During a recent SBS 2003 to Windows Server 2012 r2 Essentials migration I was having a number of problems setting up anywhere access. The two problems that I was getting were:
- Anywhere access to your server is blocked
- VPN was not configured successfully
I searched all over but couldn’t come up with anything useful and the dashboard.log file didn’t show anything useful, just the error:
ConnectivityCenter: RemoteAccessAnalyzer: VPN server deployment result: InstallationFailed
The ports 80 and 443 required for Remote Web Access and VPN were without doubt open but it seemed to me that the wizard always failed at the verification stage, so I figured that it must be seeing something different than I was picking up in my tests.
Now, this server has been configured with a custom domain name of remote.companydomain.com and externally this resolved no problems at all.
But if I tried to ping that address or use nslookup on the essentials server to resolve the external DNS address that had been assigned to the server it failed. Then I realized what was going on.
This system/customer is new to me and I had made the new essentials server a domain controller which also meant that it would have replicated DNS records from the source domain controller.
Although the internal domain name was companyname.local there was also a forward lookup zone for the main companyname.com address space, because the essentials server was rightly using itself as a DNS server it was unable to resolve remote.companyname.com using its own DNS forward lookup zone for the external domain as the record for remote.companyname.com had not been added.
So I had two choices, remove the forward lookup zone or add the remote.companyname.com address to it as an A record. It was obvious after a little investigation why the external domain name had been added to the active directory DNS, but the reason it was include it was no longer appropriate, so I deleted the whole zone.
I then restarted the DNS server and cleared the DNS cache from a command prompt with:
ipconfig /flushdns
and tried to resolve the external domain name again and this now resolved correctly.
So I once again tried to run the Windows Essentials Anywhere Access Wizard and it still failed with the exact same issue! Infuriating! .
It did show one less error this time though and I was down to “VPN was not configured successfully” only this time. This error took make ages to figure out though.
Fixing the VPN was not configured successfully error
I ended up looking through the CBS.LOG file in quite some detail and figured out what the problem was. Thanks to someone on this thread on technet for the inspiration to look there.
When the VPN is being enabled Windows Essentials attempts to setup a Windows Internal Database and in my case failed to do so due to a logon failure as it did not have the “Logon as a Service” privilege. Here is a sample of the log:
2014-08-29 23:04:32, Error CSI 00000001 (F) Logged @2014/8/29:22:04:32.768 : [ml:130{65},l:128{64}]"Attempting to start service {MSSQL$MICROSOFT##WID} synchronously" [gle=0x80004005] 2014-08-29 23:04:32, Error CSI 00000002 (F) Logged @2014/8/29:22:04:32.893 : [ml:84{42},l:82{41}]"start service MSSQL$MICROSOFT##WID (1069)" [gle=0x80004005] 2014-08-29 23:04:32, Error CSI 00000003@2014/8/29:22:04:32.893 (F) CMIADAPTER: Inner Error Message from AI HRESULT = HRESULT_FROM_WIN32(1069) [ [51]"The service did not start due to a logon failure.
Now I felt I was getting close to the source of the issue, light at the end of the tunnel (you have no idea how long I spent trying to figure this out).
The knowledge base article 2832204 from Microsoft explains how to make the appropriate group policy changes to stop this prevent the specific issue of being unable to start the MSSQL$MICROSOFT##WID service, which is actually the root cause of the wizard not being able to finish.
Here is a screenshot of how I the corrected group policy looked on the system I was working on:
After you have changed the policy run a quick “gpupdate” to make the new policy active and then try running the VPN wizard again, the wizard took a long time to complete but I was fairly confident it was going to work as I could see the SQL database becoming active in Task Manager:
The wizard finally finished with success! I spent a long time trying to figure out the cause of this problem and I am so glad to see the back of it. I really hope this helps you if you are having the same issue.
Adam says
Very nice and very well done!!
The Logon as a service issues was what was stopping VPN installing. Still reports that remote access is blocked even though the remote URL resolves fine on the box and works fine remotely – so just ignoring that!
Paulie says
Cheers Adam – glad it was also fixed for you.
JD says
Thanks. This was driving me crazy. I’ve been throwing the WSE role on standard 2012 R2 servers to provide the RD Gateway service and the very cool Client Restore Services for about a year now. The CRS feature in particular is one of my new fav’s. With CRS active throughout the environment and high-frequency backups on the server shares, I’m no longer concerned about the Crypto BS.
Thanks for the NT SERVICE\All SERVICES GP tip.
Here’s another tip. If you want to take advantage of the WSE role on standard servers, make sure the DHCP role is active before installing the WSE role. Otherwise, you’ll run into a port conflict.
Joost says
Hey this worked for me… I also had found something about changing Security options for FIPs but that ended up breaking me as well. Great post and thanks for sharing!