Administrators configuring a Windows Server Routing and Remote Access Service (RRAS) server to support Windows 10 Always On VPN connections may encounter an issue where the RemoteAccess service fails to start. Attempts to start the service might seem to work at first, but the service immediately stops again.
Troubleshooting
On the RRAS server, the Services management console (services.msc) or PowerShell Get-Service command shows the RemoteAccess service as being stopped. Attempts to start the service result in failure.
Event Log
Looking at the System event log on the RRAS server shows an error with event ID 7024 from the Service Control Manager source indicating “The Routing and Remote Access service terminated with the following service-specific error: A device attached to the system is not functioning.”
Resolution
This issue is commonly caused when IPv6 is disabled on the server via the registry. To verify, open the registry editor on the RRAS server and navigate to the following location.
HKLM\SYSTEM\CurrentControlSet\Services\Tcpip6\Parameters
If the DisabledComponents value is present and set to anything other than 0, set it to 0 or simply delete the DisabledComponents value completely and reboot the server.
The following PowerShell command can be used to remove the DisabledComponents value.
Remove-ItemProperty HKLM:\SYSTEM\CurrentControlSet\Services\Tcpip6\Parameters -Name DisabledComponents
Additional Information
IPv6 Recommended Reading for Always On VPN and DirectAccess Administrators
Guidance for Configuring IPv6 in Windows for Advanced Users (Microsoft)
shrimoutlookcom
/ May 28, 2021Hi Richard,, We have two ALWAYSONVPN Servers. We see clients connects to both servers. If we stop the Services RRAS in one server, clients are not failing over to other. Any suggestions.
Richard M. Hicks
/ May 28, 2021How are you load balancing? NLB or external load balancer? What VPN protocol are you using when this happens?
shrimoutlookcom
/ May 31, 2021We are using Fortinet firewall to load balance. User Level VPN is using SSTP.
Richard M. Hicks
/ June 4, 2021Thanks. Most load balancers will have a setting that sends a TCP reset (RST) to the client when the real server goes down. That will cause the client to restart the connection. If you can find a setting like that on the Fortinet that should help here.