Azure Security Center Troubleshooting Guide

This guide is for information technology (IT) professionals, information security analysts, and cloud administrators whose organizations are using Azure Security Center and need to troubleshoot Security Center related issues.

Security Center uses the Log Analytics agent to collect and store data. See Azure Security Center Platform Migration to learn more. The information in this article represents Security Center functionality after transition to the Log Analytics agent.

Troubleshooting guide

This guide explains how to troubleshoot Security Center related issues.

Alert types:

  • Virtual Machine Behavioral Analysis (VMBA)
  • Network Analysis
  • SQL Database and Azure Synapse Analytics Analysis
  • Contextual Information

Depending on the alert types, customers can gather the necessary information to investigate the alert by using the following resources:

  • Security logs in the Virtual Machine (VM) event viewer in Windows
  • AuditD in Linux
  • The Azure activity logs, and the enable diagnostic logs on the attack resource.

Customers can share feedback for the alert description and relevance. Navigate to the alert itself, select the Was This Useful button, select the reason, and then enter a comment to explain which explains the feedback. We consistently monitor this feedback channel to improve our alerts.

Audit log

Most of the troubleshooting done in Security Center takes place by first looking at the Audit Log records for the failed component. Through audit logs, you can determine:

  • Which operations were taken place
  • Who initiated the operation
  • When the operation occurred
  • The status of the operation
  • The values of other properties that might help you research the operation

The audit log contains all write operations (PUT, POST, DELETE) performed on your resources, however it does not include read operations (GET).

Log Analytics agent

Security Center uses the Log Analytics agent – this is the same agent used by the Azure Monitor service – to collect security data from your Azure virtual machines. After data collection is enabled and the agent is correctly installed in the target machine, the process below should be in execution:

  • HealthService.exe

If you open the services management console (services.msc), you will also see the Log Analytics agent service running as shown below:


To see which version of the agent you have, open Task Manager, in the Processes tab locate the Log Analytics agent Service, right-click on it and click Properties. In the Details tab, look the file version as shown below:


Log Analytics agent installation scenarios

There are two installation scenarios that can produce different results when installing the Log Analytics agent on your computer. The supported scenarios are:

  • Agent installed automatically by Security Center: in this scenario you will be able to view the alerts in both locations, Security Center and Log search. You will receive email notifications to the email address that was configured in the security policy for the subscription the resource belongs to.

  • Agent manually installed on a VM located in Azure: in this scenario, if you are using agents downloaded and installed manually prior to February 2017, you can view the alerts in the Security Center portal only if you filter on the subscription the workspace belongs to. If you filter on the subscription the resource belongs to, you won't see any alerts. You'll receive email notifications to the email address that was configured in the security policy for the subscription the workspace belongs to.


To avoid the behavior explained in the second scenario, make sure you download the latest version of the agent.

Monitoring agent health issues

Monitoring state defines the reason Security Center is unable to successfully monitor VMs and computers initialized for automatic provisioning. The following table shows the Monitoring state values, descriptions, and resolution steps.

