismatch you face may be caused by portal issue. You may open Azure support ticket via portal and quote the correlation ID.