Azure App Service metrics unavailable: empty metric definitions and resource mapping errors

Platform metrics are unavailable for our Azure App Service plan. The issue occurs in the Azure portal, Azure mobile app, Azure CLI, and a direct REST API request.

Environment

• Region: Brazil South
• OS: Windows
• Tier: Basic B2, one instance
• Plan status: Ready
• Web app state: Running

The issue was observed on October 30, 2026. The exact start time is unknown.

Symptoms

CPU Percentage, Memory Percentage, Data In, and Data Out are unavailable. Portal charts display “Error retrieving data”, and the plan’s metric selector shows “No results found.”

Troubleshooting performed

Subscription and providers: Confirmed the correct subscription. Both Microsoft.Insights and Microsoft.Web are Registered.

Permissions: Queried Microsoft.Authorization/permissions on the plan. The response includes actions: ["*"] and actions: ["*/read", ...], with empty notActions arrays.

Resource identity: Verified the web app’s resource ID and serverFarmId through az webapp show. The app is Running and references the expected plan.

Plan metric definitions: az monitor metrics list-definitions returns an empty array ([]), both with and without the explicit Microsoft.Web/serverfarms namespace.

Plan metric values: Queried CpuPercentage with Average aggregation over:

• The last 24 hours, using a 1-hour interval.
  • The last 30 days, using a 1-day interval.
  Both return:

        (BadRequest) Failed to find metric configuration for provider:
  

Microsoft.Web, resource Type: serverfarms, metric: CpuPercentage, Valid metrics: ```

  The valid metrics list is empty.
  
  **Direct REST request:** A GET request to:
  
  ```yaml
  {plan-resource-id}/providers/Microsoft.Insights/metrics
  ```
  
  Using API version `2018-01-01`, explicit namespace `Microsoft.Web/serverfarms`, `CpuPercentage`, `Average`, `PT1H`, and an explicit 24-hour UTC timespan returns the same `BadRequest`.
  
  **Web app metric definitions:** For `Microsoft.Web/sites`, the API returns only three definitions:
  
  ```sql
  MemoryWorkingSet

AverageMemoryWorkingSet InstanceCount ```

  **Web app metric values:** Querying the advertised `MemoryWorkingSet` metric with `Average` aggregation over 24 hours returns:
  
  ```yaml
  (NotFound) Failed to find resource mapping for provider:

Microsoft.Web, resourceType: sites, id: {web-app-resource-id} ```

  **Service Health:** The portal displayed a resolved App Service maintenance notification for Brazil South affecting resources in our subscription. However, a separate Service Health Events API query with `queryStartTime` set to October 23 returned an empty event list and no next page. We have not established whether the maintenance is related.
  

No provider re-registration, plan resizing, application restarts, or configuration changes were performed during these tests. We have not established whether historical metrics are missing or simply inaccessible.

Questions

• What could cause empty metric definitions for an existing B2 plan and a resource-mapping error for its running web app?
• Is there a documented, non-disruptive way to validate or repair the monitoring mapping/configuration?
• Could this be related to the recent App Service maintenance in Brazil South?

I do not have a paid technical support plan. Any additional read-only diagnostic steps would be appreciated. Screenshots and sanitized command output are available.Platform metrics are unavailable for our Azure App Service plan. The issue occurs in the Azure portal, Azure mobile app, Azure CLI, and a direct REST API request.