403. Access Denied after ADFS migration

Hello all,

I have completed an ADFS migration from a WIndow server 2008 ENterprise R2 to WIndows SErver 2012 STandard. I performed an in place upgrade, restore and configure the ADFS services.

I have followed Microsoft preparation and migration instructions but it is obvious that I am missing something.

AFter migration, my users were being prompted constantly for credentials on their Outlook client. Also when trying to log into the Office 365 portal, they are unable to authenticate to it, they are instantly redirected to an access denied error as soon as they type their email address. The error is: "403 Forbidden- Access Denied. You do not have permission to view this directory or page with the credentials you provided."

It will be good to mention that before migration, when trying to log into the portal, our users were prompted by our adfs for credentials, but now it redirect us straight to the Access Denied error describe.

Any help on this will be greatly appreciated!
Who is Participating?

Improve company productivity with a Business Account.Sign Up

Vasil Michev (MVP)Connect With a Mentor Commented:
I don't think you must use the exact same OS version, support agents are delirious as usual. Even the official technet documentation supports this:


If you simply upgrade all nodes in a farm one by one, you shouldn't need to update the trust settings later. If you only have a single server and update it, either export/import the token certificates as well or better, run the update cmdlet afterwards.
Vasil Michev (MVP)Commented:
Update the trust settings, as described here:


If it's still not working, check if this happens for both internal and external users. Do the tests on ExRCA: https://testconnectivity.microsoft.com/
LuiLui77Author Commented:
Hello Vasilcho, I was able to revert to the functional ADFS. Thank God I had a DR planned. The upgraded ADFS vm is up but disconnected at the meantime we figure out this issue.

Checking on the certificates of the new ADFS I have found a discrepancy. The information for the Token-Decryption and the Token signing don't match with the certificates that I had on my original server.

Should I try to recreate and apply this certificates first before updating the trust settings?

What do you think?
Has Powershell sent you back into the Stone Age?

If managing Active Directory using Windows Powershell® is making you feel like you stepped back in time, you are not alone.  For nearly 20 years, AD admins around the world have used one tool for day-to-day AD management: Hyena. Discover why.

Vasil Michev (MVP)Commented:
The cmdlet will update the trust with the correct certificate info. If you are using self signed certificates for token signing/issuing, might be a good idea to set it to auto-renew. It's explained in the article above.

If it's still not working after updating the trust, we need to check few more things.
LuiLui77Author Commented:
Hi Vasilcho,

I am planning my DR in case anything goes wrong. To proceed with the cmdlet I will have to turn back on my snapshot that has the upgrade to 2012 and disconnect my current 2008 DR clone (the one currently working).

Now, when issuing the cmdlet on the 2012, and assuming that things don't go too well, I am planning to turn back on my old snapshot with 2008, but since the certificates configuration will be now outdated (since I issued the update federation cmdlet in the new one), what will be my DR plan?

I believe that I will be able to resync federation again with my 2008 by issuing the same cmdlet, but correct me if i am missing something.

I also believe that trust relationship with the DC will fail, so I will have to change the machine password on my DR snapshot as well, but this is another story.

Any help on what to include in my DR plan will be great.
Vasil Michev (MVP)Commented:
If you are simply replacing the server, basically you are building a new farm. And the new certificates and settings you have there will result in different configuration, that you need to sync with O365.

A better approach will be to bring up additional server to the farm, verify connectivity, set it as primary, and then remove the old one. In such scenario, you will not need to update the trust settings, as the certificates and the rest of the configuration will be the same.

Of course in such scenario you will need to have at least one active ADFS server. You *should* be running at least two AD FS servers for redundancy anyway. In case of a disaster, you just bring another one to the farm. There is lot of info regarding this, for example here:

LuiLui77Author Commented:
Yeah that's the thing.
My ultimate purpose is to create a farm between my on-premises ADFS server and my DR cloud server, the thing is that in order to create a farm, both servers will need to have the same version OS (as stated by Office 365 support, correct me if you know that this is not necessary). My cloud server is a Windows Standard 2012 and this was why I performed an in-Place upgrade of my on-premises ADFS server from 2008 R2 to 2012.

I believe that to setup and transfer over the ADFS role to a new server 2012 from scratch will be more time consuming and will involve more downtime.

With all this in mind, Do you think that my best DR option is to connect back my old snapshot and issue the same update federation cmdlet on it?
LuiLui77Author Commented:
Thank you Vasilcho, I will be Updating the Federation with the cmdlet.
Question has a verified solution.

Are you are experiencing a similar issue? Get a personalized answer when you ask a related question.

Have a better answer? Share it in a comment.

All Courses

From novice to tech pro — start learning today.