Edit

Share via

Troubleshoot VM Insights

  • Article

This article provides troubleshooting information to help you with problems that you might experience when you try to enable or use the VM Insights feature in Azure Monitor.

Can't enable VM Insights on a machine

When you onboard an Azure virtual machine (VM) from the Azure portal, the following actions occur:

  • A default Log Analytics workspace is created if you selected that option.
  • The Azure Monitor Agent is installed on the virtual machine through an extension, if the agent isn't already installed.
  • The Dependency Agent is installed on the virtual machine through an extension, if you chose to enable processes and dependencies.

During the onboarding process, each of these steps is verified and a notification status appears in the portal. Configuration of the workspace and the agent installation typically takes 5 to 10 minutes. It takes another 5 to 10 minutes for data to become available to view in the portal.

If you receive a message that the virtual machine needs to be onboarded after you perform the onboarding process, allow up to 30 minutes for the process to finish. If the problem persists, see the following sections for possible causes.

Is the virtual machine running?

If the virtual machine has been turned off for a while, is off currently, or was only recently turned on, you won't have data to display until data arrives.

Is the operating system supported?

If the operating system isn't in the list of supported operating systems, installation of the extension fails and you get a message about waiting for data to arrive.

Important

If a virtual machine that you onboarded on or after April 11, 2022, doesn't appear in VM Insights, you might be running an older version of the Dependency Agent. For more information, see the blog post Potential breaking changes for VM Insights Linux customers. This consideration doesn't apply for Windows machines and for virtual machines that you onboarded before April 11, 2022.

Was the extension installed properly?

If you still see a message that the virtual machine needs to be onboarded, it might mean that one or both of the extensions failed to be installed correctly. In the Azure portal, on the Extensions pane for your virtual machine, check whether the following extensions appear:

Operating system Agents
Windows AzureMonitorWindowsAgent
DependencyAgentWindows
Linux AzureMonitorLinuxAgent
DependencyAgentLinux

If you don't see both extensions for your operating system in the list of installed extensions, you must install them. If the extensions are listed but their status doesn't appear as Provisioning succeeded, remove the extensions and reinstall them.

VM insights DCR was modified

Identify the problem

To identify whether a data collection rule (DCR) was modified, go to the Azure Monitor dashboard and locate the DCR. View the JSON properties by using the JSON View link on the upper-right side of the overview pane.

Screenshot of JSON for a data collection rule.

In this example, the original stream name was changed to reflect the name of the performance counter stream.

Screenshot that shows a changed stream name in a data collection rule.

Although the counter sections point to the Microsoft-Perf table, the stream dataflow is still configured for the proper destination Microsoft-InsightsMetrics.

Resolve the problem

You can't resolve this problem by using the Azure Monitor dashboard directly. But you can resolve it by exporting the template, normalizing the name, and then importing the modified rule over itself.

Export the DCR and save locally

  1. On the Azure Monitor dashboard, go to the DCR.

  2. Select Export template.

    Screenshot of the menu option to export a template for a data collection rule.

  3. On the Export template pane for the selected DCR, the portal creates the template file and a matching parameter file. After this process is complete, select Download to download the template package and save it locally.

    Screenshot that shows the button for downloading a template for a data collection rule.

  4. Select Open file.

    Screenshot that shows the link for opening a downloaded file package.

  5. Copy the two files to a local folder.

    Screenshot that shows parameter and template files copied to a local folder.

Modify the template

  1. Open the template file in the editor of your choice. Then locate the invalid stream name under the performance counter data source.

    Screenshot of a template file.

  2. By using the valid stream name from the dataflow node, fix the invalid reference. Then save and close your file.

    Screenshot that shows an updated stream name in a template file.

Import the template by using the custom deployment feature

  1. Back in the portal, search for and then select Deploy a custom template.

    Screenshot of search results for the service that deploys a custom template.

  2. On the Custom deployment pane, select Build your own template in the editor.

    Screenshot that shows the option to build a template in the editor.

  3. Select Load file, and then browse to your saved template and parameter files.

    Screenshot that shows the button for loading a file.

  4. Visually inspect the template to validate that the change is in place, and then select Save.

    Screenshot of the pane for editing a template.

  5. The portal uses the parameter file to fill in the deployment options (which can be changed or left intact to overwrite the existing DCR). When the portal completes the process, select Review + create.

    Screenshot of the pane for custom deployment.

  6. After validation, select Create to finish the deployment.

    Screenshot that shows the button for creating a custom deployment.

  7. After the deployment is complete, browse to the DCR again and review the JSON in the overview pane.

    Screenshot that shows the pane for reviewing JSON.

  8. The agent detects the change and downloads the new configuration, which restores ingestion to the Microsoft-InsightsMetrics table.

Performance view has no data

If the agents appear to be installed correctly but no data appears in the Performance view, see the following sections for possible causes.

Has your Log Analytics workspace reached its data limit?

Check the capacity reservations and the pricing for data ingestion.

Is your virtual machine agent connected to Azure Monitor Logs?

In the Azure portal, on the Azure Monitor menu, select Logs to open the Log Analytics workspace. Run the following query for your computer:

Heartbeat
| where Computer == "my-computer"
| sort by TimeGenerated desc

If no data appears or if the computer didn't send a heartbeat recently, you might have problems with your agent. For agent troubleshooting information, see these articles:

Virtual machine doesn't appear in the Map view

The following sections help you resolve problems with the Map view.

Is the Dependency Agent installed?

Use the information in the preceding sections to determine if the Dependency Agent is installed and working properly.

Are you on the Log Analytics free tier?

The Log Analytics free tier is a legacy pricing plan that allows for up to five unique Service Map machines. Any subsequent machines won't appear in Service Map, even if the prior five are no longer sending data.

Is your virtual machine sending log and performance data to Azure Monitor Logs?

Use the log query in the Performance view has no data section to determine if data is being collected for the virtual machine. If no data is being collected, use the TestCloudConnectivity tool to determine if you have connectivity problems.

Virtual machine appears in the Map view but has missing data

If the virtual machine is in the Map view and the Dependency Agent is installed and running, but the kernel driver didn't load, check the log file at the following locations:

Operating system Log
Windows C:\Program Files\Microsoft Dependency Agent\logs\wrapper.log
Linux /var/opt/microsoft/dependency-agent/log/service.log

The last lines of the file should indicate why the kernel didn't load. For example, the kernel might not be supported on Linux if you updated your kernel.

Related content