Monitoring state Description Resolution steps
Pending agent installation The Log Analytics agent installation is still running. Installation can take up to a few hours. Wait until automatic installation is complete.
Power state off The VM is stopped. The Log Analytics agent can only be installed on a VM that is running. Restart the VM.
Missing or invalid Azure VM agent The Log Analytics agent is not installed yet. For Security Center to install the extension a valid Azure VM agent is required. Install, reinstall or upgrade the Azure VM agent on the VM.
VM state not ready for installation The Log Analytics agent is not installed yet because the VM is not ready for installation. The VM is not ready for installation due to a problem with the VM agent or VM provisioning. Check the status of your VM. Return to Virtual Machines in the portal and select the VM for status information.
Installation failed - general error The Log Analytics agent was installed but failed due to an error. Manually install the extension or uninstall the extension so Security Center will try to install again.
Installation failed - local agent already installed Log Analytics agent install failed. Security Center identified a local agent (Log Analytics or System Center Operations Manager) already installed on the VM. To avoid multi-homing configuration, where the VM is reporting to two separate workspaces, the Log Analytics agent installation stopped. There are two ways to resolve: manually install the extension and connect it to your desired workspace. Or, set your desired workspace as your default workspace and enable automatic provisioning of the agent. See enable automatic provisioning.
Agent cannot connect to workspace Log Analytics agent installed but failed due to network connectivity. Check that there is internet access or that a valid HTTP proxy has been configured for the agent. See monitoring agent network requirements.
Agent connected to missing or unknown workspace Security Center identified that the Log Analytics agent installed on the VM is connected to a workspace which it doesn't have access to. This can happen in two cases. The workspace was deleted and no longer exists. Reinstall the agent with the correct workspace or uninstall the agent and allow Security Center to complete its automatic provisioning installation. The second case is where the workspace is part of a subscription that Security Center does not have permissions to. Security Center requires subscriptions to allow the Microsoft Security Resource Provider to access them. To enable, register the subscription to the Microsoft Security Resource Provider. This can be done by API, PowerShell, portal or by simply filtering on the subscription in the Security Center Overview dashboard. See Resource providers and types for more information.
Agent not responsive or missing ID Security Center is unable to retrieve security data scanned from the VM, even though the agent is installed. The agent is not reporting any data, including heartbeat. The agent might be damaged or something is blocking traffic. Or, the agent is reporting data but is missing an Azure resource ID so it's impossible to match the data to the Azure VM. To troubleshoot Linux, see Troubleshooting Guide for Log Analytics Agent for Linux. To troubleshoot Windows, see Troubleshooting Windows Virtual Machines.
Agent not installed Data collection is disabled. Turn on data collection in the security policy or manually install the Log Analytics agent.

Troubleshooting monitoring agent network requirements

For agents to connect to and register with Security Center, they must have access to network resources, including the port numbers and domain URLs.

  • For proxy servers, you need to ensure that the appropriate proxy server resources are configured in agent settings. Read this article for more information on how to change the proxy settings.
  • For firewalls that restrict access to the Internet, you need to configure your firewall to permit access to Log Analytics. No action is needed in agent settings.

The following table shows resources needed for communication.

Agent Resource Ports Bypass HTTPS inspection
* 443 Yes
* 443 Yes
* 443 Yes
* 443 Yes

If you encounter onboarding issues with the agent, make sure to read the article How to troubleshoot Operations Management Suite onboarding issues.

Troubleshooting endpoint protection not working properly

The guest agent is the parent process of everything the Microsoft Antimalware extension does. When the guest agent process fails, the Microsoft Antimalware that runs as a child process of the guest agent may also fail. In scenarios like that is recommended to verify the following options:

  • If the target VM is a custom image and the creator of the VM never installed guest agent.
  • If the target is a Linux VM instead of a Windows VM then installing the Windows version of the antimalware extension on a Linux VM will fail. The Linux guest agent has specific requirements in terms of OS version and required packages, and if those requirements are not met the VM agent will not work there either.
  • If the VM was created with an old version of guest agent. If it was, you should be aware that some old agents could not auto-update itself to the newer version and this could lead to this problem. Always use the latest version of guest agent if creating your own images.
  • Some third-party administration software may disable the guest agent, or block access to certain file locations. If you have third-party installed on your VM, make sure that the agent is on the exclusion list.
  • Certain firewall settings or Network Security Group (NSG) may block network traffic to and from guest agent.
  • Certain Access Control List (ACL) may prevent disk access.
  • Lack of disk space can block the guest agent from functioning properly.

By default the Microsoft Antimalware User Interface is disabled, read Enabling Microsoft Antimalware User Interface on Azure Resource Manager VMs Post Deployment for more information on how to enable it if you need.

Troubleshooting problems loading the dashboard

If you experience issues loading the Security Center dashboard, ensure that the user that registers the subscription to Security Center (i.e. the first user one who opened Security Center with the subscription) and the user who would like to turn on data collection should be Owner or Contributor on the subscription. From that moment on also users with Reader on the subscription can see the dashboard/alerts/recommendation/policy.

Contacting Microsoft Support

Some issues can be identified using the guidelines provided in this article, others you can also find documented at the Security Center public Microsoft Q&A page. However if you need further troubleshooting, you can open a new support request using Azure portal as shown below:

Microsoft Support

See also

In this document, you learned how to configure security policies in Azure Security Center. To learn more about Azure Security Center, see the following: