Known issues with Azure Monitor for VMs (preview)
This article covers known issues with Azure Monitor for VMs, a solution in Azure that combines health, discovery of application components, and performance monitoring of the Azure VM operating system.
The following are known issues with the current release of the Health feature:
- If an Azure VM is removed or deleted, it's displayed in the VM list view for sometime. Additionally, clicking the state of a removed or deleted VM opens the Health Diagnostics view and then initiates a loading loop. Selecting the name of the deleted VM opens a pane with a message stating that the VM has been deleted.
- Configuration changes, such as updating a threshold, take up to 30 minutes even if the portal or Workload Monitor API might update them immediately.
- The Health Diagnostics experience updates faster than the other views. The information might be delayed when you switch between them.
- Shutting down VMs updates some of health criteria to critical and others to healthy. The net VM state is displayed as critical.
- For Linux VMs, the title of the page listing the health criteria for a single VM view has the entire domain name of the VM instead of the user-defined VM name.
- After you disable monitoring for a VM using one of the supported methods and you try deploying it again, you should deploy it in the same workspace. If you choose a different workspace and try to view the health state for that VM, it might show inconsistent behavior.
- Total CPU Utilization health criterion for Windows shows a threshold of not equal to 4, meaning CPU utilization is greater than 95% and system queue length is greater than 15. This health criterion is not configurable in this preview.
- After removing the solution components from your workspace, you may continue to see health state from your Azure VMs; specifically performance and map data when you navigate to either view in the portal. Data will eventually stop appearing from the Performance and Map view after sometime; however the Health view will continue to show health status for your VMs. The Try now option will be available to re-onboard from Performance and Map views only.
To understand the requirements and methods for enabling monitoring of your virtual machines, review Deploy Azure Monitor for VMs.